DS Capstone.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.7.1"
    },
    "colab": {
      "name": "DS Capstone.ipynb",
      "provenance": []
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PS--vcQoGel3"
      },
      "source": [
        "# Capstone Project (Real or Not? NLP with Disaster Tweets)\n",
        "\n",
        "Predict which Tweets are about real disasters and which ones are not."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "h1R3xCs_Gel4"
      },
      "source": [
        "## Project Overview\n",
        "\n",
        "Twitter has become an important communication channel in times of emergency.\n",
        "The ubiquitousness of smartphones enables people to announce an emergency they’re observing in real-time. Because of this, more agencies are interested in programatically monitoring tweets(i.e. disaster relief organizations and news agencies) but, it’s not always clear whether a person’s words are actually announcing a disaster.\n",
        "\n",
        "Take an example of a tweet *'On plus side LOOK AT THE SKY LAST NIGHT IT WAS ABLAZE'*.The author of the tweet explicitly uses the word “ABLAZE” but means it metaphorically. This is clear to a human right away, especially with the visual aid but it’s less clear to a machine.\n",
        "\n",
        "Kaggle hosted this challenge on their platform and dataset was created by the company figure-eight originally shared on their ‘Data For Everyone’ website here."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "L33CGTcEGel5"
      },
      "source": [
        "\n",
        "## The Problem\n",
        "\n",
        "The problem to be solved by this project is how to identify which tweets are about real disasters and which one’s aren’t.This project will build a machine learning model to predict which Tweets are about real disasters and which one’s aren’t.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dcdIrTwYGel6"
      },
      "source": [
        "# 1. Analysis "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ft0KdNbBGel6"
      },
      "source": [
        "## Part I: Business understanding\n",
        "\n",
        "Objective of the capstone project is to understand the business domain, analyze and explore the given data set then train and fit the a machine learning model and finally classify given test tweets data into announcing disaster or not.Disaster relief organizations and News Agencies would be the main users of this kind of application. The data is provided by company figure-eight which contains about 10000 tweets and they are hand labeled."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "shYMTv5cGel7"
      },
      "source": [
        "## Part II: Data Exploration\n",
        "\n",
        "Given dataset contains 3 files:\n",
        "- `train.csv` - the training set\n",
        "- `test.csv` - the test set\n",
        "- `sample_submission.csv` - a sample submission file in the correct format\n",
        "\n",
        "Each sample in the train and test set has the following information:\n",
        "- The text of a tweet\n",
        "- A keyword from that tweet (although this may be blank!)\n",
        "- The location the tweet was sent from (may also be blank)\n",
        "\n",
        "Features in the files:\n",
        "- `id` - a unique identifier for each tweet\n",
        "- `text` - the text of the tweet\n",
        "- `location` - the location the tweet was sent from (may be blank)\n",
        "- `keyword` - a particular keyword from the tweet (may be blank)\n",
        "- `target` - in train.csv only, this denotes whether a tweet is about a real disaster (1) or not (0)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wC3UGEtGGel8",
        "outputId": "0bd87660-805b-4727-e0aa-9378106ca1d3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "#Importing the required libraries to read,visualize and model the givn dataset files\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "import plotly.graph_objects as go\n",
        "import plotly.express as px\n",
        "import missingno as msno \n",
        "import warnings\n",
        "warnings.filterwarnings(\"ignore\")\n",
        "import re\n",
        "import re\n",
        "import string\n",
        "import nltk\n",
        "from nltk.corpus import stopwords\n",
        "from sklearn import model_selection\n",
        "from sklearn.feature_extraction.text import CountVectorizer,TfidfVectorizer\n",
        "from sklearn.model_selection import train_test_split,GridSearchCV\n",
        "from sklearn.naive_bayes import MultinomialNB\n",
        "from sklearn.metrics import f1_score\n",
        "from sklearn.metrics import classification_report, confusion_matrix, accuracy_score\n",
        "from sklearn.pipeline import Pipeline\n",
        "from sklearn.feature_extraction.text import TfidfTransformer\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.svm import SVC\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "from sklearn.tree import DecisionTreeClassifier\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from sklearn.multioutput import MultiOutputClassifier\n",
        "from sklearn.model_selection import RepeatedStratifiedKFold\n",
        "import nltk\n",
        "from nltk.tokenize import word_tokenize,RegexpTokenizer\n",
        "from nltk.stem import WordNetLemmatizer\n",
        "from platform import python_version\n",
        "print (python_version())"
      ],
      "execution_count": 131,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "3.6.9\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9C8fM6dEK-Xd",
        "outputId": "02567fd6-b3a7-411d-9806-4225a1cd6b84",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 850
        }
      },
      "source": [
        "nltk.download(\"popular\")"
      ],
      "execution_count": 132,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[nltk_data] Downloading collection 'popular'\n",
            "[nltk_data]    | \n",
            "[nltk_data]    | Downloading package cmudict to /root/nltk_data...\n",
            "[nltk_data]    |   Package cmudict is already up-to-date!\n",
            "[nltk_data]    | Downloading package gazetteers to /root/nltk_data...\n",
            "[nltk_data]    |   Package gazetteers is already up-to-date!\n",
            "[nltk_data]    | Downloading package genesis to /root/nltk_data...\n",
            "[nltk_data]    |   Package genesis is already up-to-date!\n",
            "[nltk_data]    | Downloading package gutenberg to /root/nltk_data...\n",
            "[nltk_data]    |   Package gutenberg is already up-to-date!\n",
            "[nltk_data]    | Downloading package inaugural to /root/nltk_data...\n",
            "[nltk_data]    |   Package inaugural is already up-to-date!\n",
            "[nltk_data]    | Downloading package movie_reviews to\n",
            "[nltk_data]    |     /root/nltk_data...\n",
            "[nltk_data]    |   Package movie_reviews is already up-to-date!\n",
            "[nltk_data]    | Downloading package names to /root/nltk_data...\n",
            "[nltk_data]    |   Package names is already up-to-date!\n",
            "[nltk_data]    | Downloading package shakespeare to /root/nltk_data...\n",
            "[nltk_data]    |   Package shakespeare is already up-to-date!\n",
            "[nltk_data]    | Downloading package stopwords to /root/nltk_data...\n",
            "[nltk_data]    |   Package stopwords is already up-to-date!\n",
            "[nltk_data]    | Downloading package treebank to /root/nltk_data...\n",
            "[nltk_data]    |   Package treebank is already up-to-date!\n",
            "[nltk_data]    | Downloading package twitter_samples to\n",
            "[nltk_data]    |     /root/nltk_data...\n",
            "[nltk_data]    |   Package twitter_samples is already up-to-date!\n",
            "[nltk_data]    | Downloading package omw to /root/nltk_data...\n",
            "[nltk_data]    |   Package omw is already up-to-date!\n",
            "[nltk_data]    | Downloading package wordnet to /root/nltk_data...\n",
            "[nltk_data]    |   Package wordnet is already up-to-date!\n",
            "[nltk_data]    | Downloading package wordnet_ic to /root/nltk_data...\n",
            "[nltk_data]    |   Package wordnet_ic is already up-to-date!\n",
            "[nltk_data]    | Downloading package words to /root/nltk_data...\n",
            "[nltk_data]    |   Package words is already up-to-date!\n",
            "[nltk_data]    | Downloading package maxent_ne_chunker to\n",
            "[nltk_data]    |     /root/nltk_data...\n",
            "[nltk_data]    |   Package maxent_ne_chunker is already up-to-date!\n",
            "[nltk_data]    | Downloading package punkt to /root/nltk_data...\n",
            "[nltk_data]    |   Package punkt is already up-to-date!\n",
            "[nltk_data]    | Downloading package snowball_data to\n",
            "[nltk_data]    |     /root/nltk_data...\n",
            "[nltk_data]    |   Package snowball_data is already up-to-date!\n",
            "[nltk_data]    | Downloading package averaged_perceptron_tagger to\n",
            "[nltk_data]    |     /root/nltk_data...\n",
            "[nltk_data]    |   Package averaged_perceptron_tagger is already up-\n",
            "[nltk_data]    |       to-date!\n",
            "[nltk_data]    | \n",
            "[nltk_data]  Done downloading collection popular\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "True"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 132
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ewMqBDHdGemA",
        "outputId": "64995bc5-cc03-4796-c6dc-624e24108f92",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "# Read the dataset csv files and create pandas datframes\n",
        "train_df=pd.read_csv(\"https://raw.githubusercontent.com/pavstar619/Data-Science-Capstone/master/train.csv\")\n",
        "test_df=pd.read_csv(\"https://raw.githubusercontent.com/pavstar619/Data-Science-Capstone/master/test.csv\")\n",
        "print(\"Train and Test data sets are imported successfully\")\n",
        "#X=train_df.text\n",
        "#y=train_df.target"
      ],
      "execution_count": 133,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train and Test data sets are imported successfully\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yx1H3KTvGemG"
      },
      "source": [
        "# Define a function to explore the train and test dataframes\n",
        "def explore_data(df):\n",
        "    \n",
        "    '''Input- df= pandas dataframes to be explored\n",
        "       Output- print shape, info and first 5 records of the dataframe \n",
        "    '''\n",
        "    \n",
        "    print(\"-\"*50)\n",
        "    print('Shape of the dataframe:',df.shape)\n",
        "    print(\"Number of records in train data set:\",df.shape[0])\n",
        "    print(\"Information of the dataset:\")\n",
        "    df.info()\n",
        "    print(\"-\"*50)\n",
        "    print(\"First 5 records of the dataset:\")\n",
        "    return df.head()\n",
        "    print(\"-\"*50)"
      ],
      "execution_count": 134,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EnTrd1loGemJ",
        "outputId": "233db846-abf5-4dfb-d0ac-f6402f3b4aff",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 510
        }
      },
      "source": [
        "# Lets use explore_data() function to explore train data\n",
        "explore_data(train_df)"
      ],
      "execution_count": 135,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--------------------------------------------------\n",
            "Shape of the dataframe: (7613, 5)\n",
            "Number of records in train data set: 7613\n",
            "Information of the dataset:\n",
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 7613 entries, 0 to 7612\n",
            "Data columns (total 5 columns):\n",
            " #   Column    Non-Null Count  Dtype \n",
            "---  ------    --------------  ----- \n",
            " 0   id        7613 non-null   int64 \n",
            " 1   keyword   7552 non-null   object\n",
            " 2   location  5080 non-null   object\n",
            " 3   text      7613 non-null   object\n",
            " 4   target    7613 non-null   int64 \n",
            "dtypes: int64(2), object(3)\n",
            "memory usage: 297.5+ KB\n",
            "--------------------------------------------------\n",
            "First 5 records of the dataset:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>id</th>\n",
              "      <th>keyword</th>\n",
              "      <th>location</th>\n",
              "      <th>text</th>\n",
              "      <th>target</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Our Deeds are the Reason of this #earthquake M...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>4</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Forest fire near La Ronge Sask. Canada</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>5</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>All residents asked to 'shelter in place' are ...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>6</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>13,000 people receive #wildfires evacuation or...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>7</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Just got sent this photo from Ruby #Alaska as ...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   id keyword  ...                                               text target\n",
              "0   1     NaN  ...  Our Deeds are the Reason of this #earthquake M...      1\n",
              "1   4     NaN  ...             Forest fire near La Ronge Sask. Canada      1\n",
              "2   5     NaN  ...  All residents asked to 'shelter in place' are ...      1\n",
              "3   6     NaN  ...  13,000 people receive #wildfires evacuation or...      1\n",
              "4   7     NaN  ...  Just got sent this photo from Ruby #Alaska as ...      1\n",
              "\n",
              "[5 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 135
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_mUZeld6GemM",
        "outputId": "38b06ba2-f91a-493a-9cc2-5de7f58459cf",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 493
        }
      },
      "source": [
        "# Lets use explore_data() function to explore test data\n",
        "explore_data(test_df)"
      ],
      "execution_count": 136,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--------------------------------------------------\n",
            "Shape of the dataframe: (3263, 4)\n",
            "Number of records in train data set: 3263\n",
            "Information of the dataset:\n",
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 3263 entries, 0 to 3262\n",
            "Data columns (total 4 columns):\n",
            " #   Column    Non-Null Count  Dtype \n",
            "---  ------    --------------  ----- \n",
            " 0   id        3263 non-null   int64 \n",
            " 1   keyword   3237 non-null   object\n",
            " 2   location  2158 non-null   object\n",
            " 3   text      3263 non-null   object\n",
            "dtypes: int64(1), object(3)\n",
            "memory usage: 102.1+ KB\n",
            "--------------------------------------------------\n",
            "First 5 records of the dataset:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>id</th>\n",
              "      <th>keyword</th>\n",
              "      <th>location</th>\n",
              "      <th>text</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Just happened a terrible car crash</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Heard about #earthquake is different cities, s...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>3</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>there is a forest fire at spot pond, geese are...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>9</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Apocalypse lighting. #Spokane #wildfires</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>11</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Typhoon Soudelor kills 28 in China and Taiwan</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   id keyword location                                               text\n",
              "0   0     NaN      NaN                 Just happened a terrible car crash\n",
              "1   2     NaN      NaN  Heard about #earthquake is different cities, s...\n",
              "2   3     NaN      NaN  there is a forest fire at spot pond, geese are...\n",
              "3   9     NaN      NaN           Apocalypse lighting. #Spokane #wildfires\n",
              "4  11     NaN      NaN      Typhoon Soudelor kills 28 in China and Taiwan"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 136
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7BNboGutGemQ"
      },
      "source": [
        "It seems train and test dataset has 'NaN' values for `keyword` and `location` columns. Let's explore the dataset further for missing data values."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "l3Cq7JjCGemQ"
      },
      "source": [
        "#Calculate count and percentage of missing values in the dataframe\n",
        "\n",
        "def missing_values(df):\n",
        "    \n",
        "    '''Input- df=pandas dataframe\n",
        "       Output- print missing records count and % of the input dataframe and visualize using MSNO\n",
        "    '''\n",
        "    \n",
        "    print(\"Number of records with missing location:\",df.location.isnull().sum())\n",
        "    print(\"Number of records with missing keywords:\",df.keyword.isnull().sum())\n",
        "    print('{}% of location values are missing from Total Number of Records.'.format(round((df.location.isnull().sum())/(df.shape[0])*100),2))\n",
        "    print('{}% of keywords values are missing from Total Number of Records.'.format(round((df.keyword.isnull().sum())/(df.shape[0])*100),2))\n",
        "    msno.matrix(df);"
      ],
      "execution_count": 137,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "V9gc5QnNGemU",
        "outputId": "4a1b0d49-6330-4833-c82b-7ac86ecaeac3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 619
        }
      },
      "source": [
        "# Lets use missing_values function to explore train dataset\n",
        "missing_values(train_df)"
      ],
      "execution_count": 138,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Number of records with missing location: 2533\n",
            "Number of records with missing keywords: 61\n",
            "33.0% of location values are missing from Total Number of Records.\n",
            "1.0% of keywords values are missing from Total Number of Records.\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1800x720 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_WmtxQ_dGemX",
        "outputId": "f3266d15-7071-46dc-be9c-983674b2f055",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 619
        }
      },
      "source": [
        "# Lets use missing_values function to explore train dataset\n",
        "missing_values(test_df);"
      ],
      "execution_count": 139,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Number of records with missing location: 1105\n",
            "Number of records with missing keywords: 26\n",
            "34.0% of location values are missing from Total Number of Records.\n",
            "1.0% of keywords values are missing from Total Number of Records.\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1800x720 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zPi_952hGema"
      },
      "source": [
        "From train data frame description and analysis, it is obivous that:\n",
        "- `location` feature is not automatically generated feature, and it seems it is user input hence too many records are missing.In my opinion, `location` column from our dataset could be dropped.\n",
        "- We will keep the `keyword` feature because there are a lot of unique keywords and missing values are very insignificant (< 1 percentage)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yB_wrUtLGema"
      },
      "source": [
        "#Lets visulaize dataframe features using charts\n",
        "def feature_viz(df,feature):\n",
        "    \n",
        "    '''Input- df=pandas dataframe\n",
        "              feature= column to be charted\n",
        "       Output- bar and scatter chart using plotly       \n",
        "    \n",
        "    '''\n",
        "    #Visualize the feature\n",
        "    if feature=='target':\n",
        "        sns.countplot(feature, data=df)\n",
        "        print('Target of 0 is {} % of total'.format(round(df[feature].value_counts()[0]/len(df[feature])*100)))\n",
        "        print('Target of 1 is {} % of total'.format(round(df[feature].value_counts()[1]/len(df[feature])*100)))\n",
        "    else:\n",
        "        #Distinct keywords in train dataset\n",
        "        feat=df[feature].value_counts()\n",
        "        print(feat.head())\n",
        "        fig = px.scatter(feat, x=feat.values, y=feat.index,size=feat.values)\n",
        "        fig.show()"
      ],
      "execution_count": 140,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "adTl9GKdGemd",
        "outputId": "bfaaa8f1-2312-436c-e2c7-968505c65064",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 313
        }
      },
      "source": [
        "#Lets use feature_viz function to create charts for 'target' column\n",
        "feature_viz(train_df,'target')"
      ],
      "execution_count": 141,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Target of 0 is 57.0 % of total\n",
            "Target of 1 is 43.0 % of total\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o6_ubmjbGemg",
        "outputId": "be241640-4e3e-44ed-9400-f5e492b6537c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 644
        }
      },
      "source": [
        "#Lets use feature_viz function to create charts for 'keyword' column\n",
        "feature_viz(train_df,'keyword')"
      ],
      "execution_count": 142,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "fatalities    45\n",
            "deluge        42\n",
            "armageddon    42\n",
            "damage        41\n",
            "sinking       41\n",
            "Name: keyword, dtype: int64\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<html>\n",
              "<head><meta charset=\"utf-8\" /></head>\n",
              "<body>\n",
              "    <div>\n",
              "            <script src=\"https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.5/MathJax.js?config=TeX-AMS-MML_SVG\"></script><script type=\"text/javascript\">if (window.MathJax) {MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}</script>\n",
              "                <script type=\"text/javascript\">window.PlotlyConfig = {MathJaxConfig: 'local'};</script>\n",
              "        <script src=\"https://cdn.plot.ly/plotly-latest.min.js\"></script>    \n",
              "            <div id=\"f00dc10c-4ea4-4a8b-a755-fce43a807760\" class=\"plotly-graph-div\" style=\"height:525px; width:100%;\"></div>\n",
              "            <script type=\"text/javascript\">\n",
              "                \n",
              "                    window.PLOTLYENV=window.PLOTLYENV || {};\n",
              "                    \n",
              "                if (document.getElementById(\"f00dc10c-4ea4-4a8b-a755-fce43a807760\")) {\n",
              "                    Plotly.newPlot(\n",
              "                        'f00dc10c-4ea4-4a8b-a755-fce43a807760',\n",
              "                        [{\"hoverlabel\": {\"namelength\": 0}, \"hovertemplate\": \"x=%{x}<br>y=%{y}<br>size=%{marker.size}\", \"legendgroup\": \"\", \"marker\": {\"color\": \"#636efa\", \"size\": [45, 42, 42, 41, 41, 41, 41, 40, 40, 40, 40, 40, 40, 40, 39, 39, 39, 39, 39, 39, 39, 39, 39, 39, 39, 39, 39, 39, 38, 38, 38, 38, 38, 38, 38, 38, 38, 38, 38, 38, 38, 38, 38, 38, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 31, 31, 31, 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 30, 29, 29, 29, 29, 29, 29, 29, 29, 28, 28, 27, 27, 26, 25, 24, 22, 19, 12, 11, 10, 9], \"sizemode\": \"area\", \"sizeref\": 0.1125, \"symbol\": \"circle\"}, \"mode\": \"markers\", \"name\": \"\", \"showlegend\": false, \"type\": \"scatter\", \"x\": [45, 42, 42, 41, 41, 41, 41, 40, 40, 40, 40, 40, 40, 40, 39, 39, 39, 39, 39, 39, 39, 39, 39, 39, 39, 39, 39, 39, 38, 38, 38, 38, 38, 38, 38, 38, 38, 38, 38, 38, 38, 38, 38, 38, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 37, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 36, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 34, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 32, 31, 31, 31, 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 30, 29, 29, 29, 29, 29, 29, 29, 29, 28, 28, 27, 27, 26, 25, 24, 22, 19, 12, 11, 10, 9], \"xaxis\": \"x\", \"y\": [\"fatalities\", \"deluge\", \"armageddon\", \"damage\", \"sinking\", \"harm\", \"body%20bags\", \"fear\", \"windstorm\", \"outbreak\", \"twister\", \"evacuate\", \"siren\", \"collided\", \"weapons\", \"collision\", \"sunk\", \"derailment\", \"hellfire\", \"weapon\", \"earthquake\", \"famine\", \"flames\", \"wrecked\", \"wreckage\", \"sinkhole\", \"whirlwind\", \"explosion\", \"upheaval\", \"typhoon\", \"derailed\", \"flooding\", \"fatal\", \"ambulance\", \"drowned\", \"explode\", \"fire\", \"oil%20spill\", \"thunder\", \"blaze\", \"hurricane\", \"deaths\", \"bombed\", \"injury\", \"pandemonium\", \"curfew\", \"debris\", \"panic\", \"destroy\", \"crush\", \"bioterror\", \"sandstorm\", \"emergency\", \"mudslide\", \"hostages\", \"blizzard\", \"police\", \"quarantined\", \"wreck\", \"ruin\", \"burning%20buildings\", \"fatality\", \"wounded\", \"displaced\", \"screaming\", \"catastrophe\", \"detonate\", \"devastation\", \"engulfed\", \"ablaze\", \"desolation\", \"evacuated\", \"danger\", \"attack\", \"cliff%20fall\", \"evacuation\", \"massacre\", \"nuclear%20reactor\", \"dust%20storm\", \"refugees\", \"floods\", \"tragedy\", \"death\", \"screams\", \"accident\", \"rescued\", \"tornado\", \"buildings%20burning\", \"airplane%20accident\", \"derail\", \"rioting\", \"bleeding\", \"inundated\", \"drought\", \"hijacker\", \"attacked\", \"bridge%20collapse\", \"hazardous\", \"bloody\", \"demolition\", \"rescuers\", \"collapsed\", \"injured\", \"structural%20failure\", \"emergency%20plan\", \"hail\", \"blood\", \"storm\", \"casualties\", \"disaster\", \"traumatised\", \"suicide%20bomb\", \"razed\", \"flood\", \"hazard\", \"terrorism\", \"rainstorm\", \"drowning\", \"bomb\", \"smoke\", \"riot\", \"tsunami\", \"collide\", \"aftershock\", \"nuclear%20disaster\", \"burning\", \"lava\", \"crashed\", \"annihilated\", \"casualty\", \"military\", \"destruction\", \"demolish\", \"loud%20bang\", \"electrocuted\", \"screamed\", \"quarantine\", \"blazing\", \"army\", \"natural%20disaster\", \"flattened\", \"collapse\", \"heat%20wave\", \"arsonist\", \"exploded\", \"violent%20storm\", \"wounds\", \"emergency%20services\", \"survived\", \"mass%20murder\", \"stretcher\", \"blew%20up\", \"trouble\", \"chemical%20emergency\", \"crash\", \"burned\", \"thunderstorm\", \"wildfire\", \"body%20bag\", \"body%20bagging\", \"meltdown\", \"suicide%20bombing\", \"lightning\", \"injuries\", \"hijack\", \"blown%20up\", \"fire%20truck\", \"landslide\", \"buildings%20on%20fire\", \"panicking\", \"hijacking\", \"cyclone\", \"eyewitness\", \"hailstorm\", \"survive\", \"forest%20fires\", \"blight\", \"arson\", \"apocalypse\", \"mass%20murderer\", \"drown\", \"trapped\", \"destroyed\", \"detonation\", \"electrocute\", \"wild%20fires\", \"obliterated\", \"terrorist\", \"obliterate\", \"crushed\", \"hostage\", \"suicide%20bomber\", \"devastated\", \"trauma\", \"bioterrorism\", \"survivors\", \"dead\", \"avalanche\", \"catastrophic\", \"mayhem\", \"obliteration\", \"snowstorm\", \"first%20responders\", \"annihilation\", \"desolate\", \"seismic\", \"bombing\", \"sirens\", \"rubble\", \"demolished\", \"volcano\", \"deluged\", \"battle\", \"bush%20fires\", \"war%20zone\", \"rescue\", \"forest%20fire\", \"epicentre\", \"threat\", \"inundation\", \"radiation%20emergency\"], \"yaxis\": \"y\"}],\n",
              "                        {\"legend\": {\"itemsizing\": \"constant\", \"tracegroupgap\": 0}, \"margin\": {\"t\": 60}, \"template\": {\"data\": {\"bar\": [{\"error_x\": {\"color\": \"#2a3f5f\"}, \"error_y\": {\"color\": \"#2a3f5f\"}, \"marker\": {\"line\": {\"color\": \"#E5ECF6\", \"width\": 0.5}}, \"type\": \"bar\"}], \"barpolar\": [{\"marker\": {\"line\": {\"color\": \"#E5ECF6\", \"width\": 0.5}}, \"type\": \"barpolar\"}], \"carpet\": [{\"aaxis\": {\"endlinecolor\": \"#2a3f5f\", \"gridcolor\": \"white\", \"linecolor\": \"white\", \"minorgridcolor\": \"white\", \"startlinecolor\": \"#2a3f5f\"}, \"baxis\": {\"endlinecolor\": \"#2a3f5f\", \"gridcolor\": \"white\", \"linecolor\": \"white\", \"minorgridcolor\": \"white\", \"startlinecolor\": \"#2a3f5f\"}, \"type\": \"carpet\"}], \"choropleth\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"type\": \"choropleth\"}], \"contour\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"contour\"}], \"contourcarpet\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"type\": \"contourcarpet\"}], \"heatmap\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"heatmap\"}], \"heatmapgl\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"heatmapgl\"}], \"histogram\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"histogram\"}], \"histogram2d\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"histogram2d\"}], \"histogram2dcontour\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"histogram2dcontour\"}], \"mesh3d\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"type\": \"mesh3d\"}], \"parcoords\": [{\"line\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"parcoords\"}], \"pie\": [{\"automargin\": true, \"type\": \"pie\"}], \"scatter\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatter\"}], \"scatter3d\": [{\"line\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatter3d\"}], \"scattercarpet\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scattercarpet\"}], \"scattergeo\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scattergeo\"}], \"scattergl\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scattergl\"}], \"scattermapbox\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scattermapbox\"}], \"scatterpolar\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatterpolar\"}], \"scatterpolargl\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatterpolargl\"}], \"scatterternary\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatterternary\"}], \"surface\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"surface\"}], \"table\": [{\"cells\": {\"fill\": {\"color\": \"#EBF0F8\"}, \"line\": {\"color\": \"white\"}}, \"header\": {\"fill\": {\"color\": \"#C8D4E3\"}, \"line\": {\"color\": \"white\"}}, \"type\": \"table\"}]}, \"layout\": {\"annotationdefaults\": {\"arrowcolor\": \"#2a3f5f\", \"arrowhead\": 0, \"arrowwidth\": 1}, \"coloraxis\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"colorscale\": {\"diverging\": [[0, \"#8e0152\"], [0.1, \"#c51b7d\"], [0.2, \"#de77ae\"], [0.3, \"#f1b6da\"], [0.4, \"#fde0ef\"], [0.5, \"#f7f7f7\"], [0.6, \"#e6f5d0\"], [0.7, \"#b8e186\"], [0.8, \"#7fbc41\"], [0.9, \"#4d9221\"], [1, \"#276419\"]], \"sequential\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"sequentialminus\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]]}, \"colorway\": [\"#636efa\", \"#EF553B\", \"#00cc96\", \"#ab63fa\", \"#FFA15A\", \"#19d3f3\", \"#FF6692\", \"#B6E880\", \"#FF97FF\", \"#FECB52\"], \"font\": {\"color\": \"#2a3f5f\"}, \"geo\": {\"bgcolor\": \"white\", \"lakecolor\": \"white\", \"landcolor\": \"#E5ECF6\", \"showlakes\": true, \"showland\": true, \"subunitcolor\": \"white\"}, \"hoverlabel\": {\"align\": \"left\"}, \"hovermode\": \"closest\", \"mapbox\": {\"style\": \"light\"}, \"paper_bgcolor\": \"white\", \"plot_bgcolor\": \"#E5ECF6\", \"polar\": {\"angularaxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}, \"bgcolor\": \"#E5ECF6\", \"radialaxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}}, \"scene\": {\"xaxis\": {\"backgroundcolor\": \"#E5ECF6\", \"gridcolor\": \"white\", \"gridwidth\": 2, \"linecolor\": \"white\", \"showbackground\": true, \"ticks\": \"\", \"zerolinecolor\": \"white\"}, \"yaxis\": {\"backgroundcolor\": \"#E5ECF6\", \"gridcolor\": \"white\", \"gridwidth\": 2, \"linecolor\": \"white\", \"showbackground\": true, \"ticks\": \"\", \"zerolinecolor\": \"white\"}, \"zaxis\": {\"backgroundcolor\": \"#E5ECF6\", \"gridcolor\": \"white\", \"gridwidth\": 2, \"linecolor\": \"white\", \"showbackground\": true, \"ticks\": \"\", \"zerolinecolor\": \"white\"}}, \"shapedefaults\": {\"line\": {\"color\": \"#2a3f5f\"}}, \"ternary\": {\"aaxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}, \"baxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}, \"bgcolor\": \"#E5ECF6\", \"caxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}}, \"title\": {\"x\": 0.05}, \"xaxis\": {\"automargin\": true, \"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\", \"title\": {\"standoff\": 15}, \"zerolinecolor\": \"white\", \"zerolinewidth\": 2}, \"yaxis\": {\"automargin\": true, \"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\", \"title\": {\"standoff\": 15}, \"zerolinecolor\": \"white\", \"zerolinewidth\": 2}}}, \"xaxis\": {\"anchor\": \"y\", \"domain\": [0.0, 1.0], \"title\": {\"text\": \"x\"}}, \"yaxis\": {\"anchor\": \"x\", \"domain\": [0.0, 1.0], \"title\": {\"text\": \"y\"}}},\n",
              "                        {\"responsive\": true}\n",
              "                    ).then(function(){\n",
              "                            \n",
              "var gd = document.getElementById('f00dc10c-4ea4-4a8b-a755-fce43a807760');\n",
              "var x = new MutationObserver(function (mutations, observer) {{\n",
              "        var display = window.getComputedStyle(gd).display;\n",
              "        if (!display || display === 'none') {{\n",
              "            console.log([gd, 'removed!']);\n",
              "            Plotly.purge(gd);\n",
              "            observer.disconnect();\n",
              "        }}\n",
              "}});\n",
              "\n",
              "// Listen for the removal of the full notebook cells\n",
              "var notebookContainer = gd.closest('#notebook-container');\n",
              "if (notebookContainer) {{\n",
              "    x.observe(notebookContainer, {childList: true});\n",
              "}}\n",
              "\n",
              "// Listen for the clearing of the current output cell\n",
              "var outputEl = gd.closest('.output');\n",
              "if (outputEl) {{\n",
              "    x.observe(outputEl, {childList: true});\n",
              "}}\n",
              "\n",
              "                        })\n",
              "                };\n",
              "                \n",
              "            </script>\n",
              "        </div>\n",
              "</body>\n",
              "</html>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "N9MPLEDWGemw"
      },
      "source": [
        "Let's check how often the word 'disaster' come in the 'text' feature in the dataset and whether this help us in determining whether a tweet belongs to a disaster' category or not."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0KKV6GsVGemx",
        "outputId": "81688139-dc19-41cc-81d6-c80e1460068a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        }
      },
      "source": [
        "train_df.loc[train_df['text'].str.contains('disaster', na=False, case=False)].target.value_counts()"
      ],
      "execution_count": 143,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1    102\n",
              "0     40\n",
              "Name: target, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 143
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "M0U5lGTDGemz"
      },
      "source": [
        "It seems even though word disaster is part of the tweets but that doesn't mean tweet announces a real disaster."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "M0z6-bPvGem0",
        "outputId": "bba6dfda-7daf-441e-ff53-ad100a42f613",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 644
        }
      },
      "source": [
        "#Lets use feature_viz function to create charts for 'location' column\n",
        "feature_viz(train_df,'location')"
      ],
      "execution_count": 144,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "USA              104\n",
            "New York          71\n",
            "United States     50\n",
            "London            45\n",
            "Canada            29\n",
            "Name: location, dtype: int64\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<html>\n",
              "<head><meta charset=\"utf-8\" /></head>\n",
              "<body>\n",
              "    <div>\n",
              "            <script src=\"https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.5/MathJax.js?config=TeX-AMS-MML_SVG\"></script><script type=\"text/javascript\">if (window.MathJax) {MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}</script>\n",
              "                <script type=\"text/javascript\">window.PlotlyConfig = {MathJaxConfig: 'local'};</script>\n",
              "        <script src=\"https://cdn.plot.ly/plotly-latest.min.js\"></script>    \n",
              "            <div id=\"b54cb36f-1851-495b-b9ba-016dec55d405\" class=\"plotly-graph-div\" style=\"height:525px; width:100%;\"></div>\n",
              "            <script type=\"text/javascript\">\n",
              "                \n",
              "                    window.PLOTLYENV=window.PLOTLYENV || {};\n",
              "                    \n",
              "                if (document.getElementById(\"b54cb36f-1851-495b-b9ba-016dec55d405\")) {\n",
              "                    Plotly.newPlot(\n",
              "                        'b54cb36f-1851-495b-b9ba-016dec55d405',\n",
              "                        [{\"hoverlabel\": {\"namelength\": 0}, \"hovertemplate\": \"x=%{x}<br>y=%{y}<br>size=%{marker.size}\", \"legendgroup\": \"\", \"marker\": {\"color\": \"#636efa\", \"size\": [104, 71, 50, 45, 29, 28, 27, 26, 24, 22, 21, 20, 19, 18, 18, 17, 15, 15, 15, 14, 14, 14, 13, 13, 13, 12, 12, 12, 11, 11, 11, 11, 10, 10, 10, 10, 10, 10, 10, 9, 9, 9, 9, 9, 9, 9, 9, 8, 8, 8, 7, 7, 7, 7, 7, 7, 7, 7, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], \"sizemode\": \"area\", \"sizeref\": 0.26, \"symbol\": \"circle\"}, \"mode\": \"markers\", \"name\": \"\", \"showlegend\": false, \"type\": \"scattergl\", \"x\": [104, 71, 50, 45, 29, 28, 27, 26, 24, 22, 21, 20, 19, 18, 18, 17, 15, 15, 15, 14, 14, 14, 13, 13, 13, 12, 12, 12, 11, 11, 11, 11, 10, 10, 10, 10, 10, 10, 10, 9, 9, 9, 9, 9, 9, 9, 9, 8, 8, 8, 7, 7, 7, 7, 7, 7, 7, 7, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], \"xaxis\": \"x\", \"y\": [\"USA\", \"New York\", \"United States\", \"London\", \"Canada\", \"Nigeria\", \"UK\", \"Los Angeles, CA\", \"India\", \"Mumbai\", \"Washington, DC\", \"Kenya\", \"Worldwide\", \"Australia\", \"Chicago, IL\", \"California\", \"Everywhere\", \"California, USA\", \"New York, NY\", \"Florida\", \"United Kingdom\", \"San Francisco\", \"Indonesia\", \"Los Angeles\", \"Washington, D.C.\", \"NYC\", \"Toronto\", \"Ireland\", \"Earth\", \"Seattle\", \"San Francisco, CA\", \"Chicago\", \"Sacramento, CA\", \"London, England\", \"New York City\", \"Texas\", \"London, UK\", \"Atlanta, GA\", \"ss\", \"304\", \"World\", \"San Diego, CA\", \"US\", \"Manchester\", \"Denver, Colorado\", \"Nashville, TN\", \"Dallas, TX\", \"Houston, TX\", \"Scotland\", \"South Africa\", \"Tennessee\", \"Memphis, TN\", \"Austin, TX\", \"Denver, CO\", \"Sydney\", \"Seattle, WA\", \"worldwide\", \"Pennsylvania, USA\", \"Oklahoma City, OK\", \"Charlotte, NC\", \"Global\", \"Portland, OR\", \"Atlanta\", \" Road to the Billionaires Club\", \"Paterson, New Jersey \", \"California, United States\", \"Massachusetts\", \"Morioh, Japan\", \"Singapore\", \"Planet Earth\", \"Pedophile hunting ground\", \"Calgary, Alberta\", \"Orlando, FL\", \"Colorado\", \"Brooklyn, NY\", \"??????\", \"Newcastle\", \"Florida, USA\", \"Vancouver, BC\", \"Calgary, AB\", \"in the Word of God\", \"Pakistan\", \"Texas, USA\", \"New York, USA\", \"San Jose, CA\", \"Melbourne, Australia\", \"Cleveland, OH\", \"NY\", \"New Hampshire\", \"WorldWide\", \"The Netherlands\", \"Wisconsin\", \"Indiana\", \"Philippines\", \"Tampa, FL\", \"Puerto Rico\", \"Coventry\", \"North Carolina\", \"Boston, MA\", \"Pennsylvania\", \"Lagos\", \"Lagos, Nigeria\", \"New Jersey\", \"Brasil\", \"Leeds, England\", \"Asheville, NC\", \"Southern California\", \"Midwest\", \"Switzerland\", \"Maryland\", \"Oregon\", \"Republic of Texas\", \"Sacramento\", \"?\", \"Bend, Oregon\", \"Haddonfield, NJ\", \"Tokyo\", \"Edinburgh\", \"?????\", \"Sydney, Australia\", \"Nigeria \", \"Philadelphia, PA\", \"Oakland, CA\", \"Jamaica\", \"Seattle, Washington\", \"Port Harcourt, Nigeria\", \"Japan\", \"Happily Married with 2 kids \", \"Portland, Oregon\", \"MAD as Hell\", \"Kansas City\", \"Nairobi-KENYA\", \"Calgary\", \"Mumbai, Maharashtra\", \"china\", \"San Francisco Bay Area\", \"North Carolina, USA\", \"Sydney, New South Wales\", \"??\", \"Phoenix, AZ\", \"Birmingham\", \"British Columbia, Canada\", \"Cape Town\", \"canada\", \"Paignton\", \"Leicester\", \"Michigan\", \"Jakarta/Kuala Lumpur/S'pore\", \"Barbados\", \"london\", \"Melbourne\", \"Geneva\", \"Manchester, England\", \"Massachusetts, USA\", \"Las Vegas, Nevada\", \"Atlanta Georgia \", \"U.S.A\", \"Victoria, BC\", \"Nairobi, Kenya\", \"Anchorage, AK\", \"MA\", \"Kansas City, MO\", \"Germany\", \"China\", \"World Wide\", \"Stockholm, Sweden\", \"Vancouver, British Columbia\", \"Kama | 18 | France \", \"Madison, WI\", \"Iraq|Afghanistan| RSA |Baghdad\", \"Virginia\", \"CA\", \"Brazil\", \"Karachi Pakistan\", \"Baltimore, MD\", \"Nowhere. Everywhere.\", \"Buffalo, NY\", \"Malaysia\", \"Italy\", \"Raleigh, NC\", \"Five down from the Coffeeshop\", \"England\", \"Rio de Janeiro\", \"Upstairs.\", \"State College, PA\", \"Birmingham, England\", \"Winston-Salem, NC\", \"New Orleans ,Louisiana\", \"America\", \"Naperville\", \"Nottingham, England\", \"#FLIGHTCITY UK  \", \"Oklahoma City\", \"Africa\", \"nyc\", \"Taylor Swift\", \"Georgia\", \"???\", \"Buy Give Me My Money \", \"Illinois, USA\", \"Boston\", \"Narnia\", \"Financial News and Views\", \"uk\", \"Toronto, Ontario\", \"Washington DC\", \"Austin, Texas\", \"Manhattan, NY\", \"Calgary, AB, Canada\", \"New Orleans, LA\", \"Milwaukee, WI\", \"Hawaii, USA\", \"South, USA\", \"Inexpressible Island \", \"Victoria, British Columbia\", \"PA\", \"Arizona\", \"Glasgow\", \"Las Vegas\", \"Oregon, USA\", \"Nairobi\", \"St. Louis, MO\", \"Haysville, KS\", \"india\", \"Hong Kong\", \"Indianapolis, IN\", \"Des Moines, IA\", \"Johannesburg, South Africa\", \"Virginia, USA\", \"Gotham City\", \"East Coast\", \"Sweden\", \"Tulsa, Oklahoma\", \"Subconscious LA\", \"Ontario, Canada\", \"ARGENTINA\", \"21.462446,-158.022017\", \"Washington D.C.\", \"Dublin, Ireland\", \"San Diego\", \"Adelaide, Australia\", \"Karachi\", \"Columbus, OH\", \"new york\", \"iTunes\", \"Los Angeles, California\", \"Arlington, TX\", \"Anonymous\", \"California \", \"London UK\", \"NJ\", \"America of Founding Fathers\", \"Asia\", \"CA via Brum\", \"Memphis\", \"Utah\", \"NYC :) Ex- #Islamophobe\", \"St Paul, MN\", \"Toronto, Canada\", \"Decatur, GA\", \"Does it really matter!\", \"302\", \"Charlotte NC\", \"Spokane, WA\", \"Lincoln, NE\", \"Instagram - @heyimginog \", \"Boston MA\", \"Minna, Nigeria\", \"New Hanover County, NC\", \"Sydney Australia\", \"Bronx NY\", \"Olympia, WA\", \"online \", \"Arizona \", \"Norway\", \"Victoria, Australia, Earth\", \"Washington, DC & Charlotte, NC\", \"Cumming, GA\", \"Winnipeg, Manitoba\", \"The Universe\", \"Winnipeg\", \"Phoenix\", \"Scotland, United Kingdom\", \"Louisiana\", \"Niagara Falls, Ontario\", \"Colombia\", \"Chicago, Illinois\", \"Silver Spring, MD\", \"Columbia, SC\", \"Helsinki\", \"Brisbane\", \"Leesburg, FL\", \"Miami, FL\", \"Detroit, MI\", \"Your screen\", \"Unknown\", \"To The Right of You!\", \"S\\u00cc\\u00a3o Paulo\", \"EIC\", \"Los Angeles \", \"United States of America\", \"Live On Webcam\", \"Blackpool\", \"Ukraine\", \"los angeles, ca\", \"Vietnam\", \"Pretoria\", \"Wilmington, NC\", \"New Delhi, India\", \"Oakland\", \"Ohio, USA\", \"hell\", \"Liverpool\", \"Karachi, Pakistan\", \"Portugal\", \"Yamaku Academy, Class 3-4\", \"Israel\", \"Suplex City\", \"Alexandria, VA\", \"Moscow\", \"Fresno, CA\", \"vancouver usa\", \"Buenos Aires, Argentina\", \"Thailand\", \"Dubai\", \"Gold Coast, Australia\", \"REPUBLICA DOMINICANA\", \"Kashmir!\", \"Bandung\", \"World Wide Web\", \"Alaska\", \"Muntinlupa City, Philippines\", \"SWMO\", \"San Jose, California\", \"anzio,italy\", \"New England\", \"New Delhi, Delhi\", \"Wales\", \"Macclesfield\", \"Epic City, BB.\", \"far away\", \"Knoxville, TN\", \"New Zealand\", \"Orlando, Fl\", \"Manchester, UK\", \"Newcastle Upon Tyne, England\", \"Lisbon, Portugal\", \"Ottawa, Canada\", \"Ashburn, VA\", \"Bakersfield, California\", \"Orange County, California\", \"Greensboro, North Carolina\", \"Rock Hill, SC\", \"Connecticut\", \"Spare 'Oom\", \"world\", \"Cleveland, OH - San Diego, CA\", \"USA (Formerly @usNOAAgov)\", \"Ewa Beach, HI\", \"Moncton, New Brunswick\", \"Swaning Around\", \"Boise, Idaho\", \"???  Dreamz\", \"Buenos Aires\", \"Istanbul\", \"West\", \"Trinidad and Tobago\", \"ph\", \"Houston TX\", \"Desde Republica Argentina\", \"SF Bay Area\", \"Sarasota, FL\", \"Melbourne Australia\", \"South Stand\", \"Glasgow, Scotland\", \"\\u00cc\\u00cfT: 10.614817868480726,12.195582811791382\", \"Maldives\", \"Roaming around the world\", \"New Sweden\", \"scandinavia\", \"Roanoke, VA\", \"Houston\", \"In Hell\", \"tx\", \"Eldoret, kenya\", \"texas\", \"HTX\", \"Kingston, Jamaica\", \"Macon, GA\", \"GLOBAL\", \"Mackay, QLD, Australia\", \"Somewhere in the Canada\", \"somewhere USA \", \"#HarleyChick#PJNT#RunBenRun\", \"New York \", \"Washington, USA\", \"Ontario Canada\", \"Europe\", \"Charlotte\", \"NIGERIA\", \"England, United Kingdom\", \"Philadelphia, PA \", \"Kent\", \"Evanston, IL\", \"Rocky Mountains\", \"rome\", \"????\", \"Somalia\", \"Alabama\", \"TX\", \"Newark, NJ\", \"Mongolia\", \"El Dorado, Arkansas\", \"Venezuela\", \"Accra,Ghana\", \"Dime's Palace\", \"they/them \", \"Detroit\", \"Jakarta\", \"Selma2Oakland\", \"Halifax\", \"Belgium\", \"Lagos Nigeria\", \"Street of Dallas\", \"Portland, Ore. \", \"Chile\", \"19.600858, -99.047821\", \"Adelaide, South Australia\", \"Lyallpur, Pakistan\", \"? \", \"Honduras\", \"Odawara, Japan\", \"va\", \"Riyadh\", \"Incognito\", \"The American Wasteland (MV)\", \"Baltimore\", \"Reddit \", \"dallas\", \"Dallas Fort-Worth\", \"Sand springs oklahoma\", \"Chevy Chase, MD\", \"Palo Alto, California\", \"Bangalore, India\", \"mumbai\", \"Hackney, London\", \"everywhere\", \"Here.\", \"Na:tinixw / Hoopa, Berkeley\", \"Redding, California, USA\", \"West Virginia, USA\", \"107-18 79TH STREET\", \"Federal Capital Territory\", \"Silicon Valley\", \"Madison, GA\", \"San Antonio, TX\", \"San Diego California 92101\", \" Queensland, Australia\", \"434\", \"Erie, PA\", \"Breaking News\", \"DC\", \"Mesa, AZ\", \"St Charles, MD\", \"Jupiter\", \"Dublin City, Ireland\", \"Cherry Creek Denver CO\", \"Columbus\", \"Bukittinggi  ?? Sumatera Barat\", \"U.K.\", \"http://www.amazon.com/dp/B00HR\", \"Costa Rica\", \"Above the snake line - #YoNews\", \"Melbourne, Victoria\", \"Berlin, Germany\", \"Perth, Western Australia\", \"Heaven\", \"A little house in the outback.\", \"Trinidad & Tobago\", \"Durham, NC\", \"Paris\", \"EastCarolina\", \"New York, New York\", \"Korea\", \"Baton Rouge, LA\", \"NC\", \"Gainesville, FL\", \"Manchester, NH\", \"Jerusalem\", \"Vancouver BC\", \"Auckland\", \"Pittsburgh\", \"Dubai, UAE\", \"Dundee\", \"Washington State\", \"Philadelphia\", \"Global-NoLocation\", \"Mo.City\", \"WA State\", \"Location\", \"Yogya Berhati Nyaman\", \"Renfrew, Scotland\", \"Auckland, New Zealand\", \"CA physically- Boston Strong?\", \"Manila, Philippines\", \"617-BTOWN-BEATDOWN\", \"Georgia, USA\", \"Michigan, USA\", \"In the potters hands\", \"Huntsville, AL\", \"the insane asylum. \", \"Budapest, Hungary\", \"Palo Alto, CA\", \"Enfield, UK\", \"Santa Cruz, CA\", \"Washington\", \"Derby\", \"Mumbai , India\", \"140920-21 & 150718-19 BEIJING\", \"Somewhere Only We Know ?\", \"Berlin - Germany\", \"Bournemouth\", \"Kolkata\", \"LiVE M\\u00cc\\u0081S\", \"OES 4th Point. sisSTAR & TI\", \"rural ohio (fuck)\", \"Ideally under a big tree\", \"Chasing My Dreams w/Jass??\", \"Depok\", \"Tri-Cities, Wash.\", \"ANDY 4 LEADER X\", \"Nashville, Tennessee\", \"Greensburg, PA\", \"Amarillo\", \"staggering on tenement roofs\", \"Paranaque City\", \"Fountain City, IN \", \"New Jersey, usually\", \"Nowhere Islands/Smash Manor\", \"PARACHUTE\", \"Fort Wayne, IN\", \"Raleigh (Garner/Cleveland) NC\", \"Blogland\", \"SEC Country\", \"by a piano probably. \", \" Little Rock, AR\", \"Ogba, Lagos, Nigeria\", \"CT & NY\", \"Innerhalb der L\\u00cc_cke\", \"Palm Desert, CA\", \"Upper St Clair, PA\", \"Waverly, IA\", \"Downtown Churubusco, Indiana\", \"#UNITE THE BLUE  \", \"Asunci\\u00cc_n-PY / T\\u00cc_bingen-GER\", \"Toronto, ON, Canada\", \"Santa Maria, CA\", \"Chicora ?? Oakland\", \"Michel Delving.\", \"Unite. Bless. Wallahi \", \"Dudetown\", \"Home of the Takers.\", \" Tropical SE FLorida\", \"North East, England\", \"Los Angeles for now\", \"DC, frequently NYC/San Diego\", \"Evansville, IN\", \"Paulton, England\", \"Chicago, IL \", \"Somewhere in China.\", \"antioch, california\", \"pissing off antis\", \"ny\", \"tri state\", \"a feminist, modernist hag.\", \"Vancouver, BC.\", \"412 NW 5th Ave. Portland OR\", \"London/Bristol/Guildford\", \"Bokaro Steel City, Jharkhand\", \"Plano,TX\", \"KSU 2017\", \"FCT, Abuja \", \"Minority Privilege, USA\", \"Greensboro, NC\", \"#SandraBland\", \"we?it \\u0089\\u00db\\u00a2 ixwin\", \"california\", \"\\u00cc\\u00cfT: 33.209923,-87.545328\", \"Asheboro, NC\", \"Rotterdam, Zuid-Holland\", \"Cairo, Egypt.\", \"Halifax, Nova Scotia\", \"Marysville, MI\", \"Halton, Ontario\", \"Westchester\", \"Score More Goals Buying @\", \"Charlottetown\", \"New Orleans, Louisiana\", \"Kodiak, AK\", \"Behind The Obama Curtain\", \"Islamabad\", \"CLT\", \"PROUD INDIANS\", \"Brisbane.\", \"YA MOTHA BED\", \"Manhattan\", \"Ontario, Canada. \", \"Hartford,  connecticut\", \"LYNBROOK\", \"Huber Heights, OH\", \"Delhi \", \"L. A.\", \"Flushing, Queens\", \"www.aprylpooley.com\", \"1/3 of the blam squad \", \"Jammu | Kashmir | Delhi\", \"... -.- -.--\", \"Did anybody see me here ??\", \"Duncan\", \"Aurora, IL\", \"Just Happy to Be Anywhere\", \"DaKounty, Pa\", \"??????? ??????? ????????\", \"Lethbridge, AB, Canada\", \"#ForeverWithBAP 8 \", \"Alliston Ontario\", \"????s ?? ????\\u00cc\\u00f8????\\u00cc\\u00a1a\", \"Westerland\", \"Cleveland, TN\", \"Konoha\", \"Eastlake, OH\", \"Livingston, MT\", \"Colorado, USA\", \"Woodcreek HS, Roseville, CA\", \"Glendale, CA\", \"Honolulu,Hawaii \", \"Unnamed City\", \"Guelph Ontario Canada\", \"Frostburg\", \"New Jersey/New York\", \"\\u00cc\\u00cfT: 0.0,0.0\", \"Blackpool, England, UK.\", \"btwn a rock and a hard place\", \"Huntley, IL\", \"Metro Manila\", \"texas a&m university\", \"#BossNation!\", \"Biloxi, Mississippi\", \"Newcastle, England\", \"St PetersburgFL\", \"South 37\", \"Hartford  London Hong Kong\", \"Temecula, CA\", \"Chicopee MA\", \"Regalo Island\", \"glasgow\", \"Enterprise, NV\", \"perth, australia \", \"Jubail IC, Saudi Arabia.\", \"CPT & JHB, South Africa\", \"Jerseyville, IL\", \"FILM OUT LATE 2015\", \"Miami via Lima\", \"Florida USA\", \"Manavadar, Gujarat\", \"PROV\", \"Dallas, Texas. \", \"Jackson TN\", \"Hattiesburg, MS\", \"Crouch End, London\", \"todaysbigstock.com\", \"Kirkwall\", \"#LemonGang \", \"Texas \", \"Makai\", \"Utah, USA\", \"(Spain)\", \"Darnley, Prince Edward Island\", \"back in japan ??????????\", \"North Memphis/Global Citizen\", \"ljp/4\", \"God.Family.Money\", \"snapchat~ maddzz_babby \", \"manaus\", \"Asia Pacific   \", \"Santa Clara, CA\", \"The Citadel, Oldtown, Westeros\", \"plymouth\", \"taken by piper curda\", \"Loughton, Essex, UK\", \"#KaumElite;#F?VOR;#SMOFC\", \"Hammersmith, London\", \"Okanagan Valley, BC\", \"West Hollywood\", \"Berlin, NY, DC, Malibu\", \"Chiyoda Ward, Tokyo\", \"Somewhere in Jersey\", \"Screwston, TX\", \"Uyo, Akwa Ibom State, Nigeria\", \"Jammu and Kashmir\", \"london town..\", \"Hooters on Peachtree\", \"Aix-en-Provence, France\", \"South, England\", \"rio de janeiro | brazil\", \"don't buy the s*n\", \"Fort Worth, Texas\", \"White Plains, NY\", \"NYC-LA-MIAMI\", \"Kansas City, Mo.\", \"Bandar Lampung, Indonesia\", \"Kalimantan Timur, Indonesia\", \"MNL\", \"Perthshire \", \"North East USA\", \"Palm Bay, FL (Kissimmee)\", \"kissimmee,fl.\", \"UAE,Sharjah/ AbuDhabi\", \"buhh\", \"Bracknell\", \"Gotham\", \"London/New York\", \"the azure cloud\", \"Halifax, Nouvelle-\\u00cc\\u00e4cosse\", \"Canada BC\", \"Leicester, England\", \"Long Island, NY\", \"Area 8 \", \"Homs- Syria\", \"In The Mansion\", \"Vilnius\", \"Des Moines, Iowa \", \"lesa * she/her\", \"Lancaster, Pennsylvania, USA\", \"wisco\", \"Harpurhey, Manchester, UK\", \"big boy \\u0089\\u00db\\u00a2 0802\", \"The TARDIS\", \"Top Secret\", \"Eddyville, Oregon 97343\", \"Chicagoland\", \"Santiago de Cmpostela Galicia\", \"Fakefams\", \"I O W A\", \"Paducah, KY\", \"The Harbinger.\", \"tripoli international airport\", \"County Durham, United Kingdom\", \"Tripsburg, ms.\", \"11202\", \"Santo Domingo Alma Rosa \", \"port matilda pa\", \"Gotham City,USA\", \"Ashford, Kent, United Kingdom\", \"Charleston, WV\", \"South africa\", \"Bristol\", \"Indonesia\\n\", \"#otrakansascity\", \"Huntington, WV\", \"Santa Monica, CA\", \"Iowa, USA\", \"Rochelle, GA\", \"Northern Ireland\", \"Mass\", \"the Dirty D\", \"North West London\", \"Washington, Krasnodar (Russia)\", \"africa\", \"Around the world\", \"Den Helder, Rijkswerf\", \"BrowardCounty // Florida \", \"Kansas, The Free State! ~ KC\", \"Ukraine and Ireland\", \"portland, oregon\", \"Adventist - Lesson Sabbath\", \"somewhere too cold for me\", \"Sevier County.\", \"Orbost, Victoria, Australia\", \"high way 99\", \"IG : Sincerely_TSUNAMI\", \"?Gangsta OC / MV RP; 18+.?\", \"Neath, South Wales\", \"Perenjori, WA\", \"On the court \", \"Rocketing through the galaxy\", \"Albany/NY\", \"Proudly frozen Canuck eh !!\", \"Not where I want to be, yet\", \"Waco, Texas\", \"Halfrica\", \"Madrid\", \"Colorado Springs\", \"Burbank,CA\", \"Coolidge, AZ\", \"Twitterville\", \"Newton Centre, Massachusetts\", \"Australia \", \"English Midlands\", \"Mooseknuckle, Maine\", \"Baker City Oregon\", \"North Hastings Ontario\", \"East London\", \"Asia European Continent Korea \", \"DRAW A CIRCLE THAT'S THE EARTH\", \"Buffalo NY\", \"OK\", \"Weyburn\", \"@ ForSL/RP\", \"FLYEST HIPPIE YOU KNOW \", \"Menlo Park. SFO. The World.\", \"Augusta, GA\", \"Lima, Ohio\", \"Chicago Heights, IL\", \"V\\u00cc_a L\\u00cc\\u00c1ctea\", \"Lancashire, United Kingdom\", \"The World\", \"Geneva, Switzerland\", \"somewhere over a rainbow\", \"Peterborough, On\", \"Baydestrian\", \"Harlem, New York\", \"Virginia, United States\", \"houstn\", \"NY || live easy? \", \"Hollywood, CA\", \"Chester, IL\", \"Loughborough.\", \"Ebola\", \"253\", \"INDIA\", \"Spokane, Washington\", \"Bras\\u00cc_lia\", \"In My Lab Creating \", \"Saipan, CNMI\", \"SoCal\", \"LP, MN USA\", \"Dicky Beach\", \"Meereen \", \"Kenosha, WI 53143\", \"Sligo and Galway, Ireland\", \"McLean, VA\", \"The Desert of the Real\", \"ayr\", \"Nigeria, WORLDWIDE\", \"Ormond By The Sea, FL\", \"Use #TMW in tweets get #RT\", \"Kauai, Hawaii\", \"Nelspruit, South Africa\", \"Kwara, Nigeria\", \"East Atlanta, Georgia\", \"miami x dallas \", \"The Howling\", \"Crato - CE \", \"wrapped arnd hyuk's finger\", \"Books Published, USA\", \"North Ferriby, East Yorkshire\", \"65\", \"Vineyard\", \"justin & ari follow || tvd\", \"Tractor land aka Bristol\", \"British girl in Texas\", \"Lima, Per\\u00cc\\u00bc\", \"South Asia\", \"Pontefract UK\", \"S.F. Bay area\", \"Okuma Town, Fukushima\", \"Dhaka, Bangladsh\", \"Kuala Lumpur, Malaysia\", \"Dindigul,TamilNadu.\", \"Chicago, but Philly is home\", \"Hemel Hempstead\", \"5/5 access / rt link please x\", \"honeymoon avenue\", \"FIMAK A.S Ist Bolge Muduru\", \"Greg's place\", \"probably petting an animal\", \"The Empire/First Order\", \"Tampa, Fl\", \"A.A.S my Aztec Princess\", \"cognitive dissonance town\", \"Nakhon Si Thammarat\", \"Jerusalem, Israel\", \"SE London(heart is by the sea)\", \"Beirut/Toronto\", \"Milton keynes\", \"Telangana\", \"In a crazy genius mind\", \"Haiku, Maui, Hawaii\", \"Manila\", \"Camberwell, Melbourne\", \"Vit\\u00cc_ria (ES)\", \"Wanderlust\", \"Evergreen Colorado\", \"Tipperary (Long Way) \", \"San Diego, California\", \"Williamsbridge, Bronx, New Yor\", \"Mentor OH\", \"Purfleet\", \"sheffield // rotherham\", \"IUPUI '19\", \"Oldenburg // London\", \"Port Orange, FL\", \"Freeport Ny\", \"60th St (SS)\", \"Stamford & Cork (& Shropshire)\", \"Bucharest\", \"westwestwestwestwestwestwest\", \"Canterbury kent\", \"In my own world!!!\", \"germany\", \"hatena bookmark\", \"Beautiful British Columbia\", \"Melbourne, Florida\", \"Paradise, NV\", \"blackfalds.\", \"West Richland, WA\", \"London, Greater London, UK\", \"In Space\", \"brooklyn, NYC\", \"Minneapolis, MN\", \"West Hollywood, CA\", \"Alger-New York-San Francisco\", \"Northern Colorado\", \"Rio\", \"Waialua, Hawaii\", \"Suburban Detroit, Michigan\", \"Savage States of America\", \"The Shady Hyenatown of Finland\", \"Coimbatore\", \"probably watching survivor\", \"In your mind\", \"Your notifications\", \"Orlando\", \"Roppongi, Minato, Tokyo \", \"Azeroth\", \"Port Williams NS\", \"San Jose, CA, USA\", \"Gwersyllt, Wales\", \"rowyso dallas \", \"right here\", \"Central Coast, California\", \"it's a journey \", \"Cheshire. London. #allover\", \"beacon hills \", \"Leaving Bikini Bottom\", \"NH via Boston, MA\", \"Im In Route \", \"the local dump\", \"Based in CA - Serve Nationwide\", \"Windsor,Ontario\", \"{Detailed}\", \"$ad $hawty\", \"Cimahi,West Java,Indonesia\", \"\\u00cc\\u00cfT: 27.9136024,-81.6078532\", \"Sherwood, Brisbane, Australia\", \"Rhyme Or Reason?\", \"Fort Knox, KY 40121\", \"Ecuador\", \"Avon\", \"JDB/LJC/AGB/TW/PLL\", \"Colonial Heights, VA\", \"Fruit Bowl\", \"Wolverhampton\", \"but i love kaylen ??\", \"Orlando \", \"Detroit Tigers Dugout\", \"Norf Carolina\", \"victoria moz\\u00cc\\u00a3o \", \"Playa\", \"\\u0089\\u00db\\u00a25\\u0089\\u00db\\u00a212\\u0089\\u00db\\u00a214\\u0089\\u00db\\u00a2 | \\u00e5\\u00c8#SaviourSquad\\u00e5\\u00c7\", \"Hartford, Connecticut\", \"Louisville, KY \", \"New Chicago\", \"Piedmont Triad, NC\", \"heart of darkness, unholy ?\", \"Victoria, BC  Canada\", \"Canberra, Australian Capital Territory\", \"La Grange Park, IL\", \"Boulder\", \"Littleton, CO, USA\", \"Thornton  Colorado\", \"#Bummerville otw\", \"Netherlands,Amsterdam-Virtual \", \"Norwich\", \"Chorley, Lancashire, UK\", \"Toronto, Worldwide \", \"Karachi \", \"Wilmington, Delaware\", \"UK  & Germany\", \"U.S.A and Canada\", \"Get our App\", \"LA\", \"Sanganer, Rajasthan\", \"LA/OC/Vegas\", \"Laventillemoorings \", \"Coventry, UK\", \"Rochester Hills, MI\", \"South West, England\", \"Annapolis, MD\", \"???????? ?????????.\", \"Paradise City\", \"keli x\", \"satan's colon\", \"'Merica\", \"nc\", \"Bangalore\", \"\\u0089\\u00db\\u00a2FLG\\u0089\\u00db\\u00a2\", \"Bangalore City, India\", \"Fresno, California\", \"San Mateo County, CA\", \"Somewhere between here & there\", \"The South & WestCoast \", \"#????? Libya#\", \"Hinterestland\", \"Bloomington, Indiana\", \"central chazifornia\", \"PSN: Pipbois \", \"New York 2099\", \"Calgary, Canada\", \"Patra-Greece.\", \"El Paso, TX\", \"iPhone: -27.499212,153.011072\", \"Roadside\", \"Vancouver\", \" 45\\u00e5\\u00a1 5'12.53N   14\\u00e5\\u00a1 7'24.93E\", \"\\u00cc\\u00cfT: 30.307558,-81.403118\", \"Pawnee\", \"The Globe\", \"God is Love. \", \"the Refrigerator \", \"Carterville\", \"Storybrooke \", \"Aix-en-Provence/Utrecht\", \"St.Cloud, MN\", \"Republic of the Philippines\", \"Nantes, France\", \"New Mexico, USA\", \"Higher Places\", \"Belleville, Illinois\", \"LITTLETON, CO, USA, TERRAN\", \"eating strawberry shitsickles\", \"Definitely NOT the stables\", \"Coasts of Maine & California\", \"Earth-616\", \"Kuwait \", \"boston\", \"Malibu/SantaFe/Winning!\", \"Guayaquil\", \"Tennessee/Gallifrey\", \"Buscame EL tu Melte\", \"taco bell\", \"Everett, WA\", \"The Land of MAss Stupidity\", \"QLD Australia\", \"Gaborone, Botswana\", \"828/704(Soufside)/while looking goofy in NJ\", \"Antarctica\", \"St. Joseph, Minnesota\", \"where the wild things are\", \"Aveiro, Portugal\", \"ohio\", \"Milton Keynes \", \"Montr\\u00cc\\u00a9al, Qu\\u00cc\\u00a9bec\", \"University of Toronto\", \"Global Edition\", \"Bolton & Tewkesbury, UK\", \"Aro Diaspora\", \"Holly Springs, NC \", \"Anaheim\", \"Newcastle, OK\", \"Live Oak, TX\", \"Washington, DC NATIVE\", \" Nevada Carson City,Freeman St\", \"Jogja, Indonesia Slowly Asia\", \"Me mammy's belly\", \"Bay Area, CA\", \" The World\", \"mpls. \", \"North London\", \"CAMARILLO, CA\", \"Tarragona\", \"Georgia ? Tennessee\", \"Scotts Valley, CA\", \"golborne, north west england.\", \"New York, United States\", \"The Jewfnited State\", \" Alex/Mika/Leo|18|he/she/they\", \"Quantico, VA\", \"nbc washington\", \"ELVY\", \"Varies \", \"oxford\", \"Mumbai india\", \"Riverside, California.\", \"Pig Symbol, Alabama\", \"Uruguay / Westeros / Gallifrey\", \"Whole World \", \"Charleston, IL\", \"Sydney & Worldwide\", \"Moore, OK\", \"Cincinnati, OH\", \"mi\", \"The 5th Dimension. \", \"????? ???? ????\", \"#MadeInNorthumberland\", \"NAIROBI  KENYA \", \"Based out of Portland, Oregon\", \"Columbia Heights, MN\", \"Virgo Supercluster\", \"Pontevedra, Galicia\", \"A Hoop Somewhere\", \"Missouri, USA\", \"Auburn \", \"Arnhem, the Netherlands\", \"US-PR\", \"EVERYWHERE\", \"toronto\", \"dorito land\", \"atx\", \"en el pais de los arrechos\", \"Pocatello, ID\", \"Uppsala, Sweden\", \"Milwaukee County\", \"followurDREAMS(& my instagram)\", \"My subconscious\", \"The Great State of Maine \", \"V-RP @OZRP_ ?MV, AU, R18+?\", \"Tulalip, Washington\", \"The Kingdom of Fife, Scotland\", \"Where ever i please\", \"Geneva. And beyond. \", \"Escondido, CA\", \"Between the worlds \", \"CCH \", \"Reading a romance novel\", \"y/e/l\", \"Ventura, Ca\", \"Greenville,SC\", \"wherever there's netflix\", \"?@symbolicjensen?\", \"Bombardment Bay\", \"Ile-Ife,Osun state, Nigeria\", \"North\", \"Indiana, USA\", \"Charlotte, N.C.\", \"Suitland\", \"Hoxton, London\", \"Groton, CT\", \"Minneapolis - St. Paul\", \"NYC&NJ\", \"Shirley, NY\", \"DFW, Texas\", \"Killarney\", \"#SOUTHAMPTON ENGLAND\", \"UK & Ibiza\", \"Palma, Islas Baleares\", \"Espoo, Finland\", \"Forging my Story\", \"Hampshire UK\", \"Hamilton County, IN\", \"fluffy cloud\", \"Kokomo, In\", \"Porto Alegre, Rio Grande do Sul\", \"New Your\", \"Still. ??S.A.N.D.O.S??\", \"Ikeja, Nigeria\", \"USA \", \"antoine fisher \", \"Middle Earth / Asgard / Berk\", \"Online 24/7. Not even kidding.\", \"nj\", \"Cornwall\", \"KLA,Uganda\", \"Near Warrington\", \"the own zone layer \", \"lakewood colorado\", \"Bedford, England\", \"paradise\", \"Earth \", \"BC\", \"University of Limerick\", \"Banbridge\", \"hkXfYMhEx\", \"toronto \\u0089\\u00db\\u00a2 dallas\", \"14/cis/istj \", \"Enniscrone & Aughris, Sligo \", \"Richardson TX\", \"Afghanistan, USA\", \"#Gladiator \\u0089\\u00db\\u00a2860\\u0089\\u00db\\u00a2757\\u0089\\u00db\\u00a2\", \"Unites States\", \"Lima-Peru\", \"Waterloo, ON\", \"Cloud 9\", \" Neverland \", \"Ylisse\", \"San Diego CA\", \"labyrinthia\", \"qosqo\", \"Aarhus, Central Jutland\", \"CA, AZ & NV\", \"Intermountain West\", \"Alphen aan den Rijn, Holland\", \"London / Berlin / Online\", \"Brizzle City !\", \"Brentwood, NY\", \"everywhere \", \"D(M)V  \", \"GREENSBORO,NORTH CAROLINA\", \"Montgomery, AL\", \"T-Ville\", \"lugo\", \"Nottingham\", \"sitting on Eddie Vedders lap,\", \"All around the world baby\", \"Bhubneshwar\", \"houston\", \"Ontario\", \"Cimerak - Pangandaran\", \"Londonstan\", \"Denver Colorado. Fun Times\", \"IJmuiden, The Netherlands\", \"ARIZONA\", \"Montreal\", \"Smash Manor/Kanto\", \"Diamondville\", \"indiana\", \"Mexico! ^_^\", \"infj \", \"Gainesville/Tampa, FL\", \"Leitchfield Kentucky\", \"My heart is a ghost town!\", \"Aurora, Ontario \", \"#HAMont\", \"Piedmont Area, North Carolina\", \"south africa eastern cape\", \"Not Of This World\", \"Malaysia/Jordan\", \"City of Angels, CA\", \"Galapa / Atl\\u00cc\\u00c1ntico\", \"Here & There\", \"Somewhere else...\", \"Elchilicitanierraversal \", \"Liberty Lake, WA\", \"NIFC\", \"Athens, Greece\", \"Here, there and everywhere\", \"guaravitas\", \"Salt Lake City, UT\", \"Suginami-ku, Tokyo, Japan\", \"New Haven, Connecticut\", \"Thrissur\", \"Spokane, Washington 99206\", \"One World\", \"Your Conversation\", \"santo domingo\", \"Wailuku, Maui\", \"Made in America\", \"Leduc, Alberta, Canada\", \"50% Queanbeyan - 50% Sydney\", \"Cedar Island, Clinton CT 06413\", \"Saint Louis, Missouri\", \"jersey \", \"Shrewsbury\", \"bajaur\", \"Lincoln\", \"Oxford / bristol\", \"Bakersfield, CA\", \"Chiswick, London\", \"Plain O' Texas\", \"Illumination \", \"sneaking glances at Thancred\", \"St. Catharines, Ontario\", \"chillin at ceder rapids\", \"Waddesdon\", \"LEALMAN, FLORIDA\", \"Dundas, Ontario\", \"Corpus Christi\", \"The Weird Part of Wonderland\", \"Rock Springs, WY\", \"Slateport City, Hoenn\", \"Absecon, NJ\", \"Greenpoint\", \"Ondo\", \"Fredonia,NY\", \"Sumter, SC\", \"Nadiad ,Gujarat , India!!\", \"District 12 - Orange County\", \"Columbus ?? North Carolina\", \"planeta H2o\", \"Fairfax, VA\", \"Arvada, CO\", \"LIVERPOOL\", \"Trost District\", \"Mostly Yuin.\", \"Arkansas\", \"JKT48-Muse-A7X\", \"United Kingdom,Fraserburgh\", \"Holly, MI\", \"-?s?s?j??s-\", \"\\u00cc\\u00cfT: 39.982988,-75.261624\", \"Li\\u00cc\\u00acge\", \"UPTOWN \", \"Tucson, Az\", \"Ashxjonespr@gmail.com\", \"#GDJB #ASOT\", \"T\\u00cc\\u00c1chira - Venezuela\", \"The Shire\", \"Hannover, Germany\", \"Where I Need To Be\", \"Amman,Jordan\", \"Contoocook Valley Region of Ne\", \"\\u00e5\\u00ca(?\\u0089\\u00db\\u00a2`?\\u0089\\u00db\\u00a2\\u00e5\\u00ab)??\", \"Ventura\", \"Elizabeth, NJ\", \"Host of #MindMoversPodcast\", \"My mind is my world\", \"NOLA ?? TX\", \"ATL ? SEA \", \"amsterdayum 120615 062415\", \"jayankondacholapuram.tamilnadu\", \"204, 555 11 Ave. S.W.\", \"Santiago,Rep\\u00cc\\u00bcblica Dominicana\", \"Greenfield, Massachusetts\", \"Oakland, Ca\", \"\\u00cc\\u00cfT: 6.4682,3.18287\", \"[marvel\\u0089\\u00db\\u00a2dragon age\\u0089\\u00db\\u00a2wicdiv]\", \"Shity land of Northern Ireland\", \"Eureka, California, USA\", \"Winston Salem, North Carolina\", \"Melbourne, FL\", \"Austin\", \"We're All Mad Here\", \"Los Angeles New York\", \"IG/SC:bjfordiani\", \"wny\", \"In the clouds...\", \"Bloomington, IN\", \"between ideas & 3-5pm AEST\", \"Trapped in my Conscience \", \"Birmingham and the Marches\", \"Dil's Campsite\", \"Daddy Kink Central\", \"The Low-Cal Calzone Zone\", \"ngapain?\", \"R'lyeh, South Pacific\", \"V\\u00cc_ster\\u00cc\\u00b4s, Sweden\", \"North West England UK\", \"Terlingua, Texas\", \"ChicagoRObotz\", \"Bristol, UK\", \"Varanasi\", \"Auburn, AL\", \"Gloucestershire , UK\", \"neil's kitchen  |  32215\", \"Croydon\", \"Fukushima city Fukushima.pref\", \"Morganville, Texas.\", \"Somerset, UK\", \"West Chester, PA\", \"Cyprus\", \"Devon/London \", \"Chester\", \"ona block w/ my BOY ??\", \"Killa Hill, CO\", \"Newport, Wales, UK\", \"Belgrade\", \"??????????? ???????????..? \", \"www.tmgcgart.com\", \"Argentina\", \"Zeerust, South Africa\", \"Mysore, Karnataka\", \"Also follow ?\", \"Australian Capital Territory\", \"Antigua ?? NYC \", \"Phila.\", \"maryland\", \"Leeds, United Kingdom\", \"Brisbane Australia\", \"Brum/Lestah \", \"seoul\", \"Melton, GA\", \"New York ? ATL\", \"Himalayan Mountains\", \"Durand, MI\", \"Romania\", \"Pueblo, Colorado\", \"Surabaya\", \"Lytham St Anne's \", \"Western Washington\", \"Jersey City, NJ\", \"all over the world\", \"don't run\", \"Colorado/WorldWide\", \"Tampa\", \"Victoria, Canada\", \"Black Canyon New River, AZ\", \"Hatteras, North Carolina\", \"Yobe State\", \"Orange County, Calif.\", \"\\u0089\\u00db\\u00ca\\u0089\\u00db\\u00ca\\u0089\\u00db\\u00ca\", \"San Antonio-ish, TX\", \"Iliff,Colorado  \", \"livin in a plastic world\", \"??9?\", \"Erbil\", \"Hampton Roads, VA\", \"Alberta Pack\", \"w. Nykae \", \"am everywhere\", \"Palestine \", \"cleveland, oh\", \"Alexandria, VA, USA\", \"Atlanta(ish), GA\", \"Scotland \", \"BIG D  HOUSTON/BOSTON/DENVER\", \"Memphis, in the Tennessees\", \"Greeley, CO\", \"on to the next adventure\", \"Some where\", \"The Pumpkin Carriage of Dreams\", \"Leeds, U.K.\", \"Fashion Heaven. IG: TMId_\", \"chicago\", \"Walthamstow, London\", \"Tornado Alley, USA \", \"NV\", \"oman muscat al seeb \", \"England,UK,Europe,Sol 3.\", \"Fort Fizz, Ohio\", \"Light and dark, form and void\", \"Dover, DE\", \"Guildford, UK\", \"\\u00cc\\u00cfT: 41.252426,-96.072013\", \"Elkhart, IN\", \"Hamilton, ON\", \"New Delhi,India\", \"hyderabad\", \"?? ?+254? ? \\\\??\\u00e5\\u00a1_??\\u00e5\\u00a1_???\\u00e5\\u00a1_?/??\", \"#NewcastleuponTyne #UK\", \"Jonesboro, Arkansas USA\", \"The Memesphere\", \"Hyrule\", \"Waterfront\", \"Hillsville/Lynchburg, VA\", \"The Epicenter, and Beyond\", \"Sandton, South Africa\", \"phuket thailand\", \"11th dimension, los angeles\", \"Gold Coast, Qld, Australia\", \"WV, love the blue and gold\", \"Elsewhere, NZ\", \"City Of Joy\", \"front row at a show\", \"Houston, Texas\", \"Kolkata, India\", \"Am International\", \"i luv raquel\", \"WESTSIDE OF PHILLY 7? BLOCK??\", \"Mumbai (India)\", \"Pratt-on-Wye\", \"Niall's arms\", \"Detroit/Windsor\", \"please H? ?:??\", \"Rochester, NY\", \"Grimsby, England\", \"Bartholomew County, Indiana\", \"Adventuring in Narnia\", \"State of Dreaming\", \"http://twitch.tv/jcmonkey\", \"Uganda\", \"Trackside California\", \"Edmonton, Alberta\", \"Inang Pamantasan\", \" Blood Indian Reserve\", \"Anderson, SC\", \"taking pain like pleasure\", \"Carol Stream, Illinois\", \"Aberdeenshire\", \"Joshua Tree, CA\", \"KADUNA, NIGERIA\", \"leyland\", \"AZ\", \"Very SW CA, USA....Draenor\", \"3.28.15|7.20.15|7.25.15\", \"M\\u00cc\\u00a9xico\", \"Liberty Township, Ohio\", \"trapped in America\", \"USA, Alabama\", \"Rockford, IL\", \"?????? in Yokohama Japan\", \"From a torn up town MANCHESTER\", \"Dallas, Tejas\", \"Kleenex factory\", \"Quincy\", \"Bug Forest\", \"The Pig Sty\", \"Norwalk, CT\", \"Ealing, London\", \"Kamloops, BC\", \"Arkansas, Jonesboro\", \"Quincy MA\", \"POFFIN\", \"Somewhere \", \"Overland Park, KS\", \"marysville ca \", \"All around the world!\", \"Email: Lovethterry@gmail.com\", \"Whiterun, Skyrim\", \"Wakanda\", \"MI\", \"Home is where we park it!\", \"Alvin, TX\", \"Rockville, Maryland\", \"khartoum sudan\", \"Hawaii USA\", \"Jakarta, Indonesia\", \"Philippines \", \"Every where\", \"Not a U.S resident\", \"Htx\", \"Bangalore, INDIA\", \"Las Vegas, NV\", \"Some other mansion\", \"London.\", \"Gloucester\", \"Raleigh Durham, NC\", \"@cockerelshoes\", \"Gettysburg, PA\", \"Lawrence, KS via Emporia, KS\", \"Peterborough, Ont.\", \"Palestine Texas\", \"??????????????\", \"Kaneohe\", \"they/them\", \"Rome, Italy\", \"Asgard\", \"Kualar Lumpur, Malaysia\", \"nigeria\", \" New Delhi \", \"London, Sydney\", \"Beside Basketball\", \"Cleveland, Ohio\", \"Norman, Oklahoma\", \"Bangalore. India\", \"Cassadaga Florida\", \"650/559\", \"kano\", \"Somewhere out there\", \"Fort Lauderdale, FL\", \"?? ??\", \"?????? ???? ??????\", \"2B Hindhede Rd, Singapore\", \"Hertfordshire \", \"seattle grace mercy death\", \"South Bloomfield, OH\", \"Team Slytherin\", \"Eugene, Oregon\", \"Chippenham/Bath, UK\", \"Charlotte County Florida\", \"1D | 5SOS | AG\", \"liverpool \", \"W.I.T.S Academy\", \"West Bank, Gaza Strip\", \"Fort Walton Beach, Fl\", \"Fort Calhoun, NE\", \"probably the strip club\", \"Semarang, Indonesia\", \"Sioux Falls, S.D. \", \"Flipadelphia\", \"Mars\", \"Wrex\", \"Idaho\", \"Inverness, Nova Scotia\", \"Scituate, MA\", \"Kuwait\", \"The D\", \"Shah Alam,Malaysia\", \"yorkshire\\n\", \"Hensley Street, Portland\", \"36 & 38\", \"Realville\", \"Palermo, Sicily\", \"GO BLUE! HAIL YES!!\", \"Rhode Island\", \"Kernow\", \"Greenwich Meridian\", \"Bournemouth, Dorset, UK\", \"Durban, South Africa\", \"i got 1/13 menpa replies, omg\", \"Madisonville TN\", \"NC || OR\", \"in my own personal hell (:\", \"FOLLOWS YOU everywhere you go\", \"bk. \", \"Brighton and Hove\", \"The Peach State\", \"IRAQ\", \"Sydney, NSW\", \"The Great State of Texas\", \"NYHC\", \"MY RTs ARE NOT ENDORSEMENTS\", \"Bishops Stortford, England\", \"Instagram: trillrebel_\", \"russia\", \"ilford\", \"L/S/Z/L/T/H/C/H/R/A/S/C\", \"Cape Neddick, ME\", \"EARTH\", \"ColoRADo\", \"Adelaide\", \"BOSTON-LONDON\", \"Queen Creek AZ\", \"ill yorker\", \"Mexico City\", \"Kingswinford\", \"Arlington, VA\", \"NEW YORK\", \"Vancouver, BC, Canada\", \"Atlanta - FAU class of '18\", \"Brasil,SP\", \"Newton, NJ 07860\", \"Bronx, New York\", \"America | New Zealand \", \"Boulder, CO\", \"Paris (France)\", \"ducked off . . . \", \"Tulsa, OK\", \"Lubbock, Texas\", \"Somewhere in Spain\", \"Birmingham & Bristol\", \"ALWAYS DYING NEVER RESTING\", \"us\", \"Houston \", \"H / pez & sophia \", \"Spying on your thoughts\", \"California or Colorado\", \"South east of U.K\", \"Hustletown\", \"Jamshedpur, Jharkhand\", \"Milton/Tallahassee\", \"Nairobi , Kenya\", \"Wynne, AR\", \"North Cack/919\", \"Official Website\", \"Trancy Manor\", \"Honeymoon \\u00cc\\u00a3ve.\", \"Alexandria, Egypt.\", \"Saline, MI\", \"Right next to Compton\", \"Centurion \", \"Tallahassee Florida\", \"Viejo\", \"Planet Eyal, Shandral System\", \"on twitter \", \"Clean World\", \"Wausau, Wisconsin\", \"Columbus, Georgia\", \"London, United Kingdom\", \"lagos nigeria\", \"Statute Of Limitations_\", \"East London. \", \"Dubai, United Arab Emirates\", \"Head Office: United Kingdom\", \"North East Unsigned Radio\", \"eARth 3\", \"Clayton, NC\", \"Jersey Shore\", \"N. California USA\", \"Here And There\", \"www.twitch.tv/PKSparkxx\", \"the moon\", \"Lizzy's Knee\", \"Esp\\u00cc_rito Santo\", \"Lansing, Michigan\", \"Benedict College\", \" Somewhere.\", \"TechFish \", \"570 Vanderbilt; Brooklyn, NY\", \"Youngstown, OH\", \"Afghanistan\", \"Houston, TX  \", \"Spain - China - Latin America.\", \"St Joseph de Beauce\", \"Montana \", \"Peoria\", \"Hampstead, London.\", \"Wandsworth, London\", \"Some Where in this World\", \"not so cool KY\", \"whs '17\", \"Garden Grove\", \"Le Moyne '16\", \"Freeport IL. USA\", \"Eww, I'm not Paul Elam\", \"ARBAILO\", \"Melbourne-ish\", \"New Hampshire, USA\", \"Panam\\u00cc\\u00c1 \", \"taking bath do not disturb\", \"LONG ISLAND, NY\", \"Dallas, TX \", \"K\\u00cc\\u00a6ln, Nordrhein-Westfalen\", \"Cape Cod\", \"Loveland Colorado\", \"Ibadan,Oyo state\", \"Stay Fly?\", \"Bacon\", \"Bolivar, MO\", \"Malang\", \"ATX\", \"Tema,Accra\", \"St. John's, NL, Canada\", \"? miranda ? 521 mi\", \"?205?478?\", \"Avon, OH\", \"West Wales\", \"Hughes, AR\", \"Espa\\u00cc\\u00b1a - Spain - Espagne\", \"Numa casa de old yellow bricks\", \"New South Wales, Australia\", \"italy\", \"??? ???? ??????\", \"Lincoln, IL\", \"winston-salem north carolina\", \"Murray Hill, New Jersey\", \"New York. NY\", \"Kawartha Lakes, Ontario, Canad\", \"i love the smurfs 2\", \"Serva Fidem\", \"Maryland,Baltimore\", \"Port Charlotte, FL\", \"Proudly Canadian!\", \"??????????????????\", \"Antioch, CA \", \"NJ/NYC\", \"Victoria, Tx.\", \"Cochrane, Alberta, Canada\", \"  \\u00e5_ \", \"Born in Baltimore Living in PA\", \"\\u00cc\\u00cfT: 43.631838,-79.55807\", \"St Austell, Cornwall\", \"Wyoming, MI (Grand Rapids)\", \"Danbury, CT\", \"nowhere\", \"Williamstown, VT\", \"Lahar & Gwalior\", \"956\", \"Dreieich, Germany\", \"Calgary/Airdrie/RedDeer/AB\", \"Bucks County, Pa\", \"South Florida\", \"AFRICA\", \"Polmont \", \"  Melbourne, Australia\", \"in Dimitri's arms\", \"Aperture Science Test Facility\", \"Nicoma Park, OK\", \"walking the tightrope\", \"Lucknow, India\", \"L'Enfant Plaza Metro Station\", \"Oslo, Norway\", \"At your back\", \"Earth 0\", \"ava\", \"  News\", \"southern california\", \"im definitely taller than you.\", \"San Diego, Calif.\", \"Canadian bread\", \"San Juan, Puerto Rico\", \"THANJAVUR\", \"#keepthefaith J&J\", \"Searching for Bae \", \"Score Team Goals Buying @\", \"Mount Vernon, NY\", \"xiumin's nonexistent solos\", \"Vancouver, Colombie-Britannique\", \"Hinton, W.Va.\", \"Pelham, AL\", \"Isle of Man\", \"Pennsylvania, PA\", \"eBooks, North America\", \"Minneapolis/St. Paul\", \"on the web\", \"s\\u00cc\\u00a3o luis\", \"a botanical garden probably\", \"Spring Grove, IL\", \"Inside the Beltway (DC Area)\", \"Cambridge, MA\", \"BROOKLYN, NYC\", \"48.870833,2.399227\", \"Melrose\", \"Athens - Nicosia\", \"Concord, N.C.\", \"1/10 Taron squad\", \"Laguna Beach, Calif. \", \"Texas af\", \"92\", \"RSN: Tru\", \"Prehistoric Earth\", \"cereal aisle #17:i4\", \"iTunes - RSS\", \"Own planet!!\", \"under the blanket\", \"i'm a Citizen of the World\", \"Ad Majorem Dei Glorium\", \"Charleston S.C.\", \"United States where it's warm\", \"Wakefield MA\", \"Houma La\", \"The Windy City\", \"KOLKATA\", \"St. Patrick's Purgatory\", \"?????? ??? ?????? ????????\", \"??t?a\", \"London/Lagos/FL \\u00cc\\u00cfT: 6.6200132,\", \"Balikesir - Eskisehir\", \"Greater Los Angeles Bearia\", \"Thibodaux, LA\", \"Overton NV\", \"France\", \"Long Island NY & San Francisco\", \"cedar rapids ia\", \"Orange County, CA\", \"Fredericksburg, Virginia\", \"Davidson, NC\", \"Jacksonville Beach, FL\", \"Portage, IN / Worldwide\", \"Morocco\", \"MI,USA\", \"SaudI arabia - riyadh \", \"Bhopal, Madhya Pradesh, India.\", \"Cardiff, Wales\", \"17-Feb\", \"Manchester UK\", \"Freddy Fazbears pizzeria\", \"Yuuko-san's shop\", \"The barn\", \"Cypress, CA 90630\", \" ?currently writing a book?\", \"Long Beach, CA\", \"Amman, Jordan\", \"Colombo,Sri Lanka.\", \"my house\", \"#WashingtonState #Seattle\", \"Garden City, NY\", \"Peterborough, Ontario, Canada\", \"Bon Temps Louisiana\", \"Extraterrestrial Highway\", \"401 livin'\", \"Albuquerque\", \"Rheinbach / Germany\", \"Boksburg\", \"?????????????, Thailand \", \"midwest\", \"Basketball City, USA \", \"Killafornia made me \", \"South of D.C.\", \"WorldWideWeb\", \"Outside The Matrix, I Think.\", \"missouri USA\", \"Leeds\", \"Newcastle, England \", \"PURPLE BOOTH STUDIO\\u0089\\u00e3\\u00a2\", \"Harris County, Texas\", \"Cardiff, UK\", \" Bouvet Island\", \"Holland MI via Houston, CLE\", \"MA via PA\", \"Long Island\", \"Hermosa Beach, CA\", \"Kingston, Pennsylvania\", \"cork\", \"ON\", \"Cavite, Philippines\", \"Watertown, Mass.\", \"st.louis county missouri \", \"Portsmouth, VA\", \"Mountains\", \"Lowell, MA\", \"Phoenix, Arizona, USA\", \"Memphis,TN/ World Wide\", \"Hell\", \"17th Dimension\", \"Nicola Valley\", \"melbourne\", \"AEP\", \"All around the world\", \"IN\", \"State of Georgia\", \"Cape Cod, Massachusetts USA\", \"ITALY\", \"h+l\", \"Jacksonville, FL\", \"IL\", \"Pon Di Gully\", \"Enterprise, Alabama\", \"The windy plains of Denver\", \"Jackson, MS\", \"Chattanooga TN\", \"Attock\", \"Charlotte \", \"Warwick, RI @Dollarocracy also\", \"Ellensburg to Spokane\", \"Punjab\", \"//??//\", \"Downtown Oklahoma City\", \"Wonderland\\u0089\\u00db\\u00d3 ?????? ???? ??????\", \"Duval, WV 25573, USA ?\", \"Purgatory, USA\", \"The Orwellion police-state\", \"Brentwood,TN\", \"Free State, South Africa\", \"i love you zayn\", \"England, Great Britain.\", \"Atlanta, Georgia USA\", \"Thane\", \"EGYPT\", \"Sunrise Manor, NV\", \"The Desert\", \"buenos aires argentina\", \"Wisconsin, USA\", \"Dagenham, Essex\", \"876 Jamrock.\", \"Spinning through time.\", \"Washington, D.C., area\", \"ATLANTA , GEORGIA \", \"@UntmdOutdoors #T.O.R.K \", \"Frankfort, KY\", \"austin, texas\", \"Tamworth\", \"Caribbean\", \"Mackem in Bolton\", \"The Hammock, FL, USA\", \"Nashville, Tn\", \"????????????\", \"Vancouver (HQ) and worldwide\", \"828??864??803\", \"LAGOS\", \"Sheffield Township, Ohio\", \"East Kilbride\", \"Arthas US\", \"The shores of Lake Kilby\", \"Frome, Somerset, England\", \"1648 Queen St. West, Toronto.\", \"Park Ridge, Illinois\", \"Queens, NY\", \"Greenpoint, Brooklyn\", \"Pensacola, FL\", \"Football Field\", \"Espa\\u00cc\\u00b1a, Spain\", \"Dayton, Ohio\", \"Level 3 Garrison, Sector G\", \"Earth, Milky Way, Universe\", \" |IG: imaginedragoner\", \"Vermont, USA\", \"Beacon Hills\", \"Cottonwood Arizona\", \"toledo\", \"Beaumont, TX\", \"Gumptown\", \"Paris \", \"North Carolina \", \"Southern Califorina\", \"Internet\", \"Selangor\", \"Orange County, NY\", \"The Waystone Inn\", \"Fairy Tail! \", \"Alaska, USA\", \"CT, USA\", \"Brooklyn, New York\", \"Nairobi, Kenya \", \"Lexington\", \"Ted&Qz Inc, Ireland, Europe\", \" Jariana Town\", \"Manchester, The World, England\", \"Skyport de la Rosa\", \"NEWCASTLE\", \"I ACCEPT SONG REQUESTS\", \"Greenville, S.C.\", \"japon\", \"Hilton Head, SC  \", \"Slatina,Romania\", \"where I'm supposed to be\", \"London/Outlaw Country \", \"Glenview to Knoxville \", \"dublin \", \"Valle Del Sol\", \"New Jersey \", \"#MayGodHelpUS\", \"Ktx\", \"LOCAL ATLANTA NEWS 4/28/00 - 4/28/15 FREELANCER\", \"music.\", \"Cobblestone\", \"Sylacauga, Alabama\", \"WORLDWIDE-BOSTON\", \"WAISTDEEP, TX\", \"Chandler, AZ\", \"West Lancashire, UK.\", \"United Hoods of the Globe\", \"Surrey & Manchester\", \"Positive 852\", \"Tunbridge Wells\", \"Paname City\", \"Toledo, OH\", \"Hudson Valley, NY\", \"Greater Manchester, UK\", \"Pioneer Village, KY\", \"94123\", \"Republica Dominicana\", \"RhodeIsland\", \"Ljubljana, Slovenia\", \"[ kate + they/them + infp-t ]\", \"BOT ACCOUNT\", \"\\u0089\\u00db\\u00a2OlderCandyBloom\\u0089\\u00db\\u00a2\", \"somewhere outside\", \"Va Beach, Virginia\", \"Petaluma, CA\", \"Dhaka\", \"Lincoln City Oregon\", \"The Internet\", \"Frisco, TX\", \"@notoriousD12\", \"austin tx\", \"garowe puntland somalia\", \"NYC metro\", \"Bikini bottom\", \"Whippany, NJ\", \"Palm Beach County, FL\", \"NEPA/570\", \"Shelby County\", \"Marrakech M\\u00cc\\u00a9dina, Marrakech - Tensift - Al Haouz\", \"Eagle Pass, Texas\", \"NJ/NY/NM/NE/ND\", \"St. Louis, Mo\", \"Lynnfield, MA\", \"south of heaven \", \"Los Angeles, Calif.\", \"Atlanta, Ga\", \"Mogadishu, Somalia\", \"Justin and Ariana follow\", \"Danville, VA\", \"Davis, California\", \"Sacramento, California\", \"Brazil \", \"Russia\", \"Somewhere with Clyde\", \"Miami ??\", \"Gages Lake, IL\", \"Yuba City, CA\", \"52.479722, 62.184971\", \"Hendersonville, NC\", \"Yewa zone\", \"Stage with Trey Songz\", \"Alameda, CA\", \"Spring Tx\", \"Atlanta g.a.\", \"Thailand Malaysia Indonesia \", \"dope show\", \"An eight-sided polygon\", \"NY Capital District\", \"Bridport, England\", \"Weston super mare\", \"MUM-DEL\", \"Greenville\", \"?s????ss? a?????\", \"Vancouver, Canada\", \"kisumu\", \"the void, U.S.A\", \"Chicago, IL 60607\", \"M\\u00cc\\u00a9rida, Yucat\\u00cc\\u00c1n\", \"US: 44.414510,8.942499\", \"Louisville, KY\", \"Benicia, CA \", \"Huntsville AL\", \"Quilmes , Arg\", \"Lynchburg, VA\", \"San Diego, Texas.\", \"everydaynigerian@gmail.com\", \"North Dartmouth, Massachusetts\", \"Rural Northern Nevada\", \"Bronx, NY\", \"Estados Unidos\", \"nj/ny\", \"Sugarhouse, UT\", \"Traverse City, MI\", \"brisbane, australia\", \"Valparaiso \", \"Morris, IL\", \"Amsterdam\", \"D.C. - Baltimore - Annapolis\", \"BestCoast\", \"World news\", \"Henderson, NV\", \"Jerusalem!\", \"London, Kent & SE England.\", \"Quezon City, Philippines\", \"Pompano Beach, FL\", \"Pacific Northwest\", \"eritrean\", \"Ocean City, NJ\", \"Live4Heed??\", \"Top secret bunker \", \"Delhi\", \"beijing .China\", \"The P (South Philly)\", \"Harlingen, TX\", \"Narnia, Maryland\", \"Los Angles, CA\", \"San Francisco , CA\", \"instagram- Chloe_Bellx\", \"Welt\", \"Doghouse\", \"Beirut, Lebanon\", \"Nagpur\", \"Central Illinois\", \"Selena | Britney | Hilary\", \"SOUTHERN CALIFORNIA DESERT\", \"Otsego, MI\", \"Back East in PA\", \" Nxgerxa\", \"buffalo / madrid / granada\", \"Central Florida\", \"  \", \"9/1/13\", \"Fort Collins, CO\", \"prob turning up with sheen\", \"Alberta | Sask. | Montana\", \"Vista, CA\", \" Indiana\", \"shoujo hell \", \"\\u00cc\\u00cfT: 40.707762,-74.014213\", \"Hailing from Dayton \", \"University Heights, Ohio\", \"Pembroke NH\", \"Canada Eh! \", \"Temporary Towers\", \"Sheffield/Leeds\", \"#BlackLivesMatter\", \"Hawthorne, NE\", \"Van Buren, MO\", \"death star\", \"Pocatello, Idaho\", \"I-75 in Florida\", \"Ames, IA\", \"texasss\", \"USA - Canada - Europe - Asia\", \"USA, WA\", \"fujo garbage heaven \", \"Waterloo, Ont\", \"Ziam af \", \"Fukuoka, Japan\", \"Johannesburg \", \"\\u00cc\\u00cfT: 40.562796,-75.488849\", \"IN our hearts  Earth Global \", \"Eastern Iowa\", \"cody, austin follows ?*?\", \"Big NorthEast Litter Box\", \"See the barn of bleakness\", \"WORLDWIDE!\", \"Pomfret/Providence\", \"Medford, NJ\", \"Bellevue NE\", \"\\u00cc\\u00cfT: 42.910975,-78.865828\", \"Miami Beach, Fl\", \"Crayford, London\", \"turner fenton\", \"Displaced Son of TEXAS!\", \"St. Louis Mo.\", \"Zero Branco\", \"Wellington\", \"Frascati\", \"18 \\u0089\\u00db\\u00a2 CC\", \"Louisiana, USA\", \"Burlington, VT\", \"\\u00e5_\\u00e5_Los Mina City\\u0089\\u00e3\\u00a2\", \"STL ?NOLA\", \"Montgomery County, MD\", \"?^\\u00e5\\u00e1??\\u00e5\\u00e1?^?? ??\", \"Ohio\", \"10-Jul\", \"Kalamazoo, Michigan\", \"AUS\", \"518\", \"ANYWEHERE !!\", \"louisville, kentucky\", \"Penn Hills, PA\", \"statesboro/vidalia\", \"Poplar, London\", \"CLVLND\", \"??? ???? ?f glory. ?\", \"Milwaukee WI\", \"aggressive cannoli eater \", \"World Wide!!\", \"Le Memenet\", \"Harper Woods, MI\", \"Starling City\", \"Walker County, Alabama\", \"Oklahoma, USA\", \"Davao City\", \"Sunshine Coast, Queensland\", \"milky way\", \"Victorville, CA\", \"elizabeth king\", \"Austin TX\", \"NAWF SIDE POKING OUT \", \"Michigan \", \"Gloucester, MA\", \"Fort Myers, Florida\", \"In the middle of no where\", \"\\u00cc\\u00cfT: 1.50225,103.742992\", \"6\", \"Lehigh Valley, PA\", \"Wema building\", \"Bagalkote Karnataka \", \"Lancaster, CA\", \"NBO\", \"Yadkinville, NC\", \"In Your Notifications \", \"di langit 7 bidadari (^,^ )\", \"Mogadishu, New Jersey\", \"Sicamous, British Columbia\", \"heccfidmss@gmail.com\", \"El Dorado, KS\", \"2,360 miles away\", \"Port Jervis, NY\", \"Pluto\", \"Isolated City In World Perth\", \"Highland Park, CA\", \"Peru\", \"Milton Keynes, England\", \"SWinfo@dot.state.al.us\", \"We are global!\", \"Castaic, CA\", \"Multinational *****\", \"USA , AZ\", \"Cuernavaca, Morelos, M\\u00cc\\u00a9xico.\", \"SURROUNDED BY WEEABOOS\", \"Kensington, MD\", \"(he/him)\", \"Reston, VA, USA\", \"Lake Highlands\", \"Upstate New York\", \"New York - Connecticut\", \"Cairo, Egypt\", \"Novi, MI\", \"SEA Server\", \"Dundee, UK\", \"Hospital, bc of SKH vid.\", \"\\u00cc\\u00cfT: 35.223347,-80.827834\", \"Vero Beach , FL\", \"my deli\", \"Kansas\", \"USA, Haiti, Nepal\", \"Mississauga, Ontario\", \"US, PA\", \"Vidalia GA\", \"Moscow, Russia\", \"www.youtube.com?Malkavius2\", \"Riverdale, GA \", \"he/him or she/her (ask)\", \"Florida Forever\", \"Pueblo, CO\", \"Sao Paulo, Brazil\", \"Silesia, Poland\", \"dubai \", \"Reading UK\", \"-6.152261,106.775995\", \"Kwajalein/Virginia/Dayton, OH\", \"Sutton, London UK\", \"21 | PNW\", \"Timaru District, New Zealand\", \"??+ ... ??+\", \"WORLD\", \"Passamaquoddy\", \"www.facebook.com/stuntfm\", \"potters bar\", \"\\u0081\\u00cawagger!\\u00cc\\u0090ominican\\u00cc\\u00f7\", \"call me peach or sam lo\", \"Land Of The Kings\", \"Dappar (Mohali) Punjab\", \"DMV\", \"Atlantic Highlands, NJ\", \"Hampshire, UK\", \"Photo : Blue Mountains \", \"Belbroughton, England\", \"Arundel \", \"Tucson, AZ\", \"Los Angeles,CA, USA\", \"In the spirit world\", \"Methville, CA\", \"Fort Worth,  Texas \", \"Insula Barataria\", \"Bleak House\", \"Southern Maine\", \"Fresno\", \"#937??#734\", \"london / st catharines ?\", \"Torrance, CA\", \"Havenford\", \"Oneonta, NY/ Staten Island, NY\", \"Fairgrounds Resident\", \"T E X A S | wwat 8.24.14\", \"Reality\", \"Jeddah_Saudi Arabia.\", \"England \", \"The green and pleasant land.\", \"Your Sister's Bedroom\", \"sitting on the fence, New York\", \"Edinburgh, Scotland\", \"The ?? below ???\", \"In the Shadows\", \"West Midlands\", \"@potteratthedisc\", \"Wakefield, West Yorkshire\", \"#iminchina\", \"Halton Region\", \"WORDLDWIDE\", \"khanna\", \"Menasha, WI\", \"2005 |-/\", \"Tafekop Ga-Matsepe\", \"Worldwide.\", \"Funtua, Nigeria\", \"LA ??\", \"Co. Tyrone Northern Ireland\", \"Wood Buffalo, Alberta\", \"Nova Scotia, Canada\", \"3000 miles from everyone\", \"? icon by @Hashiren_3 ?\", \"Pittsburgh PA\", \"ATL, GA\", \"Nigeria, Global\", \"Pro-American and Anti-#Occupy\", \"West Africa\", \"Basking Ridge, NJ\", \"THE 6IX\", \"Sunny South florida \", \"Tama, Iowa\", \"Worcester, MA\", \"Illinois\", \"Lurking\", \"peekskill. new york, 10566 \", \"Baton Rouge\", \"Ab, Canada\", \"icon: cheese3d\", \"access to njh/5 and cth/4\", \"Charlotte, NC | K\\u00cc\\u00a6ln, NRW\", \"U.S.A.   FEMA Region 5\", \"Fountain Valley, CA\", \"swindon\", \"1313 W.Patrick St, Frederick\", \"Alabama, USA\", \"UK, Republic of Ireland and Australia\", \"VISIT MY YOUTUBE CHANNEL.\", \"Stanford University\", \"have car; will travel\", \"#EngleWood CHICAGO \", \"Warsaw\", \"Robin Hood's County \", \"College Station, TX\", \"LFC x GSW\", \"Someday I'll live in England. \", \"Travelling around the world \", \"West Palm Beach, Florida\", \"ca(NADA) \", \"Barcelona, Spain\", \"Anywhere\", \"New York, NY \", \"Fleet/Oxford, UK\", \"St Louis, MO\", \"Stalybridge, Tameside\", \"Mid West\", \"Riyadh ')\", \"Abuja,Nigeria\", \"Alberta, VA\", \"Abuja, Nigeria\", \"La Puente, CA\", \"Bronx NYC / M-City NY\", \"zboyer@washingtontimes.com\", \"Utica NY\", \"Principality of Zeron\", \"Every Where in the World\", \"Franklin, TN near Nashville\", \"Cascadia\", \"Rapid City, Black Hills, SD\", \"washington, d.c.\", \"dundalk ireland\", \"Salt Lake City, Utah\", \"SEATTLE, WA USA\", \"???????, ??'??????\", \"Giddy, Greenland\", \"JamDung\", \"Layang-Layang, Perak\", \"\\u0089\\u00db\\u00a2III.XII.MMXI\\u0089\\u00db\\u00a2\", \"Headed To The Top\", \"cuba\", \"Somecity, Somerset, MD\", \"West Coast, Cali USA\", \"come here in 20 minutes for an ass kicking\", \"New York / Worldwide\", \"Srinagar,Kashmir\", \"Amsterdam & Worldwide\", \"Washington, DC 20009\", \"San Luis Obispo, CA\", \"NYC,US - Cali, Colombia\", \"manchester, uk.\", \"'SAN ANTONIOOOOO'\", \"Wiltshire\", \"??? ?????????????\", \"Athens,Greece\", \"Amazon Seller , Propagandist\", \"Campo Grande-MS\", \"South Carolina, USA\", \"Bozeman, Montana\", \"dreamy lake\", \"WestEnd, Puritan Ave \", \"Lahore\", \"Charleston, SC\", \"D\\u00cc_sseldorf, Germany\", \"A sofa\", \"Jonesboro, AR MO, IOWA USA\", \"Saltillo, Coahuila de Zaragoza\", \"PunPunl\\u00cc\\u00a2ndia\", \"Somewhere Powerbraking A Chevy\", \"#RedSoxNation\", \"Ivano-Frankivsk\", \"On the toilet having a dump \", \"balvanera\", \"elena's bed // info on link\", \"The Wood\", \"Heinz Field \", \"518 \\u00e5\\u00e1 NY\", \"Helsinki, Finland\", \"\\u00e5\\u00a1\\u00e5\\u00a1Midwest \\u0089\\u00db\\u00a2\\u0089\\u00db\\u00a2\", \"Yulee, FL\", \"SoDak\", \"InterplanetaryZone\", \"PS4, now stop asking\", \"S\\u00cc\\u00a3o Paulo, Brasil\", \"Miami, Florida\", \"wherever-the-fuck washington\", \"Buxton, Venice, and Nottingham\", \"Tips on my blog at\", \"kansas\", \"Broadview Heights, Ohio\", \"North America\", \"USA/SO FLORIDA via BROOKLYN NY\", \"#goingdownthetoilet Illinois\", \"This Is Paradise. Relax. \", \"Espa\\u00cc\\u00b1a\", \"ACCRA GHANA\", \"Brazos Valley, Texas\", \"Somewhere Around You\", \"twitch.tv/naturalemblem26\", \"Bathtub de Bett \", \"marvel | books | hp | tmr\", \"GOT7SupportPH\", \"Cameroon\", \"Conversing In Janet's Caf\\u00cc\\u00ac\", \"Noida, NCR, India\", \"IG: AyshBanaysh\", \"Some pum pum\", \"Seattle WA\", \"norway\", \"Austin/Los Angeles\", \"Littleton, CO\", \"Wild Wild Web\", \"Tallahassee, FL\", \"The Canopy Kingdom\", \"Waterford MI\", \"Cymru araul\", \"GLOBAL/WORLDWIDE\", \"Haveli, Maharashtra\", \"Northern Kentucky, USA\", \"Swan River\", \"Pune, mostly \", \"Tyler, TX\", \"Soufside\", \"Bushkill pa\", \"san gabriel la union\", \"Palmyra, NJ\", \"Chicago,Illinois\", \"VCU\", \"EARTH \", \"Live m\\u00cc\\u00c1s\", \"israel\", \"Halifax, NS, Canada\", \"Paris, France\", \"The UK\", \"University of South Florida\", \"EastAtlanta ??#WestGeorgia'18\", \"Nevada (wishing for Colorado)\", \"Atlanta,Ga\", \"Sunbury, Ohio\", \"Neverland\", \"CA ??DC\", \"GCC\", \"nearest trash can \", \"Hollywood\", \"Lahti, Finland\", \"\\u00e5\\u00f8\\\\_(?)_/\\u00e5\\u00f8\", \"On a beach \", \"Rochester\", \"Alicante, Valencia\", \"Astley, Manchester\", \"Hamburg, DE\", \"PSA Nursing \", \"Goa, India\", \"In my studio\", \"NYC, New York\", \"PA, USA\", \"Trumann, Arkansas\", \"on the go\", \"belleville\", \"Tucson, Arizona \", \"Oshawa, Canada\", \"Fort Walton Beach, FL\", \"Orlando,FL  USA\", \"probably not home\", \"timeline kamu\", \"Lynwood, CA\", \"Wolverhampton/Brum/Jersey\", \"TV5, Philippines\", \"( ?\\u00e5\\u00a1 ?? ?\\u00e5\\u00a1), \", \"The land of New Jersey. \", \"Whitby, ON\", \"Northampton, MA\", \"nap central\", \"Magnolia\", \"Maryland \", \"Eaubonne, 95, France\", \"Americas Newsroom\", \"they/her\", \"somewhere in cali \", \"Two Up Two Down\", \"Topeka, KS\", \"Fort Smith, AR\", \"East Islip, NY\", \"Pleasanton, CA\", \"Las Vegas aka Hell\", \"Across the Atlantic\", \"Bedford IN \", \"NYC area\", \"Stay Tuned ;) \", \"5-Feb\", \"PA.USA\", \"Saudi Arabia\", \"Hamilton, Ontario CA\", \"Kenton, Ohio\", \"Skyhold\", \"261 5th Avenue New York, NY \", \"gamertag: bexrayandvav \", \"Chennai\", \"Scottsdale. AZ\", \"Jubail IC, Saudi Arabia\", \"Carregado\", \"Queens New York\", \"???? ???????\", \"I rap to burn shame.\", \"Over the Moon...\", \"Surulere Lagos,Home Of Swagg\", \"Orlando/Cocoa Beach, FL\", \"Inside your mind.\", \"Atlantic, IA\", \"Chicago - Lake Buena Vista\", \"Slappin and Smackin \", \"Lansdale,Pennsylvania\", \"Miami?Gainesville\", \"Washington, D.C. \", \"U.S.\", \"U.S\", \"Christiana,Tennessee\", \"Louavul, KY\", \"Broomfield, CO\", \"Aztec NM\", \"Philadelphia, Pennsylvania USA\", \"The Triskelion\", \"Land of Lincoln\", \"Nevada, USA\", \"LA - everywhere\", \"philly \", \"Atlanta, Georgia\", \"Plano, Texas\", \"Keighley, England\", \"The North\", \"Los Angeles... CA... USA\", \"Canada \", \"alberta, canada\", \"Tampa-St. Petersburg, FL\", \"Bangkok\", \"Bulgaria\", \"The Internetz\", \"NY, NY\", \"Bayonne, NJ\", \"M!$$!$$!PP!\", \"Wherever I'm sent\", \"Brasil, Fortaleza ce\", \"Saskatchewan, Canada\", \" Miami Beach\", \"Room 234\", \"Brooklyn\", \"South Korea GMT+9\", \"UK,singer,songwriter,?2 act\", \"Cape Town, Khayelitsha\", \"Dublin\", \"Deployed in the Middle East\", \"Warri\", \"miami\", \"M!A: None\", \"[ Blonde Bi Fry. ]\", \"mnl\", \"Washington DC / Nantes, France\", \"New York City ,NY\", \"Bangor, Co.Down\", \"The dark\", \"Winnipeg, MB, Canada\", \"Nirvana\", \"Kajang ? UiTM Puncak Alam\", \"The Internet & NYC\", \"Nanaimo, BC, Canada\", \"City of London, London\", \"Island Lake, IL\", \"Portoviejo-Manabi-Ecuador\", \"Swag Francisco\", \"Lima, OH\", \"CHICAGO\", \"i beg vines sorry \", \"jeddah | Khartoum\", \"Ames, Iowa\", \"cigarknub@gmail.com\", \"Studio\", \"Cuttack, Orissa\", \"Sharkatraz/Bindle's Cleft, PA\", \"Viterbo BFA Acting '18\", \"Pune, Maharashtra\", \"kenya\", \"No ID, No VOTE!!!\", \"m3, k, a, d\", \"Kuala Lumpur\", \"Kilkenny\", \"USAoV\", \"London / Birmingham\", \"Like us on Face \", \"BROKE NIGGAS DREAM!!\", \"Rogersville, MO\", \"middle eastern palace\", \"Bod\\u00cc\\u00fc, Norge\", \"Medford, Oregon\", \"Rapid City, South Dakota\", \"In a graveyard \", \"Calgary, Alberta, Canada\", \"Scottsdale, AZ\", \"laying on the bass\", \"In the Shadows...\", \"scumbernauld\", \"PG Chillin!\", \"West Coast, USA\", \"Up a hill\", \"Nice places \", \"Georgia, U.S.A.\", \"ECSU16\", \"Cambridge, Massachusetts, U.S.\", \"somewhere in Portugal\", \"Birdland, New Meridian, FD\", \"Predjama, Eslovenia.\", \"gaffney, sc \", \"Hame\", \"Here there and everywhere\", \"???????, Texas\", \"pittsboro\", \"Copenhagen, Capital Region of Denmark\", \"Argus Industries \\\\m/666\\\\m/\", \"New Delhi\", \"oklahoma\", \"Yeezy Taught Me , NV\", \"canberra\", \"627\", \"Wrigley Field\", \"??? ??? ????? ??? ???.\", \"on twitter\", \"watford\", \"Jaipur, India\", \"Gold Coast\", \"New Britain, CT\", \"Pekanbaru\\u00e5\\u00a1Batam Island\\u00e5\\u00a1Medan\", \"Houston |??| Corsicana\", \"cyprus\", \"\\u0089\\u00db\\u00a2901\\u0089\\u00db\\u00a2\", \"\\u00cc\\u00cfT: 19.123127,72.825133\", \"PH\", \"M\\u00cc\\u00a9xico D.F.\", \"MD\", \"Reddit\", \"#1 Vacation Destination,HAWAII\", \"Hueco Mundo\", \"Waco TX\", \"My old New England home\", \"WORLD WIDE\", \"Bremerton, WA\", \"Brackley Beach, PE, Canada\", \"Corpus - Las Vegas - Houston\", \"Lebanon, Tennessee\", \"Im Around ... Jersey\", \"#PhanTrash\", \"GrC Founder, 8,000 Subscribers\", \"Watch Those Videos -\", \"Largo, MD\", \"Wilbraham, MA\", \"Midwest City, OK\", \"Seattle native in Prescott, AZ\", \"Elmwood Park, NJ\", \"Ottawa,Ontario Canada\", \"Freeport il \", \"Henderson, Nevada\", \"HOMRA.\", \"SF Bay Area, California / Greater Phoenix, AZ\", \"Gurgaon, Haryana. \", \"Former Yugoslav Republic of Macedonia\", \"Soul Somalia/Body Montreal\", \"Sochi, KDA, RU\", \"the burrow\", \"Newcastle upon Tyne\", \"Friendswood, TX\", \"[Gia.] | #KardashianEmpire\", \"Yooooooo\", \"Warrandyte, Australia\", \"in my head\", \"Tacoma,Washington\", \"Colchester Essex \", \"Amsterdam | San Francisco\", \"St. Louis, Missouri\", \"Webster, TX\", \"You're not 19 forever   \", \"AUSTRALIA-SOUTHAFRICA-CAMBODIA\", \"Providence RI / Lisnaskea \", \"East Lansing, MI\", \"New Brunswick, NJ\", \"Birmingham, UK\", \"Las Vegas, NV USA\", \"QUEENS.\", \"Napa, CA\", \"Stratford, CT\", \"Bossland\", \"Shipwreck Cove\", \"Arlington, VA and DC\", \"Porthcawl\", \"Warm Heart Of Africa\", \"The Meadow\", \"Toronto-Citizen of Canada & US\", \"The Main \", \"Gameday\", \"Cambridge, Massachusetts\", \"DC Metro area\", \"U.S.A. - Global Members Site\", \"fl\", \"Ely, Cambridgeshire\", \"Tokyo & Osaka\", \"@protectingtitan's side.\", \"nashville, tn \", \"Miami\", \"Hogsmeade\", \"Lancaster California\", \"#ODU\", \"Akure city in ondo state \", \"Sunnyvale, CA\", \"Ciudad Aut\\u00cc_noma de Buenos Aires, Argentina\", \"Saint Paul\", \"Magnolia, Fiore \", \"sweden\", \"Between Dire and Radiant\", \"?? Made in the Philippines ??\", \"Bi\\u00cc\\u00b1an,Laguna\", \"Honolulu, Hawaii\", \"Saint Lucia\", \"WORLDWI$E \", \"los angeles\", \"Vancouver Canada\", \"rzl ?\", \"23 countries and counting!\", \"Bellville, Ohio\", \"Near Richmond, VA\", \"Manila City\", \"Yellowknife\", \"Dayton, OH\", \" Baku & Erzurum \", \"  Glasgow \", \"Philadelphia, Pennsylvania\", \"709?\", \"Netherlands\", \"Far Away From Home\", \"Essex, England\", \"#partsunknown\", \"Haarlem\", \"PG County, MD\", \"Medan,Indonesia\", \"Oblivion?\", \"Hagerstown, MD\", \"AKRON OHIO USA\", \"Harlem, NY or Chocolate City\", \" snapchat // fvck_casper \", \"Jump City\", \"Boston \\u0089\\u00db\\u00a2 Cape Cod ?\", \"Johns Creek, GA\", \"lrhcthband;four - bournemouth\", \"London \", \"Cosmic Oneness\", \"Lima, Peru\", \"a van down by the river\", \"lowestoft\", \"Made Here In Detroit \", \"Laredo, TX\", \"Charlotte, North Carolina\", \"Rockland County, NY\", \"Chamblee, Georgia\", \"Florida but I wanna be n Texas\", \"Johannesburg, South Africa \", \"Pittsburgh \", \"Madison, Wisconsin, USA\", \"Tennessee, USA\", \"Brecksville, OH\", \"BILASPUR,CHHATTISGARH,495001\", \"Vail Valley\", \"In the moment\", \"peshawar pakistan \", \"Las Cruces, NM\", \"The Multiverse\", \"Namjoon's pants\", \"Derry, 17 \", \"Houston, Texas ! \", \"right next to you\", \"Minneapolis,MN,US\", \"Elk Grove, CA, USA\", \"HI\\u0089\\u00db\\u00a2UT\\u0089\\u00db\\u00a2AS\", \"Numenor\", \"[@blackparavde is my frankie]\", \"?? Cloud Mafia ??\", \"?semekeepschanging@soyeh?\", \"Wolmers Trust School for Boys \", \"3???2???????\", \"Right here\", \"3rd Eye Chakra\", \"Somewhere between Chicago & Milwaukee\", \"denver colorado\", \"Trinity, Bailiwick of Jersey\", \"Why should you know?\", \"Cydia\", \"Camaqu\\u00cc\\u00a3/Pelotas\", \"Western New York\", \"atlanta\", \"Hickville, USA\", \"Hope Road, Jamaica \", \"3?3?7?SLOPelousas??2?2?5?\", \"Richmond, VA\", \"with Doflamingo\", \"Boston, Massachusetts\", \"18 | 509 \", \"{GoT | Modern AU | Lizz}\", \"Chester Football Club\", \"y(our) boyfriends legs \", \"Chatham, IL\", \"contactSimpleNews@gmail.com\", \"Stockton on tees Teesside UK\", \"In #Fairie, where else? ;-)\", \"Concord, NH \", \"In @4SkinChan 's arms\", \"B&B near Alton Towers\", \"lia\\u0089\\u00db\\u00a2dani\\u0089\\u00db\\u00a2laura\", \"Monterrey, M\\u00cc\\u00a9xico\", \"hertfordshire.\", \"Alameda and Pleasanton, CA\", \"Saint Marys, GA\", \"IDN\", \"dmv ?? fashion school @ KSU. \", \"Shady Pines \", \"\\u00cc\\u00cfT: -26.695807,27.837865\", \"Darlington\", \"Storybrooke / The Moors\", \"ATL ??\", \"Detroit, Michigan\", \"Visit our  dedicated website @\", \"a box\", \"seattle wa\", \"??? ?? ??????? \", \"sri lanka\", \"denmark\", \"Wilmington, DE\", \"Swindon,England \", \"wherever the $$$ at\", \"sindria\", \"Birmingham, United Kingdom\", \"Torry Alvarez love forever ? ?\", \"S\\u00cc\\u00a3o Paulo SP,  Brasil\", \"West Vancouver, B.C.\", \"At Da Laundry Mat Wit Nivea \", \"East Aurora, NY\", \"Planet of da Bathing Apes\", \"Nashua NH\", \"Mostly Wellington, NZ \", \"Chappaqua NY and Redlands CA\", \"Sugar Land, TX\", \"Making Worldwide Change Near U\", \"302???? 815\", \"Leeds, UK\", \"Dammam- KSA\", \"At Work\", \"illinois. united state \", \"us-east-1a\", \"IndiLand \", \" Quantico Marine Base, VA.\", \" 616 \\u0089\\u00db\\u00a2 Kentwood , MI \", \"Not Steven Yeun / AMC.\", \"Dalkeith, Scotland\", \"Notts\", \"Heathrow\", \"San Fransokyo\", \"emily | helen | shelley \", \"Proud @BuckMasonUSA supporter!\", \"Belfast\", \"MIchigan\", \"Aracaju - Sergipe\", \"Roanoke VA\", \"the road to success\", \"mind ya business\", \"Third rock from the Sun\", \"new york, ny\", \"too far\", \"proudly South African\", \"Lives in London\", \"Santiago de Chile\", \"The Sanctuary Network, Rome\", \"Derbyshire, United Kingdom\", \"Stowmarket\", \"Yellowknife, NT\", \"Irving , Texas\", \"Silang, Cavite / Para\\u00cc\\u00b1aque\", \"Buffalo/DC\", \"Not Los Angeles, Not New York.\", \"Loading...\", \"Kutztown, PA\", \"USA, North Dakota\", \"iPhone: 33.104393,-96.628624\", \"Spain but Opa-Locka, FL\", \"Kettering, OH\", \"sydney, australia\", \"Essex\", \"Quito, Ecuador.\", \"Kicking Horse Pass\", \"Innsmouth, Mass.\", \"21, Porto\", \"Warszawa\", \"youtube.com/channel/UCHWTLC9B4ZjUGh7yDlb55Iw\", \"Abuja\", \"Hermitage, PA\", \"Wellington, New Zealand\", \"Sale, England\", \"Royton\", \"\\u00cc\\u00cfT: 6.488400524109015,3.352798039832285\", \"South Carolina\", \"MI - CA\", \"Queensland\", \"EIU  Chucktown/LaSalle IL\", \"05/04/2014 18:23 ?\", \"Docker container\", \"Silvermoon or Ironforge\", \"Gidi\", \"Mooresville, NC\", \"CT ? NYC\", \"california | oregon | peru |\", \"July 11th, 2015. ?\", \"mainly California\", \"Mankato, MN\", \"Galveston, Texas\", \"FSC '19\", \"Jersey\", \"Catalonia, Spain\", \"| CA \\u0089\\u00db\\u00a2 GA  |\", \" New England\", \"Shanghai\", \"Caserta-Roma, Italy \", \"Augusta, Maine, 04330\", \"Albuquerque New Mexico\", \"11/4/14\", \"Oregon and Washington\", \"Ottawa, Ontario\", \"Suva, Fiji Islands.\", \"U.S. Northern Virginia\", \"Paris.\", \"Brisbane, Queensland\", \"Maryland, USA\", \"Near Yosemite\", \"Maricopa, AZ\", \"calgary,ab\", \"'soooota\", \"North Port, FL\", \"Kabul, Tuebingen, Innsbruck\", \"Merica!\", \"Speaking the Truth in Love\", \"Surry Hills, Sydney\", \"lost in history\", \"Williamsburg, VA\", \"Chicago, Il\", \"Charter Member of the VRWC\", \"Sheff/Bangor/Salamanca/Madrid\", \"10 Steps Ahead.  Cloud 9\", \"Earthling (For now!)\", \"Gander NF\", \"New York NYC\", \"home \", \"Old Blighty\", \"NYC / International\", \"Coventry, Rhode Island\", \"International\", \"Medell\\u00cc_n, Antioquia\", \"Playa del Carmen, Mexico\", \"Long Eaton \\u00e5\\u00e1 Derbyshire \\u00e5\\u00e1 UK\", \"california mermaid ? \", \"Olathe, KS\", \"England & Wales Border, UK\", \"Inside your webcam. Stop that.\", \"southwest, Tx\", \"Wales, United Kingdom\", \"Savannah, GA\", \"highlands&slands scotland\", \"london essex england uk\", \"VONT ISLAND, LAGOS\", \"New Jersey, USA\", \"International \", \"PDX\", \"The Forever Girl\", \"Edmonton, Alberta - Treaty 6\", \"Paonia, Colorado \", \"lost in my thoughts\", \"nor*cal\", \"\\u00cc\\u00f8\\u00e5\\u00c0\\u00e5_T: 40.736324,-73.990062\", \"???????????\", \"46.950109,7.439469\", \"EPTX\", \"MAURITIUS\", \"Lake Monticello, VA\", \"England.\", \"Anywhere Safe\", \"Fairfield, California\", \"Livingston, IL  U.S.A.\", \"she/her/your majesty/empress\", \"Austin | San Diego\", \"Orm\", \"Your Six\", \"Hawaii\", \"Jackson\", \"Ojodu,Lagos\", \"Hearts & Minds\", \"316\", \"Maracay y Nirgua, Venezuela\", \"Waukesha, WI\", \"Voorhees, NJ\", \"Oxford, MS\", \"#WhereverI'mAt\", \"teh internets\", \"Lindenhurst\", \"Harbour Heights, FL\", \"Wildomar, CA\", \"International Action\", \"Santiago Bernabeau\", \"Ikorodu\", \"South Central Wales\", \"#freegucci\", \"COMING SOON\", \"sisterhood\", \"Massachusetts \", \"Atmosphere\", \"505 W. Maple, Suite 100\", \"columbus ohio\", \"Riverside, CA\", \"|| c h i c a g o ||\", \"North Jersey\", \"iamdigitalent.com\", \"BKI-KUA\", \"Fife, WA\", \"Loughborough, England\", \"Lubbock, TX\", \"? Jet Life ?\", \"Konoha Village\", \"Sacae Plains\", \"Our Empire State\", \"The Sun's Corona\", \"hey Georgia\", \"Detroit, MI, United States\", \"Dorset, UK\", \"toronto, ontario\", \"TN\", \"Concord, CA\", \"Tn\", \"someplace living my life\", \"Eau Claire, Wisconsin\", \"Desert Storm?? |BCHS|\", \"global\", \"Northern California U.S.A.\", \"Hyderabad Telangana INDIA\", \"Hogwarts\", \"Daruka (near Tamworth) NSW\", \"ATL??AL??\", \"Dutch/English/German\", \"Alberta \", \"Dunwoody, GA\", \"\\u00e5_: ?? \\u00cc\\u00d1 ? : ?\", \"Turkmenistan\", \"Chicago IL\", \"Richmond Heights, OH\", \"In your head\", \" Eugene, Oregon\", \"labuan, malaysia\", \"PLFD cuh..\", \"South Pasadena, CA\", \"Durham N.C \", \"Ankara - Malatya - ad Orontem\", \"Maharashtra\", \"(a) property of the universe\", \"Chester \", \"Portsmouth, UK\", \"I'm standing behind you\", \"Winter Park, Colorado\", \"Montana, USA\", \"Somewhere\", \"District of Gentrification/ DC\", \"kediri,,jawa timur\", \"Oklahoma\", \"Ashland, Oregon\", \"Marietta, GA\", \"Carry On Jutta!!!\", \"London, Riyadh\", \"Rotterdam, The Netherlands\", \"SouthEast Asia\", \"Eagle River Alaska\", \"All Around the World\", \"\\u00cc\\u00cfT: 36.142163,-95.979189\", \"Pakistan, Islamabad\", \"N?? Y???.\", \"Nomad, USA\", \"Kyiv, Ukraine\", \"Eastbourne England\", \"UK Great Britain \", \"The Grey Area\", \"Miami,FL\", \"bangalore\", \"Dakar\", \"Mid north coast of NSW\", \"73101\", \"Clearwater, FL\", \"Livonia, MI\", \"Calgary,AB, Canada\", \"//RP\\\\ ot @Mort3mer\\\\\\\\\", \"Inglewood, CA\", \"facebook.com/tradcatknights\", \"on a catwalk somewhere\", \"Toronto, ON\", \"Toronto, Bob-Lo, Miami Beach\", \"New Jersey/ D.R.\", \"Bangkok Thailand\", \"Bay Area\", \"instagram: bribriony\", \"Riverview, FL \", \"Philadelphia, PA USA\", \"Macon, Georgia\", \"Center for Domestic Preparedness\", \"Earth: Senseless nonsense\", \"A small federal enclave\", \"Zac Newsome loves me\", \"The Web\", \"Wahpeton, ND\", \"philly\", \"KurveZ@GearHeadCentral.net\", \"bahstun/porta reeko\", \"lee london\", \"Nashville\", \"East TN.\", \"Bristol, England\", \"Est. September 2012 - Bristol\", \"(RP)\", \"Fort Valley,GA/Fayetteville,AR\", \"Emirates\", \"Grand Rapids MI\", \"Melbourne, Australia.\", \"Campinas Sp\", \"Roads/Trails Everywhere\", \"Anywhere I like\", \"Sao Paulo\", \"Twitter Lockout in progress\", \"San Jose\", \" ? ??????? ? ( ?? \\u00e5\\u00a1 ? ? ? \\u00e5\\u00a1)\", \"Groningen, Netherlands, Europe\", \"Hamilton, Ontario Canada\", \"Eagle Mountain, Texas \", \"playing soccer & eating pizza\", \"Benton City, Washington\", \"Nunya\", \"Wasington, DC\", \"Texas-USA\\u0089\\u00e3\\u00a2 ?\", \"#expelcl*y\", \"UGA '15 Alumnus - Economics \", \"Play For Ryan ??\", \"Bow, NH\", \"Houston,  TX\", \"somewhere in Indiana \", \"Garrett\", \"New York City, NY\", \"Where the money at\", \"Los Angeles, London, Kent\", \"Denton, Texas\", \"Edappally,Kochi\", \"Essex/Brighton\", \"Mpela'zwe \", \"at my home\", \"Reading MA\", \"Lethbridge, Alberta, Canada\", \"St. Louis\", \"North Highlands, CA\", \"Dorset, United Kingdom\", \"Rutherfordton, NC\", \"Xi'an, China\", \"Deadend, UK\", \"Linton Hall, VA\", \"Stateless Global Citizen\", \"Madison, WI & St. Louis MO\", \"Primum non nocere\", \"Queensland, Australia\", \"EveryWhere\", \"Redondo Beach, CA\", \"nap queen\", \"livin life in the 610\", \"North East / Middlesbrough \", \"Conroe, TX\", \"Griffin :3\", \"Itirapina, S\\u00cc\\u00a3o Paulo\", \"North-East Region, Singapore\", \"Poconos\", \"Christchurch New Zealand\", \"THE WORLD T.G.G / M.M.M \", \"Milky Way galaxy \", \" BC, US, Asia or Europe.\", \"CORNFIELDS\", \"Corpus Christi, Texas\", \"801 SL,UT\", \"Oxford, OH\", \"Rafael castillo\", \"\\u0089\\u00db\\u00a2 Views From The Six \\u0089\\u00db\\u00a2\", \"hollywoodland \", \"pettyville, usa\", \"Chapel Hill, NC\", \"Madrid, Comunidad de Madrid\", \"Guatemala\", \"Worldwide - Global\", \"My contac 27B80F7E 08170156520\", \"Phoenix Az\", \"Dalston, Hackney\", \"The Circle of Life\", \"TonyJ@Centralizedhockey.com\", \"Catskills\", \"IM LOST \", \"SD |Norway| KSA\", \"Intramuros, Manila\", \"Birmingham UK\", \"mexico\", \"Finland\", \"Tring \", \"Upper manhattan, New York\", \"Jersey - C.I\", \"Terre Haute, IN\", \"Na waffi\", \"El Paso, Texas\", \"Coconut Creek, Florida\", \"moss chamber b\", \"Mariveles, Bataan\", \"ur\\u00cc\\u00a3nus\", \"Spain\", \"Marbella. Spain\", \"Alicante, Spain\", \"Njoro, Kenya\", \"Mumbai, India\", \"Huntsville, Alabama\", \"Here, unless there.  \", \"Somewhere Out There\", \"Chicago Area\", \"Jersey City, New Jersey\", \"Dimapur\", \"Tring, UK\", \"Atlanta Georgia\", \"Peshawar\", \"Unknown \", \"State College, Pa\", \"London/Surrey \", \"CHICAGO (312)\", \"Bishops Lydeard, England\", \"Sioux Falls, SD\", \"Miami,Fla\", \"Beit El - Israel\", \"Nottingham, United Kingdom\", \" Alberta\", \"Click the link below, okay \", \"Boston/Montreal \", \"Las Vegas, NV \", \"? Philly Baby ?\", \"New York Brooklyn\", \"Wherever I'm needed\", \"WASHINGTON,DC\", \"Under Santa Barbara Skies\", \"az\", \"Bahrain\", \"Woosley\", \"Gujranwala, Pakistan\", \"Anna Maria, FL\", \"Galatians 2:20 \", \"lagos. Unilag\", \"Caracas, Venezuela.\", \"|-/\", \"Fiore, Lamia Scale\", \"Horsemind, MI\", \"University of Chicago\", \"Plano, IL\", \"travelling to tae's pants\", \"Perth, Australia\", \"Karolinska v\\u00cc_gen 18, Solna\", \"planet earth\", \"Midwestern USA\", \"From NY. In Scranton, PA\", \"I Heard #2MBikers\", \"Lakewood, Tennessee\", \"2 high 2 come down \", \"tokyo\", \"Oshawa/Toronto\", \"Kentucky, USA\", \"772 Temperance Permenence\", \"Kelowna, BC\", \"North Vancouver, BC\", \"Washington state\", \"Nebraska, Colorado & The GLOBE\", \"Zimbabwe\", \"NY, CT & Greece\", \"Jaipur, Rajasthan, India\"], \"yaxis\": \"y\"}],\n",
              "                        {\"legend\": {\"itemsizing\": \"constant\", \"tracegroupgap\": 0}, \"margin\": {\"t\": 60}, \"template\": {\"data\": {\"bar\": [{\"error_x\": {\"color\": \"#2a3f5f\"}, \"error_y\": {\"color\": \"#2a3f5f\"}, \"marker\": {\"line\": {\"color\": \"#E5ECF6\", \"width\": 0.5}}, \"type\": \"bar\"}], \"barpolar\": [{\"marker\": {\"line\": {\"color\": \"#E5ECF6\", \"width\": 0.5}}, \"type\": \"barpolar\"}], \"carpet\": [{\"aaxis\": {\"endlinecolor\": \"#2a3f5f\", \"gridcolor\": \"white\", \"linecolor\": \"white\", \"minorgridcolor\": \"white\", \"startlinecolor\": \"#2a3f5f\"}, \"baxis\": {\"endlinecolor\": \"#2a3f5f\", \"gridcolor\": \"white\", \"linecolor\": \"white\", \"minorgridcolor\": \"white\", \"startlinecolor\": \"#2a3f5f\"}, \"type\": \"carpet\"}], \"choropleth\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"type\": \"choropleth\"}], \"contour\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"contour\"}], \"contourcarpet\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"type\": \"contourcarpet\"}], \"heatmap\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"heatmap\"}], \"heatmapgl\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"heatmapgl\"}], \"histogram\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"histogram\"}], \"histogram2d\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"histogram2d\"}], \"histogram2dcontour\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"histogram2dcontour\"}], \"mesh3d\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"type\": \"mesh3d\"}], \"parcoords\": [{\"line\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"parcoords\"}], \"pie\": [{\"automargin\": true, \"type\": \"pie\"}], \"scatter\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatter\"}], \"scatter3d\": [{\"line\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatter3d\"}], \"scattercarpet\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scattercarpet\"}], \"scattergeo\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scattergeo\"}], \"scattergl\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scattergl\"}], \"scattermapbox\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scattermapbox\"}], \"scatterpolar\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatterpolar\"}], \"scatterpolargl\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatterpolargl\"}], \"scatterternary\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatterternary\"}], \"surface\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"surface\"}], \"table\": [{\"cells\": {\"fill\": {\"color\": \"#EBF0F8\"}, \"line\": {\"color\": \"white\"}}, \"header\": {\"fill\": {\"color\": \"#C8D4E3\"}, \"line\": {\"color\": \"white\"}}, \"type\": \"table\"}]}, \"layout\": {\"annotationdefaults\": {\"arrowcolor\": \"#2a3f5f\", \"arrowhead\": 0, \"arrowwidth\": 1}, \"coloraxis\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"colorscale\": {\"diverging\": [[0, \"#8e0152\"], [0.1, \"#c51b7d\"], [0.2, \"#de77ae\"], [0.3, \"#f1b6da\"], [0.4, \"#fde0ef\"], [0.5, \"#f7f7f7\"], [0.6, \"#e6f5d0\"], [0.7, \"#b8e186\"], [0.8, \"#7fbc41\"], [0.9, \"#4d9221\"], [1, \"#276419\"]], \"sequential\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"sequentialminus\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]]}, \"colorway\": [\"#636efa\", \"#EF553B\", \"#00cc96\", \"#ab63fa\", \"#FFA15A\", \"#19d3f3\", \"#FF6692\", \"#B6E880\", \"#FF97FF\", \"#FECB52\"], \"font\": {\"color\": \"#2a3f5f\"}, \"geo\": {\"bgcolor\": \"white\", \"lakecolor\": \"white\", \"landcolor\": \"#E5ECF6\", \"showlakes\": true, \"showland\": true, \"subunitcolor\": \"white\"}, \"hoverlabel\": {\"align\": \"left\"}, \"hovermode\": \"closest\", \"mapbox\": {\"style\": \"light\"}, \"paper_bgcolor\": \"white\", \"plot_bgcolor\": \"#E5ECF6\", \"polar\": {\"angularaxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}, \"bgcolor\": \"#E5ECF6\", \"radialaxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}}, \"scene\": {\"xaxis\": {\"backgroundcolor\": \"#E5ECF6\", \"gridcolor\": \"white\", \"gridwidth\": 2, \"linecolor\": \"white\", \"showbackground\": true, \"ticks\": \"\", \"zerolinecolor\": \"white\"}, \"yaxis\": {\"backgroundcolor\": \"#E5ECF6\", \"gridcolor\": \"white\", \"gridwidth\": 2, \"linecolor\": \"white\", \"showbackground\": true, \"ticks\": \"\", \"zerolinecolor\": \"white\"}, \"zaxis\": {\"backgroundcolor\": \"#E5ECF6\", \"gridcolor\": \"white\", \"gridwidth\": 2, \"linecolor\": \"white\", \"showbackground\": true, \"ticks\": \"\", \"zerolinecolor\": \"white\"}}, \"shapedefaults\": {\"line\": {\"color\": \"#2a3f5f\"}}, \"ternary\": {\"aaxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}, \"baxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}, \"bgcolor\": \"#E5ECF6\", \"caxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}}, \"title\": {\"x\": 0.05}, \"xaxis\": {\"automargin\": true, \"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\", \"title\": {\"standoff\": 15}, \"zerolinecolor\": \"white\", \"zerolinewidth\": 2}, \"yaxis\": {\"automargin\": true, \"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\", \"title\": {\"standoff\": 15}, \"zerolinecolor\": \"white\", \"zerolinewidth\": 2}}}, \"xaxis\": {\"anchor\": \"y\", \"domain\": [0.0, 1.0], \"title\": {\"text\": \"x\"}}, \"yaxis\": {\"anchor\": \"x\", \"domain\": [0.0, 1.0], \"title\": {\"text\": \"y\"}}},\n",
              "                        {\"responsive\": true}\n",
              "                    ).then(function(){\n",
              "                            \n",
              "var gd = document.getElementById('b54cb36f-1851-495b-b9ba-016dec55d405');\n",
              "var x = new MutationObserver(function (mutations, observer) {{\n",
              "        var display = window.getComputedStyle(gd).display;\n",
              "        if (!display || display === 'none') {{\n",
              "            console.log([gd, 'removed!']);\n",
              "            Plotly.purge(gd);\n",
              "            observer.disconnect();\n",
              "        }}\n",
              "}});\n",
              "\n",
              "// Listen for the removal of the full notebook cells\n",
              "var notebookContainer = gd.closest('#notebook-container');\n",
              "if (notebookContainer) {{\n",
              "    x.observe(notebookContainer, {childList: true});\n",
              "}}\n",
              "\n",
              "// Listen for the clearing of the current output cell\n",
              "var outputEl = gd.closest('.output');\n",
              "if (outputEl) {{\n",
              "    x.observe(outputEl, {childList: true});\n",
              "}}\n",
              "\n",
              "                        })\n",
              "                };\n",
              "                \n",
              "            </script>\n",
              "        </div>\n",
              "</body>\n",
              "</html>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "E7NeNwoMGem2"
      },
      "source": [
        "### Exploring the 'location' column\n",
        "\n",
        "Even though the column 'location' has lots of missing values, let's check the top 20 locations present in the dataset. Since some of the locations are duplicated, lets clean the duplicate locations and plot them."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "21VOr-XQGem3",
        "outputId": "2cddd261-b83f-4a45-c7a4-bd6e2a66bbaf",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "source": [
        "# Replacing the ambigious locations name with standard names and create a dictionary object\n",
        "\n",
        "loc_dict={'United States':'USA','New York':'USA',\"London\":'UK',\"Los Angeles, CA\":'USA',\"Washington, D.C.\":'USA',\n",
        "          \"California\":'USA',\"Chicago, IL\":'USA',\"Chicago\":'USA',\"New York, NY\":'USA',\"California, USA\":'USA',\n",
        "          \"FLorida\":'USA',\"Nigeria\":'Africa',\"Kenya\":'Africa',\"Everywhere\":'Worldwide',\"San Francisco\":'USA',\n",
        "          \"Florida\":'USA',\"United Kingdom\":'UK',\"Los Angeles\":'USA',\"Toronto\":'Canada',\"San Francisco, CA\":'USA',\n",
        "          \"NYC\":'USA',\"Seattle\":'USA',\"Earth\":'Worldwide',\"Ireland\":'UK',\"London, England\":'UK',\"New York City\":'USA',\n",
        "          \"Texas\":'USA',\"London, UK\":'UK',\"Atlanta, GA\":'USA',\"Mumbai\":\"India\"}\n",
        "\n",
        "train_df['location'].replace(loc_dict,inplace=True)\n",
        "\n",
        "#Create barchart for top 10 locations using seaborn\n",
        "sns.barplot(y=train_df['location'].value_counts()[:10].index,x=train_df['location'].value_counts()[:10],\n",
        "            orient='h');"
      ],
      "execution_count": 145,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAbQAAAEGCAYAAAANNmA4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAfBUlEQVR4nO3deZQfVYH28e8jAiEkBgIIyGITDEtCWBsGgmJYfEUFQScKEQUcNO6ICi7jiMFxxu1FlEU0oIKIBgFxEHwBBaKsSTqQkAABhDAjEgfCmoAghOf9o27Dj6Y73Z10+peufj7n9OmqW7du3d89OXnOrar+XdkmIiJioHtVszsQERHRFxJoERFRCwm0iIiohQRaRETUQgItIiJq4dXN7sBgteGGG7qlpaXZ3YiIGFBmz5692PZGnR1LoDVJS0sLbW1tze5GRMSAIum/uzqWW44REVELmaE1yZ0PPMJuJ/ys2d2IiOhXs79z5CprOzO0iIiohQRaRETUQgItIiJqIYEWERG1kEDrQFKLpPkdyqZIOl7SnpJmSJoj6U5JUzrU+56kv0rKuEZE9LO85dg75wLvtT1X0hrAtu0HSoi9C/gL8Gbg2uZ0MSJicMpMondeCywCsL3M9h0NxyYAtwNnApP6v2sREYNbAq13TgHuknSJpI9IGtJwbBLwS+AS4B2S1ux4sqTJktoktT3/9JJ+6nJExOCQQHulrpbwtu2vAa3AVcD7gCsAJK0FvB34je0ngRnAWztpYKrtVtutrx46fJV0PiJisMoztFd6BFi/Q9lIYCGA7XuBMyWdBTwsaQNgPLAeME8SwFDg78Bl/dXpiIjBLjO0DmwvBRZJ2g9A0kjgQOB6Se9QSSxgNLAMeJzqduOHbLfYbgG2At4iaWi/f4CIiEEqgda5I4GvSJoDXAOcVGZmH6B6hjYHOA84AlibKvAubz/Z9lPA9cDB/d3xiIjBKrccO1HeXty3k/LDuzhlZCd1393X/YqIiK5lhhYREbWQQIuIiFpIoEVERC3kGVqTbL/5BrStwoXuIiIGm8zQIiKiFhJoERFRCwm0iIiohTxDa5J/LLqd//nauGZ3o1a2PHFes7sQEU2UGVpERNRCAi0iImohgRYREbWQQIuIiFpIoEVERC0k0HpIUouk+R3Kpkg6XtI5kiaWspGSbpX0web0NCJicEqg9SFJI4Argam2f9rs/kREDCYJtL4zDPh/wC9sn9nszkREDDYJtL7zXeB626d0VUHSZEltktoefWpZP3YtIqL+Emg9527KrwEOkfTaLhuwp9putd06ct01+ryDERGDWQKt5x4B1u9QNhJYXLanAT8EfidpeH92LCIiEmg9ZnspsEjSflC9zQgcCFzfUOcU4Grg15LWakpHIyIGqQRa7xwJfEXSHKpbjCfZvrexgu0vAA8A50nK+EZE9JN8234v2L4D2LeT8qM77Odv0CIi+llmEBERUQsJtIiIqIUEWkRE1EKeoTXJWpuOZcsT25rdjYiI2sgMLSIiaiGBFhERtZBAi4iIWsgztCZZ8NAC9j5t72Z342Vu+NQNze5CRMQKywwtIiJqIYEWERG1kECLiIhaSKBFREQtJNAiIqIWBmWgSTpUkiVtV/Y3kjRD0q2S3tRJ/bMljen/nkZERE8NykADJlEtzDmp7O8PzLO9i+3rGitKWsP2h8rSMRERsZoadIEmaRjwRuAY4HBJOwPfBg6RNEfSOpKWSjpZ0lxgL0nTJbWW8w+UdIukuZKuLmV7SLqpzPBulLRtsz5fRMRgNRj/sPoQ4Arbd0t6BFgDOBFotf1JAEnrAjNsf67sU35vBJwF7GN7oaSRpc0FwJtsPy/pAOA/gX/ueGFJk4HJAGutv9Yq/IgREYPPYAy0ScD3y/a0sj+/Q51lwMWdnLsn8CfbCwFsP1rKRwDnShoNGFizswvbngpMBRi25TCvxGeIiIgOBlWglRnVfsA4SaaanRm4vUPVZ2wv60XT/w5ca/tdklqA6Svf24iI6I3B9gxtInCe7dfbbrG9BbAQ2KKH598M7CNpK3gxIKGaof21bB/dh/2NiIgeGmyBNgm4pEPZxcCXenKy7YepnoH9urwwckE59G3gG5JuZZDNeiMiVhey8yinGYZtOcw7nbBTs7vxMvm2/YhY3Umabbu1s2ODbYYWERE1lUCLiIhayPOeJtnutdvlFl9ERB/KDC0iImohgRYREbWQQIuIiFpIoEVERC3kpZAmWXLXXfxxnzc3tQ9v/tMfm3r9iIi+lBlaRETUQgItIiJqIYEWERG1kECLiIhaGPSBJmlpL+tPkHRZ2X6npC+ump5FRERv5C3HlWD7UuDSZvcjIiIyQ3tRmXlNl3SRpAWSzpekcuzAUnYL8O6Gc46WdHrZPljSDEm3SvqDpI2b9FEiIgalBNrL7QIcB4wBRgF7SxoCnAUcDOwGbNLFudcDe9reBZgGfL5jBUmTJbVJanviuedWRf8jIgat3HJ8uZm2HwCQNAdoAZYCC23fU8p/TrVqdUebAxdI2hRYC1jYsYLtqcBUgG2HD8/KqhERfSgztJd7tmF7Gb0L/NOA022PAz4CDOnLjkVExPIl0Lq3AGiRtHXZn9RFvRHAX8v2Uau8VxER8TIJtG7YfobqFuPl5aWQh7qoOgW4UNJsYHE/dS8iIgrZeZTTDNsOH+6pu+za1D7ky4kjYqCRNNt2a2fHMkOLiIhaSKBFREQtJNAiIqIW8ndoTTJ8223zDCsiog9lhhYREbWQQIuIiFpIoEVERC0k0CIiohbyUkiTPPTAE5z+ud82tQ+fPPngpl4/IqIvZYYWERG1kECLiIhaSKBFREQtJNAiIqIWVrtAk3SKpOMa9q+UdHbD/smSPtvDts6RNLGT8gmSLuthG62STu3i2P2SNuxJOxERsWqtdoEG3ACMB5D0KmBDYGzD8fHAjd01ImmNvuiM7Tbbx/ZFWxERseqsjoF2I7BX2R4LzAeWSFpf0trA9sAISbdKmifpJ6W8fcb0rbIQ53saG5V0oKQF5di7G8rnSVpPlUckHVnKfybpLY2zOUkbSLpK0u1l1qiGdt4vaaakOZJ+1FeBGhERPbPaBZrtB4HnJW1JNRu7CZhBFXKtwD3A2cBhtsdR/S3dxxqaeMT2rrantRdIGgKcBRwM7AZs0lD/BmBvqvC8D3hTKd+LV84Evwpcb3sscAmwZWl/e+AwYG/bOwPLgCM6fjZJkyW1SWpb+vQTvRqXiIhYvtUu0IobqcKsPdBuath/AFho++5S91xgn4ZzL+ikve3KOfe4WqL75w3Hrivn7wOcCYyTtBnwmO2nOrSzT/u5ti8HHivl+1MF5SxJc8r+qI6dsD3Vdqvt1mFDR3Q/ChER0WOra6C1P0cbR3XL8WaqGdN4YHo353YMoe78iWpW9qbS9sPARKqg6ykB59reufxsa3tKL/sRERErYXUNtBuBg4BHbS+z/SiwHlWoXQy0SHpDqfsBoLuFxRaUc7Yu+5PaD9j+C9WLJ6Nt3wdcDxxPFXQd/Ql4H4CktwHrl/KrgYmSXluOjZT0+l583oiIWEmra6DNowqZmzuUPWH7AeCDwIWS5gEvAD9cXmO2nwEmA5eXl0Ie6lBlBtB+C/M6YDOqYOvoJGAfSbdTvVjyP6X9O4B/A66SdBvwe2DTnn3UiIjoC6oeKUV/23KT0f78Ed9tah/y5cQRMdBImm27tbNjq+sMLSIiolcSaBERUQsJtIiIqIUs8Nkkr918RJ5hRUT0oczQIiKiFhJoERFRCwm0iIiohTxDa5JFC+/lP97/iqXaVrkv//yifr9mRER/yAwtIiJqIYEWERG1kECLiIhaSKBFREQtJNAiIqIWahVokjaRNE3SvZJmS/qdpG1W4fWWrqq2IyKid2rz2r4kAZdQrRx9eCnbCdiYl9Y6i4iImqrTDG1f4DnbLy72aXsucKukqyXdImmepEMAJLVIulPSWZJul3SVpHXKsQ9LmiVprqSLJQ0t5VtJuqm08/X260ga1tk1IiKi/9Qp0HYAZndS/gzwLtu7UoXeyWU2BzAaOMP2WOBx4J9L+a9t7257J+BO4JhS/n3gTNvjgEU9vMaLJE2W1Cap7alnnl2pDxsRES9Xp0DrioD/lHQb8AdgM6rbkAALbc8p27OBlrK9g6TrJM0DjgDGlvK9gV+W7fN6eI0X2Z5qu9V267pD1u6TDxcREZXaPEMDbgc6+y6pI4CNgN1sPyfpfmBIOdY4TVoGrFO2zwEOtT1X0tHAhIZ67uU1IiKiH9RphnYNsLakye0FknYEXg88VIJm37LfneHAIklrUoVVuxuAw8t2Y/mIFbhGRET0odoEmm0D7wIOKK/t3w58A/gd0FpuHx4JLOhBc18BZlAFWGP9TwOfKG1t1lB+/gpcIyIi+pCqHIj+ttkG6/vjb9u/36+bb9uPiIFM0mzbrZ0dq80MLSIiBrcEWkRE1EICLSIiaqFOr+0PKJtutXWeZ0VE9KHM0CIiohYSaBERUQsJtIiIqIU8Q2uSZxYt4c7/uGaFzt3+y/v1cW8iIga+zNAiIqIWEmgREVELCbSIiKiFBFpERNRCAi0iImqh20CTdIqk4xr2r5R0dsP+yZI+25uLSpoi6fgujt3Ym7Y6nDtB0vgVPb+btu+XNK/83CHp65KGNBzfRtLvJN0j6RZJv5L0ilWrIyJi1ejJDO0GYDyApFcBGwJjG46PB1Y4hDqyvTKBNIHS11VkX9vjgD2AUcCPAEqwXQ6caXu07V2BH1CtYh0REf2gJ4F2I7BX2R4LzAeWSFpf0trA9sAtkk6UNEvSfElTJQlA0rFlRnObpGkN7Y6RNF3SfZKObS+UtLT8nlCOXyRpgaTzG9p8eymbLelUSZdJagE+CnxG0hxJb5LUIumacu2rJW1Zzj+nnHdjuf7E3gya7aXlWodKGgm8D7jJ9m8b6ky3Pb837UZExIrrNtBsPwg8X8JgPHAT1WrOewGtwDzb/wBOt7277R2AdYCDShNfBHaxvSNVCLTbDngr1Wznq5LW7OTyuwDHAWOoZkR7l9nQj4C32d6NMguyfT/wQ+AU2zvbvg44DTi3XPt84NSGtjcF3lj6+c3uxqGTcXkSWAiMBnYAZnd3jqTJktoktT361OO9vWRERCxHT18KuZEqzNoD7aaG/RtKnX0lzZA0D9iPl25L3gacL+n9wPMNbV5u+1nbi4GHgM6eN820/YDtF4A5QAtVEN5ne2Gp88vl9Hsv4Bdl+zyqAGv3G9sv2L6ji2v3hHpT2fZU2622W0euu94KXjIiIjrT00Brf442juqW481UYTEeuLHMmn4ATCzPmM4C2l+YeAdwBrArMEtS+9dtPdvQ/jI6/xquntRZUY1t9yqYACQNpwrYu4Hbgd36plsREbEiejNDOwh41PYy248C61GF2o28FF6LJQ0DJsKLL5FsYfta4AvACGDYSvb5LmBUeWYGcFjDsSXA8A79PrxsHwFc113jkhb0oM4wqgD/je3HqGaB4yW9o6HOPpJ26K6tiIjoGz2d8cyjervxFx3KhpVbhkg6i2r29jdgVqmzBvBzSSOoZkGn2n68vNuxQmz/XdLHgSskPdVwLYDfAhdJOgT4VPn5qaQTgIeBDy6vbUkbsvzZ2rXlxZRXAZcA/97Qp4OA70n6HvAc1a3WT6/IZ4yIiN6T7Wb3odckDbO9tITLGcA9tk/pg3YPAkbZPrXbyitph8229YUfP3OFzs237UfEYCVptu3Wzo4N1OVjPizpKGAt4FbK34OtLNuX9UU7ERHR/wZkoJXZ2ErPyCIioj4GZKDVwZBNh+fWYUREH8qXE0dERC0k0CIiohYSaBERUQsJtIiIqIW8FNIkDz74IFOmTOnVOb2tHxExmGSGFhERtZBAi4iIWkigRURELSTQIiKiFgZkoEk6VJIlbbcS549ZgfOOlnR62f6opCNX5PoREdH3BmSgAZOA68vvFXEo0GmgNSxAuly2f2j7Zyt4/YiI6GMDLtDK4ppvBI6hLN4paYKkyxrqnC7p6LL9TUl3SLpN0v+VNB54J/AdSXMkbS1puqTvSWoDPi3pYEkzJN0q6Q+SNu6kH1MkHV+2PyxplqS5ki6WNHSVD0RERLzMQPw7tEOAK2zfLekRSbt1VVHSBsC7gO1sW9J6ZYHRS4HLbF9U6gGs1b7GjqT1gT3LOR8CPg98bjl9+rXts8q5X6cK29M66c9kYDLAiBEjev3BIyKiawNuhkZ1m3Fa2Z7G8m87PgE8A/xY0ruBp5dT94KG7c2BKyXNA04AxnbTpx0kXVfqH9FVfdtTbbfabh06NJO4iIi+NKACTdJIYD/gbEn3U4XNe4FlvPyzDAGw/TywB3ARcBBwxXKaf6ph+zTgdNvjgI+0t7cc5wCfLPVP6kH9iIjoYwMq0ICJwHm2X2+7xfYWwEKqzzFG0tqS1gP2hxeft42w/TvgM8BOpZ0lwPDlXGcE8NeyfVQP+jUcWCRpTaoZWkRE9LOBFmiTgEs6lF1M9XLIr4D55fet5dhw4DJJt1G9FfnZUj4NOKG89LF1J9eZAlwoaTawuAf9+gowA7gBWNDjTxMREX1Gtpvdh0Hpda97nSdPntyrc/LlxBEx2Ema3f4CX0cDbYYWERHRqQRaRETUQgItIiJqIc/QmqS1tdVtbW3N7kZExICSZ2gREVF7CbSIiKiFBFpERNTCQPxy4lp47LE7+dWFe3Rb773vmdkPvYmIGPgyQ4uIiFpIoEVERC0k0CIiohYSaBERUQsJtIiIqIXVPtAkLe1l/QmSLltV/elwra9JOqA/rhUREcuX1/ZXgu0Tm92HiIiorPYztHZl5jVd0kWSFkg6X5LKsQNL2S3AuxvOGSnpN5Juk3SzpB1L+RRJPynt3Sfp2IZz3i9ppqQ5kn4kaY3yc46k+ZLmSfpMqXuOpIll+0RJs0qdqe19i4iI/jFgAq3YBTgOGAOMAvaWNAQ4CzgY2A3YpKH+ScCttncE/hX4WcOx7YC3AnsAX5W0pqTtgcOAvW3vDCwDjgB2BjazvYPtccBPO+nb6bZ3t70DsA5wUMcKkiZLapPU9uSTz6/4KERExCsMtECbafsB2y8Ac4AWqmBaaPseV0sH/Lyh/huB8wBsXwNsIOk15djltp+1vRh4CNgY2J8qFGdJmlP2RwH3AaMknSbpQODJTvq2r6QZkuYB+wFjO1awPdV2q+3W17wmd3sjIvrSQPtf9dmG7WWsXP87a0vAuba/1LGypJ2oZnQfBd4L/EvDsSHAD4BW23+RNAUYshJ9i4iIXhpoM7TOLABaJG1d9ic1HLuO6pYhkiYAi213NrtqdzUwUdJryzkjJb1e0obAq2xfDPwbsGuH89rDa7GkYcDElflAERHRewNthvYKtp+RNBm4XNLTVCE2vByeAvxE0m3A08BR3bR1h6R/A66S9CrgOeATwN+Bn5YygC91OO9xSWcB84G/AbP65MNFRESPZcXqJtl663X9jW++4jHbK+Tb9iMiXpIVqyMiovYSaBERUQsJtIiIqIUB/1LIQLX++tvn+VhERB/KDC0iImohgRYREbWQQIuIiFrIM7QmueOxJ9npois7PTZ34lv7uTcREQNfZmgREVELCbSIiKiFBFpERNRCAi0iImohgRYREbXQbaBJ+rKk2yXdJmmOpH/qj471NUnHSRq6km0cL2lBGYdZko5sOLahpOckfXTlexsREb213ECTtBdwELCr7R2BA4C/rMwFJTXrTwWOA1Y40EpQvQXYw/bOwP5UK1y3ew9wMy9fYDQiIvpJdzO0TalWeX4WwPZi2w8CSDqxzFLmS5oqSaX8DZL+IGmupFskbS1pgqTrJF0K3FHq/UbS7DL7m9x+QUlLJX2nlP9B0h6Spku6T9I7S501Sp1ZZeb4kVI+odS9qMykzlflWOB1wLWSri11J0maV/r/rR6M1b8CH2tf8dr2k7bPbTg+CfgcsJmkzXvQXkRE9KHuAu0qYAtJd0v6gaQ3Nxw73fbutncA1qGayQGcD5xheydgPLColO8KfNr2NmX/X2zvBrQCx0raoJSvC1xjeyywBPg61czoXcDXSp1jgCds7w7sDnxY0lbl2C5Us7ExwChgb9unAg8C+9reV9LrgG8B+wE7A7tLOrSrQZD0GmC47fu6OL4FsKntmcCvgMO6qDdZUpuktueffKKry0VExApYbqDZXgrsBkwGHgYukHR0ObyvpBmS5lEFw1hJw4HNbF9Szn/G9tOl/kzbCxuaP1bSXKrbdFsAo0v5P4AryvY84I+2nyvbLaX8/wBHSpoDzAA2aDh/pu0HbL8AzGk4p9HuwHTbD9t+niqE91neWHTjMKogA5hGF7cdbU+13Wq79dWvGbESl4uIiI66fZ5lexkwHZhewusoSdOAHwCttv8iaQowpJumnmrfkDSB6nncXrafljS94fznbLtsvwC03+58oeH5m4BP2X7Zd0eVdp9tKFrWk8/YHdtPlluho7qYpU0CNpF0RNl/naTRtu9Z2WtHRETPdPdSyLaSRjcU7Qz8Ny+Fz2JJw4CJALaXAA+0376TtHYXbxaOAB4rYbYdsGcv+30l8DFJa5brbCNp3W7OWQIML9szgTeXNxPXoAqkP5a2fiZpj07O/wZwRrn9iKRhko6UtA0wzPZmtltst5S6eTkkIqIfdTd7GQacJmk94Hngz8Bk249LOguYD/wNmNVwzgeAH0n6GvAc1dt/HV0BfFTSncBdVLcde+NsqluJt5SXUR4GunwGVkwFrpD0YHmO9kXgWqrZ3uW2/6vU25HqeVtHZ1KNxyxJz1F9tpOpguuSDnUvBi7gpWd+ERGxiumlu3tRZl8/tt1ZCPepoVtv49HfOq3TY/m2/YiIzkmabbu1s2NZPqZBeSV/lYdZRET0vXz1VURE1EICLSIiaiG3HJtkzPqvoS3PyiIi+kxmaBERUQt5y7FJJC2h+pOF6NyGwOJmd2I1lvHpWsZm+Qb6+Lze9kadHcgtx+a5q6tXTwMktWV8upbx6VrGZvnqPD655RgREbWQQIuIiFpIoDXP1GZ3YDWX8Vm+jE/XMjbLV9vxyUshERFRC5mhRURELSTQIiKiFhJoTSDpQEl3SfpzWcZm0JH0E0kPSZrfUDZS0u8l3VN+r1/KJenUMl63Sdq1eT1f9SRtIelaSXdIul3Sp0t5xgeQNETSTElzy/icVMq3kjSjjMMFktYq5WuX/T+X4y3N7H9/kLSGpFslXVb2B8XYJND6WVlQ9AzgbcAYYJKkMc3tVVOcAxzYoeyLwNW2RwNXl32oxmp0+ZlMtTZdnT0PfM72GKrFbz9R/o1kfCrPAvvZ3olq0eEDJe0JfAs4xfYbgMeAY0r9Y6gWFH4DcEqpV3efBu5s2B8UY5NA6397AH+2fZ/tfwDTgEOa3Kd+Z/tPwKMdig8Bzi3b5/LSoq2HAD9z5WZgPUmb9k9P+5/tRbZvKdtLqP5j2oyMDwDlcy4tu2uWHwP7AReV8o7j0z5uFwH7l4WBa0nS5sA7qBZCpnzWQTE2CbT+txnwl4b9B0pZwMa2F5XtvwEbl+1BO2blFtAuwAwyPi8qt9TmAA8BvwfuBR63/Xyp0jgGL45POf4EsEH/9rhffQ/4PPBC2d+AQTI2CbRYLbn6e5JB/TclkoYBFwPHlcVnXzTYx8f2Mts7A5tT3fXYrsldWi1IOgh4yPbsZvelGRJo/e+vwBYN+5uXsoD/bb9VVn4/VMoH3ZhJWpMqzM63/etSnPHpwPbjwLXAXlS3Wtu/n7ZxDF4cn3J8BPBIP3e1v+wNvFPS/VSPM/YDvs8gGZsEWv+bBYwubx2tBRwOXNrkPq0uLgWOKttHAf/VUH5keZtvT+CJhltvtVOeYfwYuNP2dxsOZXwASRtJWq9srwO8heo547XAxFKt4/i0j9tE4BrX9BslbH/J9ua2W6j+b7nG9hEMlrGxnZ9+/gHeDtxNdd//y83uT5PG4JfAIuA5qnv6x1Ddu78auAf4AzCy1BXVm6H3AvOA1mb3fxWPzRupbifeBswpP2/P+Lw4PjsCt5bxmQ+cWMpHATOBPwMXAmuX8iFl/8/l+Khmf4Z+GqcJwGWDaWzy1VcREVELueUYERG1kECLiIhaSKBFREQtJNAiIqIWEmgREVELCbSImpC0tPtavWrv0MYvzpb0NUkH9OU1IvpSXtuPqAlJS20P68P2zqH6O6aLuqsbsTrIDC2iZso3hnxH0nxJ8yQd1nDsC6VsrqRvlrIPS5pVyi6WNFTSeOCdwHckzZG0taRzJE0s5+xf1tuap2ptu7VL+f2STpJ0SzmW71iMfpNAi6ifd1OtE7YTcABVKG0q6W1Uy4X8k6u1xL5d6v/a9u6l7E7gGNs3Un0t0gm2d7Z9b3vjkoZQrWd3mO1xwKuBjzVcf7HtXanWZTt+VX7QiEYJtIj6eSPwS1ffSP+/wB+B3anC7ae2nwaw3b4e3Q6SrpM0DzgCGNtN+9sCC23fXfbPBfZpON7+ZcqzgZaV/TARPZVAi4hzgE+W2dZJVN/vtzKeLb+XUc3eIvpFAi2ifq4DDiuLYG5ENXuaSbUQ5gclDQWQNLLUHw4sKkvWHNHQzpJyrKO7gBZJbyj7H6CaBUY0VQIton4uofom+rnANcDnbf/N9hVUz8XaymrP7c+3vkK1IvYNwIKGdqYBJ5SXP7ZuL7T9DPBB4MJym/IF4Ier+DNFdCuv7UdERC1khhYREbWQQIuIiFpIoEVERC0k0CIiohYSaBERUQsJtIiIqIUEWkRE1ML/B+r1mjOUdaWcAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xjvNanSSGem5"
      },
      "source": [
        "It doesn't seem that `location` feature has any value in our analysis and could be dropped from the dataframe."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1DfnRVzuGem5"
      },
      "source": [
        "# Drop the column 'location' from the training dataset\n",
        "train_df=train_df.drop(['location'],axis=1)"
      ],
      "execution_count": 146,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "k8q1NdomGem8"
      },
      "source": [
        "Lets further analyize the `text` feature and if there is any correlation between `text` and length of the tweets itself."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "R_YHkjECGem9",
        "outputId": "49cb7492-b014-4c0c-ace6-de69c9f7073e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "source": [
        "# Lets find the length of the 'text' for each row and add a new cloumn to train dataframe \n",
        "train_df['text_length'] = train_df['text'].apply(lambda x : len(x))\n",
        "train_df.head(4)"
      ],
      "execution_count": 147,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>id</th>\n",
              "      <th>keyword</th>\n",
              "      <th>text</th>\n",
              "      <th>target</th>\n",
              "      <th>text_length</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Our Deeds are the Reason of this #earthquake M...</td>\n",
              "      <td>1</td>\n",
              "      <td>69</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>4</td>\n",
              "      <td>NaN</td>\n",
              "      <td>Forest fire near La Ronge Sask. Canada</td>\n",
              "      <td>1</td>\n",
              "      <td>38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>5</td>\n",
              "      <td>NaN</td>\n",
              "      <td>All residents asked to 'shelter in place' are ...</td>\n",
              "      <td>1</td>\n",
              "      <td>133</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>6</td>\n",
              "      <td>NaN</td>\n",
              "      <td>13,000 people receive #wildfires evacuation or...</td>\n",
              "      <td>1</td>\n",
              "      <td>65</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   id keyword  ... target  text_length\n",
              "0   1     NaN  ...      1           69\n",
              "1   4     NaN  ...      1           38\n",
              "2   5     NaN  ...      1          133\n",
              "3   6     NaN  ...      1           65\n",
              "\n",
              "[4 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 147
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "to6K9iFhGenA",
        "outputId": "49ab7505-1c93-410e-9600-a05f23f46330",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "source": [
        "train_df.text_length.describe()"
      ],
      "execution_count": 148,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "count    7613.000000\n",
              "mean      101.037436\n",
              "std        33.781325\n",
              "min         7.000000\n",
              "25%        78.000000\n",
              "50%       107.000000\n",
              "75%       133.000000\n",
              "max       157.000000\n",
              "Name: text_length, dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 148
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fJOzvcOdGenC",
        "outputId": "97ca1f41-4f6e-47b5-94db-3f130ab49505",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "source": [
        "#Create distribution chart to visualize text length distribution\n",
        "ax=sns.distplot(train_df['text_length']).set_title('Distribution of the tweet lengths');\n",
        "plt.grid(True)"
      ],
      "execution_count": 149,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY4AAAEXCAYAAAC6baP3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deXwcV5Xo8d/RLmu1LVuyJa/xbpM4togTQoJCNodJMIGEGEIIEPAwkOENDDNjhiHk5cFAZuaFDzwygCGZJIbghGwYSHA2K3scL/G+W94kW7Yla7f2Pu+PunLanZbULXe7u6Xz/Xz6o+pbt26drm716apbdUtUFWOMMSZUSbEOwBhjTGKxxGGMMSYsljiMMcaExRKHMcaYsFjiMMYYExZLHMYYY8JiicP0SUR+KSLfi1Bb40WkWUSS3fNyEflyJNp27T0nIrdHqr0w1vsDEakRkeoQ698tIr+NdlyxJCIPicgPYrTuAyJyVSzWPVRY4hjC3D9Yq4g0iUi9iLwpIl8VkdOfC1X9qqr+nxDb6vOfVVUPqWq2qnZHIPb3ffmq6nWq+vDZth1mHOOBfwRmqWpRkPllIlIZxfWf8y/oaL+mcMQyQQ1lljjMDaqaA0wAfgz8C/BApFciIimRbjNOjAdqVfV4rAMx5pxRVXsM0QdwALgqoOwiwAfMcc8fAn7gpguAPwP1wEngNbwfH8vdMq1AM/DPwERAgTuAQ8CrfmUprr1y4EfAO0Aj8EdghJtXBlQGixdYCHQAnW59m/za+7KbTgL+DTgIHAceAfLcvJ44bnex1QDf7WM75bnlT7j2/s21f5V7zT4Xx0MBy2UFzG8GxgJ3A4+7NpuAbUCp33JjgSfd+vYD3+glriVuG3S4tv8EfBH4k1+dPcAf/J4fBua66RnAC+693AV82q9eOvBfbvscA34JZPb2moLE9hDuc+OeXw9sxPvsvAmcH/C+fhvYDDQAjwEZfvP/GTgKHAG+7N67KcFef3/t0ctnONb/i4n2iHkA9ojhmx8kcbjyQ8DfuenTXwB4X/K/BFLd4zJAgrXFe1/Oj7gvm0yCJ44qYI6r8yTwWzevjF4Sh5u+u6eu3/xy3kscXwL2ApOBbOApYHlAbL92cV0AtAMze9lOj+AltRy37G7gjt7iDFg22Ou4G2gDPgYku+36tpuXBKwH7gLSXPwVwLW9tH/6/XHPJ7svxSS8BHSwZ/1uXp2bl4WXRL4IpAAX4iXQWa7uT4CVwAj3uv8E/CiU1xzkc3MhXvJe4F7v7e69TPd7X99x8Y4AdgBfdfMWAtXAbGAY8Fv33k0J9vpDaK/Xz7A9Qn/YoSoTzBG8f7hAncAYYIKqdqrqa+r+G/twt6q2qGprL/OXq+pWVW0Bvgd8uqfz/CzdCtynqhWq2gx8B1gccMjsf6tqq6puAjbhJZAzuFgWA99R1SZVPQD8X+C2s4zvdVV9Vr3+nuV+6/4gMEpV71HVDlWtwEtwi0Np1NVvAuYClwOrgCMiMgP4CPCaqvrw9gAOqOr/qGqXqr6Ll7hvFhHB+zX/TVU9qapNwL+HGkMQS4BfqeoaVe1Wrx+qHbjYr87PVPWIqp7ES1JzXfmngf9R1W2qegov6Yait/YG8hk2ASxxmGCK8XbjA/0n3q/450WkQkSWhtDW4TDmH8T7FVgQUpR96/m17d92ClDoV+Z/FtQpvD2TQAUupsC2is8yvsB1Z7ikNgEY605WqBeReuBfA+Luzyt4ewWXu+lyvKTxEfcct54FAeu5FSgCRuH9ul/vN++vrnwgJgD/GLCucXjvUY/e3ouxnPkZ6e/z1F97A/kMmwCDtcPSDJCIfBDvS/H1wHnul+c/4n0JzAFeFpG1qvoS3uGDYPr7NTfOb3o83i/CGqAF78urJ65kzvzi6q/dI3hfWP5td+Edry/pZ1l/NS6mCcB2v7aqQlw+3F+zh4H9qjr1LNp/BbgBmIS3p9CTFC4Bfu63nldU9erAhd1Zda3AbFUN9joH8pp+qKo/DHM58Po2/N+vcQHzw4qln8+wCZHtcRgARCRXRK4HVuD1HWwJUud6EZniDmU0AN14naTgfSFPHsCqPycis0RkGHAP8IQ7fLMb71f434hIKl6HdLrfcseAif6nDgf4PfBNEZkkItl4X6CPqWpXOMG5WB4HfigiOSIyAfgW3rH2UBwDRopIXoj13wGaRORfRCRTRJJFZI5L6L21H7jdXwGuADJVtRKvA3ghMBJ419X5MzBNRG4TkVT3+KCIzHSHsn4N/ERERgOISLGIXDvA1/Rr4KsiskA8We59zQlh2ceBL4rITPcZCbymKKzPXT+fYRMiSxzmTyLShPer8LvAfXgdpsFMBV7EO4PlLeC/VXW1m/cj4N/coYhvh7H+5XgdnNVABvANAFVtAL4G/Abv130L4H/twB/c31oR2RCk3Qdd26/inZnUBvx9GHH5+3u3/gq8PbFHXfv9UtWdeEmswm2bsf3U78brf5jr4q7B2wa9fUk/AMxybT/j2tiN9x695p43utjfcO33/PK+Bq/f4gje9r+X95Lzv+Ad0nlbRBrx3vfpA3xN64Cv4O3t1Ll2v9DXMn7LPgf8DFjdE4+b1d7b6+9HX59hEyKxfiFjTKIQkZnAVrwzssLaezSRY3scxpi4JiI3iki6iAzH2yv6kyWN2LLEYYyJd3+Ldx3IPrw+ib+LbTjGDlUZY4wJi+1xGGOMCcuQuI6joKBAJ06cGJN1t7S0kJWVFZN1hyMR4kyEGCEx4rQYIycR4hxojOvXr69R1fdf+BnrMU/OxWP+/PkaK6tXr47ZusORCHEmQoyqiRGnxRg5iRDnQGME1qmNVWWMMeZsWeIwxhgTFkscxhhjwmKJwxhjTFgscRhjjAmLJQ5jjDFhscRhjDEmLJY4jDHGhMUShzHGmLAMiSFHjDFDz6NrDoVU77MLxkc5ksHH9jiMMcaExRKHMcaYsFjiMMYYExZLHMYYY8JiicMYY0xYopo4RGShiOwSkb0isjTI/HQReczNXyMiE135RSKy0T02iciNfsscEJEtbt66aMZvjDHm/aJ2Oq6IJAP3A1cDlcBaEVmpqtv9qt0B1KnqFBFZDNwL3AJsBUpVtUtExgCbRORPqtrllrtCVWuiFbsxxpjeRXOP4yJgr6pWqGoHsAJYFFBnEfCwm34CuFJERFVP+SWJDECjGKcxxpgwiHd3wCg0LHITsFBVv+ye3wYsUNU7/epsdXUq3fN9rk6NiCwAHgQmALep6tOuzn6gDi+Z/EpVl/Wy/iXAEoDCwsL5K1asiMrr7E9zczPZ2dkxWXc4EiHORIgREiPOoRDjyZaOkOqNyEob8DpgcG/LK664Yr2qlgaWx+2V46q6BpgtIjOBh0XkOVVtAz6sqlUiMhp4QUR2quqrQZZfBiwDKC0t1bKysnMZ/mnl5eXEat3hSIQ4EyFGSIw4h0KMoV45XnaWV44PhW0ZKJqHqqqAcX7PS1xZ0DoikgLkAbX+FVR1B9AMzHHPq9zf48DTeIfEjDHGnCPRTBxrgakiMklE0oDFwMqAOiuB2930TcDLqqpumRQAEZkAzAAOiEiWiOS48izgGryOdGOMMedI1A5VuTOi7gRWAcnAg6q6TUTuAdap6krgAWC5iOwFTuIlF4APA0tFpBPwAV9z/R6TgadFpCf2R1X1r9F6DcYYY94vqn0cqvos8GxA2V1+023AzUGWWw4sD1JeAVwQ+UiNMcaEyq4cN8YYExZLHMYYY8JiicMYY0xYLHEYY4wJiyUOY4wxYbHEYYwxJiyWOIwxxoTFEocxxpiwWOIwxhgTFkscxhhjwmKJwxhjTFgscRhjjAmLJQ5jjDFhscRhjDEmLJY4jDHGhMUShzHGmLBY4jDGGBMWSxzGGGPCYonDGGNMWKKaOERkoYjsEpG9IrI0yPx0EXnMzV8jIhNd+UUistE9NonIjaG2aYwxJrqiljhEJBm4H7gOmAV8RkRmBVS7A6hT1SnAT4B7XflWoFRV5wILgV+JSEqIbRpjjImiaO5xXATsVdUKVe0AVgCLAuosAh52008AV4qIqOopVe1y5RmAhtGmMcaYKEqJYtvFwGG/55XAgt7qqGqXiDQAI4EaEVkAPAhMAG5z80NpEwARWQIsASgsLKS8vPysX9BANDc3x2zd4UiEOBMhRkiMOIdCjBktHSHVKy+vGPA6YGhsy0DRTBxnRVXXALNFZCbwsIg8F+byy4BlAKWlpVpWVhb5IENQXl5OrNYdjkSIMxFihMSIcyjE+OiaQyHVK1swfsDrgKGxLQNF81BVFTDO73mJKwtaR0RSgDyg1r+Cqu4AmoE5IbZpjDEmiqKZONYCU0VkkoikAYuBlQF1VgK3u+mbgJdVVd0yKQAiMgGYARwIsU1jjDFRFLVDVa5P4k5gFZAMPKiq20TkHmCdqq4EHgCWi8he4CReIgD4MLBURDoBH/A1Va0BCNZmtF6DMcaY94tqH4eqPgs8G1B2l990G3BzkOWWA8tDbdMYY8y5Y1eOG2OMCYslDmOMMWGxxGGMMSYsljiMMcaExRKHMcaYsFjiMMYYExZLHMYYY8JiicMYY0xYLHEYY4wJiyUOY4wxYbHEYYwxJiyWOIwxxoTFEocxxpiwWOIwxhgTFkscxhhjwmKJwxhjTFgscRhjjAmLJQ5jjDFhscRhjDEmLFFNHCKyUER2icheEVkaZH66iDzm5q8RkYmu/GoRWS8iW9zfj/otU+7a3Ogeo6P5GowxxpwpJVoNi0gycD9wNVAJrBWRlaq63a/aHUCdqk4RkcXAvcAtQA1wg6oeEZE5wCqg2G+5W1V1XbRiN8YY07to7nFcBOxV1QpV7QBWAIsC6iwCHnbTTwBXioio6ruqesSVbwMyRSQ9irEaY4wJUTQTRzFw2O95JWfuNZxRR1W7gAZgZECdTwEbVLXdr+x/3GGq74mIRDZsY4wxfRFVjU7DIjcBC1X1y+75bcACVb3Tr85WV6fSPd/n6tS457OBlcA1qrrPlRWrapWI5ABPAr9V1UeCrH8JsASgsLBw/ooVK6LyOvvT3NxMdnZ2TNYdjkSIMxFihMSIcyjEeLKlI6R6I7LSBrwOGNzb8oorrlivqqWB5VHr4wCqgHF+z0tcWbA6lSKSAuQBtQAiUgI8DXy+J2kAqGqV+9skIo/iHRJ7X+JQ1WXAMoDS0lItKyuLzKsKU3l5ObFadzgSIc5EiBESI86hEOOjaw6FVK9swfgBrwOGxrYMFM1DVWuBqSIySUTSgMV4ew/+VgK3u+mbgJdVVUUkH/gLsFRV3+ipLCIpIlLgplOB64GtUXwNxhhjAkQtcbg+izvxzojaATyuqttE5B4R+bir9gAwUkT2At8Cek7ZvROYAtwVcNptOrBKRDYDG/H2WH4drddgjDHm/aJ5qApVfRZ4NqDsLr/pNuDmIMv9APhBL83Oj2SMxhhjwmNXjhtjjAmLJQ5jjDFhscRhjDEmLJY4jDHGhMUShzHGmLBY4jDGGBMWSxzGGGPCYonDGGNMWCxxGGOMCYslDmOMMWGxxGGMMSYsISUOEXlKRP5GRCzRGGPMEBdqIvhv4LPAHhH5sYhMj2JMxhhj4lhIiUNVX1TVW4F5wAHgRRF5U0S+6O6LYYwxZogI+dCTiIwEvgB8GXgX+CleInkhKpEZY4yJSyHdj0NEngamA8uBG1T1qJv1mIisi1Zwxhhj4k+oN3L6tbsp02kikq6q7cFuZG6MMWbwCvVQVbC78b0VyUCMMcYkhj73OESkCCgGMkXkQkDcrFxgWJRjM8YYE4f6O1R1LV6HeAlwn195E/CvUYrJGGNMHOvzUJWqPqyqVwBfUNUr/B4fV9Wn+mtcRBaKyC4R2SsiS4PMTxeRx9z8NSIy0ZVfLSLrRWSL+/tRv2Xmu/K9IvIzEZHAdo0xxkRPf4eqPqeqvwUmisi3Auer6n1BFutZNhm4H7gaqATWishKVd3uV+0OoE5Vp4jIYuBe4BagBu/srSMiMgdYhXfIDOAXwFeANcCzwELguZBerTHGmLPWX+d4lvubDeQEefTlImCvqlaoagewAlgUUGcR8LCbfgK4UkREVd9V1SOufBteH0u6iIwBclX1bVVV4BHgE/3EYYwxJoLE+/6NQsMiNwELVfXL7vltwAJVvdOvzlZXp9I93+fq1AS081VVvUpESoEfq+pVbt5lwL+o6vVB1r8EWAJQWFg4f8WKFVF5nf1pbm4mOzs7JusORyLEmQgxQmLEORRiPNnSEVK9EVlpA14HDO5tecUVV6wPdslFqBcA/gfeKbmtwF+B84FvusNYUSMis/EOX10T7rKqugxYBlBaWqplZWWRDS5E5eXlxGrd4UiEOBMhRkiMOIdCjI+uORRSvbIF4we8Dhga2zJQqNdxXKOqjcD1eGNVTQH+qZ9lqoBxfs9LXFnQOiKSAuQBte55CfA08HlV3edXv6SfNo0xxkRRqImjZ8/kb4A/qGpDCMusBaaKyCQRSQMWAysD6qwEbnfTNwEvq6qKSD7wF2Cpqr7RU9kNddIoIhe7s6k+D/wxxNdgjBlC6k918OKOYzz05n52VTfGOpxBJdTE8WcR2QnMB14SkVFAW18LqGoXcCfeGVE7gMdVdZuI3CMiH3fVHgBGishe4FtAzym7d+Lt1dwlIhvdY7Sb9zXgN8BeYB92RpUxJkD9qQ6WvVrB6p3HOVLfxiNvHeS1PSdiHdagEVIfh6oudf0cDaraLSItvP8MqWDLPYt3yqx/2V1+023AzUGW+wHBhzlBVdcBc0KJ2xgz9Ph8yj8+vommti6++pHzKMzN4PF1h1m1rZqZRbkU5KTHOsSEF84d/WYAt4jI5/EOK4XdYW2MMdH23NZqXtp5nOs+UMS4EcNIS0li0dyxJCcJL+48FuvwBoVQz6paDpwHbAS6XXHPdRTGGBMXVJVfvrKPSQVZXDx55OnynIxULj2vgPLdJyib1kZRXkYMo0x8oQ6rXgrM0mhd9GGMMRHwVkUtW6oa+PcbP/C+eZdNHcXre2tYf/Akf3P+2BhEN3iEeqhqK1AUzUCMMeZsPfDafgqy0/jkvOL3zctMS2bq6Gy2HmnEZ7+Bz0qoexwFwHYReQdo7ylU1Y/3vogxxpw7Nc3tlO8+wZLLJ5ORmhy0zpziPHZUN1F58hTjR2YFrWP6F2riuDuaQRhjzNn6y+ajdPuUT8x9/95Gj5ljcklOErYeabTEcRZCOlSlqq/gXTGe6qbXAhuiGJcxxoTlmY1VzCjKYXpR7+OvZqT2HK5qwLpsBy6kxCEiX8EbvfZXrqgYeCZaQRljTDgO1Z7i3UP1LOpjb6PHtMIc6k91Uneq8xxENjiF2jn+deBSoBFAVfcAo/tcwhhjzpHnt1cDcP35Y/qtO6nAO0RVcaI5qjENZqEmjnZ3Tw3g9ICEtp9njIkLq3cdZ+robMaNGNZv3dE56WSlJbO/puUcRDY4hZo4XhGRf8W7odLVwB+AP0UvLGOMCU1zexfv7D/JR2eEdhBERJg0Kpv9NS3WzzFAoSaOpcAJYAvwt3jjT/1btIIyxphQvb7nBJ3dyhUhJg6AyQVZ1LdaP8dAhTrIoU9EngGeUVUbYtIYEzdW7zxBTkYK8ycMD3mZnn6O/TXWzzEQfe5xiOduEakBdgG7ROSEiNzV13LGGHMuqCqrdx3n8qmjSE0OfczW0TnpZKQmcehkaxSjG7z629LfxDub6oOqOkJVRwALgEtF5JtRj84YY/qw7Ugjx5vawzpMBV4/R0n+MKrqTkUpssGtv8RxG/AZVd3fU6CqFcDn8O6+Z4wxMbN653EAyqaPCnvZ4uGZVDe20dbZ3X9lc4b+EkeqqtYEFrp+jtTohGSMMaF5eddxLijJoyA7/JszlQzPxKew/ajdVjZc/SWOjgHOM8aYqKptbmfj4fqwD1P1KBnuXfOx+XB9JMMaEvo7q+oCEQmWjgWwO6EYY2Lm1T0nUCXk6zcC5WakkJOewubKhghHNvj1ucehqsmqmhvkkaOq/R6qEpGFIrJLRPaKyNIg89NF5DE3f42ITHTlI0VktYg0i8jPA5Ypd21udA8b+sSYIejlnScoyE5nzti8AS0vIhQPz2RTpe1xhCuce46HRUSSgfuB64BZwGdEZFZAtTuAOlWdAvwEuNeVtwHfA77dS/O3qupc9zge+eiNMfGsq9vHK7uOUzZ9FElJMuB2iodnsu9ECy3tXRGMbvCLWuIALgL2qmqFG+dqBbAooM4i4GE3/QRwpYiIqrao6ut4CcQYY87w7uF6Gtu6BnyYqseY3EwAdh9rikRYQ0aoN3IaiGLgsN/zSrxrQILWUdUuEWkARgLvO5MrwP+ISDfwJPCDYPdCF5ElwBKAwsJCysvLB/Iazlpzc3PM1h2ORIgzEWKExIgz0WP8w64OkgWo3kl57a6gdTJa+j9/Z0K399Xxx1fW0TBuYCeKJvq2HIhoJo5ouVVVq0QkBy9x3AY8ElhJVZcBywBKS0u1rKzsnAbZo7y8nFitOxyJEGcixAiJEWeix/jjja/ywUk5fOzqS3pd/tE1h/pdR6YqWZt3QN5YyspmRzzOeBHpGKN5qKoKGOf3vMSVBa3jhmrPA2r7alRVq9zfJuBRvENixpgh4kh9Kzurm876MBVAkgjTi3LYYddyhCWaiWMtMFVEJolIGrAYWBlQZyVwu5u+CXg52GGnHiKSIiIFbjoVuB7YGvHIjTFxa/Uu73yYSCQOgBljctlZ3WRDrIchaolDVbuAO4FVwA7gcVXdJiL3iMjHXbUHgJEishf4Ft7w7QCIyAHgPuALIlLpzshKB1aJyGZgI94ey6+j9RqMMfFn9c7jlAzP5LxR2RFpb0ZRDg2tnVQ32rk4oYpqH4eqPot37w7/srv8ptuAm3tZdmIvzc6PVHzGmMTS1tnNG3trubm0BJGBn4brb0ZRLgA7q5sYk5cZkTYHu2geqjLGmIhas/8krZ3dAx5mJJjpRTkA7Dxqp+SGyhKHMSZhvLC9mszUZC6ZPDJibeZlplKcn8nOausgD5UlDmNMQvD5lBe2H+Mj00aRkZoc0bZnFOXYHkcYLHEYYxLCpsp6jjW2c+2cwoi3Pb0oh30nmuno8kW87cHIEocxJiGs2naMlCTho9MjnzhmjMmly6fsO2H3IA+FJQ5jTNxTVZ7fVs0l540kb1jk7yE3s6eD3Po5QmKJwxgT97YfbaSipoWFc4qi0v6kgizSkpOsnyNEljiMMXFv5cYjpCQJH5szJirtpyQnMbUwmx3VljhCYYnDGBPXfD5l5aYjfGTaKIZnpUVtPdOLcthlh6pCkoij4xpjhpC1B05ytKGNpdfNOF0Wysi34ZpZlMtTG6qoa+mIaoIaDGyPwxgT155YX8mwtGSunhX5s6n8TXMd5HZTp/5Z4jDGxK2G1k7+tPkIi+YWMywtugdIphda4giVJQ5jTNx6ekMlbZ0+bl0wPurrKsxNJzcjhV2WOPplicMYE5dUlUffOcT5JXnMKc6L+vpEhGmFOeyutosA+2OJwxgTl7bUdLP7WDOfu3jCOVvntKIcdh2zmzr1xxKHMSYu/bmikzF5GXxibvE5W+f0Qu+mTseb2s/ZOhORJQ5jTNxZd+Aku+t8fOWyyaSlnLuvqWnWQR4SSxzGmLiiqvzHX3eRkwaLLxp3Ttc9rdC7He0uu4K8T5Y4jDFx5dkt1bxz4CSfmpoW9VNwA43MTqcgO932OPoR1cQhIgtFZJeI7BWRpUHmp4vIY27+GhGZ6MpHishqEWkWkZ8HLDNfRLa4ZX4mkbrxsDEm5k51dPHvz+5gRlEOl5fEZmCLaYXZ7DpmZ1b1JWqJQ0SSgfuB64BZwGdEZFZAtTuAOlWdAvwEuNeVtwHfA74dpOlfAF8BprrHwshHb4yJhf/z5+0caWjlnkVzSIrRb8JphTnsOdaEz2dnVvUmmnscFwF7VbVCVTuAFcCigDqLgIfd9BPAlSIiqtqiqq/jJZDTRGQMkKuqb6t3vtwjwCei+BqMMefIX7dW8/t3DvO3l5/HRZNGxCyO6UU5nOropqq+NWYxxLto7gsWA4f9nlcCC3qro6pdItIAjARq+mizMqDNoOfqicgSYAlAYWEh5eXlYYYfGc3NzTFbdzgSIc5EiBESI854i3F/Qzc/eqeNSblJTO46yFN/OURSVxtP/eX5oPUzIrju8vKKM5631HUD8MSLb3Lh6P6/IuNtWwYT6RgH7ei4qroMWAZQWlqqZWVlMYmjvLycWK07HIkQZyLECPETZ18jyGa07ORI5mQ+ew6G8ujPzupGvv2bdxiVk8ljX/8QL24/TheQUbOTtoIZ/S5/tsoCtsH8tk5+sOZ50kdPpKxsSr/Lx8v73ZdIxxjNQ1VVgP+5dCWuLGgdEUkB8oDaftos6adNY0yCeHNvDTf/8i2Sk+DhL13E6JxI7ksMTE5GKsX5mey2U3J7Fc3EsRaYKiKTRCQNWAysDKizErjdTd8EvKx9XOuvqkeBRhG52J1N9Xngj5EP3RgTTW2d3fz4uZ3c+sAaCnMzePLvPsSU0dmxDuu0qXZmVZ+idqjK9VncCawCkoEHVXWbiNwDrFPVlcADwHIR2QucxEsuAIjIASAXSBORTwDXqOp24GvAQ0Am8Jx7GGNiLJSbK31yXjHPvFvFT1/aw9GGNm4pHcddN8wiKz2+jppPL8zhzb21dHX7SEm2y90CRfXdUtVngWcDyu7ym24Dbu5l2Ym9lK8D5kQuSmPiSyhfwPHQNxEqnyqVda1sqaznvhd2UdPcwfklefzklrlcPHlkrMMLanpRDh3dPvbXtDDVDUNi3hNfad4YE3fCvU1rZ7eP6oY2qupbqahpYd/xZlo7u0lOEq6aOZrbL5nIJeeNJJ6v3Z1RlAvAjuomSxxBWOIwJkGoKk3tXZxoaufxtYc5XHeK6oY26k51Un+qg+b2Ljq6fXR1Kz5V2jt9pKYIqUlJpCYnkZmWTHZGCjnpKQzv9JHR1ciWygYKctIYmZUe9mCCbZ3d1LZ0cNI9apraqapv5XhTGz3XzuVmpDBrTC5TC7OZOjqHOy6bFIUtE3lTRmeTkiTsONrIxy8YG+tw4o4lDmPiVHtXN4dOnmJ/TQtVda0caREDD70AABaTSURBVGijpb3r9PwkgVE56QwflsbwYWmMHzGM1JQk0pKTEGD38WY6u3x0+ny0d3VT55JLe5fPa2DPQR5+6+Dp9nIzUijITmdEVhqZaclkpiaTnprMgZoWun1Kt09p7eympb2Llo4u2jp9Z8SbnZ7C2PwMZowZRXF+JmPzM8nPTI3rPYvepKUkMWV0NjuPNsY6lLhkicOYOKGq7DvRzKu7T7CjupHDJ0/hUy9BFOZmML0whzF5GRTmZvCFD01kTH4GqX103PZ2iKmz20fn0V3UDJvAvPH51DR3UNvcTm1LByea2znZ3EFjm7dn09bZTXN7NylJQlISZKYmMzY/k6z0ZPIz0xiR5T1GZqWRnpocrU0TEzPH5PLWvr6uDhi6LHEYE2OHT57imXereHpjFRUnWgAYm5fBZVNHMakgiwkjhr3vS3n8yGEDXl9qchI5GcKwEcO4ZnZRv/XD7eM4V21F28wxOTz9bhV1LR0Mz0qLdThxxRKHMTHQ2e1j1bZqHnnrIO/sPwnAgkkj+OKlk2hq7SR/WN9fVIn0BZyo3usgb+RD5xXEOJr4YonDmHOoub2Lh988wPK3DlLd2Ma4EZn807XT+cSFxRTnZwKWFOLFzDEucRxtssQRwBKHMRHg/2Wf0dLxvi//zm4f7+w/ydsVtdS2dHDZ1AJ+eOMcyqaPJjkp8TqPh4JROd5NnbYfsQ7yQJY4jImibp/y7qE6Xtp5nIbWTi6dMpJ/unYGc8flxzo0E4I5xblsO9IQ6zDijiUOY6LAp8q2I428uP0YJ5rbKRmeyafmlXDXDYH3MjPxbM7YPF7bU0NbZzcZg+yssbNhicOYCFJVdtb5+POWvRypb2N0Tjq3LhjPrDG5cXk9g/Wn9G1OcR7dPmXH0UYuHD881uHEDUscxkTIodoWVm0/xv6aboYPS+Km+SXMHZcfs1ugmrP3gZI8ALZWNVji8GOJw5iztLO6kUfeOsDO6iay01O48bwkLpw9LeioqvYLP7GMzctgRFYaW6qsn8OfJQ5jBmjH0Ub+38t7eHZLNRmpSVwzq5APnVdAbv1u2mwo7kFBRJg9NpctVXZmlT9LHMaEaWtVAz97aQ/Pbz9GTnoKd14xheHDvPGdzODzgeI8lr1aYR3kfixxGBMCn095Zc8JHnrjAK/sPkFuRgr/cNVUvvihSeQNS7VDUIPY+SV5dPmU7UcbmWf9HIAlDjPE9feF39bZjQg88tZB9te0MConnW9fM43Pf2giuRmp5yhKE0s9neLvHqq3xOFY4jAmgKpSVd/K2gN1bKqsp6PLx4Xj8/np4rlcN2dM2PetMImtMDeD4vxMNhyq4w4S434i0WaJwxintaObjYfrWHewjqMNbaQmCx8ozuP7N8zmArvSe0i7cHw+Gw7WxTqMuBHVxCEiC4GfAsnAb1T1xwHz04FHgPlALXCLqh5w874D3AF0A99Q1VWu/ADQ5Mq7VLU0mq/BDG6qyv6aFtYdrGNrVQNdPqU4P5NFc8dyQUk+GanJljQM88YP58+bj1Ld0EZRXkasw4m5qCUOEUkG7geuBiqBtSKyUlW3+1W7A6hT1Skishi4F7hFRGYBi4HZwFjgRRGZpqrdbrkrVLUmWrGbwe94YxtPbKjkgdf2U9vSQUZqEqUTh1M6YQRj3Si1Pazj28yb4PVtbDhUx8c+MCbG0cReNPc4LgL2qmoFgIisABYB/oljEXC3m34C+Ll44zIsAlaoajuwX0T2uvbeimK8ZpBr6+zmpR3HefrdKlbvOk63T5lUkMVHZ4xmTnFen3fTM0PbrDG5pKckseGgJQ6IbuIoBg77Pa8EFvRWR1W7RKQBGOnK3w5YtthNK/C8iCjwK1VdFoXYzSDw6JpDdPu827FuOlzPtqONdHT5yMlI4dLzCiidMJyCnPRYh2kSQFpKEheU5LP2wMlYhxIXErFz/MOqWiUio4EXRGSnqr4aWElElgBLAAoLCykvLz/HYXqam5tjtu5wJEKcocZY3+5jW003a492sqteae2CjGSYWyDMH53MeXmQJPXQXg/tkY8zqauNjJqdkW84gizG95SXV4RUb0xyBysPdPLsC6sZlvre+GOD6X8nVNFMHFXAOL/nJa4sWJ1KEUkB8vA6yXtdVlV7/h4XkafxDmG9L3G4PZFlAKWlpVpWVnb2r2gAysvLidW6wxEvcfbVn5DRspMjmZMB+OyC8XT7lCP1rVTUtFBxopktlQ2sO1jHoZOtAGSnpzBzbA4zx+QwrTDn9KGojii/hoyanbQVzIjyWs6OxfiesgXjQ6qXMb6WP+57m7TiWZTNKjxdHi//O32JdIzRTBxrgakiMgnvS38x8NmAOiuB2/H6Lm4CXlZVFZGVwKMich9e5/hU4B0RyQKSVLXJTV8D3BPF12DiQGtHNyea2qg/5uPosWpqmtt56M39HKg9RUeX73S9guw05k8Yzm0XT2DB5BFsrmywkWlNxFw4Pp/0lCTe3FfLVX6JYyiKWuJwfRZ3AqvwTsd9UFW3icg9wDpVXQk8ACx3nd8n8ZILrt7jeB3pXcDXVbVbRAqBp919DVKAR1X1r9F6DebcUlVqm9s5XNdKdUMb1Y2tHGtsp6G183SdJDnBiKx05o7Lp2z6aCYXZDF5VDaTCrIoyE47454XW21gOhNB6SnJlE4czlsVtbEOJeai2sehqs8CzwaU3eU33Qbc3MuyPwR+GFBWAVwQ+UhNrNQ2t/PyzuO8tqeGd/afpLqxDYBkEUblpDOpIIui3AxG56RT3FXFsOIZJCcJnw3x8IIx/QnldOuez9slk0fyX8/v5mRLByOy0qIdWtxKxM5xk+D2Hm/mxR3HeHH7MdYfqkMVRueks2DySASYMHIYo3LSSUk68/TYjJojtCV5exR2bYWJhUunFPBfz+/mtT0nWDS3uP8FBilLHCbqun3K+oN1p5NFRU0LALPH5vKNj07l6lmFzB7r3VrVEoKJZxeU5FOQnc4L249Z4jAm0jq6fLxVUctftx7l+W3HqG3pIDVZuOS8Ar546USunFn4viu0jYl3SUnCR2eM4rmt1XR0+YbsgJeWOEzEtHV288ruE6zaWs2LO47R2NZFVloyV8wYzbWzizjR1H76Rjjlu07EOFpjBuaqmYU8vq6StQdOcumUgliHExOWOMxZaWnv4sUdx/j1qxXsOtZEZ7eSmZrMzDG5zB6by5TR2aQmJ9HU1mV3TzODwoenFpCWksQL249Z4jAmVF3dPt7YV8vTGypZte0YrZ3d5KSnMG/8cGaPzWNSQRbJSXb9hBmchqWlcPnUUTy39Sjfu35WrMOJCUscJmT7TjTz+zWHeGbjEWqa28nLTOXGecV8Ym4xu4812cV2Zsj45LxiXtxxjDf3Dc1Bui1xmD51dvu464/bWLO/looTLSSLML0oh2tnFzK9MIeU5CT2Hm+2pGGGlI/OGE1ORgpPbahi0RC8iNwShwnqSH0rK945xIq1hzne1E5+ZirXzCpk/oTh5Ni9ts0Ql5GazPXnj+WZd6u4ZuTQG2HZEscQ13PdREZLB799+yD7jjfz9v6T7DzqDdcxrTCHhbOLmFaUY3sVxvi5aX4Jv3/nEG8e6eK6WAdzjlniMJzq6OLtym7e2LCb2pYOstKSuXzaKD44ccSQHlbBmL7MG5/PBePyWXWggbt9OqROCLHEMYRtrqznyfWVbKqsp8unTBiZwlUzvau4U+xueMb0SUT428sn87XfbeCF7dUsnDN07gxoiWOIaevs5s+bj7L87YNsOlxPWnIS88YP57LhjYyceF6swzMmoVw7u4hRmcL9q/dxzawikobIXocljiFiz7EmHl93mD+sr6T+VCfnjcri7htm4VOvoy+jpom2WAdpTIJJThJunJrGss0NPLmhkptLx/W/0CBgiWMQa2rr5C+bj/LYusO8e6ielCTh6lmF3HbJBC6ZPNIGFTQmAi4ek8za+nzu/esuFs4pGhJnHVriGGR8PmXtgZP8YX0lf9l8lNbObqaMzua7H5vJjfOKKcgeeqcOGhNNSSJ8/4bZ3Pjfb/D9ldu479NzYx1S1FniGARUla1VjazcVMWfNx/laEMb6SlJnF+SR+mEEZQMz0REeH7bsViHasygNHdcPt/46FR++tIeLp48kk8P8kNWljgSVLdPefdQHS/uOM6qbdXsr2khNVn4yLRRLL1uBnUtnUN2yGdjYuEbV05l7YGTfPfpLYzMSuPKmYP3knJLHAnkRFM7a/bX8sBr+9l1rIlTHd0kCUwqyOLGC4uZPTaXYWkptLR3W9Iw5hxLThJ+8bn53PbAGv7utxv48ac+wCfnlcQ6rKiwxBGnVJXKulY2HKpjzf6TrKmoZd8J7855manJTC/KYUZRDlNH55CZZsOVGxMP8jJTWf6lBSxZvo5vPb6J1/fUsPS6GYzOzYh1aBEV1cQhIguBnwLJwG9U9ccB89OBR4D5QC1wi6oecPO+A9wBdAPfUNVVobSZaFSVulOdHKxtoeJEC09tqORIQxtHG1pp6/QBkJ6SxMSRWVw7u4hJBVkU52cOqatUjYm1vs4+zGjp4NE1h/jsgvEA5A1L5XdfXsDPXtrDf5fv47mt1dw0v4SbS0uYMzZvUFzrEbXEISLJwP3A1UAlsFZEVqrqdr9qdwB1qjpFRBYD9wK3iMgsYDEwGxgLvCgi09wy/bUZUz6fcqqzm1PtXTS3d3GgoZs1FbU0tnVR09zOiaZ2aprbOd7YTmX9KQ7WnqKprev08ilJQlFeBucX5zMmP4OS/GGMyc+wcaKMSSApyUl865rpfHJeCT97aQ+PrTvM8rcPUpCdxvkl+UwtzGbq6BzG5mUwPCuNEVlpDEtLJi0libTkJCTO/9+jucdxEbBXVSsARGQFsAjw/5JfBNztpp8Afi7eFlsErFDVdmC/iOx17RFCmxHz+QffYX9NMz6f1xnt056He+7KulXx+cCnSpdP39/QW2+f8TQvM5WC7DSKhw9j3vjhjB8xjIkjs5hYMIy39p20vQljBomJBVncd8tc7rphFi/tOM7re2vYcbSR1/fU0NHt63W5tOQkL4mkJJ3+PhCgJ58IgohXBpxONKfny3t1Vv3D5RF/XdFMHMXAYb/nlcCC3uqoapeINAAjXfnbAcsWu+n+2gRARJYAS9zTZhHZNYDXEAkFQCLc7SUR4kyEGCEx4rQYI6cAqLk11lH0IvOfgYFvywnBCgdt57iqLgOWxToOEVmnqqWxjqM/iRBnIsQIiRGnxRg5iRBnpGOM5jmbVYD/VTAlrixoHRFJAfLwOsl7WzaUNo0xxkRRNBPHWmCqiEwSkTS8zu6VAXVWAre76ZuAl1VVXfliEUkXkUnAVOCdENs0xhgTRVE7VOX6LO4EVuGdOvugqm4TkXuAdaq6EngAWO46v0/iJQJcvcfxOr27gK+rajdAsDaj9RoiJOaHy0KUCHEmQoyQGHFajJGTCHFGNEbxfuAbY4wxobFxKYwxxoTFEocxxpiwWOKIIhFZKCK7RGSviCyNdTwAIjJORFaLyHYR2SYi/8uV3y0iVSKy0T0+FgexHhCRLS6eda5shIi8ICJ73N/hMYxvut/22igijSLyD/GwLUXkQRE5LiJb/cqCbjvx/Mx9TjeLyLwYxvifIrLTxfG0iOS78oki0uq3TX8Zwxh7fX9F5DtuO+4SkWvPRYx9xPmYX4wHRGSjKz/7bamq9ojCA6/zfh8wGUgDNgGz4iCuMcA8N50D7AZm4V3B/+1YxxcQ6wGgIKDsP4ClbnopcG+s4/R7v6vxLpiK+bYELgfmAVv723bAx4Dn8C5EvhhYE8MYrwFS3PS9fjFO9K8X4+0Y9P11/0ebgHRgkvv/T45VnAHz/y9wV6S2pe1xRM/pIVdUtQPoGR4lplT1qKpucNNNwA7euyo/ESwCHnbTDwOfiGEs/q4E9qnqwVgHAqCqr+Kdqeivt223CHhEPW8D+SIyJhYxqurzqtozeNvbeNdqxUwv27E3p4dKUtX9gP9QSVHVV5xuGKdPA7+P1PoscURPsCFX4uoLWkQmAhcCa1zRne4QwYOxPATkR4HnRWS9G0IGoFBVj7rpaiBe7pazmDP/MeNtW0Lv2y5eP6tfwtsT6jFJRN4VkVdE5LJYBeUEe3/jdTteBhxT1T1+ZWe1LS1xDFEikg08CfyDqjYCvwDOA+YCR/F2bWPtw6o6D7gO+LqInDFam3r73TE/n9xdjPpx4A+uKB635RniZdv1RkS+i3cN1+9c0VFgvKpeCHwLeFREcmMUXty/vwE+w5k/as56W1riiJ64HR5FRFLxksbvVPUpAFU9pqrdquoDfs052sXui6pWub/HgafxYjrWcxjF/T0euwhPuw7YoKrHID63pdPbtourz6qIfAG4HrjVJTjc4Z9aN70er/9gWq+NRFEf729cbUc4PZTTJ4HHesoisS0tcURPXA6P4o53PgDsUNX7/Mr9j2nfCGwNXPZcEpEsEcnpmcbrNN3KmcPU3A78MTYRnuGMX3Txti399LbtVgKfd2dXXQw0+B3SOqfEu1HbPwMfV9VTfuWjxLvHDyIyGW8YoooYxdjb+9vbUEmxdBWwU1Urewoisi3PRY//UH3gna2yGy+jfzfW8biYPox3iGIzsNE9PgYsB7a48pXAmBjHORnvDJVNwLae7Yc37P5LwB7gRWBEjOPMwhuYM8+vLObbEi+RHQU68Y6139HbtsM7m+p+9zndApTGMMa9eP0EPZ/NX7q6n3Kfg43ABuCGGMbY6/sLfNdtx13AdbF8v135Q8BXA+qe9ba0IUeMMcaExQ5VGWOMCYslDmOMMWGxxGGMMSYsljiMMcaExRKHMcaYsFjiMMYYExZLHMb4EZF8EfnaAJed298Q6iLyBRH5+cCi67fdsX7PD4hIQaTXYwxY4jAmUD4woMSBN3ZRrO5j8gVgbH+VjIkESxzGnOnHwHnuBjf/KSL/JCJr3Uio/xtARG4UkZfcEB1jRGS3iIwH7gFuccve0t+K3NAPT7r214rIpa78bjfqarmIVIjIN/yW+Z67SdDrIvJ7Efm2iNwElAK/c+vOdNX/XkQ2iHczrBmR3lBm6LLEYcyZluLdV2Mu8ALeOD4X4e1NzBeRy1X1abzhHb6ON8jd91X1EHAX8JiqzlXVx4I3f4afAj9R1Q/iDQPxG795M4Br3bq/LyKpItJT7wK8gRVLAVT1CWAd3qCAc1W11bVRo97owr8Avj3A7WHM+6TEOgBj4tg17vGue56Nl0heBf4eb3C7t1V1oDfIuQqY5Y07CUCuG+4e4C+q2g60i8hxvHtnXAr8UVXbgDYR+VM/7T/l/q7HGyHVmIiwxGFM7wT4kar+Ksi8EsAHFIpIknpDbIcrCbjYJYL3Vuolkna/om4G9r/a08ZAlzcmKDtUZcyZmvDuxQ6wCvhSz16AiBSLyGh3j4MH8YZT34F3M5zAZUPxPN6eC679uf3UfwO4QUQyXEzX9xK3MVFlicMYP+rd4OYNEdkKXA08CrwlIluAJ/C+nP8VeE1VX8dLGl8WkZnAarxDTyF1jgPfAEpdx/t24Kv9xLYWbxjvzXi3VN0CNLjZDwG/DOgcNyYqbFh1YxKIiGSrarOIDMPra1miqhtiHZcZWuy4pzGJZZmIzAIygIctaZhYsD0OY6JARL4I/K+A4jdU9euxiMeYSLLEYYwxJizWOW6MMSYsljiMMcaExRKHMcaYsFjiMMYYE5b/D+pG67pdIIswAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tSjYqrPzGenE",
        "outputId": "698250de-1372-45ec-f3d8-ead615b6b341",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "source": [
        "#Create visualization of the distribution of text length in comparision to target feature\n",
        "f, (ax1, ax2) = plt.subplots(1, 2, sharex=True,figsize=(10,6))\n",
        "sns.distplot(train_df[(train_df['target'] == 1)]['text_length'], ax=ax1, kde=False, color='green',label='Disater Tweets')\n",
        "sns.distplot(train_df[(train_df['target'] == 0)]['text_length'],ax=ax2, kde=False, color='red',label='Non-Disater Tweets');\n",
        "f.suptitle('Tweet length distribution')\n",
        "f.legend(loc='upper right')\n",
        "ax1.grid()\n",
        "ax2.grid()\n",
        "plt.show()"
      ],
      "execution_count": 150,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x432 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "r7L28p8FGenG"
      },
      "source": [
        "The distribution of both kind of tweets seems to be almost same.About 120 to 140 characters in a tweet are the most common among both."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "x2Jvbmh4GenG",
        "outputId": "eb620759-af6d-4edb-d483-0d923c74019a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "source": [
        "#Create visualization of the distribution of the word counts in comparision to target feature\n",
        "fig,(ax1,ax2)=plt.subplots(1,2,figsize=(10,5))\n",
        "dis_tweet=train_df[train_df['target']==1]['text'].str.split().map(lambda x: len(x))\n",
        "ax1.hist(dis_tweet,color='blue')\n",
        "ax1.set_title('Disaster tweets')\n",
        "ax1.grid()\n",
        "nondis_tweet=train_df[train_df['target']==0]['text'].str.split().map(lambda x: len(x))\n",
        "ax2.hist(nondis_tweet,color='red')\n",
        "ax2.set_title('Non-disaster tweets')\n",
        "ax2.grid()\n",
        "fig.suptitle('Words in a tweet')\n",
        "plt.show()"
      ],
      "execution_count": 151,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x360 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Tv4glIFWGenI"
      },
      "source": [
        "### Exploring the Target Column\n",
        "\n",
        "- Distribution of the Target Column-\n",
        "   - We have to predict whether a given tweet is about a real disaster or not.\n",
        "   - If real disaster, predict a `1`. If not, predict a `0`.\n",
        "- Exploring the Target Column- Let's look at what the disaster and the non disaster tweets look like."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OBK7QDjMGenJ",
        "outputId": "10c0eab9-d2d4-4866-9024-8545abd056a8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "source": [
        "# A disaster tweet exmaple\n",
        "train_df[train_df['target']==1]['text'][10:20]"
      ],
      "execution_count": 152,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "10          Three people died from the heat wave so far\n",
              "11    Haha South Tampa is getting flooded hah- WAIT ...\n",
              "12    #raining #flooding #Florida #TampaBay #Tampa 1...\n",
              "13              #Flood in Bago Myanmar #We arrived Bago\n",
              "14    Damage to school bus on 80 in multi car crash ...\n",
              "31    @bbcmtd Wholesale Markets ablaze http://t.co/l...\n",
              "33    #AFRICANBAZE: Breaking news:Nigeria flag set a...\n",
              "37    INEC Office in Abia Set Ablaze - http://t.co/3...\n",
              "38    Barbados #Bridgetown JAMAICA ÛÒ Two cars set ...\n",
              "46    How the West was burned: Thousands of wildfire...\n",
              "Name: text, dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 152
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "L1JMtuvMGenL",
        "outputId": "dbcb96c1-0ac8-4b1e-b61d-5f5c9a3095d1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "source": [
        "#A non-disaster tweet example\n",
        "train_df[train_df['target']==0]['text'][10:20]"
      ],
      "execution_count": 153,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "25                       No way...I can't eat that shit\n",
              "26                                Was in NYC last week!\n",
              "27                                   Love my girlfriend\n",
              "28                                            Cooool :)\n",
              "29                                   Do you like pasta?\n",
              "30                                             The end!\n",
              "32    We always try to bring the heavy. #metal #RT h...\n",
              "34                   Crying out for more! Set me ablaze\n",
              "35    On plus side LOOK AT THE SKY LAST NIGHT IT WAS...\n",
              "36    @PhDSquares #mufc they've built so much hype a...\n",
              "Name: text, dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 153
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4Vg1c7itGenN"
      },
      "source": [
        "# 2. Data Preprocessing"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OVW0TItuGenO"
      },
      "source": [
        "### Data Cleaning\n",
        "\n",
        "Before starting any NLP project, text data needs to be pre-processed to convert it into in a consistent format.Text will be cleaned, tokneized and converted into a matrix.\n",
        "\n",
        "Some of the basic text pre-processing techniques includes:\n",
        "### Make text all lower or uppercase\n",
        "Algorithms does not treat the same word different in different cases.\n",
        "### Removing Noise\n",
        "Everything in the text that isn’t a standard number or letter i.e. Punctuation, Numerical values,etc.\n",
        "###  Tokenization\n",
        "Tokenization is just the term used to describe the process of converting the normal text strings into a list of tokens i.e. words.Sentence tokenizer can be used to find the list of sentences and Word tokenizer can be used to find the list of words in strings.\n",
        "### Stopword Removal\n",
        "Sometimes, some extremely common words which would appear to be of little value in helping select documents matching a user need are excluded from the vocabulary entirely. These words are called stop words\n",
        "\n",
        "### Stemming\n",
        "Stemming is the process of reducing inflected (or sometimes derived) words to their stem, base or root form — generally a written word form. Example if we were to stem the following words: “Stems”, “Stemming”, “Stemmed”, “and Stemtization”, the result would be a single word “stem”.\n",
        "### Lemmatization\n",
        "A slight variant of stemming is lemmatization. The major difference between these is, that, stemming can often create non-existent words, whereas lemmas are actual words. So, your root stem, meaning the word you end up with, is not something you can just look up in a dictionary, but you can look up a lemma. Examples of Lemmatization are that “run” is a base form for words like “running” or “ran” or that the word “better” and “good” are in the same lemma so they are considered the same."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CUlD_ut1GenO",
        "outputId": "741304ce-9ee0-416c-eefd-153a310e390a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "source": [
        "#Lets have a quick look of the text data\n",
        "train_df['text'][:5]"
      ],
      "execution_count": 154,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0    Our Deeds are the Reason of this #earthquake M...\n",
              "1               Forest fire near La Ronge Sask. Canada\n",
              "2    All residents asked to 'shelter in place' are ...\n",
              "3    13,000 people receive #wildfires evacuation or...\n",
              "4    Just got sent this photo from Ruby #Alaska as ...\n",
              "Name: text, dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 154
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "I2f7Yvv5GenR"
      },
      "source": [
        "# Create a function to clean the text\n",
        "\n",
        "def clean_text(text):\n",
        "\n",
        "    '''\n",
        "    Input- 'text' to be cleaned\n",
        "       \n",
        "       Output- Convert input 'text' to lowercase,remove square brackets,links,punctuation\n",
        "       and words containing numbers. Return clean text.\n",
        "    \n",
        "    '''\n",
        "    text = text.lower()\n",
        "    text = re.sub('\\[.*?\\]', '', text)\n",
        "    text = re.sub('https?://\\S+|www\\.\\S+', '', text)\n",
        "    text = re.sub('<.*?>+', '', text)\n",
        "    text = re.sub('[%s]' % re.escape(string.punctuation), '', text)\n",
        "    text = re.sub('\\n', '', text)\n",
        "    text = re.sub('\\w*\\d\\w*', '', text)\n",
        "    return text"
      ],
      "execution_count": 155,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "t4Xy3SB3GenT"
      },
      "source": [
        "# Lets apply the clean_text function to both test and training datasets copies\n",
        "train_df1=train_df.copy()\n",
        "test_df1=test_df.copy()\n",
        "train_df1['text'] = train_df1['text'].apply(lambda x: clean_text(x))\n",
        "test_df1['text'] = test_df1['text'].apply(lambda x: clean_text(x))"
      ],
      "execution_count": 156,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GhX4l3YTkotJ",
        "outputId": "314691f2-4e15-459d-8999-fb9435e64706",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "source": [
        "train_df1['text'][:5]"
      ],
      "execution_count": 157,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0    our deeds are the reason of this earthquake ma...\n",
              "1                forest fire near la ronge sask canada\n",
              "2    all residents asked to shelter in place are be...\n",
              "3     people receive wildfires evacuation orders in...\n",
              "4    just got sent this photo from ruby alaska as s...\n",
              "Name: text, dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 157
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jecLk_0SGenV"
      },
      "source": [
        "#Lets look cleaned text data\n",
        "def text_after_preprocess(before_text,after_text):\n",
        "    \n",
        "    '''\n",
        "    Input- before_text=text column before cleanup\n",
        "              after_text= text column after cleanup\n",
        "       Output- print before and after text to compare how it looks after cleanup\n",
        "       \n",
        "    '''\n",
        "    print('-'*60)\n",
        "    print('Text before cleanup')\n",
        "    print('-'*60)\n",
        "    print(before_text.head(5))\n",
        "    print('-'*60)\n",
        "    print('Text after cleanup')\n",
        "    print('-'*60)\n",
        "    print(after_text.head(5))"
      ],
      "execution_count": 158,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JITjpv--GenX",
        "outputId": "079237bb-f908-49a1-ee8b-30b963df25b3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "source": [
        "text_after_preprocess(train_df.text,train_df1.text)"
      ],
      "execution_count": 159,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "------------------------------------------------------------\n",
            "Text before cleanup\n",
            "------------------------------------------------------------\n",
            "0    Our Deeds are the Reason of this #earthquake M...\n",
            "1               Forest fire near La Ronge Sask. Canada\n",
            "2    All residents asked to 'shelter in place' are ...\n",
            "3    13,000 people receive #wildfires evacuation or...\n",
            "4    Just got sent this photo from Ruby #Alaska as ...\n",
            "Name: text, dtype: object\n",
            "------------------------------------------------------------\n",
            "Text after cleanup\n",
            "------------------------------------------------------------\n",
            "0    our deeds are the reason of this earthquake ma...\n",
            "1                forest fire near la ronge sask canada\n",
            "2    all residents asked to shelter in place are be...\n",
            "3     people receive wildfires evacuation orders in...\n",
            "4    just got sent this photo from ruby alaska as s...\n",
            "Name: text, dtype: object\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Xc7Xp983GenZ",
        "outputId": "23b1bdcf-602a-41da-89ef-e523c48673dc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "source": [
        "text_after_preprocess(test_df.text,test_df1.text)"
      ],
      "execution_count": 160,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "------------------------------------------------------------\n",
            "Text before cleanup\n",
            "------------------------------------------------------------\n",
            "0                   Just happened a terrible car crash\n",
            "1    Heard about #earthquake is different cities, s...\n",
            "2    there is a forest fire at spot pond, geese are...\n",
            "3             Apocalypse lighting. #Spokane #wildfires\n",
            "4        Typhoon Soudelor kills 28 in China and Taiwan\n",
            "Name: text, dtype: object\n",
            "------------------------------------------------------------\n",
            "Text after cleanup\n",
            "------------------------------------------------------------\n",
            "0                   just happened a terrible car crash\n",
            "1    heard about earthquake is different cities sta...\n",
            "2    there is a forest fire at spot pond geese are ...\n",
            "3                apocalypse lighting spokane wildfires\n",
            "4          typhoon soudelor kills  in china and taiwan\n",
            "Name: text, dtype: object\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9vr7oc0EGenc",
        "outputId": "cccf4c21-02ec-410d-d7e1-080396a87966",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "source": [
        "test_df.text[1]"
      ],
      "execution_count": 161,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'Heard about #earthquake is different cities, stay safe everyone.'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 161
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "N565rdbFGene"
      },
      "source": [
        "### Tokenization\n",
        "Tokenization is a process which splits an input text into tokens and the tokens can be a word, sentence, paragraph etc.\n",
        "\n",
        "Following code will show how tokenization of text works:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9_ELvCk1Genf",
        "outputId": "f9d515f4-5233-484d-e8eb-5eb7a0d26fc9",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "source": [
        "# Example how tokenization of text works\n",
        "text = \"Heard about #earthquake is different cities, stay safe everyone.\"\n",
        "tokenizer1 = nltk.tokenize.WhitespaceTokenizer()\n",
        "tokenizer2 = nltk.tokenize.TreebankWordTokenizer()\n",
        "tokenizer3 = nltk.tokenize.WordPunctTokenizer()\n",
        "tokenizer4 = nltk.tokenize.RegexpTokenizer(r'\\w+')\n",
        "print(\"-\"*100)\n",
        "print(\"Example Text: \",text)\n",
        "print(\"-\"*100)\n",
        "print(\"Tokenization by whitespace:- \",tokenizer1.tokenize(text))\n",
        "print(\"Tokenization by words using Treebank Word Tokenizer:- \",tokenizer2.tokenize(text))\n",
        "print(\"Tokenization by punctuation:- \",tokenizer3.tokenize(text))\n",
        "print(\"Tokenization by regular expression:- \",tokenizer4.tokenize(text))"
      ],
      "execution_count": 162,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "----------------------------------------------------------------------------------------------------\n",
            "Example Text:  Heard about #earthquake is different cities, stay safe everyone.\n",
            "----------------------------------------------------------------------------------------------------\n",
            "Tokenization by whitespace:-  ['Heard', 'about', '#earthquake', 'is', 'different', 'cities,', 'stay', 'safe', 'everyone.']\n",
            "Tokenization by words using Treebank Word Tokenizer:-  ['Heard', 'about', '#', 'earthquake', 'is', 'different', 'cities', ',', 'stay', 'safe', 'everyone', '.']\n",
            "Tokenization by punctuation:-  ['Heard', 'about', '#', 'earthquake', 'is', 'different', 'cities', ',', 'stay', 'safe', 'everyone', '.']\n",
            "Tokenization by regular expression:-  ['Heard', 'about', 'earthquake', 'is', 'different', 'cities', 'stay', 'safe', 'everyone']\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SjArD3p-Genh"
      },
      "source": [
        "# Lets Tokenize the training and the test dataset copies with RegEx tokenizer\n",
        "tokenizer = nltk.tokenize.RegexpTokenizer(r'\\w+')\n",
        "train_df1['text'] = train_df1['text'].apply(lambda x: tokenizer.tokenize(x))\n",
        "test_df1['text'] = test_df1['text'].apply(lambda x: tokenizer.tokenize(x))"
      ],
      "execution_count": 163,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jAYzdnvDGenj",
        "outputId": "116d10c8-499b-4cc0-ef4f-1d3768325728",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "source": [
        "#Lets checy tokenized text\n",
        "train_df1['text'].head()"
      ],
      "execution_count": 164,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0    [our, deeds, are, the, reason, of, this, earth...\n",
              "1        [forest, fire, near, la, ronge, sask, canada]\n",
              "2    [all, residents, asked, to, shelter, in, place...\n",
              "3    [people, receive, wildfires, evacuation, order...\n",
              "4    [just, got, sent, this, photo, from, ruby, ala...\n",
              "Name: text, dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 164
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0deNDQpiGenl",
        "outputId": "1b53b608-99dc-4f5e-d35f-0e1c8d9ea289",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "source": [
        "test_df1['text'].head()"
      ],
      "execution_count": 165,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0            [just, happened, a, terrible, car, crash]\n",
              "1    [heard, about, earthquake, is, different, citi...\n",
              "2    [there, is, a, forest, fire, at, spot, pond, g...\n",
              "3           [apocalypse, lighting, spokane, wildfires]\n",
              "4    [typhoon, soudelor, kills, in, china, and, tai...\n",
              "Name: text, dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 165
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bAlw79dqGenn"
      },
      "source": [
        "### Stopwords Removal\n",
        "\n",
        "Now, let's get rid of the stopwords i.e words which occur very frequently and have possible value like a, an, the, are etc."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qGve1r6-Genn"
      },
      "source": [
        "#Create a funtion to remove stopwords\n",
        "def remove_stopwords(text):\n",
        "    \n",
        "    \"\"\"\n",
        "    Input- text=text from which english stopwprds will be removed\n",
        "    Output- return text without english stopwords \n",
        "    \n",
        "    \"\"\"\n",
        "    words = [w for w in text if w not in stopwords.words('english')]\n",
        "    return words"
      ],
      "execution_count": 166,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hwmvvzV1Genp"
      },
      "source": [
        "train_df1['text'] = train_df1['text'].apply(lambda x : remove_stopwords(x))\n",
        "test_df1['text'] = test_df1['text'].apply(lambda x : remove_stopwords(x))"
      ],
      "execution_count": 167,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "42W3vowEGenr",
        "outputId": "4c47d5e5-a575-4774-a0d9-612b7e53ab1c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "source": [
        "train_df1.text.head()"
      ],
      "execution_count": 168,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0    [deeds, reason, earthquake, may, allah, forgiv...\n",
              "1        [forest, fire, near, la, ronge, sask, canada]\n",
              "2    [residents, asked, shelter, place, notified, o...\n",
              "3    [people, receive, wildfires, evacuation, order...\n",
              "4    [got, sent, photo, ruby, alaska, smoke, wildfi...\n",
              "Name: text, dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 168
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7so9peq6Genu",
        "outputId": "5a43b602-3ab7-4ed7-f869-422553e06d07",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "source": [
        "test_df1.text.head()"
      ],
      "execution_count": 169,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0                     [happened, terrible, car, crash]\n",
              "1    [heard, earthquake, different, cities, stay, s...\n",
              "2    [forest, fire, spot, pond, geese, fleeing, acr...\n",
              "3           [apocalypse, lighting, spokane, wildfires]\n",
              "4            [typhoon, soudelor, kills, china, taiwan]\n",
              "Name: text, dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 169
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Q_GZFkFBGenx"
      },
      "source": [
        "### Stemming and Lemmatization\n",
        "- **Stemming**: Removing and replacing suffixes to get to the root form of the word, which is called the stem for instance cats - cat, wolves - wolv\n",
        "- **Lemmatization** : Returns the base or dictionary form of a word, which is known as the lemma\n",
        "\n",
        "It is important to note that stemming and lemmatization sometimes doesnt necessarily improve results as sometimes we dont want to trim words rather preserve their original form.Its usage from problem to problem and for this problem it wouldnt be good idea to use it."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "t474E1ixGenx",
        "outputId": "89d82e22-3496-4969-d368-e7045556456f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "source": [
        "# Stemming and Lemmatization examples\n",
        "text =  \"ran deduced dogs talking studies\"\n",
        "\n",
        "tokenizer = nltk.tokenize.TreebankWordTokenizer()\n",
        "tokens = tokenizer.tokenize(text)\n",
        "\n",
        "# Stemmer\n",
        "stemmer = nltk.stem.PorterStemmer()\n",
        "print(\"Stemming the sentence: \", \" \".join(stemmer.stem(token) for token in tokens))\n",
        "\n",
        "# Lemmatizer\n",
        "lemmatizer=nltk.stem.WordNetLemmatizer()\n",
        "print(\"Lemmatizing the sentence: \", \" \".join(lemmatizer.lemmatize(token) for token in tokens))"
      ],
      "execution_count": 170,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Stemming the sentence:  ran deduc dog talk studi\n",
            "Lemmatizing the sentence:  ran deduced dog talking study\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fEFZlw8dGen0"
      },
      "source": [
        "# Lets combine text after processing it\n",
        "def combine_text(text):\n",
        "    \n",
        "    '''\n",
        "    Input-text= list cleand and tokenized text\n",
        "    Output- Takes a list of text and returns combined one large chunk of text.\n",
        "    \n",
        "    '''\n",
        "    all_text = ' '.join(text)\n",
        "    return all_text"
      ],
      "execution_count": 171,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6RHna4roGen3"
      },
      "source": [
        "train_df1['text'] = train_df1['text'].apply(lambda x : combine_text(x))\n",
        "test_df1['text'] = test_df1['text'].apply(lambda x : combine_text(x))"
      ],
      "execution_count": 172,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "r8iMd9TvGen5",
        "outputId": "45f660ce-34f4-4857-d62d-387cffa19684",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "source": [
        "train_df1.head()"
      ],
      "execution_count": 173,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>id</th>\n",
              "      <th>keyword</th>\n",
              "      <th>text</th>\n",
              "      <th>target</th>\n",
              "      <th>text_length</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>NaN</td>\n",
              "      <td>deeds reason earthquake may allah forgive us</td>\n",
              "      <td>1</td>\n",
              "      <td>69</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>4</td>\n",
              "      <td>NaN</td>\n",
              "      <td>forest fire near la ronge sask canada</td>\n",
              "      <td>1</td>\n",
              "      <td>38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>5</td>\n",
              "      <td>NaN</td>\n",
              "      <td>residents asked shelter place notified officer...</td>\n",
              "      <td>1</td>\n",
              "      <td>133</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>6</td>\n",
              "      <td>NaN</td>\n",
              "      <td>people receive wildfires evacuation orders cal...</td>\n",
              "      <td>1</td>\n",
              "      <td>65</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>7</td>\n",
              "      <td>NaN</td>\n",
              "      <td>got sent photo ruby alaska smoke wildfires pou...</td>\n",
              "      <td>1</td>\n",
              "      <td>88</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   id keyword  ... target  text_length\n",
              "0   1     NaN  ...      1           69\n",
              "1   4     NaN  ...      1           38\n",
              "2   5     NaN  ...      1          133\n",
              "3   6     NaN  ...      1           65\n",
              "4   7     NaN  ...      1           88\n",
              "\n",
              "[5 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 173
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0Dkq6IArGen8"
      },
      "source": [
        "### Text Pre-processing Function\n",
        "\n",
        "Putting all above functions together, below will be final function to pre-process the text before modeling with ML algorithms."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MRGGBOp4Gen9"
      },
      "source": [
        "# Create a function to pre-process the tweets\n",
        "def pre_process_text_combined(text):\n",
        "    \n",
        "    \"\"\"\n",
        "    Input- text= text to be pre-processed\n",
        "    \n",
        "    Oputput- return cleaned and combined text to be vectrorized for Machine learning.\n",
        "\n",
        "    \"\"\"\n",
        "    #Initiate a tokenizer\n",
        "    tokenizer = nltk.tokenize.RegexpTokenizer(r'\\w+')\n",
        "    # Clean the text using clean_text function\n",
        "    cleaned_txt = clean_text(text)\n",
        "    tokenized_text = tokenizer.tokenize(cleaned_txt)\n",
        "    remove_stopwords = [w for w in tokenized_text if w not in stopwords.words('english')]\n",
        "    combined_text = ' '.join(remove_stopwords)\n",
        "    return  combined_text"
      ],
      "execution_count": 174,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eMFds3l7GeoB"
      },
      "source": [
        "# Create a function to pre-process the tweets\n",
        "def pre_process_text(text):\n",
        "    \"\"\"\n",
        "    Input- text= text to be pre-processed\n",
        "    \n",
        "    Oputput- return cleaned text to be vectrorized for Machine learning.\n",
        "\n",
        "    \"\"\"\n",
        "    #Initiate a tokenizer\n",
        "    tokenizer = nltk.tokenize.RegexpTokenizer(r'\\w+')\n",
        "    # Clean the text using clean_text function\n",
        "    cleaned_txt = clean_text(text)\n",
        "    tokenized_text = tokenizer.tokenize(cleaned_txt)\n",
        "    remove_stopwords = [w for w in tokenized_text if w not in stopwords.words('english')]\n",
        "    return remove_stopwords"
      ],
      "execution_count": 175,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "khzXeecPGeoE",
        "outputId": "1f590aba-25d7-4029-f930-e091b24b2773",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        }
      },
      "source": [
        "# Text before pre-processing\n",
        "train_df.text.head()"
      ],
      "execution_count": 176,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0    Our Deeds are the Reason of this #earthquake M...\n",
              "1               Forest fire near La Ronge Sask. Canada\n",
              "2    All residents asked to 'shelter in place' are ...\n",
              "3    13,000 people receive #wildfires evacuation or...\n",
              "4    Just got sent this photo from Ruby #Alaska as ...\n",
              "Name: text, dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 176
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "A7AZPAlTGeoH"
      },
      "source": [
        "# Lets pre-process train data text\n",
        "train_df2=train_df.copy()\n",
        "train_df2['text'] = train_df2['text'].apply(lambda x : pre_process_text_combined(x))"
      ],
      "execution_count": 177,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "21Lhy5V1GeoK",
        "outputId": "66124393-49ac-422a-8a0b-254eee056b16",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "# Text after pre-processing the text column\n",
        "train_df2.head()"
      ],
      "execution_count": 178,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>id</th>\n",
              "      <th>keyword</th>\n",
              "      <th>text</th>\n",
              "      <th>target</th>\n",
              "      <th>text_length</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>NaN</td>\n",
              "      <td>deeds reason earthquake may allah forgive us</td>\n",
              "      <td>1</td>\n",
              "      <td>69</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>4</td>\n",
              "      <td>NaN</td>\n",
              "      <td>forest fire near la ronge sask canada</td>\n",
              "      <td>1</td>\n",
              "      <td>38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>5</td>\n",
              "      <td>NaN</td>\n",
              "      <td>residents asked shelter place notified officer...</td>\n",
              "      <td>1</td>\n",
              "      <td>133</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>6</td>\n",
              "      <td>NaN</td>\n",
              "      <td>people receive wildfires evacuation orders cal...</td>\n",
              "      <td>1</td>\n",
              "      <td>65</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>7</td>\n",
              "      <td>NaN</td>\n",
              "      <td>got sent photo ruby alaska smoke wildfires pou...</td>\n",
              "      <td>1</td>\n",
              "      <td>88</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   id keyword  ... target  text_length\n",
              "0   1     NaN  ...      1           69\n",
              "1   4     NaN  ...      1           38\n",
              "2   5     NaN  ...      1          133\n",
              "3   6     NaN  ...      1           65\n",
              "4   7     NaN  ...      1           88\n",
              "\n",
              "[5 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 178
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LWFPWpIZGeoO"
      },
      "source": [
        "# Lets pre-process test data text\n",
        "test_df2=test_df.copy()\n",
        "test_df2['text'] = test_df2['text'].apply(lambda x : pre_process_text_combined(x))"
      ],
      "execution_count": 179,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yb6CQRoyGeoQ",
        "outputId": "3450ea31-be4b-48ba-8acf-d0d7b4833f46",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 359
        }
      },
      "source": [
        "# Text after pre-processing the text column\n",
        "test_df2.head(10)"
      ],
      "execution_count": 180,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>id</th>\n",
              "      <th>keyword</th>\n",
              "      <th>location</th>\n",
              "      <th>text</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>happened terrible car crash</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>heard earthquake different cities stay safe ev...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>3</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>forest fire spot pond geese fleeing across str...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>9</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>apocalypse lighting spokane wildfires</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>11</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>typhoon soudelor kills china taiwan</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>12</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>shakingits earthquake</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>21</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>theyd probably still show life arsenal yesterd...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>22</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>hey</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>27</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>nice hat</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>29</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>fuck</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   id keyword location                                               text\n",
              "0   0     NaN      NaN                        happened terrible car crash\n",
              "1   2     NaN      NaN  heard earthquake different cities stay safe ev...\n",
              "2   3     NaN      NaN  forest fire spot pond geese fleeing across str...\n",
              "3   9     NaN      NaN              apocalypse lighting spokane wildfires\n",
              "4  11     NaN      NaN                typhoon soudelor kills china taiwan\n",
              "5  12     NaN      NaN                              shakingits earthquake\n",
              "6  21     NaN      NaN  theyd probably still show life arsenal yesterd...\n",
              "7  22     NaN      NaN                                                hey\n",
              "8  27     NaN      NaN                                           nice hat\n",
              "9  29     NaN      NaN                                               fuck"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 180
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5bHqNJL_GeoT"
      },
      "source": [
        "# Lets pre-process train data text\n",
        "train_df3=train_df.copy()\n",
        "train_df3['text'] = train_df3['text'].apply(lambda x : pre_process_text(x))"
      ],
      "execution_count": 181,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gtC3LEQBGeoX",
        "outputId": "d4a370f6-8ce3-4473-d032-d2a9cffa278a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "train_df3.head()"
      ],
      "execution_count": 182,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>id</th>\n",
              "      <th>keyword</th>\n",
              "      <th>text</th>\n",
              "      <th>target</th>\n",
              "      <th>text_length</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[deeds, reason, earthquake, may, allah, forgiv...</td>\n",
              "      <td>1</td>\n",
              "      <td>69</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>4</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[forest, fire, near, la, ronge, sask, canada]</td>\n",
              "      <td>1</td>\n",
              "      <td>38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>5</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[residents, asked, shelter, place, notified, o...</td>\n",
              "      <td>1</td>\n",
              "      <td>133</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>6</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[people, receive, wildfires, evacuation, order...</td>\n",
              "      <td>1</td>\n",
              "      <td>65</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>7</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[got, sent, photo, ruby, alaska, smoke, wildfi...</td>\n",
              "      <td>1</td>\n",
              "      <td>88</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   id keyword  ... target  text_length\n",
              "0   1     NaN  ...      1           69\n",
              "1   4     NaN  ...      1           38\n",
              "2   5     NaN  ...      1          133\n",
              "3   6     NaN  ...      1           65\n",
              "4   7     NaN  ...      1           88\n",
              "\n",
              "[5 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 182
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gKx6zZu7Geoa"
      },
      "source": [
        "# Lets pre-process train data text\n",
        "test_df3=test_df.copy()\n",
        "test_df3['text'] = test_df3['text'].apply(lambda x : pre_process_text(x))"
      ],
      "execution_count": 183,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WRR_dxL5Geod",
        "outputId": "79a6246a-cca2-4afa-fdd0-fe0689ce984d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "test_df3.head()"
      ],
      "execution_count": 184,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>id</th>\n",
              "      <th>keyword</th>\n",
              "      <th>location</th>\n",
              "      <th>text</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[happened, terrible, car, crash]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[heard, earthquake, different, cities, stay, s...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>3</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[forest, fire, spot, pond, geese, fleeing, acr...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>9</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[apocalypse, lighting, spokane, wildfires]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>11</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[typhoon, soudelor, kills, china, taiwan]</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   id keyword location                                               text\n",
              "0   0     NaN      NaN                   [happened, terrible, car, crash]\n",
              "1   2     NaN      NaN  [heard, earthquake, different, cities, stay, s...\n",
              "2   3     NaN      NaN  [forest, fire, spot, pond, geese, fleeing, acr...\n",
              "3   9     NaN      NaN         [apocalypse, lighting, spokane, wildfires]\n",
              "4  11     NaN      NaN          [typhoon, soudelor, kills, china, taiwan]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 184
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zewa8pGYGeog",
        "outputId": "4a45bd20-6800-4b9e-8bf1-c706af45ebe5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 397
        }
      },
      "source": [
        "#Lets have a quick look of the tweets in wordcloud\n",
        "from wordcloud import WordCloud\n",
        "fig, ax = plt.subplots(figsize=[10, 6])\n",
        "wordcloud = WordCloud( background_color='red',\n",
        "                        width=600,\n",
        "                        height=400).generate(\" \".join(train_df2.text))\n",
        "ax.imshow(wordcloud)\n",
        "ax.axis('off')\n",
        "ax.set_title('Disaster Tweets',fontsize=40);"
      ],
      "execution_count": 185,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfcAAAF8CAYAAADSAvI8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOy9d5wdV3n4/T0zc+vuvdt716606l2y5CL3grENBptmQgukQfJCQhJIAjEhPyD83gQCgQR4QwyYEoxtcDduki3ZapZk9bIrbe/13r39zpz3j7m7unv37u5daVUQ8/185rN3dk6bmTPnOec5z3mOkFJiYWFhYWFhceWgXOoCWFhYWFhYWMwvlnC3sLCwsLC4wrCEu4WFhYWFxRWGJdwtLCwsLCyuMCzhbmFhYWFhcYVhCXcLCwsLC4srDEu4W1hYWFhYXGFYwt0CIcSDQgiZdNRe6jJZWFhYWJw72qUugIVJQqCemSVYHIgAo0BvIvwRYCfwmpTSfwGLaGExBSHER4D/ucDZ3Cil3HqB87CwuKKwRu6/W2hAFlAOrAHeBXwBeBroE0I8KoS4/hKW7/cGIURtirbjwUtdJguLVIQQW5PqaMulLo/FxcMS7lcOTkxhv1UI8VshxIJLXSALCwsLi0uDpZa/fOkErk35nwC8QC5QDGwArgY2M7mjdiuwTwhxv5TyhYtQVovfX34FbM0wbOq0U12G8XoyLo2FhQVgCffLmbiUsmWWMI8ACCEagE8DfwKoiWs5wBNCiNullK/OlIiU8kHgwfMprMXvJ1LKMWAsk7BCiNS4LRegSBYWFlhq+SsCKWWTlPJTmCP2vqRLTuAXQoiCS1MyCwsLC4tLgSXcryCklK8At2Na1I9ThjUqt7CwsPi9wlLLX2FIKQ8IIf4W+GbSvz8hhPiSlHJgPvMSQijAcmAFpg1AFhAFfEAbcEJK2XwO6RYl0l2IaV+gAsNAF/CGlLL/PMtdDqwFajBtGACCQDdwGjgkpYxME/2CIoTIAq4BKoEizI5aH7BXSnlynvNaDizF7ABmAS1Syp/NZx5XGsKcW1iF+dyKMbVjfUAz8LqUMjYPeVy0OjDHcl22341FGqSU1nEZHEAtIJOOlvNIy4HZGCSn9+kZwj+YErZ2lvSzgS9jGv3JWY4+4CfAxlnSXA/8K+a6/dnS3AXcew7P5Z3A6xmkHwG2AX+aJo2tGcRPPWZ9l5jGkU8B4RnSOQl8BFAyvN+WpLhbk/7/UeBwmvRHLnAdn5TfNGEeTwrTk0Gan0hzH1WzxFmTEv6PMsgnD/gXzA7mdO/HB3wHKDrH5zMvdSBxfa51VAI3XKjvxjou/nHJC2AdiRcxj8I9kd5XU9L77QxhH0wJWztD2AZMq+e5NhzfnCHNm86xMfoZ4MrgWajAj84h/SnCjnkW7oAN+P4c09sK5GZw3y0pcRzAL+dyv/NcxyflN02YT6WEWzZLmr9Icx8fniXOZ1PC188S/p2YmqOM6w2m451Mn8u81gHmSbgzj9+NdVz8w1LLX7m8DHwu6XyTEELIxFd7LgghnMCzmB2RZNqAo8AQZoOQg6lSX4C5fG82Um0/osAxoAPTG58d03HPasCdFO79ifTfP0v6XwI+lPK/ALAfU6UYBTyJPJam5HHBSDzPJ4FbUi75gb2YXggdwGJgSdL164FtQojNUsrgHLL8JnB/4rfEvP+WxO8FZL407ULycsr5zZjanCkkVOQ3prl0M6ZQmo6bk363yRmmjoQQfwx8l8l11AAOYj67MGa9uQrzXYFZ/58VQrxdSvnSDOW4FHVgLlyW341Fhlzq3oV1mAfzP3LPxWyEktOsmybsgynhaqcJ98mUcLuBDbOU4X2YHvT+bYZwt2COdr6FOYq3TRPODfwRMJBSjvtnSDsfU104HtaPqcq1TxNexfQv8G/AmTTXSxPv6tqUMnwz8f90R+U0ef1XShptmB0VLU3YlcD2lPDfn6UOtKTc9/jvn6Qr03TvfR7r+Kwj90S4ZNX3b2YItyI1zcTROUMcG+bSvfGwP5wh7LWYLp/Hw8aBrwMlacJ6MbVlyd9cF1A4yzOZ9zqAOW02Xvd2JoXtmKGO1gLOC/XdWMfFPy55Aawj8SLmWbgn0uxOSXPLNOEeTAlXO024Z5PCDJCBajgprnOGa/lA9hzSWsRkAb9rhrDvTbm3D85TmVPf14NzfDe3p8Q/AOTPEscOPJ8Sb/kM4VtSwkrgq5ewjmcq3H+aFG4EUKcJ9+mkcK9iGneNny+eJk5qpyxtfUg86+TnFwXuyOAeP5KS/v+9xHVg67m0KRfqu7GOi3dYS+GubEZSzgvPM73qpN+vSClT058WKWV4hmtD0nSGkmlaJzFVhuNsnGEnu+qU88fnkM+0ZZ4H/i7pdwh4p5RyaJbyRIE/wBRi4/zFHPI8APzDHMJfKpJV2TmYxpbpSFavP4M5qk13bbo4MHUaYJwHMK3Cx/kHKeVz04SdQEr5EKbXvnE+LoSYTl19KepAplyu341FhljC/comVfi65jHtonlM61z4Tcr5VRnGu9TlRgixFNiS9K//lhl6a5NS9mEakY3z9jlk/U0ppT6H8JeKdPPukxBCqEx+hi8xuVMwnXC/Ken3cSll1zTh/jTp9yCTl5bOxreSfucy1Y30pawD58ol/24s5oYl3K9s5vv9nkj6fa0Q4p55Tn8SwiRLCFGW2IVt4sCcO01m8TTJnEg5/1pCMFxKUo3AfpU21PS8lvS7XAiRqSHck3PM55KQEHKnk/51U5pgGzi71noE2Mdk4X5Dwg/DBIkR9Kakf6U1dhNCeIF1Sf96MjFizpRdQPJ692vShLlUdSBTLsfvxmIOWNbyVzY5Keeh80zvF8C9id8q8GshxOOYBlovzkW1Ph1CiGswjfA2AcvIXNuQN83/X8Kcnx+fkngvsEoI8T3g15mOluaZ1MZ+dIZphXTIlPNapm7KkkrbbCrfy4yXMS34Aa4RQjhT1L3JAn+rlFIXQuzDFPS5mPVhDfBmUrhrMeesk/NIxyYmd4w75/h+wFzlMV7n0sW9FHVgLlyO343FHLCE+5VNbsr5eXl2wxxdPAXclTgXmNvMvguICyHeBHZgGjdtm8ucvBBiCfA94LpzLJs33T+llAEhxCcxOybjy/IWA98AvpHY43o75khoq7w4HsAqU873n2d6+RmEOd93f7F5Cfh44rcTc/fDZGGcrHZ/EUBKaQghtmKuSx8P8+Y0cQzglWnyTn0/f584zpV07+dS1IGMuUy/G4s5YKnlr1CEEHmY7jGTaT+fNKWUBvBuzPnHeMplDXPe+y+BXwN9QognhRCp63fTlXUtZqfgXAU7zFCXpZS/BN6BuRQolVrgg5gdixNCiCNCiL+cwQhqPpjXhhhz6dNsnLdW5SLzMpNHpxOCObE2/Oqkay9N8zt13j35/ICUcniavC/G+7kUdWBOXIbfjcUcsIT7lctGJjuQ8QGt55uolDIqpfwM5nK0L2NaYBtpgtowR/gvCCEeFUJ40qUnhLADP2eyWr0P0xXtOzBV8/mYnuhE8jHHcj+J6Vjnw5iW1dMJu6WJvI8LITbPJY85kGovcL7M6Vn8LpAwGkt2XpMsmK/GHM0DdEkpjyddSxbu1ybq13hnd8004VK5GO/nd6IOXGbfjcUcsNTyVy6pRkg7pZSp83TnjJTyDPBF4IuJhnMTpvXvzZhLl5Ibm3dhbkxyR5qk3oPZURhnK/AOKaVvpvyn6yzMUuYw8GPgx0IIDdPj3TXADZiOdJJHP1XAc0KIjVLKVOOi8yV17tstpTxfe4grkZcwNxACWC+EyJFSjjJZ0E8S0lLKY0KIbszNcNzAZkx/5zcweTAzk3BPfT9/JqX8z7kXf0Z+Z+rAZfTdWMwBa+R+BZJQW34s5d9PX6j8pJTDUspnpZSfl1JuxFwf/K9A8rKr24UQb0sTPXkZj4HpF3xGwZ6g9NxLDFLKuJRyr5Ty36WU92IaDv0Bk6cuvMA/nU8+09CXcn6+/geuVJLn2FVMl6uQZr59hng3p/wF05I92do8lYvxfn4n68Al/m4s5oAl3K9M/oTJjUUEc5OVi4KUsl1K+VngH1Mu3Z0meEPS72NSyrYMs9k0e5DMkVJGpJQPY45Ekh2E3DnNEqDz0YLsTDnPdI3+7xvbmNxBvCmxTC3ZqU26EXjy/25K+QumFmsmf+y7Us4vxPu5WHVg3rR1aROf+3djcZGwhPsVhhBiDfC1lH//QM7zXu4Zkrp5R22aMMnL9TIZsY/zwJxLkwEJq983kv6VDRSkCZq6b7U9TZjpSB1tvmcOcX9vSKjgU63dr8ccxQOckFJ2pomaLNw3CiEWMnnTlemWwI3n24W5EdI4Nwoh5ntkfbHqQHI9nUsdnRNz+G4sLhKWcL+CEELcADzH2d2pwPQv/+ClKA9ThXU6RyDJy+UaUh2PpEMIcT1w2/kUbBYyKfdoynnG0wRSyj2YTlfGebcQwhq9pydZUC9n8g6AaefNE9qfpsSpjcmuiqeNl8L3kn6706RxXlzEOpBcTwsv8Gg6k+/G4iJhCfcrACFEvRDi25ijgeTlb2HgvVLKwXnK5y/naMiWOrpOZ2BzKOl3EebympnK0AA8TIbWwUKI+xKuPjNCCFHC5PnZ3nTr9RPGT8nzjFvm2HAmCwsFeFwIsWIO8RFCNCQ6dFcyqaPs9yX9nklIJ8dLjhNkqko8Hd/H3NVtnD8TQvxVBvEmEEK4hRAfmCHIxagDyd+cjTSucKdJ94J8NxYXkUu9c411mAdTdxlLtz1jHea2j1uA+4B/wTQM0lPiju+mdVuGeT+YErd2mnDj6X4fc+TsmiacC3PHrnBSmgawKk3Yt6XkHQQ+SspOYJgN04cw97eWmE5ZkuM9NE1ZHkrk/SLmlpVlMzyH6zCXXyWn+68zhH84JezjwJ1AY8p7m27L1++lufd/mqWMxZg7jz2VeO9fmyFsS1LaWy+DOp7RrnBp6lI4NW7i3vNmiPeeNHEk8NwcyrsFc/SZHP8ZTEtxMU0cB+b8/rcxPbz1zJLHha4Dt6Sk3wL8IabFex3Tb/l6wb4b67g4h0i8HItLTML15Hy5j3wB+FMpZXOGeT/IZOO3OpnGvaQQIrWyxIFjmCPYYcy50EpMv9ypbmO/Kc318eny38bkTTQAeoA9mPtIF2Ku2x/3uGdgeiF7Iin8j6SUH0mT9kOYa3ST6QSOYy5HimKuo18JVKSEa8XskKSq4MfT3ozpfGc2LUKrlLI2TXw78ChnPf4lcxxTtTyKuaY7D9NDWHlKuH+RUn5umvK1cHZns21SyhtmKecFJbX+yAx9FQghXsFcdpXMXinlhhniFGF2BFPz+Fsp5dczyTeRzoeBHzB1Xfogpo+H8a2HczCXgS1m8hLjXinltFM2F6EOCOAw5jr02bhRSrk1Ee8hLtB3Y3GRuNS9C+swD6aO3Od6hDAbiRvOIe8HU9KqnSbcuZTLwHRZqcyQfwlmJyGT9KLAR9KU56Fp0n7oHMt9EKjO4Nn9JWYnZ6a0WmaIrwD/jLk861zK+bkZ0m5JCrf1Mqjjk8o+h3j/kOa+px2tJsU7kCbeunMo93WYmrRzeT/HMkj/gtWBRPrLMQXubOnckBTngn431nHhj0teAOtIvIjMhLuOqbbrxvRF/ViiUbgT8JxH3g+m5FM7TbhrgK8n8p5NoEUwt2XdnGEZPJhbZQZnSO8xklT7KdcfmibdCuBTwLOYI6BMGqdPAdocnt/SxHPZgbl+OVWN3JJBGgswVbQDs5TPAN7CnJJZPkuaLUnxtl4GdfxchfvmNM/h1gzi/WtKnCFm6GTOkpYzUS8OZVCHeoGfYm6yZJtDHvNeB5LS9gJ/hrkz4BlMjZiRku4NF/O7sY4Le1hqeYtzImFYtxyox5wDdGMK4BFMI579Ukr/OaZ7Heb692zMhq4TeF1O7ws807QVTLXmQkwV6rhxoB9zauGAlPK8XfSeDwk16irMDkMhZqMcxJz2OAUclb9bu7tdcQghSjH9LBRjLvcyMC3F2zGX0J2R59GwXm514Hfhu7GYiiXcLSwsLCwsrjCspXAWFhYWFhZXGJZwt7CwsLCwuMKwhLuFhYWFhcUVhiXcLSwsLCwsrjAs4W5hYWFhYXGFYQl3CwsLCwuLKwxtxqtT3Y1aXGTiNoWB2mx8xS4ibg2pgBozcPpjePvD5HYH0SLGFB+bEoi6VYbLs/AXOYlkaxiKQIsauEajFHQE8PSFJ8UzBHQuz2OoMou8riDlR0fQYkbacsUcKm0r8xgrdJLfHqDq8OxL0AN2O0PZWRT7fDji+qzh5wMJjHpd7FpbhzsUZdWRDrxj4YuS94WkuaaQprpiCofGWHqiG1ckNm9pSyBgdxCyOcgNBVANHb/DhUTgjYSQAp5pXMugO5uP7NuGBF5esJwjpVX8xevPms/c6SaqauSFAqb3JUVlzOEkJxTAbuiENBujziyyoyGyoxGiisqIK4uccBCHHieiavgdLh5bfhW3nzhAzejMOxbHhWA0KwsJ5ASDPLFuLSWjoyzt7CInGESREp/LhRQCbyhEwOHAFY0SttlQDQN7PI56jsuCDU0QyrOhRgwc/jgxt8pgYzZ2f5zcliCGKhhYko0WNsg7HUSNGITzbUgBzpE4QpcYNkEo34bdrxP1qBy/t5S1P2gj7lDRwjpK+s/Q4jJEVwRCgnKxlplP48Z5ZuFucUkZLXax+/5amjYXM1yRRchjQyoCLarjHomS1xVkyStdbHykBXv4rLCUQOuafPbfXUPvQi/DZW5COTYMVcEW0ckajFDS7GPDr86w+LXesxkKQfuKfJ7+6xVUHR7mfZ/dTW5vKG3Z+hZ4ePavVtDdmMNd/3IwI+E+6Mlm94IFLOjvp2ZgkPyxMfwuJ905uRT7feQGgoRtNnwuF36nk+rBQYaz3ChSErI7KBkdxR6PZ7YdXBIdZXn84+fuoaJrmK/8n8evCOG+7epFfP9D17HpzTP87beew9U3P8JdAgNZHrbVLSMnHGB9RzMBu5PdVQuxx2Os6G2neqSfVd0tbFtguisXwNqu0xwvNl2M60JwsLQGv8OFJxKiwjfE/vI68kJjSCHYcuYYLzWswB2NUBTwsWigm53VC9GFyqjTzW1Nb/Fm+QLCmp2WvOJpy5pc5rbCQvYsWEDZyAjrz5whZLNxqKqK1sJCrj9+AgN4Y1EDjlicVW1tvFlby7LOLk6UlVIyOkpjdw/5gcA5Pa9wjkbvKi9jpQ4WPt1HOFej7bp8clqCZPVFiGWptF5fgLc9RFZ/BIch6V6bQ8Sjkd8coPDIGMfuK8M1ECW3NYR9NA5CMNiYTcSrUb5n3EHcxacjJ589VQ3ceWwfDj1Ob3YOu6obuKH5CN6I+R2dLCyjZGyUnHBwIt7hkipyQwEqfENpv9eW3CLiqkLdUN85d6ouV37+3g24g1He9ZsDGYU3BJyuK+LIkjIiThvFfX5WHu4gbzg457YumctWuJ/L6073IDJJJ92o93zynA+iTpXXPrKQve+uASlYsKefwhY/ii7xlbjoXJpL6+oCat8cQKT5OEI5dt68twZ7IE7F0RGKm31oMYO+BR5aVxdw/PpSBquyyOkNU3bS3N9BGJIFu/spbBuja3EOXUtzyekLkaq/MQR0LclhoCYb51iM5S90Znxfp0uK0QyD51eu5NPPPkdXbh4thQX84upNfOGxX3OgtoZd9Quo6+snPxDg1xvWE1NUdEVhVVsb1x87fl7P9XJlNNvJ7rV11LYPUt/Sh3IB27uZkpZAU34prliUdZ2nyY6Gebl+BdUjA8RUleb8Esp8MztHE0DA7uRUYRnOeIzsaBhdUVg00M3WBcvwOVx0egvQDJ3F/V10efPYUbMYZzxGa24R5b4hBrM8rO08w1tlNdOWO/nbi9hsnCgvo3pwANUwsOk6Db29+J1OenJyOFVaQl1/P1FV5VRJCSfLSpFCYSgri4im0dDbi+EVjP6TF71Kw/tlH/YDmXWYpBAEC+20X51HzatDOEfjZPVEKDw+hn0sjhKXZPVGKDgZwDESN8vr1ehb7kExJFpQJ5KtUv/sKGrUIOpR8VU6ab6tiFU/6kDoF6YySCCqahwrqSAnFKRmuB+/w8WZ/GIaBnvIikbIDQU4VlzBbSfewqHH8URCtOQVE7KdwhsJM+J0M5DloXhs1PQRrWq05RWyr6KOFT1tlI6N0JFTAFJSOTqEQNKaV8RbZbXkB/3UDPdPEe7n0mafT7y5xJ8tDYBXrm8ECfcmhLuuKiiGMdGOJseXwOub6vnxA5vorMgjriq4QlEWNvfx/3znZSo6R85ZxlyWwt0AvnDv/ZwuKubzTz/B/ppaHt50DZ5ImI9s38bthw9xoqyMr7/tbtrzCth0+hSffOm3VA1P7iWGNRsnS0v57bKV7K2tozMvn6iqUjg2xsr2Nu7dv4dVbW3Y9cmjwYDdwT+8635OlZTylUd/ScXwED/bdA1bG5fQk5OLIx5jZXsb792zk/Utp7Hr869i7lvg4cz6QuI2lXd8eT+rnulAMUw3zlII4naFgVoPnoEwWmSqzm7B7gHe9cV91L45QPZQBCXRQBiq4MjN5Tz72RUMVWVx6LaKs8IdyO0O0vhaDzseaGDvvTUseq0HLT652gfzHJzeUETMpbLy8Q6yhqMZ31dN/wCbTzXRkZ/PgMdDTFXpzc3hSEUFuqIQVVUWdfdwy+EjhO02DCGo6+9HAl15eef6OC9rJNBWmc/3PnwdDzy6m7q2ART9wulhdxc08m9L3klQdUz6/yeanuOO7jcBgZASRzyGIiWKlKjSIC4VFGlgCEFEsxFTNGKKgiIlYc1OXFWJKSr7y2sZdmVxU/Mh9pfXTTgfVw0DISWKNHj/W9s5VVDGT9ds4b5Dr9PY38XG9lNkxaIEbXa6vXloho6S1OQ+VbGRn9TdhDcW5KsH/oeiyFnvxgv6+vjg9h08fM3VLOjrR9N1Cn3mdVNNKlF1AyVR3oqhEVoL8qnrHyDosOOORJHZgthyG/FFNgxvZk2qoQnar8nD0xWm8NgYGBI1YmAL64iE53Y1amALmY27MCRtW/KRiqBqxxCGTTGfN2Y4NWYAKlpQxxbW8Zc5cI7ELsjAPapqfG/TrTT2dxHS7HgjIZ5avI6q0QH+a9HtfHLHc2RHIzji8Yk47lgUZzx6VlBJycGyGkr8o3giIfZUN9BUUMrpghIWDXRzuKSaN2oXEVNUbjl1EEMovLRwhfk+5NQ6LoG4Q6F/sYdjby+ja00OY0VOFF3iHohQdNJP3asDLHqpz3y+SfEiXo2uVTk03VRM98ocAkUOhCHx9ESo3DvMykc6yG0PTjvF8asfrGVoQRbv+Iu3KGwao31DPgfvq6BvsQfdruDtDlPzxiCb/us0anzmF9Jb4uV/71vHti2NdFTkIgxYv6+V+x97k0WnelENM35PaQ5P3LWKkwtLyB8KUNXvp6fUy961Nfz7J2/m//zjr7HHzk2+XJbCHSBst9OVm8ePr76OUyXmjonHyir491vexpjTxS82biZssxFXFR5buwHVMPjiE4/hTKqIT6xeyz++8z5UwyAvECArHCYbGHG7eWzdBp5euZq/ee4pPrBzx+TekYCQ3U5bfiF7a+v40j3voqWwiMIxP95wCL/TyTMrV7OzfiGffuFZ3v3m7nkX8IamYKgCkGQPRtCietJoTmKLGtOqwgXgCMZZ+2Tb1IsxWPl8B3veXUvrmgIGarORnO1NOgNxFuwe4ODtlTRtLmGgNpvSprONqARGyty0rCtEixqserYdYWTe8tjjcWy6jioNBjzZHKqqYlF3D2/WjZoNnCFxxOPYdZ2wtKEYEkcsRlxRyWyD0N89DCHoKM+jsywXXbnwN+mJBVno78JncxNRbLS7CwlrDmKKhpCSxoEuHl+2kV+uvIbbTx5gy5mjPL78KuzxGDecPorf4WJn9SKG3B6aCsooGhvltbqlBG0OjhZXUjzmY0/VQg6U11EU8AHQmlvESw0axWOjOOJxXqlfht/hYmlvO9UjAzQVlPFyw0rc0Qj3HNuDJxLipfoVOOJxnHFzBB1TVAKaE03qyKQvVgpBU0kxu+vrKR8ZxhmLUeT34wmH0VWFrEiEm48c5ZGNG3HEY9x85ChVQ0Psra2j1DdKXFFxRaPIc7AvFoYkqy9K72ovUhWoEQM1YuAaitJ6QwHOkRiuwSjezhBt1+bjGInhbQ/TsyaHcJ6N4kM+8psDtF+Tx5H3l1N0xE/+qQAlh/xUvj5M56ZcU1UfuAADiOwcigI+tpw+iise43BJFeX+YTa1naIvO5e23AIaB7pnTCMnEiIvNAaY76cvO4drW47jjYTQDJ23ymvx25049Dgjriy6vPlcd+YYQZudmDpV/AQL7Oz5WC37P1CFVAXO0Ri2oA4C/OVO+hd7GKrNouGV/gkBCWDYBK9/sp79H6jGFtLN5z4cw1AFI9UuepZ7OXZXKW/7/GFq3kg/VRBza4S9NoZqszhzXSH7HqhG0WWioybpX5gNSDb+4Myswt3ndfHdP74BW0zHFY4Rtyu8dGMjTfVFfObbL7HqYAeKlHSXeDnZUEx59wj/+rlfUdrj48SiEr7x57ewZ10NR5eUsfpgx6zvMh2XrXAH6Pd4aSou4WOvbaVyZIj/uOk2dixs5D9vvIUbjh/lvr27OVpewXduupUXly7ns889jSNpTnZ9y2luOXqYpV0dNPZ0Uzo6aqqFCgr51fqN7Gho5Ds33crNxw5TNjp16+Gw3c73r7+JIr+fv3r+aZZ1dmAzdDry8vnppmt4o2ERv1q/kQ1nmmno75vXe8/rCpDfHqC/zsPWP2okkGen9sAQeR0B1PNU06lxSW53iLY1pmGcVMQkAV1+dITyYyOcuK6U3ffVcffXDk48U0MVnFlXgK/YRe2+AYpOj2WsNvKGwizs6cUZi7G6pZXy4RFCdjvDWVlcffIUQkoqhofQFbORtcfjrGpto2BsDEMISny+87rvy5WoQ+PQ0gqkcnEWryz1tfPlt35CRLEx4PDyzyvex0lv5cT1/NAYH937MuaI0hSjH9vzEggxYST0B/tfnZTmew69Pun8T3Y+jyLNuAfKaljfeZotZ45iM0whdY6ZXX0AACAASURBVOfx/eiKQDNMY9Bbmw4SFwqqNM9vbj6MnshvtvqlSMnSzi4au3smNA1bTpycEu5PXpp8D43dPZOun4v4VAyofGOYil3DCONsJ7nqtWEqd4wgDLP8ZXtGKX3TN/Gdbfz3M5PCr/pRB1JhYlTZ+IRpC7PoqfltV5LxhoNENBs7axaRFxyjKODjRFE5e6oaGHW6yA0FOJ1fTI8nl0Nl1azqaqU/20unt4BDZdVsbjnJiMtNW24RrliUgqCf3FCAA2W1dOQWUj08wKKBLhzxKFUjgyzraUczDA6V1iAFlPkmD06ibpWdf7yAA++vwjUSo+HFXqr2DpPdF8FQBIEiO4P12ZQdHJ3QRI4j4pLa7QOEvTZKD/soaB7DPRQlblfoW+Jh3x/UMFSXxZ6P1VJ6xIfTFycdhk3h6D1lBAvsLHmmm8q9w7iGYsTcKsM1brL7I6jx2bVqwpBUtw+xZfspyrtHiNo1jiwp4/XN9Tx/61Lqm/vwjkUIue34clys29dG7kgIATSe7GXL9pO01BSwc+OCK1O4xzSNVR1t3HT8CFnRKPft3c32RYsJ22y8f9frLOnuIj8wxgvLVvDawkb6PR7ygmeNYuoG+vnCk49TOOafZLm4uLsLTyhES0ERHXn5vFVVQ9nowbRliGg2/vK3z3DD8aMT/fqlXZ2UjY7wsfJKTpSW05mXT31/37zOv2cPRLj2J6cI5NtpX5HPYHU2hS1jlB8bYfG2bhbsGZjWkn2cuE2hr95D59JcRsrchLx2Yk4V3a7QtjLfHAmPH0l4BsPU7+6nZV0hp64pYbjcTX6XaSyj2xQOvq3SHOG91oPbl7lKPjcYJDdoprO5qRmAYp9vQnMggLr+s1bRznicTc3NGaUdctjYu7qGvatrGM51kx2IsORkD9fvOIFqzP4xRm0qe1fVcGB5JX1FXoSUlPT72XCgheXHOnFE0zf9hhB0l+RwYHklTXXFDOVlYQiBdyxEVecwV+9pprpzqoZlKNfNW8sqOVNdwJnqQvatNOeWn7hjNXvW1E6qr8KAL/6/T+GMpmmQEuGaawt57aoGOiryiWkqeSNBVhzr5OrdTbgiU+MJwGnEyIuOYTOmXjfrupx8PgfDp+R51LqhPip8w2jG2WeoIBPTTGfRUtS0czG0EoA2y3ue6z3MJW9hpPufnP48bfh5L9qMeCMhbmg+gs/pIjsSpsQ/yvqOZvxOFzc1HaYgOIYUCvcdfIOCoB9FGjhjUe469ibeSAhVGjjjMW5uOoQzFsURj7G24zTtuYUsHOyhergfZzxGViSMQKIZBiu6W3HEYwgkRWO+Sd9m21X5nLytBDVqsObhVtb8vB2Hb/KUqa6Zluip0xSKhOqdQ5Qd9OH0xSbZCZUc8WFoClv/ZhG+ChfDNVmUHZo6mAPQbYK+JR7W/aiV1b9oxx7QJ/KXr4FUMntPrlCUDz38Bte/dgp7XEcC177ehJBwYGUVfo8T71gEXRHEVQXPWHjiexDAkuM9uENRmuqLZs9sGmYU7mN2B4+t24AzFqO5qJhPvLqVlsJCTpaWMZidzb1v7sEdjbKvpo7DFZW8c/9ecoMBvnrnO7j+5DHWt5yhyO87Z6GnGAblI8NkRyIAlI2OoOk6VUOD5AcCCMAdjeAJh0AIRt3uSfEFUOyfOtoTwIKBfor8PjryC+j3eKctw8YzzaxrPTPlHqqGBqke7OdQZTUD2R4MIebV6lMANfsHeeAzuzh0eyV73l1L57JcupbkTMyTb/nhSer2DqDok0c3Ehioyeb5zyynfXkecYeCoSpIxVRhCsyPZNq8JSx9uZtd71mAv8jJsRvLuPqnzQigc2ku3YtyyO0KUnloGDV6fi1Smr7FnJDAUF4WP/zANby0ZTEBlx1DEaiG5IXrl/LMLcv5w59un7ZRl8Bwrpt/+fM7OLC8krDThq4q5lypYfCbO1Zx67ZjfPzh7XjGJi8djGoqj7xjHb+6ey3+bCdRm2o+Z8zRpBY3eOSedXzi4de4/eWjk+IeW1TGdz92A0O5WUTtKjFNBaCprojm2sJJZVQMia4pkKYfpRqSF25YypO3rWQgP5u4TcFIjE6fuWU5m95cyF98/2UKhgMXzPhzNnIiIYikX3Xxu4QEUDF7CRLQmRAiUiSuAYz3lcbDjkcejzNbHkriEEn/NMwj9TvHlqYsqeWMT05qotXXoXa4f0JOCqBmZGDSeenYCKVjIxN5Fgd8FAfOtqmOUJz8zjMT5654jNyeton4ACt6zk4PCmBld+uk6wCGAq2bCwjm28lrDbL2p+04xqZ2OmdSh2sxiTY61QhSkVC5bxg1JonbFcLemce03q4wy57sxpEyFZJxB0yCMxzjqj1nsCeW/AqgYHCMpce62LZlIVG7drZ/IsSEPdU4hYNjaDGdobysDDJMz4x3KYUgOxJmZUc7MVWjMy+X3XX1DHg8RDSNtoJCqgcHGMzKpqWwiJDNRi6C/MAYdxx667wbE5uu445GJtLRDB1HPEZuMDgxClCknOj9RdPM4RiY8+cBh4OoqhFXVQwhGHW50RUViTlXNB1LuzpxxGKT7sWcGzYoCARACCI2G1KIeR8VKAZkD0bY/LNm1v26hZa1hey/u5q21fmcWVdId2MOd3zjMKufbp9U6YfL3fzoP69mtMRFQVuA5Y91UL97gLyOAC5fFC1i8OiX13HwbZXT5p3bFaRxWzdvPNDAyetKWP10O+6RKPveWYNUBRVHhiluzqzjJoGoTSOmKdjiBrZYnKhdwxACRzSOVARxRUE1DAxFwR6LoyuCiN2GMxKbsdMUtWs8cs86Hr9zNc5wjGt3NXHN7mZU3eBoYxnbr2rgK5++c8oocZyB/Gy+8Pl3cGhpBRXdw9z54mEam3owhODAiiq2b1rII/esI2rT+Mv/emGScYtqGOSNBtFVhYbTfaw60kFd+yCGEBxbVMprmxbSWZbLj9+zmeqOIZaePKsGXnG0g3/+yq/RFYVAloNvf+JGTtaX8sCju9jy+inUJIM6gTkSSMexRWUcX1iKxx/mgUd3UdU5zKjXxatXL+LAskq2XtNIZdcwH/3Z69jmyUhPArpQ8NnchFU7caEiMExr6lgQpxGbVYAFVQdjmpOoakNijtpd8Qg5seAkI7q5EBcKAw4vEdWGzdApCo9ik/M3V22UKIx8I5fIVXYcu6LkfGEUrdlMP7rRzujXchA+g4L3DxFdbyf4gIvoWjsyT0HpM3C8GsH98yC2wzFEmmJJFfRKldBdTsJ3OonXaiBAOx3H+VwE1+Mh1C797KjUBT27StDa4+T+7Si2w6YwNPIVfF/0ELrXhdplUHxt39k5BwX6XihCSEnO5304dkWnvKvzbbdnSy9d+qFcO75yJ1KFqj1D2NMI9kwwVIi6NWIuFd2mIFWBVCCYa5/QVEp1hoGNAXktAbL6IueUP4AzYtb/dKuYBBBXFXMAARN/zYdytlz2aBxFSiLOc1euzxrzrPA01WgL+3qoGh6kcmiIRb3dvLB0BbqqUDU0mFCLSWy6Pi+u71TDQEvTIGm6PmV5Fkw1KB1zONi5oIHdC+o5VlZBjzcXn8tFVDOXVkU026xlyA0FJxlupMvxQq7SHH/djqBO4/ZeFm3vpWNFHts+tojj15ex4w8aWLS9F8/g2cq49101jJa4yO0KctfX3qJ+V/+UHn/Mqc5YbgGs+00bb95bS3+th/aV+ZQdG+HU5mIcgTg1B4YytpIPOWy8sqERXVEoHvKz+kQ7by6pxu92UtM9RMSu4Xc7GfG4yPUHWXq6h6bqIgJOOzljIVaf6Egr4CXQWZbL43euRtUN3vvrvXz4f9/AETVVeW97+QjX7mzii5+7B0OdWiNjmsKjd6/l2KJSKruG+btvPMPKo50Tz+qWV4+z9EQ33/vIFp68fQU3vH6CTW+eHaWohuSaXU0saOmnrm0QR5La/G0vHWbD/la+8Ll76Cn2cqSxfJJw945F8I6Z86mjHieuoDniKOnzseRUT9p6n46hXDfrD7TyN//xW6o6zxoK3fniYf7+79/JrrW1HF5SQV+Rl4qekRnTypRhezbbi5bxetESmj1ljNrcOIwY5cEhNg6e5Preg9SN9ZKuhkWFyklvJVtLVrA/r54eVx5xoZATC7LI38lNPW+xfugU3tjcRvpRobK3YCE/aLiDLncBt3ft40OnX6Qw6p89cgbE61R8f+chssmOY2cUz9d8E4IdMIWGAnqDRug+F/6/yEaEJVq7juzW0StUgh90E73KTu5nR7C9ObkDJBWzg+D/Gw/RNTbUTh3bCbM+6aUK/s9mE77dYS7R25NQO8fBdiJGvEZDL1QYb82MAoV4nQYS9EKFeK2KLVFWI1dBr1SxHY6hdl8cR1KZEHOrxNzmICunMzTnDoYExkoctG3Kp2NdHgMN2QQKHUSzVAybaZwcdyo4fTMvbxSGxDU6c+d0Niq6RjhTU8ie9XVs2X4KTTcmNIRHF5dhKAr7Vlfh9YU4U2Nq6UJO2ySD4bDThqEoOAPn3smYUbg74jFWdLSTPzY2oWIvHx3hSHkFAYe5jGZ1eytnCosoGxmmcMyPKxrjxuNHz7lAUzk30Rmw2/nPG2/h0XUbiWg2VrW3ctOxwxQGxiY8Uz2y/ipaCmee0zBHUJePkwUBVB4aZvPPmmlZW0B/rYe4Q51k8T5Q68FQBN7+MOXHp66T9Bc5GarMmrWLXtzso2FnH8duLKN9RR7DZW7CHhu53SHqd2VuYzCUk8WBxkqWNncz7HWhGgaDOdmcriykt8Cc3156upuntqzgun2n2Lmijp0r6yjvH6Ev30tt9xAFo+kdjLyxfgGjOW4WNvdy82vHJwT7OBsOtHD1nmaevWXFlLhdJbnsW1lN1KZxx0uHJwl2AHtc5+o9p/ntjUvZt7KaZ25ePkm4A+T4w+T4pzrFEcDVe5ooGvTTWZbHcK570juaLzyBMPc8/xYV3cOT0s4ORrjjpcPsWlvLQH42ox4nFT3TJpMxAw4PP6y/jZdLVqFJnXp/NysjZwirdpo9pfyiZgv78xbwx6eeZflo66QyxYXC7sJGHlpwCy3ZJZSEhlk71IRN6vQ5ctlTsJCDubXc3bGb+9q2kxfLzKmMjmBvwUL+u/52ul35vKP9DT7Qso3cDOPPRrxOxf/XHsI3OXHsjOL9im9ilJyK9Ar8n8rG8XoU9yNB1E4daRPEVtgIfDyL2HIbY5/IIm//iKlWJKF1L1Xwfc5DbIUN11NhXL8KobWa+nS9UiX4Hjehtzvx/7WHnH/wmYLfANuhGNE1dozCs22Aka+gV6tox+PotSqxVXZszWZnKb5YAxXUQQO19/IR7ghTUwzAHFbfjBMotPPy5xbTenUBtpBOxf4RanYO4RyNoUV0wl4bO/9kQWZFOc/Hcu2OJl6+fjE/+uBmWmoKqOwcIWpXObSsnO1XN1Dc7+eRd6/nwKoqmhYUg5ScaigmatdwRMz260xtIWGHxqKT6W0DMmFG4W7XdeoTVuA54bM96etOnZj4XTM4QPXgwKS50zXtredcoPnilcXLeGT9VcQVlT97+be8/eABckJBnDFTzduX7WFr49JZhfulYqAmi6hLo+SUb4plqFQEvhIXMZeGezSKkmK96fKZ61DDWRq+IheupHmoSLbGzvctYLBq9rkcIeGqX57h2I1ldC7NI5KtoWsKFUeGKWwZy/heCkYD6IpCVijKqlMdHGqoIGpTWXesjY7iXCJ2jeqeIQpGx/AGw3QU57GwrY/r9jehGJLs4PQe5Q4uNb2i1bUNpJ1XFhLWvdWWVrh3lOfRU2x2LjbvNYV2arNSMDxG7qhpBNi0oJiYqkxRb0/XFKm6pHBojI6KfOKqato7zPPUTVmvj+qOobTapepO09lMxK4Rs00/9ZQpMaHySPV1vFS6Cm80xB82P8+q4TM4jShxodLjyuN/a7aws3Ax32m8i6/t/x9yY+azk0CvM5f/r+F22t2F3NJzgPvadpAX8aMgCat2DuTV8R+Nd/Nk5VUURUZ5e+eeadXqyXPI+/Ib+I9FdzPsyOb+tte4r20H3lgwbbyM0CXIhNAtUfD9lYfwbU7s+6Lk/P0o2pmZW39l2MD7VR9q99l6op2OY+QrxOtVIlscSJdABM6+s+C9LmJrbNj3RMn+zhja8fiEdlI7raMdjxNbrBFdYyd8uwOtOQ462A7HwAF6mQp2kHHQq1SkTeDcGibwQBaxlTZ4zGy/Y0s00EFtiSNSBoUXYwgzXedWCxtoCS+bYyXOOXeE33pvJU23FJN/JsA1326iYv8IDn8cNWr6FxiqzWL3x+vOt/gZsXFvCzdvPc6zty2jq3wDrlAMXREEshyU9Pn4xA+3c3hpOY+/YzX2qM7mXadprcrnBx+7lrufPkjQZeeJt69kLNvBugNpljNnyDkp9Od7juZC8GZtHX6ni5qBfj64cweu2GR1zKjbTdBhv0Slm532Ffk895nluHwxavYNUtDmxxHUCWdrdKzIo2lzCVIRrHq6Had/shpp6SvdHLq9kr4GL09+fiUrn+/E0x9muNzNsRvLGKj1UHFshNbVBbOWo/TkKLV7B2hbnU/cpqDGzbXtc1mOF1NVWsvyUQ2JOxIlfzTAgcYq8nwBaruHTMcmhsQdimKL6aw81cmOVfU8uWUF5f2j3PvK9G4c+wo9AOSOBnGF008TlPalX0Ln8zoZy3JgKII//b8fmHZePuwwP5OIXSPksmNLuK+VQFxT6C/wsHd1NQeXVtFekYfP4yLospnTDR6XmcgF+ki8vhBZ06jutIQxjxTi7KjoPDjurWRXYSMGCu9u386t3fsnqd4LIz40YyudrgJOZ5fxTPkG3t+6beLWny1fT5u7iGWjrXzgzFaqg5Oni4rDIwzZvfyw/lZeLF3N+sFTVIQG03TYEk51hMK+/Aa+tux+YkLjPa2vcX/ra2TpkfN63CIiEXFzFD76JS/hW504dkbI+bwPtW2WYZ0E1xMhlO7JHUARSwjUEYlRpGAUKCjjBlsCQu92gQH2nTG0k/Ep045Kn4H7f0P4vughusGOXhpG7dCxnYpDBPQaFekSEJXElmuoPTqON6IE3+smtkJDaiDiEF9mQ0TlhMo/maiqsa12Ge25hazuPsPartN0e/IYcbrxOdys72xm0O3hVGEZxWOj1A330Z/lJSccZNCdjScSotuThyIlfoeLhsEeHPEYh0qrccTjLOnvwKGn13i4hqJ4eiMgoX2DaQRsS+OcazqabyxGKuZ8ff0r/VMM7/ylDuRF2ibNEYnz5999mdqWQZ67bRmD+VmoumTT7jO8/5d7WH6kky3bT3HXswfRFYXy7lG+9ckbee62ZTx/67JEu6JS2TnMrS8eO+dyXNZL4c6HuGKa5ajSmNIljagauxc00J43u3C7VGQNR8nrDDJc4ebAXVWmcYgCii6xB+Pk9IZYuKOXax5uwhGa3OAs2t7LtT86xYG7qulamkfr2kKEYcbL7whw67eOkNMX4sff3jxjGQTg9MdY8UInrWsL0O0qBU2jLNjdP6d72bphER95cif2WJwRj5s1x9t58HtPoejGJNuMT/1y28TvZc1dxDQVLT6z/UbUblZhTTemdaZjT7eEDIir6oRBS3G/H2WWpVQl/X6S7SzCThu/uX0VP3/3BobysigcHCPHH8LrD5E/HEDVDY4tKj0r4C8Amp7eLmW+kUCTp5w+Zy4OI8ZtKYIdzPqyyNdB/Vg37VlFvFG0hHe378BhxNGFwrbiFQgkq4fPUBaa6khElQY39B7kkZprOemtpNudT0VocEpZFCSaNHi9cAnfXfR2dKFwX9trvLf1NZzG+fvYF2GJzBaMfsFL+E4n9j0xvF/yo80m2AEk2A+kn7NVAhIRTtQf59kQRpGCXq4i/BK1LT6tWti2LwpCoFepGPkKWoeO8Em01jjxahXpFggBsRU21A4d7VQctUvHyFfRq1TUdp3YQg0RlWgnp34TEkFU1XDEo+SGzCmNF+tXokhJ40AnAsmwK4sRZxY/XbmFr/72YfaVL2B5bxu7KxeycLCbbbXLcMRj6IpClyefa1qP0Zudx77yBWQfCFM/1JP22ai6pGrPEKduKWak2s3hd1Ww7Ndd2FPaNgnEnQpaePJmWYYmQKa3Zo9ka5y4o5S4/eJIdwFkB6K8/5E93P/YXsayHDiiOs7w2XqhGjoNp88u+/3ED3cAcKqhBEMRlPb6+OiPXyf7Qs25/y6zqLcHZzxGrzeHR9dv5NpTJ3BHowxlZbGrrp4nV68jbLMhLqP59GTqd/WR0xukpyEHX4mLSFZiV7eYQfZgmNKTo5Se9E3aMGYcIeH6/zaXyXUtySWYa0eNSzwDYSoPDVHc7CfmVLnuf05R0BGYcY5L1SVFZ/xkjUTxFbtY80Q7amxuz2zD4RaOLijDEILGlh5c4eisBpcCJpaRzMR45Q/bbcQ1FUcaV40hV3rDSUckhj0aJ2ZT+fT3XiJ/eOapBntcxx08qx3YsbGe//roFmwxnbueP8j1b5yiumMIrz+EMxJH1Q3+6N8+yKGl069K+F0hLlQGHF7Cio2aQB950fTPSkVSHeg3d5Kzueh35FAZGmRMc9HnzMWpxygOj2BPo24XgEuPUBEc5HhOFe3uQtYMNU9Z/46Eo94qHqq/hT5nLu9tfZX3tb6KI816/XNCEQQ+lEVks90UGMMGwp95B0oZyiBskmQy8hQQIKISZWyGb7HfTNfIUpAJz8EiINHOxIkts5kjdwGxRTZcvwkhRiW2/THCdziJ12sQM/MSQYl2euqzcuoxigKjuGMR6oZ7zfcRi7Kqp4VFg90ENTuDbg9Z0QiaoU/SBo2vVMqOhlg42E1U1Rh2ZdOZU4BNj+OMR2dtaWt3DFK7Y5Djd5ay++N1jBU7qNg/gmsoiqEJIl4bo+VO4k6VdT9unSTIy94aZWBhNl2rczl9fRElR30IXeIrd9J8YzHt6/MuiYZZ0yW5vtk3qioe8PPX33iB9so84ppCWc8oHn+EoN1hevU0dOKKQl92DsX+0anfRLq85+MGLkduPXKIl5Ys4/X6hXz75tv4zZp12ONxxhxOBrOzufXIYcpGR3hl8dJLWs6A6uDR8qtpGOvm6uGzm6KocUlJk5+SpsnWvgM2D49UXMtdPbvTCvaJ+Lqkbt8gdfumjnwA1ECcW787u8rHUMBX5CLkteMaibDs5a4M7+wsZYM+ygYvjHe5qs4h3lpexUBBNgG3naw0S8bay9P7pC8YDpDrCyVU89B4em7ewJ66dQVhp52N+07ykV+8QfGAf8qqhBGve7rok7gcp7aSiSkqIc0OQpAzy3x2diyEgiQuVIKaKYXGbC6kwGzo9elXWahSkhU3G0OfzU06/3SjdjcP1d9CS1YJAH2uXHQxf6MywyuIXGPHsTVCfLmN6LV2gh90k/3dAEpg9o6tmMU16dQMxyMyc0VINjRIZCGCEvW0TvhmJ4ZXQZYLpEugteuIoMS+L0boPhfxBRoiJJFOgXY8juJPX8bCoJ/Xa5aAgJtOH0aRxsQAKKZqnM4rRVcUBBKbEUeVBi80rKItp5Al/R2m/3w5rt0SdHnyOVFYQVTVZq3jrpEY13yrCWFIjr29jL0fqeHo3WU4xuIYiiDmVgnn2sg/HWTtw22TBiVrft5O21X59C32sPVvF5HdG0EYEMqzEXcqrH+olYP3V866xv1SYo/p1J85O5qXwIHyWvZU1bOp9RQjLjcBu4O8YICbmo/Mml5Gd6oLMfHxCMz15he6MVJ13dxjOdmrk5TY4nE0XWe8dgtpeqayxeOTvHqV+Eb5+iM/5ydXX8uLS5bTVFyCIxansbebD+94lVuOHuaZlavZU7sARRpTepWabqaZbnOD1DCKPKuqnShrhvcZEypHPNW49CjMvmsqYdXOIW8NWwYOA1MFtwReLVjGCl8L+edpLSyBqEuj6epiYg6FpS/24ukPm4ZGwvQUJZBmW5PoxQspkcL0niExDcgUJLpQEAlXoiJxrkh51sI34XhFCnP5pZEIP34+3fPcuL+Vp25byfGFpXSX5FA0ONkdrq4IXtu8MG3curYBqjuH6CjP49d3rmH9gVbssal1e/z+REo5BgrM+f6CoQBef3iKYD/SWE5/Qfasz1lIiT1mjqRGvW4MRZybL9QLiCKl6VtCyhn9QoC5dzsk5sYT/ihsCXW5kdSWpGN8DT2YHYF0Zl4BzYlfc3Fzz1vsy69nR+FSiutG+FjzC2hyHtomVeD9ug/XL0NEr7Ix8vVcAh/LQjuj43osNO+e5NQBHQyQDoHhFdMak+ll5nNV/MaEel+E5Fmr+moVvVJFGTZQW815e/u+KNIliNeoiIAp3O0Hp+9c1Yz088G3tiWePdxx6gD2xG9PJMQ9x/cQUxTuOrEXdzTC9WeOELA5EInriwa6sBmm739dCBQk67qaUQzDdGY0CzmdIW766nEan+/h2NvL6FmRw2iFCy2kkz0QoeaNQRY/1zvFyLjwhJ97PvMW+z9QRdumAvoWe3COxig/MMrKRzuoeHOE4Ro3zdcXTaupVGLmxj2paZ8Pc00p9b2HbHbWdpwmYHewt6qeD+/Zxuu1jRmllZFw31a9jB+vvInu7DwKgz6+8spPqPENzB4xDRLw2V1kx8LTOicRwFd++0t4BtSgMRFvSbiL17/7JdShs73JnFCQrz3yc776q19MFgIa5NvG+Mzzz/CZ3z47yevSeM/yPXt2ce3xU+Z6UKEwZncx5MymdrSPb//0RxOCJR2ecJjvPPwQUsCII5uoYsNlRAj//+y9d3hc133m/7l1+gxm0DFoJAAS7GAVSYmSSEq2pLjbimvi2E6cvnGS3ewm8Tq7KZv493McJ48TpziOW1zkIsuS1XslKRaRYAFJ9N5mBtNnbt0/LggSBEEAJKiS7Ps8fIiZOffec9v5nvMt7yspTqzaXB597csRLcT48ol/mjeckJNcjm4kNQAAIABJREFU3BfdQ00hft3GHWCsOcjJO2pwZQ3aHh5ALpqcD0V5JrqRTbEeNk12c7K0ka5gNQAb4r08XbOJ2uwE50pqWR/vY2VqhIfrtxPNxogU0zSkx3m+ej0tyWFysouyQornajawe/Q0vYFK9g0d50crd7MqOUx3sJpf6njCmfxcAduO9dDUM0HXinK+8cFd/Ma/PUt5LAu2TcGt8PStrZxYU3tFgqHwVI67njzJ2ZWVHNzSyN/82h184MGjlKRyiKbtrBYUiamQl8NtDdz57GmqJi56UhoGYnSurKC7sYzOFeWOmptlo6kyfbUR/vbT+xc1y1N0i/rBOIc2N/Lc7hb2HDhP5UQKwXYmJ5YkEnkDGeYAVMsgqOcQbYsxdwmGIF0xk/1CVrwpSA697fQzWKJl8ZgaeUklqfiwYE5oxgZ0UWbCHQLbpqKYvOL7Fymm+ezJ77E6NcR99bfw7ZX7eCS6nfJikncMvopqX6d73gb5pI44aeF+qkjg7zKkP+Mn+WdBpCET9YB2RZ6Na4WQsFHO6GhbVYxmGVSuyEZYvNkJE8i9JlLMGRcFG6RRCzFmYTTJ6GtkpJiF3DtN8jVqIg2ZmFWSw1rnElCPzT82ybZF2SUMdD79YsxXxCZYvOi1ueC29+jazOfLE+bs6TYXfl/wWgCulMHKZydZ+dw0W970hjPX/ArXXrCh4kyat3/u9MVt7OlNp9vv/auz3P5XZ+e9dx/8xGFsgWW5t5YAUyEvibCXRNhH1qtc0Qt1KXYf7J6jAFeVmuKn67fh0TRSLi9dpZWLDiUvyrjvGThN21gPP2vexiPNW7keJ6IpiPzTlrv45dceJ1KY3/jk9qtIMQv3CR0x5dyl7D4Xtk+g5Os5J7GiVAQRpISFaNjYCuhlIlLaxvQLpO/1Ev5KBtsFYsGelahiCgITnhD/uPFudo50cOtgjimXj2+s288fHvwBblNHFyQSbi9+rYDL1DEFkaTLi9dwZA8tQSClenmmfgOtsUFWx4foiNRSUszSmBqnKCkOK54gEdKcgTHpcli9vEYRv3ZxJqsJMmf8tYy6SlBsk7r8JNF8DHnaZ5eUvZwO1JGR3ci2xcZkD6X6xbhnUZA5549yOlDHhBriYHg1vd4KADYleyjXFmaTs4F0mZuiX8YWIFXp4alfW4Pultjw2BDVZ5MINpQVUlTn4kypPkxBpM9fwc6xDqZcfuIuP5YgsGWiC1OQkE2DjpI63NPXzKcXOFtSy47xs1TlErxW1kRGcdM22cWaxABxV4CuYBWCDZooU5ONUZSUKxp3AYcM5hPfe5kvf2ovL+9ooq+ulPUdw0imRX80zHB1Cfc82c5j+9Zdcfu9L55jsjTA99+zjQfubuOlm5ppGIjhzWsUXTITpQGGqkowZJFbDnbO2v6dj5/g2IY6TqyN8oXfvJP1Z4bxFHVGK0KcbK0hOjLFbS+f47F966963V2awc2Hunh+Zwvnmir54z9+D2vPDqPoJnmPSlGR+cs/v3/ZGOauBSI2jdlxIlqGKdXHyZIGNie657SbUnx0+Z38iktj84plsDnexfOVG+gKVJNSfFesQ+/1VTLqDhPRMtTlJq/IVqfYJqXFFCI2Hxh4iYQrwE/qdvHjut1UFJLsnOxYVExyIQgABfD8KI+xQib3AQ/JPw9S8rtTqCeWKb4/De9382ibVbRdLrSNBdQjs/nRjWaJ/Hs8CDkL5bCGOHHx/MRxC2nCwqiX0NcpyN0m0uA07akFygkds0bCKnWScpVTCxC5LOG3hcaUa7EUM9tcZpyvdZtkpRt3WseVm+vVMRSB8eYg3imNkpH8shh2Gzi2qZ4fvncLr22qI+tzLepCPPD+v0dNXrQJAtA6PkTkUBrJtnAZBu3V9ezpXlwG/aKMu2JZhAtZIoXF1zbPh96SCl6oX8svtD9z1Xa2KlDcoGBUy3gPFJH7TNyv6eR3qjM9L2xTMYMCareB+7hO8iNe5HELpc+YSWoprlEwwwLel2YbB02U6QlVIlsmdelJfHqBoJYnpOUQsCmKMscrVpBWvSRdXu7uPsLZSJRxbwkZ1c1dPUfpC5bTVVJNV6ialvgwmiQz6gvPrNqPVaxk2B9BtG02j3cT0PI82bAJxTKpS02yZdwRRRGweTXcQq+3AsU2mVKcGvRP9D9Jc3YEgemEJjVIR6CWV8KtfPbc9ylNXrwfuijR7ymny1dFQVLo95STkp0s7ebsKOXawjFvSxJof3uUU3fUAA6NbbrCQ3VHkpvu68E/zYJXkFRcps6ZcD1bJzppSo1wqGI1piDQFutBsi1E7JkZZsvUEAmXj4CeJ5qNUZON81L1Wgb85XgMDa9eJK9cyBCCmlyc/kAFbkMjrGUIavPHeGXT4uZDXciGxSP719G+ppan9rTizWu0dI/z6W++wI6jvZxcGyXjdc3d3rK496dHqB+M8+JNzZxYF+XMqmpyHgVVN4kksrSdHGDj6SHKYrOf/y3H+/ntrz7Do/vX09FcyfmVlbiLOlVjKfa90ME7Hz/BSEVoQeMu2jYbTw3yG//2LI/csZ6zzVU8cdtaZNMimM7T2H/lvInXG2uS/TSnRzhU2sJ9DXuoLExRk4/P/J4XFZ6qbqPHX4lq6tw1fHhmTBOAe4Ze5dXSFo5EmjlQ1sodo8dmGeFRd5jvrLgdSxDZFjtPVT6x4Jgo2RYf6X2WjOzmsZqtfGPlfsqKKVanBpfN0yFNWvi/ksGsESnucZH+/QChP72Moe464X6igPsxF4W3u0n/QQDPj/IoZ5wJhLFKJvd+D2athPqKhueRAsIlcwtpwkQcN9E3qVgVItJTxYuJeSYox3W0nSq2KCCNmIiTb9wk8UZD80gkqz1YkkBoNA8WTDQFqDqbwpUzMRSRXImK5pMJjeQQLBhvCeLK6pSMLI/+wUBdmL/9rX30NpbhLmjUD8QcxbcFOC4ur3qxgeFQmOdWrsWnOWPvO84cWbS8+LJkF9iAhcDzDet4pGkLSbeP5vgI9555iYakU8vaXVLJj1p3caxqJZPeIL9/5ydnYjlffOJrhAuzY6WCaaN2GiAKaI0S8qA556BmyJkACAKIaRvLJ+B5pYigOUkxep1E5h1uwv+UmVNeopoGtZkYlbkkzYmROe6/uCfAhDfEpvEeXoqu4UR5I+O+EjZM9HK0sonTpXWM+sLUp8aJu/2I2HgMDZepk1KdJKq4O0Ck4MiVTnqCmIJIzBOkJh2fdvlfCDkIaKLMPeNHaMyNMeoK87X6O3gttJKG3DiqbRLR09w1fpSVuTFeLZkbQ/aaRfZNnqBUT9MebOSdo4dYmXMoyVxLCBEYLomxpiCa1yHIaXuon53f66a6w1m120CJlmFdvJ/18T7CxTQhLUtF3qE2Lc8niRRShIsZ9g4dR7ItfHqRsOZch3Axg1cv4jI1bEEgqOWma5YlvEaBPcMnCeh5/Hp+xsNxqRvKxpGpHW0IEZgqEIzlKQQUbn61k+aBccYqAki6zVhjkIqRFGuPjyBaNr/z9acZjwYoT2ZIl7iQDBvBsrFkAXdGoy4Z5+6DJ3nfI0fJe1QO3bkC2TBZe3iEhs4YkmAxvjKI0pnAlTdmsvnvfO4MW070kwx60BWJkUAYd0HHny5SOpajdDLL7/7V01RPJNFFiUl/gPJ0ClsQyLjchAo5ZMvCl9fY/0IHm9sHnH3JEqJto+gm3rw258V/5+Mn2HG0h0C2SGn8yh6w+sE4X/3MN1F1k+jI1KwFkCbKZGU3MVeAouhMmONqgDFPmKCew2NoyPZF0QuAiJbhg33PcT5QzdFIM59f+wF2TXZQm5skL7k4EmniUNlqUoqXD/U9z6ZEz6z+rE/28a7Bg/yw/ha+2vy2mdW/amoM+Mp5sXwd54JR6rPj3DV8eFFENAIQ1HN8uO85EqqfA2Wt/N3qd/Fnx79FREsvm4GX+0yC/ztF/OsRire4yHzCR+Cv04iJhSVpF4IAkLQJfCENIhTudKO3KohJRynGDohYJQLKUZ3Qn6WQ+mYPZuKkhTRmUdwjIRRtlNOXvO8mqO066RLHyym/YvBSWxPlExlW9YzN259ryR+6HFcyZZfmBBqSyJc+sJ/W/lFuae+kJJPnoV0buPvgKWRrbh7U5TAkkZTXjT9fRDVMLFmg56Yy8iGVwQ1h9v59B4JlM94SIDScxx8rEq/3cey9dVSfSeKbLODOGIRGchT9C1ORLxZP7FtDf12E6FCCT37jJVadH5+3HPdSeLNzvZN94XIaExM0Tzpj+VLKXpfFuFuCwI9ad3Hf2lu4o+c45dkUL9av4U9u/TCff/obVGaTRPJp7ux5DY+uMe4N8QvtzxLJO/FLnzZPqYDthEoF0+Fetj0CtgtsGXK3ukAW8D1XxPI5NY7OEhcwp5O5ck5piV4nISZmE0NI2LgMHcU08BjOrKgoyU5WsOTEtkTbRphO7pItc7oO2sZCQLIsJ5N0Wj/6wnWwBQFNUmYSxsKFDHlZdfTIc1MkVR+t2iCt8dkavZuSPaxJD+C2dFTLoD4/yZjLiW2q00lCim2izlPHKwIeS8NtaojYuC0dn7m0GknRtLn5m+fZ+d0upzTHBtGwkfTZNaUeo0i1oc1cR2yLqmx85nN5PomATbh4caVbmnW+U3BW9TU5p/3lA0dYcwxVRf4i7eKlbQxF5MSeWlIRN4MtEe75Wjs968uo7pmiZ305jacmCSQKFGsUTu+PUjs8hS3A+T0VtBwbR7Jtzm2pxJcqohZMskEXza+NM9wcZqrMQ+lIltZDI/RmywhN5FnRN4limTz+C2up74hzbksld3z3DMq0BKxsWlROpqmcTGOIIj/bt51AMU8s6md9aIDbzp/GEzN5tbKZQLJIVnVxpipKeSZFzBfglq6LFRKKcXFfC6E8lqE8dnVPmqdosP7syMxnG+j2V/H7W34ZXZRnEgU10RkGftBwCz+u3z2TBLkqNcQfnP4hVYWpmfuwYaqXPz3xbb6w5v2cD9Rw9hIdeAEbj6nx8e6nuLf/hZnJwYVtPabGh/qeR7ItHqnZxlNVm3iyqm2mjWhbNKVH+M1zD7F+qm/RAjICEM3F+MWep0mofs4Fo3xxzXv5g9M/IKgvnqdcsGzEhIU4YSJcIeYtd5mEf3eKxJdKKN7hQj2m4/lJ3lFk051tbd9VkiF1h73OdgsXleMuHNsGudOk5PeTFO4okH+PxylfA5Tjzmrd/dMCQnbuZEIwQD5rII2aCAUb9cTsMUIcMZHPGFhVIq5XNWJBH5NBP2mvi9buUVyaQWdDBXm3wqqeMbwFndPNVUwFPLR2j1E2dW25Ow9V7+Cfm+7CRmDzVBe/c+6BGa7/gipzeHUjgxVhbmnvJJAr0l1dhq9QnFnhjpSGiAV9TPm97DrVTdLn4XRjNb5CkY1dQ3TUV/P8pmZWjMTYf+QMkmijuWXUrIFkWLhTusMLkjWxFOeqWZJAeCjPuseGEZda1bBInFpbg6qZvOuh4+x/5uycRNylIJLL8NDarWRUN5JtUZY9hWi9jiv3gWAZT63YxEfbn+Nd5w8h2Tabxnv4k9s+zDONG/nQqRcoKeZoG+tl2B9BNQ3aRrupzs4vZCHFbLRVjviB/2EdKyKS36FglkpoTTLygElhi4IVEHCd0lE7DfI3q6Q+4nU+nzXwvKbhfaFI9hYXSpeJdFkNqWrqRDMOA1ZRkmkvX4Fg27xSs4Y9g6cozyV5oXY9km3RNtHDGVPnlZo1AKyL9eE1ChysXu3EhPUiw/4IfcFKZ5CaGqG0kCKg5VFNHdG2GfOVYIoCXeEq/HqeLWMXtcpDRm7GcF9g4NJEaVmYxRYLAZANe4bZ7EqwgU4hwpTgwRAENppj2AicFUsBgVXWJAVBJiF48No6MhZuDB6UVxGyi9xkDlJpL5wcNt/vBZ/CSGMIWxDwpjXyfoXSkQx9a0pJh91ERrMkyz0Yskjep2LKIpPVfppfG2fLsw6VoyUKWKKIKduYskAupGJKAgWvgu6S8CeLlA2lqRxIUz6c4dTOanSXTKrUg1I0KXjlGeN+ORTTYFtfF4PhUjRJZsrrwxAlsi43miSztb+bL9/2djYN9RHOZRftYlsuqJZBS3px5Yw1+dgsDXZwJpGtqUH+5ui/8EL5Ok6H6ki4AqjoVHni3H6ynebMyBVj3gIQ0nN8susJdk52cKCilXPNNUgxm9J0irXJAW4ba593UlpRSLI50YXPKKBeVtMuTPfr491P8VDtjhla2tvH2me100PTK9gpay4L3JRN2b1x5oMAqId1Km+ZS+KkHtEpf8/VwyeuQxrl75q/jQAIaRvv/QW89y9cG30p/F/N4v/qXCMsAEqXScXdTgK0Dehvk5iM+JkKeKbfeZOf7t9IJJljMuxn52s9fPcd27nj5Y7roi02RImc5AJBoCgqs8YywQa3rhPMFqhMpBFtm2CuwCM3beDm9i4kw+TpLa3kVYX1PcPYwJHV9XTXlLOuZxgEZ+UuWTalqYwjcKZZqDmDnp1ltDw/BgIUggq5sEqyykOkL+uMrZqJZDiVOppXIlXpQfPKFL0S6hXi8ktFzqOiGAarzo/Nm5C9WJRl09zS48TYDVFiKfn3y2LcR31hRvxhzpZF+YZnHwB5WSWtejgXqVny/gTA90wR7zPFmRUkQPirF910NhD5YmZWWUro6zkQL7IUBX7ivCDBB678opRoOe7uPQKA2zTYPnqe7aPnZ37fNtbJ5vGumSz89bF+1sYGEKbzHlcnhmlJDM8qT/3kySdmtq/Jzh4onmhoY9uok5RliI707AU4ceo3PywEOqQy3LZBQvBQLuYYE/wclaqdMiZTYI05yRG5GpdtstscQLFN3LZJpZ0hYC9OSW4+qAWDmp4kmlvCP1WkuifJVIWX47fWEe1MoGgmfa2l5IIufMkigmVTMpHj+K11nLg5SkNHjFCswPm2ClwFg0C8wNDKEsbrgoQm80i68/D4p4p0bSgnEC9QNpihdCRDTfcUgUQBf/LqHpELHh8EgdNVtYwFgoRzGSTLRDUNGuKTnKuo5oNHXrmua7FUCEBdbpL//9jXZn1vyVAsd+RF3WMGpltE0G1sCQRhWrGsRMJyCSgpC0uBQCrH3WOv8vbkEaS8RbZRpRBUKEtn0QMiVt55D02PiFi0sFQRrUzCNWogaTarrUEarVEmav1UH03hHpttrI2A016dMJEyFpZbYIfZwdb4OdxDBrYEhRoZKWMhpy1Mr4CowbbCOXacOYelgOUR0Sol1OlSM71MIrnDjZSzKXkhNytu/Z8JLs1g28k+itMyzL3RCBXxDPXDccrjadxFndsPniMZ8DAV8FA9sfwcFW7doHY8QdlUhhUjjjZJeTKDr3hxfPAUNTZ2DbKxewgbWN0/RkFV6K+MsPtUF2XJNDWTU7QMjqOYFgWfTD6kUtaTYXhtCaGRPLJmERwrYCoihkvCF9eo6rjoFdS8MpJu4croaB4ZNXf9k+1IIsdgbRhbvL5pggBUZpJUZpz+HqtpXNL2y2LcdUnGFESKkkJGdc98f0fPcVbFl056cgFXy5QU4CL5w1W+u15cXq53uatwKQZ5+8g5JrwhRNumLJ9Cscyl7WARUG0D0bbISnOTx5YLbtug0ZpCFS00JJKCi1Zzgjo7RcTO48LARCQjSHhsHS86IbtAqZXHx/WVCKoFk5ajY8SrfQiWY0RLxnNsf6yHQKKAopmsPjzKVIWXphPjjsE2LNa9MgQCKEWT+o4Ykmkh6hbBeAFXwcCb0VGKjrEHaGqfwJvRUAsGwVie9S8PYagS3rQ27zMpWhbvbD9CWTZNTTKBJQiohkFNMoFkmdQlYgi2TfPEKHlFpTS7PHKk1wMbyK50kbjJi6dPQ0mYJLZ5cA/r6CUSctbCM6jT82ulBNsLhA/nyDa7UKZMpLxFsUym5GieYqVMrk4FsmRbXFgy+Lo0pto8BE8VmNgfwDVuMLVVoOyZDKPvCKLGTHIr5z6ntgD5epliVGb8fQEa/jpO8iY3uWYV16iJOpEl3eaiGJXRIhKVP8yQ2ejC06uTb5RxjZjkmhW0ShlbhNKnctjYxN/mx/QKeLpvTJnqWwUCzNJR2NQxyEN7N5D2uVjVUyDjdTEV9DIR8RNO5ljTfeXY/HKiqMgca66jq6acx7evZd/RsxcnyTir9FjIcdGfWlHDzz9zGF9BIxby8+CujXzomcNYsuDQ0AKWIkzTdBcID83O3biQGCwAgcki6x+7dht1JbQd7+fI5nra19Wy9Wj/NXsCnmpeT1k2zQsrWxEtm1PVdXz+oW+jaIsLty6LcffoRVyGzl1dR9k43jvrN+kyvm5nRYOzJPhPhhItR8lVMr/ngw1kJTdx1U+fpwJDkBjwlFGqpSnRMwSM/MwcoSYfI6xn+E7trcTH/NiCwIZULxWXZMubguN7kGxrwQfvQlzW5mLNv4QTk72QEb/RHOVBuZURAuw3uukUI4StAiXk6RIjbLJGKbXzPCo3g9FJiz2/63MhCEAwUSCQKFzynU1t18UQT3g8R3g8N9MeoKFj9jFXnpiY+d0Gmo6Pz2rvTWk0Hb/YJtp5cf/zXTMRaJ50BsKS/EU1tLJpIy4AL61YxdH6Fdx79MCbxlMjYJNrUPF1FhFM0CISStJEL3GW7rbkJLZWPJVBLDplp/2/ECH8qnOOombj69JI3ORUefjOF+n67TJqY1NoFTL5jMrUFg9q3MBSBeS0hZSzCb+aI7PqypNQWxEoRhWS29yOCzYg4u3SCR3MY7kFCg0KoQN5UlvdZDeo6KUi6piAHpGRUhamT8R7XkMPSWjlEvl6Bd/pIpZb+E859lyKfa90IFvWNBGVjWxY/OJPDmIJAoFsAcm02HegAxsBf+7auc0XQsVUmk88+jIAim6wsXuQv/77H+Auang0nbcfOjWTQCabFqv7x6idmOIdr5xANi1KMnne+/wxDFlEMQzUFKx7bAhLdt4sd/qNmcTd/dgpnti/lif3tdLSOcaOw73IxrR0+FW86pd6gAF29HdyqqqO27tOU51K8DNtyxx7ejUsaNwvMEaZgog+7UrWJBlNlJAth0ymITXBqvgwjzRvpTYVw6/lMUWRnOKmLJfi0gyTsnwaSxDoDVXg0wrYgkCgmF908sx/NEhYVBUThC6p+RWxKdVS6KKTMa0JMk+Ut/Fo5RYAyrQUD1du4+HKbeyKn+WDQy/gsRx3VsjI8dvdD/GD6C3cX7OLoJGnPj8xy7g/uH4rR+pW8psvPE5Vev68B4CiLPPEqo2kPF7uOvMakVyGfaaTBd1kXqTU+1X9sDNhwCJqp1nL7LjkVmuYTdYo0jK5VpZSi7sc7a/VHFy+3e6ec+zuOXeNe7sxcA/qRO+bYuCjJfi6NKSCjVYqU6yQkXOOp0LKWcjThFJKwsQWINeoUvZcBltyjK/hFTD8InLGQkmaTN7mI/xqHmwIH85R9lwGqWBTLJMpbpXRgxKWOvfKFqtlMhtdhA7lSbc5NcKiAVLWQspPhws0GyMkYfpEpB4dISSil0loFRLuAQHBtJEyFqZHxBZAzlkYQRFbERAL/znHGnCeR29xrtELJ2cvOiJTsyfHNwKS5cTZL+3XpX279G/ndw1v8SIhjmDbhLKz68JdOZM3mt6x4FH41Ndf5G/+y5382R/+HFte66ftxABlk1mUK2hfXMDOQ7NJbPxage0DXaRdbixBYE/3mZkKs8VgQeOeU1wcqFlFb0klp8rriHsC/Lh1J5XZJDuHzrI6NkRZLsW9Z17imxv38tc7300kn8EQJRIeP7938AHqkxcH+tbYEGsnB/ha235WJsZQLJNfO/IIfv3GzRDfzAgYBX6366ezvnNbOh8cfnHWd+8dPcB7Rw8suD8BaMmO8EfnfjBvm6O1K/nRxpv46OEXFzTuhihxqKGZVxpX0TbUS2lufsGQhZI9lOWOmbwF8aZbMwpQqFZIr3MTOlFAzloEzhSY3ON3aHEnTUTNInjqEk+JCRVPZ8jXKniGdAyPSKbVjTplkqtX8J8tUvVwmom9PgIdRbBsco0K8d0+XKM64UN5Mi0qqY0efD0aUn72cyEnTdQxk+xqFf9xRwbUNWQgZZx2UtbGf7JIus2NmLfxn9IQ8zbprW6kvIUSNx2q6piJqNuIORtvp8bkz/kRDHD3FZedQvatjjfdc3kFLHXS/kbh735zH0c215PxOyHql3Y389Lu5gW3u5zEBqA3XMbxaCMuwzHq0WQc6QrMkFfCwm55G1yWQaiYZfdgB7sHL5buXJDIFIAto91U5FKcLY2SVt2opkFFNkllZrbx8Gt5fu/ATzlZXk9eVvHpRSf2/DrAxsmUjpf4mAp5SfndpPxukkEPGZ+LgkvBUER0WUKXJQx5msvZshFtC9mwUHUTRTPwFHS8eQ1fXsOfKRBKFyhJ5ggns7iLxpvyobsWSJZFsJAn4fWRl5evFvRGo6DKDFWX0FtbylhFkKmgh7xbxRRFZNPEXTQITUuzRkenqB+KU5JaHhKL60XerdBbV8pATZjx0iDJoIeCy1EFVHUTdfr5K4tnKI+lqRpPUTM65cjeLvFYgu3Exn09zgpdsJ0Vcn1f4iLbF1D5WHrWNiXH8pQcc66XkrGo+tnspCtfz/Q+p1H1UNpZcZvOPqt/lsaejolc3mc5a1P6eHbmWACB9ouTf8EG33kdb6c+s72/Q8N3Vrvi/i6g+lupWQm61wJLEJiM+BgvDRCL+EmEvGT8LrIeF0WXjCFLGJLoXEfTRJlWEvTlNUcKeCpLaTxL9Xjqhrq8l4q06ublFatpmRhhZXxpAkr/D7PhTxdo6lqaLDYw7bqfjYTXT3t1Aw3xCQTsJVVPLWjcfUaRWwYWprsTgLrUJHULcM4LQG06Rm16+Vm3Ln9nLUGgvzZCT30p/dFS+mojDNREyPhcaIqEoThGXJ/+3ymRErAE5/+LYijO3h0j7/wvmRaS6Rh8xXBeYlU3cRV1IlNZ6oYI2pgMAAAgAElEQVQSrOiP0dQ3wcq+iTm6vG8l4y/YNkVZwZSuvSzmRuPCvY+X+Hhk/3oObFnBSGWIrFel4FLQFBlTcu6paNuIpiNC5NJMPHmNYKZAS/cYtx44z45jvXgK02WJr1PfLVGgq7Gcp25p5dW2RhIlXnIep++6ImGKjntZtGwky0IybVyagauo45429Fva+9l1pIfVnaMoxmwCmqvh8kTUqyWyXisEmEMkdTUjuxgDfHmbhbZZynnN0JgLMBnxc3xdHe2tUc43VZAIeim4ZIouxRlHpg26JYlYwkURJdF2iJLkS8YJVTNQNQNfXqNyPMWa8yNsPDPEurPDM8/cTF+vA1NuL0m3h4TXz4aRfjRJpqe0gtJsmopMiozqprOsirJcmuhUHLehUZRkxv3Bqxr3C9dlzFXCi+VrOR1sIKH4UG2DqkKCXbEONk714DG1GdGghXBhn4YgcS4Q5USokV5fJTE1QFFScJkGJXqGFdlRtia6aM4Mz5svZAgi36m/nYOlq7ERWJfq4xM9T+C9ihLhpf3o9FfzN6veC4DL0vmjM/dRVkwu6X788tdfQpeXnlHjy83to2RZbBvoYuNwn8MTsoSF8JtX/26RsAFTEtEUiYJLYaQyRPuaKMfX1XJydZSpoGfGaNvThnupCTW2ACBgiouI5kyXzR3YMj0ZsCxcmkHDYJytJ/rYfryPlb0TePMaLs1AtK6f4epGwQbSLjdd5ZW4DX1ajW9hJIIeTGlpD7dkWpSkFk84cmkfTUlkrCzIA3dt4id3t5HxuTAkcd777EzewFAk8h5H4GHEtjm/soLHbl9H1XiKn3/wCG9/9hT+TAH5Knr31wMbKLpkuhvK+ca9Ozm0uZGCS7l47a7Qf0tyhGR0BQpuBXAohvtrIxxfV8u/v/8mWjtHed/PjnLT0V4CmfkFmt5oaLJEKuBeuOE88OY1PAV92d6fC8mjBZdMPOzj5W0refLWtZxfUYGmSJiS6Cj2LXL8sAQBRDBlieLleYO2zbmVlby0ownJsvHliuw41sPtL51jQ8cQwXQBl3btHsDHVm8i5guwamKYoqTwTMs6km4vZypr+czzD5FR3QyHwnx3y8189okfEyjm8WmFqx7PBvKSymNVW/lO/W2kFN9Mci44uUJPVG5m01QPn+x5HL9ZmMN2d6V9ZmQ3j1Vt5ZGqrYx4IhiCk9t16cgoYPO8vZ5vN+xja6KTX+/6GTX5+Jz+SrZFTT7GkKeUtOIhI7vZPXmGtqnuRV3LR6q2zRAz7ZzswLWIScHlmI8x8lqgmganK2s5Xen06bdefASPsbhEwbescc+7FGIRHxMRP90N5RxbX8dr6+sYLwu8sdmw05nltiRgSQASmkvh5JooJ9dE+ea9uyiLp9nSPsBNx3po7h6nejxJIF24YdnTCY+X7tJKDNE5wrg/CMDJ6jqSbs+822mywnNNa3hxZSurxkcI5ReX6f+JL32c4aqSJd2HyvEU93/yK0uiV7SBVMDN8ztX8e/v2053Q/m133th2mOjigzURvjir97BQ3du4MP3H2L34W5C1zDxuBp0WaQvWsoDd23igbvapg31dUAQMGWJvCxxbEM9xzbUs+H0IB/70UHaTg1SkswtW/8NRIeUhGnWRPvajNDRDXX8l7/40DXfs4/86CCf/vYLeAvXnxVtigIjlSE6mqt4dvcqXt7WRMbnunFjiSBgC448rgloqsxje9fz2O3riI5OccvBTm47cJ6GgRiRqews6evFQDUN9nSfYcPoADGvn95wuVM2ms8y6QthiCK6JJNVXTPjwkLISS5+XLubH9TtISu7wbbxGwVK9AyqZaCLMlOKj0ORVcRVP7dNnFyUo6TbV8U3GveTm96ny9IJajk8pkOBbAgSScXrTCZEiVdKW4mrfj575vuztA3A8XjsiJ+nqpAgLXsYdYc5GWpgXbIPdYFYdVr28GLZWgBky2BX7Axes/iGLr4aEpPcce4E4ExwlCWQXr2ljLshiQxWl3CmpZqzzZWcW1lJd0MZiRLf68rkdj2wRYGJsiCP7V3H47etpX4ozsbTg2w+OcDW431UTSys3rZUnCuv5nP3fJBJn6M/rknObf8/d773qmIGRVnBEgT8WoE7z56gOrUIwflrRDzsY6LUT/X44gkzBmvCfOd9O3hk7zpyVxCFuR7YosDZ5iq++Kt3cnztGT78k1dpGJy7UrgW5F0yz+1exfffvZ2O5qolezkWi/Y1Uf7iM/ew/4UOfv6nhx1J2mtcxNtATPLT7q2jXy0jKXmxBAGvVaRST7IuP8jK4vjrWt43UllCzqNel3G3BOiPRnhl60pe2NlC+5ooRdcbmFsiCAxVh/n+u7fx6L71bDvex+5XO9n9ajeRqcXL/oq2NVO66tE1mmJjFGWFUD5HTTLOo61tmKJIsODkTYz5Q5wrr8FjaKweHyZUmD0ZtBA4Fm7igejOGcO+MdnD/rHjNGbHZuR8e32VvFi2luMlK3kgunPh0wXqcxPsjHVwzh9lTXqA5swwNfk4YS2Dy9IpigoD3nJeLFvL4UgLRUml21/Nz6q388nux6eTeS8iaOTYO36CTn81pihxJNzM/rHj1BSuXoL7SmnrjC5IbW6SVemhZVEXXAgZn0ra76ZiIj1nEneyqo7HV22kpJAl4fHTMjGCpGuLeg7eEsZ9KuDhcFsDL29voquhnJHKEMmA57oZgN5o2KJAX10p/dEIz+9soWEozp4D53nHE+2UXiOf85XQOj7M5x79Ie019RyrXcHxaAMplwe3rs+hF70U4XyW+vgkd55r586O4/j062OXuxosAc6vqFi0ce9oquQfP34bB7esuGHGESAV9PDg2zbR3VjOf/uHx1nVfX3JRrok8u/vu4n772ljovQGe5kEgVTAw0N3bKC7vozf+PqztJ1aulqaDQyopXyr9BbaPXWkJC8+q4iARV50IdkWdVqMd00dYV/q1JzB9kZhuCpE1uuiLLH0d8XGoQl9dO86Ht23jrMrKx3vyZtlkSAIJIMenrplNUc21vPo3nW895HXuP3lc4vybt3U1zmjxe7RNbYPdDEaKEG0bdyGzu7es0x5fKwZG8SvFcjLKnu6zyBiX7HcqigqPBDdSUJ1FggtmWF+59xPqc/NntCtTQ3QNtXNPzT/HK9GVi3qegb1HL/Y+xRZ2UNNPkbAmOslW50eZG2yn6KocDjSgi5InAnUkVD9M3z1l+KOsdf4ft0ekqqfc4Eo3f4qqgqJeUuuNUHihfL1GIIMts3a1ADR/OujxvjCzS387O4N/OX/vJ9AZnZuliUIbBgdQDV0jkW90xS0i8Oi6twv/C8Ak34/RxtWsGmgj4rUxVXm86taWTMyTHk6hS5JJHw+AoUCHk2btZ/LC/WvBAswZInB6hIefNsmnt/ZQiLsJedWsW7gQP5GwRYFkiEvJ4Iezq6s5LG9a/nojw5x+8vn8OYv1nVeK0KFPLt7z7FtoJsPHX2Z/2//u/jp+m184YFv0TwxOu92F9xAXq2IeoMrGmxB4PzKSm69TDN9Tjugs7Gcv/vlfRzdUI+1SJfi9cBQJI6vreV//f47+OO/fYS150auyUDqisSXP3E799+9meL1uuGXAF2VObG2lv/539/NX/6f+1nfMbyk/hcFhftLtnPCU8+dqXb2pU8RnB6Ai6LMaU+U75TezP3h7dRoCdYVhm7UqczC8HTC5IWxaTG4kLzYF43wxV+/k/bWKPk3k1G/HILAVMjLq22NnG2u4vmdLXzyey9TPxhDuEplQFVmtvBSaS5D5JIy1ppUgpppT5wAqGae0Ngg86HPV87x0ArA8Qr8Qu/TNOTG5xxfxCaaj/GewQOcCdY7stMLXFsJm9ppQ3o1gqhoIcaueAdnQvVkZTdpxUPMFbyicQ9rGfaOn+An0V0UJZVnyzewLX4e9zzCW52BGvp8Fdg4+gfrk714lyi8dS2wgeGaEtrXRZ08ocuwZbCHoiLTVVqJu0/Hu8hVOyzCuE/6AxRUhZ6ycrb19nAqWsukP8ChFc2UZdLc1N1JzB9Al2RUw8AWBM5WVfP0mnWsmBjnztMn0SWJ0zW1CLbN+qEBvNrcDtqApkgkSnyca6rgp2/bxMEtKyiq0118s758ywlBoOiS6Wys4C8+cw/P7l7NJ77/Ei3d4yhXKJNY0q4Bl2mgmgZNk2O4dY1ILkNFdvl5o68FtihwbmXFVQdqGxiqLuErv3Qbhzc1vO7PROeKCr706f38wd8/RlPvxJJc3Bmvi699+GZ+/HNb0K5DjONaYYsC42UBPvff3sX/+uuHWH9maNGJdgVR5lX/Su5KneAjsZdQbWdlJwC2CdX6FI3FCf44+kHOuateN+Oe9rsZKw/S2jm6qOz6Czkaj+5dx1c/cgvJ4MKG500DQSDld/p+elU1n/rOS+w52Ikvt/iY8PWc6aHS1ZiCY3xWZkZZnZ5/IiAAbVNdlBZTjnFfpr4JQF1uArepkZXdaKJMflqueG5bm7tHj/B41RZyspuXy9Yw3heiLjc551gWAkdLmohNeyWqC3E2Jntfl1i7JQoUFWk6aXsubAG6SqsoyWeJJhMzMuGLwYLLnqFwmBebV/Hwhja6yytor60j6fHgLxZ4dMNGprw+fMUCZ6uqiPt8MF0q5tWKRKcSiLbNy00tPLl2HQ+2beFktG6OY0RTJE6truF779nOZ//Hu/kff/Q+nt+1yol9XUN2+1sa0+dryBLP7V7FH/7he3nwzo1MBRf3kiy4e6AhMUE4n71qvH0+DLaUzLp/9vR3Bc/1RXhsYKAmTME1/36yXpVv3LuLl7Y3v/7PxPR9aW91kiJjYd+inc85t8JP7m7jwbdtRFPlN+55FgSGq0r4ysdvpb82suj+2zg0pdVaAtd08tyFM7jwd7mRpsxMowuvY6RPcMoHzUV4bywBeutK+Ydfup2vfPw2kiHvW29cmX4G+2tL+avfejv/+pGbGaoqeV0Ofc4fnfm7KTOC27x6lYJqmzRlRq7S4trgMvWZcctCwBKufO8FoLIwxY64wwZZlFQerdp6xbYxNUBHsI6ipCLZFuuSfVQVFs4vssHhn5hOor4WGLJ41TGhoyLKQ2u30l1aycNrNpNTXYs+1oJvolvXGQmF2d15jifWrmfV2ChFSWLNyDBdFZWk3W5KsxmCBcdNJwLhXJbKVIqG2CSSZTERCNI8Pkb11BSVqYs1g0VVor01yjM3r+bYhnp6a0sx3oBVzZsZI5UhvvypvfTUl/Hh+1+lZjy58EYLoG2oj//+1E/nsNOlw26mKjxERrJ4UxpTlV6yQZXywQyuvEGs2sdEfYDoeWe7XEBhqsJL+54oOx7pxZ2/DpktQSDrdTFSWcLK/rlcCTbw43u28PAd69/QXAtLEnluVwsreyf42I8OoiwQ/zRFgVe2reS+d20ltUwTtOuBLQqcbI3y43s28yv//iLBzMLSooptUqMnSMpeTIQ5MXUbyIhuTEQqjNfXE9TVWO6E6q5yH2zgTHM1X/ml2zi6of4/xBiT97r43nu20x+N8PH7XmHDmaEbutIcc1+cRJQXk8iLYEmrLF6d/fJKKAoyfb4KRt1h4mqAnOwiL6kYgoQhSoy7QmSUxZVPeo0Cu2JnOBhZTV528Xz5Bu4deJHwJVTfNtDvK+e831EvdVk6t0+0L+panlpTzaNvW09jf4z333905vvv3ruNibLAovpoyCInNtTO+7slCHj0IhP+IFMe75L4JxY07pFsFtkyWTU2yn3bd/LO145xMhqdrue2MUSJ47X1HFrRRNzn56OvvISvWCTu8/G9Hbv4pRefp22gj4c2bSbu87FqzJnNDdSE+ZeP3sJr6+qYKPU7SVFvtZn06wFBION185O728h6XfzG15+lNLH4rNkroSyb5o5zs3WuNZfEkx9bw+rDo4iGhTurM7AqTNErc3ZHFfv+vQPNLXF+ayVtTw2Q98kcumsFgUSBgdYIOx7pva7TBCi4FAZqwnOMuw0c3tTAfe/aii6/8QNzwa3ynfftYPfhblZ1j101jDBWHuSfP7aHsfLg69nFq0JTZX52xwa2He9jz8HzC4YXPJbGXckTPBraSG0xxo5sF/I0640N9KllfLv0Fhq0SdpyfVccf27Um93VUI55lcmeDby2vo7P/+bb6akve8sn4V4KUxJ5eXsT8RIfv/W1Z9h8sv+qcfjrQfoS97rXLC7KPew3FmZ8vPCsTCk+nqps48WytUy6ghQkFU2Up2veRSwcAqqlcDBK2LSkh2nKjHAy1EBMDfBy2RruGTk8sxdNlDkdrCfmCoBt05QZoSW9uLDSyXVRHr5rPeFEdpZxf2pvK51NFYvu59UqvVZPDJPw+hkLhNh//iReffFhmAWNe3k6xa8/8ySSZfHVr/8LqmHQGJtAsiw+/uLzSJaFLQh84b7vOApDpiN2/+EDL2OKEm5dY93QIM3jY9gIuKczrnVZYqAmzGhF8P8Z9YUgQNGl8PD+9WiKxH/9xyeWnSpV0Uxkw2S8LkB1VxJLEjBlkYlaP5mI82JX9aZm3O/xaj+erE7roVGGWpbHNVhwK/RHI3O+j5f4+NYHdt747PIlIBn08Le/so8vfe6+WWIPl0JXJL74q3fSU1/2pun3BWR8Lr557062tPcTyF49cagoKJzw1jEuh/irmnfjNYtEzCwCNinJQ1p0YyHQoE3yJ9H3Y10W7ftY7EVuynbdkPMYqAnPWw5niAJHNjXw+d+6i8HqpfEuvFVgSiKnVlfzF5+5hz/624fZ3D5wQ0iLZhHK2Isj3rpaJc6lOB2s58st76DLV40piEi2hdvSUSwDt6Xhs4r4jAIuU6cgqXT5qtClxSWkRvMx2pLddARr0USZZys2snf8BF5Tc3IwZC8HI63Yggi2zT0jh1EWGdde0TtJc/c4rWdnJyXbooBo2azpGLmqUAyAJYoM1pYwOc9KX7Btbu45i2I6Ut5LeYIXNO4CoExzyLunyesvcMpf+B7bRjJmu2RdpgmXFNx79NkvX3RkipuO9nKuqQr9P4Cb7PWAJYk8fvtaFMPkd//5qUW5VBcLUxbZ9mgfvetKeW1vLStOxhhdEaTl6DhndlVjygLJMg8Fv0q82odSNNA8EhN1AQq+5cn8Lk7zwZuCMDNA6ZLIU3taOdla8+ZadQkCJ1trePrm1bz92dNzXjpTEHjojg28vG3lm9OoCALta6I8uaeV9z56/KpNDUGkTy0nbGYJm7PLztyWTgWzXfHSZTyO1x6RXBimJNK5ooKyRM+s7w1R4OjGev7uU/sYrA6/tfielwpBYLAmzJ/+3jv47JceZtvxvpka9+WC3ywwieN9ykuqw763wCHy0sLcE6PuMJ9vfT9D3nLApjqfYFv8PDviZ1mRHaOsmEK2zZnbd6xkJX+55ueJLdK4y7bF1vh5nq7YxLAnwoCnjBOhFeyMnwWgz1/BuaDjki8vprhp+vvF4KbDvew43AvMfbyCqTx/+bn7CaYXmDirMv/w6du4/z2br/j7segKVNNgx8DSJ8dvWJ27SzfYdaSbJ29tpb+29I3qxlsPgsATt62lbjjBR+4/hLu49Dh3f0kpr0UbqZ2KsXGkH9myMGWReLUP/1SBurM5SsZzVFb7KPoUWo6MY4sC6VIPqw+NkI64qe6aoqo7RSripvnYOJ7M9dfAW5LIZMRPxu8ilC44ZSJVJTx565oZhaU3E4qqwsN3bGDHsV4il0lmdjeW84N3bsW4Bo7p1xP3vXsb+188e9WJot8q8ocjD1zzMQLmjRXk6WiuZOfRi8bdxqls+JeP7eH8yor/2Ib9EoxWBPnCr9/Jf/3KE2x/bXmzvcsLSXq9jqs57gpiCBIurj72TLoWDkU9WHMTQ54ywEmA+3T3I9w8cXpergRDWDqLwqr0MC3pIUbcYRKqn+MlK9g81YVsmTxV0YYpOIvL2yba8RpLK3+b7xrLhoUvu/CYKBsmLm3+6+jVNZ5uXs9ooATBtnn72eOLLkt+Q0lsWjtH2XhmiKHq8A0lIvmPBk2R+PE9m6kdTnDHCx1LnqUfaGzhH2++k+aJUf7hB/8KgCtvsPaV4WlxEqfd5qf6Z6lsNZ6K0XjqIrFDy7HxeZW9rhWJEh+xsJ9QuoAhiRzYuoLTq6qXae/LC1sUOL+igqMb6rnjxYtqiQVV5pF96xiIRt6cq/YLEASGqkp4eVsTdz17at5mIjblxtxa4jcLzjZVzfp8IWRyYk3tm/v6LzcEgb7aCP/8sT348kXWnV2+bPXmzDCvRloA6PZVUpQUfFepA7eBLt/C7+3RcNPM39vj59ic6LoqCVLcFUAXl2a23JbO3vETHChtpSipdATrGHOXENJzHChdDYDPKLAz1rGoRMGFEEgV8GW0RVEGi5aNqhkI87RdER9nb+dJZMtEsFnSWP+GWlSXZvBzT7TjzRXhzSJuYduz/gmW7ZReXPrvjYYgMFEW4Ifv2EJv3eJLmi5goKSUmDdATSqBdEl8SbAvGvYLnxcaGpc7gSce8hKLOGVmGZ+Ln9zV5pSKLBaX3rvX4Z7FS3wc2txIxuvU29rAmZYqXrip5drr2a/U/xt0DkVV4clbWyleZ2jMBnREjDdgSDnbVIk1/RBqisiXfmU/x9bX39gwzmXPmWC9OcYIWxQ52VrDt96/k9HywLIFRLYlzs+85+cDUQY8ZfPu+4JhH/HMzZ+5vF1S8c58LiumcF2BHe9C26IocypYT3YR7v7LcVP8rMNDb9t0+qsZ8pTxaniVkyho26xL9hHNx5ZlLPvff/4gf/ani/N0CYA3p+Gdh6/AXyywabiXTSP9NMXGZo3XC+ENp5/d0t5P28lBXti5sJj9ssF2JFvdRQOXpqPojmSrok/Lt2oGLs3ArRkouokhCWiqI/GYn5bg1KdV6AouhbxbWZJa1LKcgiBwYl0tz+5eTXTkIK4FEjcuRU51Y4oiK2JzGabeaCRKvEyG/QA8u3s1XY3lC24j2DaevEYgU8Sb1/DlivhzRSTDJOtzkQx4yPjcJIMex+Au432yRYGTa6IM1IRp7Rwj71Z44aYW+mqXsGq3bWTDIpjJO9rfBQ1f1jkXQxJJBTyk/W5Sfjdpv3tZWRptAXrqyzizqpq2U/MTkyyEgqDwr+V7KTXSfDj+yrL1b0FM07SOlQcpi2f4wTu38czNq533cTlg2yiGiSev49J0VM1E1Q0Uw0LVDNxFHVfRUXcsuGXybpWiS0ZTZIoumYIqk/eoFFXldQsPWJLIMzevonYkwae++xLuZVDOW5UeYnV6kI5gHbog8W8r7uSzp79HqZaetW8nSc3D/bW7HMO9wDsQ1rKMu5yE3BFPhLysoujGnH1qosxz5Rs4GFk9b2371aBYJu8ePsCXWt5NTnJxJlBLZ6AGGwHZNtmc6CJyBaa7a0EotbRcqA/94FV+/keHkcz/y917R8l1XWe+vxsq547VOSI0ck4kQDBTDJIoiYqUlSwHyc9j+43GfsvLM/J4Zjx+tmZsS2NLWrZkyQpUJCmKBBOYQBAEQACN3Dnn7qrqrlx1w3l/VKOBBjqiCxTX+9bq1d1V9966derW2fvu8+3vuzFdOlHdgFXX2TnQxRNb9vHx5reWzJj/jQd3CfjEkyc4vq2O7AICJit6DVPgSaQpCsUpDsXwTyUpCicIjk9ROh4jMJXAF0vjjabwxlIo89iwXlHRi3ocTHodjJT4GAz6GSzzM1AeYKDMz0iJ710znjBlmace2MLed7pyKl1L3M+fSmAxjBnFqfcSkg4rY0We3F37+7YsPDkIQWAyyZaL/Wy8PMj6tmFqBkIEJq+aXghg0uugrb6Uk1treXtbHR11JXkNkL2VBXTWFrOqa4yB8gCH9zctObBbMxrr24ZZ3zrEpsuD1PVOUD46iapfZcbq045lF9ZWcGJbLSe21DFe6M5PkiJJhAIuWhtK2XwTuvNXICSJhGwl8BtIFw1FprO2mMGgn6ce2ELKbl1ZIBUC/1SSqqEIpeNRKkYmqRkIUzIRpTCSoCCSwBtPz1l2FeQMriJ+FyMlXkaLvfRWFtBRW8JgmX+G3X+rbwSELPPzh7exrm2YO48unSQ2H2yGxgcHj/ENRyFxi5Pzvlr+cfUHuH/4FGXpMHZDIyurjNr9HC1ax5Hi9RRkY4zbfPO+VwnYFW6j1ZMTyDlesIaG+Ag7w234srlujJRiZdzm40yggUNlOzBkedopbnmxQgL2TVzmx1V3MOoIcCbQQO80h6AsFWFtbOBdMYmZC7IAeY7ADmA1DPoDRbiyGRI227KS1rxF00u1QWxZnbqhiVmFOQEk7bmSpSs9N8FgQ+sQO8728tauhjmfvxnYMhp1fROs7hqltj9M+cgkxaEYxaE4vmgKe2b52awE2DSD4nCc4nCcVT3jQE5oIBRwMVjmp7OmmHc213ByS23Oq/oWf4lHSn389P3b+Yv/9eyS38/6kX48mRRnK2ox+Q2vzVwPSeKV29fQU12UI0PNA9kwue1kJw+/eI7NlwYIzGNpKgGBaIrdzT1su9DHXW+28PQDW3j6/s15C/CaReX0xmoOvN3BK7etZbjUt/hOQrC6a4yHXzrH3ne6qByOzLtGp5qCquFJqoYnue1kJ6c2VfNvH9vH5VXBvFxfSYeVtvpS4i4bnkQGARjLvCo0FFJyfp35loqMVeWHH9pF1qLSX37zzHhFN2hqH2HHuV6a2oZnAroruXQ9bwmwGCYloRgloRhcHpxZYhosC9BVU8TxrXUc3dlwy0WNkk4b3/r0flZ1j1E1tDJHRwXB7lArHe4yfl22i7Rq463CJi55qgimI9jMXHAft/kI2bzUx4d5/9Bx/vfqRxdcGrhv5BRHitbT4y5l0urm32rv5o2i9fi1XHBPKjbGbV4GHUUEtDif7H2N54Pb6XQvn4fj1ZLcNXaWH9ccpM1TgSEpSAjqEyPUx+f32PhNYs3YIBGHi56CYnb0d+GYQ7p9Psp4i7oAACAASURBVCwY3HVZYtLjxJ3KEHXZcaaypGwWIl4XNk2jcnQSQ5HpLi/k9JpqGgbHqRyfpKWmFKuuUz8wQdpm4Zn9m/Am0uw530XxZJyRQi/jfjer+sexZzWsms6jh85wcmsNmmWZ+cY161u2rM6my4PsOt3N5ov9FIfieBIZHOks6jXa7NMcsFlmNlz3/7UX5GKDKQsxE/A3XB7kzqOtDJX6+dkj2zi8f+0tlxx9bd9qPvOTY9QOLmxpeAV7etrZ1dvBiZpG3qxfy/6uHBnsvVKib2kso7UhOHeWKgSOtMaHnj3NR585RXBs6Ra5Ft1kXfsI5aOvE5hM8L2P7sVQlLy88dObaugvD/Cr+zctfDwhUEzBnlNdfPEHb7Kqe3TWXfpi8MbT3HGsnerBMP/lK4/QVl+68mtLkuiqKWasyIsnMc6Qxc9/L3t0WYcwJYkx1cuqdP4lRxeDocic3lST+84udyyEQBKwrm2ITzx5ko0tg/inbi75nw8S4ElkWNsxwurOUfad7OSxZ07x5Pu28PKBppxxzc2c+xLQU1XIv37iNv7s689jX4CVvRR49BQf73sDh5HlF5W3kVRsRKxuIlb3zDYygo2TPXy++0XqEqP8U8PDpNW59d8BytIR/kP703yz4UFaPRUkFRsXfDVzbvfZ7pe4feIS/Y5iutzBOY62MFRhsCfUyvPB7URsub5yl55h62QnLiN/bcXXYjHOw2KfuCeT5mDHRTRFxaZry1pzV7761a/O+2Tib/76q0e2rsKVynBkayMWw+TI1lX0Bws4trGBmpEw7TWlHN3UyGihl5JInJJwlOY11RzbWI87lcGR0Xh7Qz3+eIqakQgRn5N/+eDt6KpKV0URq/vHUMzchN1VUzydeS/hIhcCW0bHF0+zoWWQx39xnD/7xgt84PmzbGkZoCwcxRPPYBM6iiZAgczHnahtGpIBxmqV1J95sZzJIiUFQgbtHjtywoSEALuEWTE98aenPwXr9O8r46sCV3hIMiByJRaHplEyEeO2dzrZfr6fwTI/oYDrlq3L66qMapjsPtO9RAMGwfaBbi6XVvDU5l0oQlCQjKMpMmnVSsoy949qmksSyHjiAzuIuVdQtZCkHBnq+v2nA/tHf/UOn3/irXnv1heDLaOzunuMlMNKS2MpIg/OcjGXjUN3b1z0fSum4MCxNv78Hw5RNRSZdwloIUhAYCrJ6q4x3tlSS8xlW/F1lbGq7DnVRcXIJCHVzROFt6FLMrqskJEti/6kZQtp2cqG1ACbUv1Les2BMj+H7t6Ql+Rk2R4U07yb+v4J/vBfXuX3v/c6a7pG8SYyWIzliYUs61TJXX8loRh7TnWzumuU/ooCJv3OW+ZwOFrsxRdLsaZjdEVVOokc83zjVC+3TVzCPc2WtwqDomyU9dE+Hu99lU/2vUZlKoSM4Jy/Dk1WqEmOsTPcjtO4sXpbmpnktolLVCXHkRHICGymRlE2yrpoP+8fOs6XOp+lKTqARRikpxnvRZkYO8PtlGaWJsktkQvwva4Sel2lIAQlmSm+0PXigsz/m8GV5ZnxEg9vHFjNE4/t5Gcf3s4zD23i6N4GBqoCuOMZnCltweT+XFkNQ/4CjtQ18eTGXezpa8dyfSvcV7/6l3Ptu+htsilJmJJE1pITyHcnM6zpHaWzspiRIh+9wQL2N7czUBLAkc4S8TgxFJmUzYqmqhRNJqgeCbO2Z4SiqTjP7V0/I7jvSmXIWlQsRpbCcJwDx9o4s6GKlGP+TM+i6RRPxKgeDLPrTA97TndT2zeBek1J06hW0GpVsEoIFSzHswi/jJRlJpVS23TkwauZrCiSEXYJNEACY41K9l47SruO9YUUwiqh77SCRUI9mkHSIXu3HSlhol7SMEsVpDEDyQCzQEbp0rFqBtvO9/FXf/M0T3xwJ8/cu+mWOFEJSeLorlV89JlTVIwsfqEfqV/LyepGrIZOyOnmv9/3If7xwPsoTMSwGPMT8/7roZ+weagvn6e+LMhCcN9rl/jMT4/hSt18X70E+GJpHn32DO11JTSvr1o5s1qSFmf1C8Hu09382ddfIBBNLrztYi8HrOkc5bFnTvGtx/eTXuA7sxREPQ6GggEMuRcJqMiG+YOxF1mbWpoUZ0Kx8felD67oHN5NeONpDr7Vyid/cZL6/qtyxwIYw0lKslAsEjjRFwz0SVSi2AiydE/5K8ezaToHjnewsWWIf/7MHbxy+xqinjyX6iWJqNfBCwfXsf1cL7UDS6vuzXs4csIwdckxvtD9IjHFQY+zhKrUBD59tiy2Kkz+9tx3Fj1mv6OIhGLnrrFzPDhyatHtD46f5+D4+UW3mwuGJJORLTPvZUekg+Jsfv0QBDm1zcN3ruWJj+ygr/pGcu1bext44rGdPPDiRT7yy1NUDE3OKQN9pSRfMRXGkGU0RUFoSys2LjgbqYaBYpqcXF9LZ3kxG9sHSdhtnF5bTcTjZH3nEBlV4dTaGqJuO1vaBmipDTJc6MMXT6GYJhICZybLyaYaAtEkjf1j9JYVUj84QfVIBOf0OrxqCrZe6GdV99h0f+rsc3GksmxoGWTzpQE2XRxkQ+vgvGthRoOKcMto99qxvJJGX29BPZtF22HB+pwE2o2jKEUFRo2CelFGipiQEUgRE6VbzwV8t4RZrKDtt6F06hA20e6yYf9uAikhQBNoDzhQOjSERULpuJo4lITi/M6/H6F8ZJJ/+9g+xvKtMy5JhP1OTmypXVRtDODVVRv4+ZY9sx6L2R3E7AtPLCnLb2ZN9QrWdIzy+R8fxZ1cuWAOQF3/BA+9dJ6O2hJinlsvktPYPc4ffOdV/PMEdgGESlz0NhRiqDKrLo0RCCUxZImMQ8WW0mety1s1g9tPdPDavtWcXbfCnm4JOuqKcwmKCTah4TQzWFhaGdAqDBxmfj6XW43g2BSf+sUJ3v/CWRyZq8qZApjEzqtKHToy9xmduWCGiY6MgsCKQQwrMclKoUgxLjlplst4wOhAR8aBtuy7Y/9Ukj/69svU9E/ws0d2MFSWf6e3llVBju1ooHx0al655JtBr7OYb9Y8wOf6D7NtqmPZFQ8DmeeLt9HuKuePu56mPLM4N2DQVsCwvYAdUx3Lei0BDNkLuOytBnJ38feOnF54p5vEs+/byHd/ay+6orD5/ADVfWG8sTSyKUg6rYyWeLi4rpxnHtzEpM/Bl7/5GiUT8RuOs2Z8iKjdwerxYZxaBps2tx/9XFgwuNuyOrsv9jDud7Oua5jyiSnaq0sIxJLsuNRL1ViE0nCUwmguaw2GchlQxcQUqm5QPjGFbAp2XexhuMiHPatRNBnHPNWKIUl4E7OVq6qGIuw800NrQ+kM49w/lWD/2x3cebSV2v4QJaHY4henBPKAjjRsIA8aCK+EHBNI6flLylJaIMVzK/GSyAV7KWwiDxkgQN9iQQ6bKP3GdC8ryEMG6vncYCtdOtlHHJBRUZuzSNedoj2r84EXzuJIa3zj83cSKnDfeBIrQNJh5eKach46fGHR8Xn8nTe4p/Xcsl9jzdi749M9FyxZnd/5wRGC4/nLsmUBdx1t5YU713FyS+0t5UW442ke/+Vxavtv9JOG3MRzcWs5T31qK/11AdzRDJ/7+lECoSQpl4V//tODPPzTczSdm038qRyKsP1sLy2NwRV3aXTUlZCxqZTEonxx7BXKskt39ZKFIGAksImVreveatT2T/DFfz/CHW+336D7LYBJyU5CslBpRnGhcV4uJShihHDiI4NfpHlRaaBCRHGIMQDSKJyRg5SKBLVi+U5oEuBIa3z4uTN44xm+9en9eb8BSNut/PrejRx4u53ykcm8LTvUJMf4Us9zVKbz0yO+FBwtWEevo3jZwV2XFE4HGmd665ui/TTeAlvarvoifvDxXVizBr/1s7fZ93YXgXAix+MQuY6rmMdOZ10x//74Ho7ubWDL2QEeee4cqjGbGxaMTfLQ5dPIQlAWi2AxjCWP84LBXQaKIzGKI7n+P01V8MVTNAxMUDeUm6SsusG67uGZkxFAwXSwv3IShVMJCqeuPramd3TWG5g5GcPkniOXeemOdSScVh5+6RwPvHqRstEpHNPGEEt6YwZIJkiaAFMgLDLZu23oTVbSn3JieyqF0aSib7WRNsH+gwRGvQV9jw2zVMHxnThSzEQEZNKfd2H/5zjCKqPdbkO4pGk2noDsNclCFtSzGtp2K7Zfzi23qeom9xxpIea28+3H9xPPw1rpFZiyTG9VIUOlvpnS25WzE1LOj1uQG7+GiVEax0dmPq/rfzPtT3xFDcmUJI43NtBaEWRHV/esJZB3BUJw8Fgbmy7dfB/2fHAnM3zw+bOc3lSDodyi6Wm6HL/zTM+cvawA42UenvnYJnSLzKM/OMMvP71tRrrWkdBIeGw076q6IbgrpuD2E508c99mxldI3OyrCJC1qPjNFBvSyxtrm9B4PPTmjNf2ew5CUDoR4ws/fJM732pFMW7kOkhAsUhQKaKsFmGsGESxESBFTLJiFQaTsp8yEWO3OYgFgylsXJaLcZsa282bDxQSYM/oPPDqBXRV5hufO7gy3soc6Kgt5rm71vP5J96atzPDQEKTVWymhoGMIcsIcvOHVVxdojCQyMoqqjCoT45gNW9M6q68gi4pGJJ8A7ns+n1MSSYjqTPzlSoM5GnLGjF9HF1SeMffiFtPk5KvJrN2c/H4ELM4eDG4DaaP/8jQ8bwo0l2PwwfXknDbePi58zz6dDMWbXZAdmR07Jk4RaE4EvA3//F+XjuwmvsOX0K9riopwQzPybrAkulcWJJxzBVYdIP7jl++4fH5/p7vsYU+gPreCf7T/3mBNZ0jeGPpRbefC5bXc+QIx9krJYzcb8vhq6QJ9UgWz5GrUqpqOIv7ndnrUfZ/S+QyHB2sh1JYX0jlyHTTV6njH6+WUYQvV+63vJOFzNxfHAmwZnUeevk8HXUlPHf3BvR8WZhKMFAWoLeykJqBMBIQcTmJuFyM+H1s7u2nPVhKSTRKezCIN5Vk3cAQA4UFFMbijHs9BBJJQh4XE243uqKwvTtn33mmtobW8iCNI2Pc2Etw6+FOZHjw5fO4E0u3O1wODhxrIzg6xWB54BYcHQojCe5+s4WicHze8x8t8zBV4OSLXztCeV+EX376qpGELARl/VNMBOd2jlrTOULpRDTX+74ChP0uom47JROxZY+zBDjfq2V5IXAnMjz+87e5943L81rcSuRavhQEKibydBk+JDkISU58IoNdaAzLbiYlOz6RRgKKRBIhYERyUyGW3r0xFyy6yQcPNRN32fjWpw8sT51xEQhZ5okP7uT9L56jdGJuwZZmXz1/1/Ao/6Pl+7ztX8PRgnVMqU6KtCh/3vZTirTcfue9tXy99mGSqo1J1cVftf6A7VM3mpt0OYM8FdxNpzNIXHUQsnqREBRlovxx19Osj+U4PJqs8mrRRpo9dYzbfHj0FHdOnOeB8VO4jQxR1cmTwT2842+k2xlEFia/s+kPZl7n6xe+jU+fn/eQVGz8sPpOxmw+QLBhqoctka4VjOb8uLwmiDVrsPtE9w2B/QokcuqeNf0hGjrHudxUlncfimVdOe9G2UUCdjX3LGsfARwraiJqcXDf8BluzBFvEiZXmfG55t/5N61UEU4J6ysLB6ArbTEff+okF9aU01WTPzvQiM/JYNCPKUsopmDc4+XY6ka6S4pwZrIcWbua8sgkRbE4r61rwq7ptJSXsaF/gObaGlYNj3C+uhLFNJlyOCmMJxj3eugvyMlIppfbppgnrG8dorb/1pX9LLrB3Udb+P5je/N/cCGo7x1n64X+Bc9fs6rY0jqOuXgkIlf1Sc7Tl2/RTbae6+Pi6vKc1v/NQpLoqSqicVq/YS4YSMQUB5qk4DIyOMV7NKBfhwdfucCHnmte1LtewaTCjOISWWQE9SLCObkUGwYFpCgSSQaFl3NSKZvFCB6R5XajDz9pkqiYSCuefyTgsV+doqO2mBfuXJ9XFn3MbeeZ+zbx2z86Ou82AolfBPdhNTUeHT6GjMmwrRCffpUr0hTv57+0/Zhz3lr+vfKuOY+jSSrfrHkAVRh8vv9l7KbGsyU7uOyu4vd6D9EUv9pV0eMowW5k2R++RKEWo9lbx9PB3VRkQuyOtOGYfm5rtIuv1X+Q0swUnxk4PLP/ta1sE1YPCdWOxTTQZIWw1cPJglU8V7YDyPW7PzR8Eq9+c902iyFttyAJgWsJNyMWzcCe0Ujb1AV93W8Gv3GFunxAIPGTmv10u0u5e+Qs8i0otSwG9YKGemHpZIeGnnEePXSG//279+TsE/MAU5EZKAuQsltxJzPYdI2Q28Xm3n5eWd/EqpFR0hYLa4eGGPd6mPC4c+V6SSJtsSAkcGWyNI6MMlAQIG2xMO7xsG5wkIjLhbrMslA+oOoG69qGKQ7dOuMSScD+tzv40aO78ldJmYYtm3M/LIgszKa2pzQ0i0zCbSMwMXvbuNdG+7oSNp+Yv8Vs8+VBfigEy2+sm42eqrkdGk0kumwlvOZpYsBagATcO3WefYl2BNBiL2dScbI52feeC/jr2ob59M/exqIvfv1aMWkSVyt6lSJKpZHjeVwZ2XvNrpllLAkoFPl3vbNldT77k2MMl/ho3lid12O/vL+JR59rpnBy7mtSkLvT/d3e5/FMB80r73fm/Eyd6vQEYasHyzw8iynVQZczyKcHXmPHVCdier//vKYeTVaxCmPGi8Bmajwwdpr94YsoCGqSY3Q5g7S5ytk+2YFV6DQkR9AlGZup4dMTbIjN3blztGgdrxdvBECXZSJWD2M2P7qsYDU07hw9x85w2y1TpCuYTNJTU8RoqZeNl4YW3DbhshHxO/FPpZDzvNz5nhInyxfEdT9XHjOv+//KY9c+zhz/z3ec6/+e77hzfWQS8NDL56nrnZjj2ZvHQLmfpCO3FuVNpXFlMtSOj3OmrpY9HZ34k0l+tG8PLeVlrB8YxJHN8rPdO2murcaU5JwJhhDTZSPBhoEBntm+lTfXriarvvu5oH+6n9uq31ppyODYFBXDK1Pxmgv2jMaBtxdnEQcHpigeifOTL+ygeXc1piIzHvTw9h11/MN/vpupgIO9r85fRmxqHc6Lh/dQme+Gu38BdNhK+XrJffzKv53TrjrOOaoZt1xdJgirbn5csI8e2+JeAO8mXIkMX/jR0ZtODqVrfq59TObWVjIloHowzKPPN+ON5TF5kCRGSry8uath3vqCBGyOds8E9iuPLRdWYYAEmnzV5y0t5+5qrxdjKc5GqUpPzDjCOY0MPj3BlMW1bC35sNXDZW8VZ/11XPTWMGQvQJdkVNPg7rGzfKLvdXzazbWiXh9X5sL6i0NkrQqv719F1GOfc3sBaKrMufUVtDeWsP7S0A0Ez5Xi/xd37tdj0uHkLx7+KO8/f4p7W3L9kIO+AP/3hz/NXxz6JU0jg7xTXc8Pdh+gs6iEiskwnzz5Jrd3tqKaJl94/Pf4yJnjPHTxDADf232A8+VVfO3JH5JWLfzWZ77Enxx+lp9v3UNbaRl7utv541eew65laSst5/vT2xfHo3zw7Enuv3RuTrcjVzLLh589zd9+6f68OVj1VxSQcNogFCeQSPDbr76BLATf+ea/YDEM7r5wif0trSimwGIY3Hv+IgcvtSABFsNga08vEoL1A4O5L7QQ/PmTzyCmn89HAFkOCiMJKnrDJFGxT/cc53tSlQBnKsu61mF6q4ryd2AhaOwep2oJyoGBUJIP/ftpfvh7u/mXP9lPymnh+1/ei6KbuKMZfvf/fZ2aztC8+xdMJSgKxxkpWYL07QIYLbqRpZ2WLDwV2EFKtvKVkV+zIdXPF2u/OGub1ekhIqqLXmsR69K/ua6KayEJwR3H2tjQMoj0HuX5LQTZFOx5p4s9p7p46UBTXsSWINdZ8/q+1dx1tBVPYi7xFkGBdmNb1nLh0tPcHrrEy0WbkYWJ20jxXMlO1sYHaIrNrkI5jAyO64RkZMFNVaIa40Nsi3Qy7AiQlVVceob6xAj3jDazebILRdycSFFCsZFSbYDAYhrzJgjve+ECTz+yhbf2NPC1/wAf+/lJgiNRlGlxJFOWSDitvHjPOn724e1YswYPHTqPLbP0yu9SsOTgnpUU+l3FhK0eNFnh2inWZmRpiA/jn36zAohanAw6ColZHMhCUJCNUZEMzbAar4UuyfQ7iwjbvGiSgtPIUJmcoCB74wVmIDFm9zPsKCArq3i15A1WfYFUkp29nTyzcTt3tl1CNQ2O163Cn0pSPzFKX6CIH+zaz+7udv70xac5U1nDD3fux5tOs62/m6yiYFwTbHPiAVeHKqNaeGLHbXz6xBsUx6LoioJV1wm73DyxfR/146N86fUX6Skq5jt7DuJPJjnQ2cJc2PdOF8GxKYaD+eltHSvykHRYZ8poqpnLkK8wLRUhcGhXEw1FCBR9rrLaNbK+1zwfrnbiHs9gTb0LJXohKArHaRv1cIhKdjBKNVFKSGJn6S0hS4Eto9PQO/9a883i4FutS1QNhJquMH/yFy/RsjlIf10BhiJTOBZnw+kh/OGF1wclAbX9oRUH97GiG0l7WVnloqOSB6bOsi/ehi7NXrqQAIswcZmZm7LjvFUoCCc4+FYb/mhq2ddKymphNOAh5HORtlrQldx7vv1cx5JUGvMBiZwvwoOHL9C8oYqxQk9++DmSRFd1ERfWlLP3dPfc2+ThLSqYPDB2mi5XkJP+VQT0BLeHL3HPeDOBBchvS8FCp7d/4hL7Jy6t6Phz4UhwA4fLN1MXG8FmaHy+/aU5ryv/VIo//D+v8I3fP8hrB1bz5r4GyocnKQgnkYUg5rYzUOEn6bTin0zy2C9Ps+Hi0KJ8kOViScE9I6s8V76TF8u2IpMrqfS5iompTspTIbaFOynIxvFrSQTQ5yzmF9W3cS5Qh4GMkCTcWoo7x87x0OBJXHp6ZlCuHPtwcAsRqxtZmJiSzPqpXh7tP8ba6FUikoHERX8NP6k5QKu3Eqeexqln2BrpJKHaZg30Pa0XeGrLLlpLy6gJT3CmqpbbO1qwGAYDgUIEcFfrBcqjk3jSKU7UNnIpWMGWgZ5Fx0MC9ne0sKNv9hdjwu3hVFUdmizz8227ySgqIbeHjpIg+ztb5uwa8MZS7Gzu4VcPbFnKR7EoNItKxOe8YY1sMaQ9KrFiO2rWwD2eQcgSsRI73tEUlrRJ1qGQKLJx8X3lrD80REHfyhTWlgJVN6ntD+HOZjlPEa9I1dSJKdYSpokQ6whTQn5IMRbdoGJkEmtWzxtDWRKCHc29y9rHntHZcmKALSeW2/YnqBoM8/b2+mXuNxuhgAtDlpGv6bc1kdAkBZ+RREEwn+O2NL3tewGSKdjUMsimS8tzutNlmYt1Zby4ax1nGyvpKy0g5rSRtahIQnDyt/8njuzVG5Rz9eW0VQfxJNNsb+mlKLqyoDUXtp3rY+OlQV65fU3eSFfjhR5aG4PsPNs7q7c6nzCROBlYhVtP8cddT1OcXVknAYAsTBymRkx15IW8uBzUxkbYN2qhMhmi3Zuzi5XmeH0J2H2yG1U3eOHe9TRvrqSvqpDemqtVQXc8ze6T3dz5eiu3H+1YkeLmfFjSLHbBX8Mvqm9jTXSAT/a8hsU0aPVW8I01j1CeDPOZrpfxaLl1oZRi5fv1d9PmreBD/W+xdmqArKxwtHgdT1buxW5oPDxwfGZt5XBwCz+sO8iWSBdf6HwBu5Fl1O7nXxvuJ17n4Ettv6YylStHhmweflqdI859ofMF6uKjTFmcPF++nUFHIZZriHTlUxH2drXyyy27+FDzCUJON1sHenKZky3HpLxSKldME2c2S0a1zPnlyajXiYMICEYnr3+IrJLzda8LjVMSz5FwmkaHWDsyf5nSkdbYeHmQX9+7KW8uZeNFnlyZf4kEjaxD4cJDFXjG0gyv97H9J7107y7CkjJou7OUrT/vo+u2YkxFYmytl/WHFiaJ5AuqblDfO84uRqgmRpfw8RblvCJV8xpVlIokTYS4jUHWEsa2RDW1uSCRy7j9U8m8CYgEx6MLiu6c315B0mlh59FeslaFpMuKbzI1by/8ghBQOr5y0mHWqhJz2ymYupq8yULgMjJEFNecYqwCGLX4ict2Akb+g9vNwJXKsLO5h8DU0pNQTZF5aWcT//zBA3RUlqCpyqw7Zck0b8iYx/1u/ufj9+NKZ/iL7z7H3aduTOJXCntW54PPN3Nkzyqy1jw5GVpVWhqDTHodFC1C9rweJjBkLySuOOh0BslKKp2uMhxGFq+eojg7hd3UkBAE0xFeKN7G3zZ8GFXoWEydmtQ490ycpTq1/EqZDGyI9vBS8RZ+UHmQsnSYlGLl/vEz2Obotc8nylIRAtk4p4saWRUdmjOwX4FqmOw81Utj5xgjQR/DpT4m/Q5MWcYbS1MyFqVsZIqS8di8mgMrxZKCe7ungrDNw0ODJ2mYtsYrT4X4ac1+xu1e7EYWizAQQKu3kmNFa3ms700eGjyJZXrAK5MhhpyF/LD2Du4aacajpwlZ3bxYthWLafB/tT6DV8vdha2ODqIIk79e/1GOF62hvP8YEoI+VwnvFK7i8e5XuGvkLFZTn7lDbfFWkVCvlgQl4GOn3ubLH/8cRfEYBck45VORnINbPEbSYiPscBFIJkjYbIRdbtaNDOYCvZYlZnNgSBJCkugqutF6VJ6DaenQshTHY6wbGWBnbxeQY6LnZHjnhmwKqoYilEzEGFmKVegSMFbomU5SlnbRGFaZVMBK8PIUulUhUWhD0UxqT4a4UGynb0cBml2h7u0JJhrzq6y3ECzG9J07Go1MUscUtzNISNg5ThmHqeZFaniFasqIc4ABDjKAj8x0n/Ly4Epm8EfzF9zXtw4tyNA+s7uKhMfG9mN9tGwq4/X7V/PpfzpGQWj5VREJKJlYuXqfQCLqmR3c7UJjR7KLNzxrWZ0eZnVmJNcZKilkeacZ7wAAIABJREFUJJUeWxFfL7mfIj1KU+rdSfwWhBAEJpPsO9m55EArgDOrq/jLzz1MyOdCFgKrlpP7NRSZ7DxtoNva+jElicEiP6fWVLP3Qieua9ZOkxYrVl2/YZ1XkCOayUIsibW9/VwfG1oGOb2pZonvaHFcXh1kvMhDYeSqJrwsTOyGhrJAopyRLXyt/oM5Brokk5VVfhncx69Kd1OSmeQL/S+xIdbHoL2QQyXbqUmOsSoxhCIMkoqd075GBhxF/F7PIYqyUSzCwGZqs+7CJQRWU5uTif/xoSPossobhRvISgqlmUnumjiHbZ6aUr5gMXUGXYXcMXx+SRUUWQgKI0kKI0maWkYw5aviYPJNGEYtF0sK7hIChJjFor1SEjHIifFfwcnC1Xj1FGujA1jMq3l+QTbGmugApwoaueyrYleonV5XKeM2HztDbTOBHXJCEk1T/QSycTo8ZcQsdlx6hnZPBXYjS2NsGKt5lVzVGBvEpyVmBXeAyskQGwf7+dWm7XzhrVfxJ3MZal1olKrIBD/eeRvb+rppCZYjCcH23hwbeXdPBy+v3YgrmybsdNNTUEz51MJMagkom4qwq6eDn23bw6jXj8UwiNod7OztZNX43H7BEjmHr9LxaN6C+3ihe1n9zvYpDUxBz+4iNj3dj6ILTItEvMiG5pRxTmqk/FYShVayjneJgykEqmZQNXh13BUETnQcxKmknQ/RTjPFvCDVcopSvi028T3Ws5chDtLPekJ4yS45yLuTGXzR/DGT17aPzipvX4+U05pTdBOgW2RSLsuKiJXFoZWToJBy3uPXwip0Hpo8w5AlwN+UvZ8CPU5GtnDIt5mXvBsZsvgp0aN8fOItKrSVGZPkA5KA7Wd7KRtdmlsYQMJu5a8++xAhnwtPMsP+s+18+PUzbOwc5Nf7NvJfP//wnPs5Mhrb2vp4bdsaLtSXk7TbZgX3L3/gi3z+5GFu753NuYnZ7Hxr132sGxvgodbF9c0Vw+Qjvz7NmQ3VeSPfDpf4aK8rYXXn6Mzd45ZoN989+w8L7mc3Nf7XpX9d9Pj/UnMfdlPjL9t+NEsJ7oy3nm/UPcS4zUdJdorP9h++Yd8CLc6fdv5izuN69BS/3/Pcoq+fb7xZuo6Xy7fyp+d+xuHyLXyk5805A7Q5/eC1a+iyEMg3U5FbAZY0U6+ZGqAwG+OZit3T5WyNVm8Fo3Y/dw8349avtkyM2v3YjeysYA3TZc9sAoupM2zPiaJELU4yipXS9I16zKowKcjGmLK4ychWnGQJ29y49Ax2Y7bQh0dLzyl/KAnBvq42WkvLWT88MDPJFyXifOHYaxxes4GzlTWUxqb4UPNJKqfCCOCjp9/Grmm0lFZQFo3wn176FT2Fubt3xTR54FIzpdEbJw53NsMnTx7lSONaLpVVAoLGsVECyYUnXV80RXE4DxPzNOIu27LW5mKldmRdYI9pdO8rZt2hITyjGUbXeinoSVB+PoIpS0zUuSnsjmOLvzv64b5YCt8cbUA6EiO46MTPeYrow4sHjSbCONDpxcvfS9vZIUb5JJepZH5luGvhSGl44nmyfhSCmoGJBbsL3NE0F7eW09FUzESJm5TDwnClj5Rzbo14e1KjaHz+Eqo3nrqxIXm5pw25botrIAEVWoTfHT/MG+61tNnL8JhpTCQcZpZ1qQHuiF1mY2phoZ53C7JpctfRpREZr+DIllV0VBZjz2p85NVTfPmXr+Of9r5QzPkTNMU0aRwY47Vta+gvLSCzRL6GLEBXFGK2pRsWbbkwQOlEdMWkyRlIEm9vq+OeNy7jXIa891LHddTqpyY1PqtqYSARtnqQBQuy1hcTArseQoJ4rRXLpIEtcmvIviXpKYKpCCPOAoQkXSOMOxuH71yLPa2z/62lad+bwHCZj76qArJWFf9kkrqeCTzxlSlyLulKbIr286nuV/mn1Q/R6yrBYWQxJIl7hpt5rPfIDSewcH4izWwvLdoxeN3giauP3oA5m8kleguKWDc8QP3E2KynyqKTPH7yzTlf0ZtO3fDcrr7cXb3VNPj9N1+e94z96SSPXDjNIxeW7jbkSaQJTCZXPDFfQdaqLotmkvao2BI6triOYZVRdJPaEyFMeVqjH6g+Hc5VA8S7o1QIUDE82+Aii8xZijlFKa0U0Ct5KREJdosRNjBBPVM40OnHw2lRwrNSPT8Va/gjTqMsYUTsWQ1nnogtrlQ2J+O6wMtuemeAE3fU8+2vHEBXFaYKHHz/y3tR5+npX3dmiMe/dXze4znSGhbdQLOsTIjn+uAOuc+8XJvko5HjTCkOYkqO0GQXGgV6HIvIb/fCSuCLptjQsrx2vDc3NmDIMrVDIR5+6/xMYF8MkhAEwzmuQ9jjRFMUTCSidgcZxUJWUZl0uBhzXROQJej3FdJVUMq60aURJyXAlUyz950unnxw66LbLxVn11eRtag40lreP78tU10cLVzHU8HdVKfG0SWFbmeQNwuaWBvvJ5jOn66EUCCy1YV1UsfdnsYxlDNpSZVZwATHqIbuknOGXkIgZAk1aZIutZANKDj7s6hxk2xAIVuo4uzLomRnf3ldWprCTIx+VzFbQ53zzilPPLYTxTBngrshSzMCNXPFyrObq3jisR10NJSQtSj4oik2Xhjksz84RsnY8qWgr2BJwd1q6kxa3bj1NF9u+zWFmRgWUyeQjeO8rjexPBXigr+GKYtrVqwSMNNGV5bKle58WgK7kWXYUXDDG9ZlhZDNS2VyfPpOXVCQjZNUbaQU66xjJ1TbdHteDrokM+l00lEc5PDaDfzJ4eew6/ntIcwnVN3EE0+jGAZGHhTSMtbluYMV9iRwhbMIWUIyBI5Ibqzk62LMu90rXD6aq+iEsXGUCl6liiHcZFBYS5g/FKdZQ5gAGWzXtMatJkINUcaFkzepwERaUnC3ZfRccBdixS1HRRMxnKmFJ8y150b4o//yEq0bg1zeFKR7dTEbTg/hTMydYJT3ze84JgGyIXCmMjOuVzcFKdeDOx9kBAEjScC49d0SN4utF/qXnaT1lRZgShLFU3HqhpcuLCUJcKZzc2DamiPUJq1WvrPjLl6vW8+QN8A/3PYQTm32+WQVlcqpCTaMzq2yNhdsWZ3dp7r49b0b0fIkBT1R4GawzD9nhWyl+PjQESzC4KWiLcRVBwomAS3GvkgL94w3L6gFfzPI+mSSFQ4im5xUPRlB8ypM7HEhGYLiN+Mkq6xYIwaSLjCcMt7WNJ2fL6T4zThqzMC0SvT8ViGu7izRNXbKnp9CvqYIUBcfJZiezHkVXFOtvgESjJZ6efHuJt68rZHBMj+SEGxt7uf9z56jcjAyMy9MFLl56pHNHN9ZhzWrUxhOMFriZeheP7qq8Gd/9/xNE+6WdIVMWVy8VrqRzZFutka6kBcop+wZb+HJqn1c8lWzLdyBbbqvfcLmpdVXiUdLsTaaEzCoi48STE1ysmAVEaubQDY+005z3l/LpMVFY2wYt55GQrA6NkhKsdDuqWBLpAubqeXUszzlTFldM+cw6XTy93e+j7bScj514k1um6fH/L2CKy1xVs0glY/gblueO5iiC1zh95ZkKDCzZvoitTzBGtxoHGCA++ilhijKdOPVXC2GdgyCJFCXwaBXDBPHtC3jijTayYnv2LILL19YdJPajhA1nSGKh2OYisz7fnGewvlK74v0V8tCYM2urCQpgPQc1rEmOTGblGzFkOZPldxmBpeZp6WNm8TtJzqWnYgm7bmERtUNHJmlLzsJKRfUAeyahmwKXNkMXzr2Ave3NfPVez7GtqFOGieucm4koCgRY+NILwWppS/HSQIqRiapGorQVZMfJUAhS5zZWMW6tvxbn/r0JJ/tP8xvDbw6c71I5Eh7V9ze8glL1CTQnCJVaSFTqDK5yUHJqzGyhSqJGhu6S0ZJCmRVYDhklJSJrEOm2IL/QorwDheJKiu6Q8YW0tFdMtbo1fmj3VtOeTKMT188EYr4nfz1Vx4AcvOKkCQ6Gko4t7GSP/76y6xpG0Uipwh5YX05haE4f/f//IKKoUnOry/nH798F4fvXMuHnzrN2rbRhV9sHiwpuCvCwJdNctlXxYvBrXj0FBICm6FRnJ6iIhVCnQ749fER7ho5yxsl6/HqSdZO9aPJKkeLmuhyBflkz2s49dyX36cleXjwBN9c9T7+cc37eXDonZlWuO/V38PmyS52hdpmWJTViTF2hto5VLEdvxanNj5KzOLk1dKNJFT7zMVSlIjz3379s5sakJuFAIQlV7eW5ygSZAsVhALWsbnLl85UFnUJ2tdLQdYye/LVy2WkhECeWvgLJQCm53XpPVDouEIQq2WKz3GB2xmigPSSJ4VqotxPz5J7YSVyrn2KYaLLK0uyCqaSWLXFg4REbtL2Taao6I1gyRo3rTEtCbFoQrEUXBFsuQITGLAW8rR/Oydd9YRUz7Qfwo2fxOMTR/hU+K0Vn8PNwpLVaWpffqAqiOY4QhmrhZjTRkFsaZUJU5bpK81VHgunEjN+23ZDY/3YANsHO7mr8wI7BxaXIF4MEjnd8tq+UN6CO8CZDdV86hcn8na8K5CYdtl7l7w+JFMgZ0XugpXAPqITb7RhOGUcgxqyJkjUWZF1gRozMS0S5c9NEd7pJLzdhX1EI9CcouBkAmtExxKdfWMw5vATtnnwZZO4tRS1ibG5T2QaBZEkW5v7qBqIkLZbuLQ2yOWmcg7dt4HKgQjuZJb4tLb8HUfaCY5GsWoGW8/2c/CNNn780Z0c3dtwa4N7UrXj1lO0WCv5Ud1BZGEikNBkhdL0JB/ue4vbxi8hI7CbGp/oeR2HkeX5su38qmI3kHPi+VD/Ue4ZaZ412d4+fpGsrHKofAd/v/YDyNO+4xsne/jAwNtUJq+WyAoyMT7W+wY/q97P9+vvniHu7Z5oRQ+rNAfqbmoQFoMAtICMPG3lalol1LhJutyCrAusozqmTSKy14mSNPFeyKDETAynRCaoYg0ZxNbb0AIK3rNpbGM6SnL2BG7Rjbz1O15xkRIW0EtkkvfasZ3TsDVr6MUymKCETBBglMoIu4QyZIAqkdpnRU4KbOc1pJjAKJLBAlJGIIcFpl/C9MuogwamXcqNiQbCJSFlBVIeCwBXgvtucnc9y50c9zDCHubuUpgPtqyOYprorCy4ByYTWJcRaCt6J/FOpXHFbv6uVxJiWa85H4zr9BaykoUn/Tt5y7OKTck+KrMX5/XB3pRaepn5VqBieBJvbOkJ4BWs7h/l9S2rGA14aK8sYfflniXtl7BbObqpYeYYjszsL8AHLp2kKLHyFsUr8EeT1PZPIBtm3nQxLjcGSdvUZVUs3muQTPBdTGENGygZEzVm4O7IEN7hRA3peFvTCAl0lxslY+Lsyy1DZgpV3F0Z3J0Z7EMa2UKVVFluXmd09ngUZGIMOQtJK1Z0WYYFVhUcaY1P/fg4D75wAXsm17I9VObjn373IKe2VfOxnztwJ7MYioyuKnijaaTp+V8WsP7SEM6URntj6U2PyaLBPWR186PaO4hY3fz5hSfwZxMzTmIJ1c7X1zzCofLtrI32U5zJKRAF0xEe736Fe0bOEFdz8rM+LUEwFcFmzl6HVIXJ7dELVDNKNm3FkGTsRpZgehJ/Ns54lYfJYgf158ZR9VyL3Jfbfs2Y3Y8mK7j1FOXJMBGrm5DNO2+W2LmpmOKBGN7w1bWSy7uC+MaTlHUvrJwkVBh9xIs1pKNGTTJlFopfihHbYCNdaaHw1QRq3CS6zY6jT8PVmUXKSvT9XgHulgyiPfeFn9zpQDIhKkHJs7FZ6zmqfvN3bDecby5lJrPRQrbJQnadirVFR2tUSe2xggWcL2eQUoLYYw5s5zXkqMC0QWa7BbXPwNKpo8QFka+4sb+dxXZRR1hM4h9yoA4YZNaoqMMGZkDG1qyROmDF8caNJJSVoGAyF9zfTaKWRcvP5+CLppZlBOFIaThSKyuXSCJXAlwpdGX2iGdklXdcddwWa+MT4bcI6Il5qyELCXu8G2joHb+p6sXBM61896F9DBb7eXbfRlb3jxKIL1x+NSWJ771vD50VxSiGwe6LPXiSs5OzpvH86uyruknNQBhvPM2kbwXcimsQd9vpri5iXfvyEuH3EiQTvG25sbdPXP38Sw/nyI5XqNnlz07N+r/kjennpy/bkldjM09eP+9sjPSyPtK3JN6zI6Vx8Egb9umESQKCo1G2Nvfxzraa2SqYkjStO3/1u1MyHsOi6UwU3ryuyKLBvc9VwonCNXyq+1W2h2frKhuSzKZID63eCqYsTooyUcZqvARGEqiKQZFzirrwKCM1XjS7itRhEnXaQQJbUifltmBLGxx/sA4BNB0fxhNOo9sU3MkMoXIXcb+NsSoPWbtKVVsY70QauyuD0pChsieKL5Irp7lSGSpTITJ2lYTPim8iRdyfY/1KAjJOFUUzc2uKTpXhBj99TYU0pHV0q8xwXY70EOyeQhIw1OBDSBLlXZMopkmmQgUZdI/AtOQCqO5TiK+z4T2TwtmrYe/X8FzIYB03iG20YZ3QKXo5AYYgU6biPZvGcz7N5E4HpkNGjl+diGUz58aWLwirhF6nYnsnm7sYTUgetJJdoyJsEsqggeNEFqNIxnTISGmBGheo/Qa28xrKWO7cTIeE61AaSYPEg3asFzQs3TqpvVaUi4LE+62oPQZ6mYIcze/EfkVdrAMfR6ngIP1UcyN7NIaFF6nFgc499GJdgVKdbJor/xyEwJ3MzMt6X/bhgOEqHz2NhexbwBkORF6MfQx19h2hiYQuKaxKD1Ok3zx7991AQ+/4TVUvtrYPcPB0Gy/uauLnB7cSc9r50i9fo3J8ctYyRVZVkIRg3Ofhuw/t5ecHt2FKEhu7h9h9qfsG0aKccqWKLisYcxi/2HUN6xymUvNBIsdF8UVTeQvuhizNCu4CGC30Ys3qeJJp0jYLLbVB1nUPr1gmNdcfNS3mgmDYFuCMp54HJ07NBM05W93IXYdZRUUgYTX1OcV/Rqq8DKwK0HRyGEc8O20+A4YiXe0zl8CQpGkHzNxDpizB9P/z8TWertrNa+WbCGTiVCbG+Vz7y3OeqzWrI5kC+3XJumwKrJqBZpFnBG1m7L6vO5A9neNvzNcWuxQsGtw1WSGrqNOl3qutaQKYsjjpdxXh0jO4ptfRJ8pcDNX7sGQMFN1kvNJN+5ZSAmMJpgrtCEVC1gUlfVH61xSw6swYaaeKN5TGltIZqfcRKXHRdHyYC7dVUDwQY7jeR8FIkpceX8edP2nl2MP1BMaSHH+ojse+9g6u2NVB1Gwy7VtLqLsY4sT9tex9pgtnPEv3hiL8Y0nsSY2WXUEmS1xMlLuouajQvrWUll1BdIvMtpd7yTosDKwOYE3rpDwWVjWPYR3TQUC2RMF9MUP4gAtZF3jOpnNr7IZASQuyhQqmPVe21z0KmVIFdXrtRo2byNr0BXXdh5n3CdMEKS0wimVMjwQSqAMGlh4D62UdZdQAGbw/SJJ42IE2aGC9qOX2KcyV6qW0QEoL5OlihxwxMYpk5JiChIQcNbFeNog/5sB5OJNXNr0jrc3c+Xbh59fUs525157s6Exi4xWquJ2BFQZ3seL3oRgmjlQ2rw56nWtKeOuu+gWDuyTIS9Xh+rK8IkwKjDhJxbbkzoPfBCQhKB+ZxHITSZVimPzHH79If0mA1ppSnrltI4f2rmdtzwj69HgISeK/feZBRgq8XKorY8rlAKA0EuOTL55kbe/IDd/jYU+OLX+2vJYpuxNJCFTTIG6148mk+cobT/Nwy6llnWvpeDSvNrCmLNFXMbtj6VRTNW9ubWTPuS76SwuoGgvTUlvKp5+7+bV5ASRlG22uckDQmBzBQCalWBm2BkCC0szknNeXicRbJev49uoHmLS5eaz7CJ/qevWGbU1FYrLYyeWdZXgiaRrOjzNR4Wa0yosvlCLYM8VojZdYwI4lbdB4doxYgZ2BxgCWjEHD+fF5k/L1k31kFZVd422cKmqc9w6+dDRGf2UBF9aXs/1MH4qZW2qOue101RVhyjKdDcUUheIMl+VaJDNWdZY2SdaiYkoS1hUslSwa3IP/H3vvHSXHeZ75/ip2DtOTcwIwyCByIgFGUaQoSqRtyVpJtmTJtuT1WW+wvWvvXXvP+q50jr137zroOkiyLFu2skhKTCIpgBkgSOQ4Oeee6dxd3ZXuHz0REwE0hqKs5xwSPdXV1V+F/t7vTc+TidCQHOWZqr3IlklJNgYIxBQ3Z0Lr6PBW8vGelynN5sMdDVcmOP7RFkIjKeovT9Cxs4wtbw4i6xbDjQEQBRTLxBYFLFnAlcgRGklTPJwkNJomUuHBFqY8YyX/o6ruiNFwOczg+iBDTQEsSUQ0bBovhjFUCZg17p54DmfaYGB9EbYoUDSWQrTBG9HyOWZJJB1wsP7MKHLORDRthtb5cSV1PLEssmHT2+hn84khsi6ZWIkLSxQoejOD6RHRQyK+q1lMj0h0tws5aaFOmAgWeK5mSWx3ohdJuLp1XP06sd0uPNeyOIYNbEVASlq4O3IIBQxfLwYhZ+M4r6PtUZAmLKQRE7nTJnNURduj4H7FxpYht0nJh+BHTLBAvWKQ3aFgFhnIQzauN2ZX6q63cqQedmLUSqitOlLcxvVGlvT9fhz/p3AkPAC+pDZjZGOoeASdKntxMhoFmzLSRAQnpn1rechCRFAcOQNndjEW9vnQZRFLElCzJraw0GOeho1AIrA6tbVCGHfruk4Lp61zKNnGeVc9zdoom7RB1J+ivvZpuDI5imKZm1pUCUD9yCR/9LVn+OvHj/DWpkayDoVLzdVzdhJ44uj8HvOG4TCffuZNHnnzwqJqcd/ecZieolI+c+olLlQ2IFsG24f7uFBZT1pxcMdQ9w2PtTiaIhhLF6RlE/I1Ov3XGXdfSuOOa/1EfS7a6st46M1LnNh26zVN53wN+bY428L0SBTnEgyrRZz1N7E73rlkuiclOzlZ2kK/twxsm6dr9/PIwKlFlUNTAQcNV8K8/b5GPPEsndvKqOqOcvqeOnYf76NtZzlV3VHO3F1HMJzm6p5KIhUeRmt9+KIaVd2LMxs6zRybov2cLW7GaeZYwMMyhYNvdfH6oXV87VcOMVbqp2Isji6LXN5cxfGjLQRjGb76q4e4uLmK89trwLbpqS9GlyUcU/NGf22IrEOmqXv1bZnXY0XjXpWZ5OPdx3m6Zj/fqj+KJQh5Kj0sKjIRfrPjOQ6PX50JkTgyBv4JjXjIRWg0TePFMBfvqsGR0ansiuGf1DjxwSbGa7xIev4zwfE0lw5Xo2omgXCG80dqyLplEkVOKrtjhKu9nL23jqxTpqYtwkSlF0MR8cRz+CLzc1yCDaUDCV7+SAu7XurDFgSGG3z0bQqhqxL7nu8hMJ7h4p3VRMvclPXFqb86SevuciTdomg0TcPlCS7eWY1o2TSfH0fSLXxX53+PDbg7cvNyM97WHN622W1lP8znb673BAPnlumRLBRsUHpM5J689zszxh/kv1uYCld5ns4veqbHqF4zUFuNmXPwPTU7VkEHz1PazDnp9RLpexz4/zmNUOBaHG86O2NkTUTUFbjiVUzMAmR8C5EayRv3lfPnr79vPW2by/jM/3mD3qYQT318aXKSodoApSO3LgyzGkQkD18tuXvetpjkottRyt+U3U99bpyQkVo0LLor3c3OdO+ajPN6BGNpPOmbL0iUbJtdbX380dee4dU71vPC3s1caqoi6Z7PIicZJrXjUY6ebeP+d66y92rvjLTy9WgrqeT9bef4xUsn0aW8d/aBa6c52NfK3xx4kM7iCmriN0bZq+p5BUPZsDBukbAI8o7UWIkPTZVxTqU0NvaMkHGqeMc06ocnOb+hhvrh5cdpIdDnKSXs9LMuPkxQX1hxNqH62ZzsIyuqRBQPliCQkF1oooLbXJqRLd9CN/vbFG1rUX0PgLL+OJU9MQLhDPFiF46MTlVXlL4NIRJBB+5EjvLeOAPNRSSKnBiqRE37JBvOjOCJL512qEpPUJWeoDibwKtnllyIHDrZyb63u3n98Dp66ksIxDMYsshkkYdgNM3Hv32KKxsreeqDd6AYJtsvDjJe6uV7j+/i3lda0RwKz79vCymPgzsu9C/6HavBisZdti12RLppSo6Qkl3ooohgg2ybeRUgPT3zI8/ndm1KBpOk/CrOlE51RxRfNJs3JPEssm5x379cQ7BsVM1A1QwaL4QJDafwxLM4Uwb3fKcVG5AMC0faoKYtgimLKDkTbyTLnhd60DwKSs6cqTCci/LeOO//+0v4ohqiZRMaTvHwVy4hGRbuRJYNZ0ap6oxiC+CNZpEMm6KRFAgCnngWT0yjaDSf7/VFNYRFniFh+oSX2bbYPmuNBT3g14/veqIaWHbMc9+XRk08z2n5yvsCw52e/ZF50ImhkkYmwMIJwAIiOFGmFgA2s2R/S+XvpnE7vE9FN1dVTJdxKUyW5CNVkVIPF/bWUDEQQ17wWYG01wGsjXFPSC5OBjcv2G4JIhN46VOLp1pfFz4oHkt714x70S0ad8gb+IaRSapeeof3n7xCxOdioKyIcMBLVlUIJDNUTsSonIhRFE/h0XLLPkOKac7k1N16lkF/CF2S8GczBDIpRnzBmX0NUWTS4yHldCDYNvXhiSWPXTMcRTbNghj3aeKiiZCX6pE8UVJxLMUdrf2kXPmxbGsfxJtZ/tpmJYWvr7ufK8Fafu/S99kzsZB+dUeim+dLdqFaBoeiV1Fsk+b0CLVamIu+OvbGOlAXKYr2GBqHxq5yqaiBuOLi33S/jF9fPDUhTEXfBNumaDTFZLmHn3x0E1mXzI7X++neUsprH15PMuDk7u+3knXJXN1biZLLh+WXwrRzUZlZnlnPm8zyO186RvVwlBfv28RIuR/Bhpb2EX75O29z4K1u7jt2jQeOXcUSBBp6J/irz9/Nt39xD889uBVLFIgE3QTTM+8cAAAgAElEQVTiGg8cu7rsdy2H1fW5YxPU0wT1lXs/Lx+qZqzWx6Ef5vs6JcsmOJb/3HQQIzScmvkbQMmZlAzOVkYHR9Pz3ndmjHn5DW88N7PCWuzhl0yb0NjsWB2aiWNkzirSMCnS5n/HtDGf3lY0Nv/9n2MhRA1E7fZoQbszuRkvuoEYHlvnn9jEb3IRNzoyU+2YiHQQ5A2q2GRPoGKSkFw8Ubafrck+die68oyHQn4SlG0TC4ExNYhkmxTriYITasimtawa3DTe/8Rl3vfk5RmJ123vDPCpv3yTwOR1k5YAL3xoMxd3Vy9ylMKjQo/y6f5/vqnPFhuFTc/cCAIJrWD0waphUhZNUBpNsKF/bIrUKD+DTS+QV/PMNE+OcKWsBoCqeITnWnZyqaIOn5ahs7iC5snZOpLOsjK+ffAAlZEIkmXz6VdfXfK4ZeFEwQo2ATIOlbES34xxP9dSyw+PbKduJO+tf+LZUziWWbDaQFpycKpkA7ooo4uLm5Y6LcyvDf4EG2Ykuqu1iamF4uL66JAvvtsfvsaeiTZAQJoiwrkelT0xKnrzRdEPfuMygmVT3h/HUCQkwyIZcFDWH6fpUpjQaBLJsNl0apj158by9RC5W7+mAvlq9899+VV+7etvEPe5cGQNvKksomUh2lNsg6dmUzKf/7tX+V///n66G0qxBajtn+S3/+ZliidunsWvoBJfArD9tYVcycISr5fadrP73AhWOt7Pjfq7C2d2tmWyhQh3MshTrGMUD3sYpZgMFgID+DghVGLYIr/KZZwYiKbBxtTgzGIyKnto9VQj2hYb0kMYgsgLxTso0RPsj7VSrBfWIMmGuaqJV7TsGW/Ak8hS1zWJK5VDvq6dzQa88bVjfVNtg/XZmyPOeDfhzmQL0uc/F9PPYN6g33gY7v72C7zZsBGAjeMDrAuP8N/v/yiyZbEhPMSW0dmwayCT4Z4rV2geHVsxPVQ6kShI2+M0cqpE1O+a+TuryBw528HmriEEGxRj5et6NVhLSnGhmounpGacues889miuOXPWWRxqe3rv2Nm8TXN526Bms1/p5o1KR1M5qPIxtT79uz7hcK0Y+vSDFzawojb9faldjDCn/7XJ2hbV4ahSDT0ThCMpm/JDq2RfufPBizAkCV0RUSXZQw5T0CgyyK6LGEoUv7fme2zf+uKhCFJ6Ep+f2Pue7JER2MZKdfSnN7/2uDMGjO/dRWLx2nHg87rVPNtoYX01KMbJMtGe5L30ctWJhbNy5/zNSDZFoao0umqoCkzgtvMUpmN4DELbzRlY3We+1w0tE9QMRBf0D4zjcb2cEFbJX8W4U7rOJaoLraBtEclHnRSFE7hyK5NQeCmsQE2jw0gAP6sxq+dPsbuoS4MUWTD+BD10dmCKU2RuVJdxbjPh2RZPHrm7JLHLZ1ILFgE3gp0RSLhma0tKI0k+Mn+jSTcKqIN73/jMuoKz/Spkg0FG8/tgiul03xp6dD7uwlHzmDblaGCHe+WjPtFtQrVNmnRb36VnxVk0oJC0Mq8K96yDaTcKuFiHxNBD9GAi0jATdznIulxkPQ4SLkdZJwKWVXGlERMMd+naEkiliBgigKWmK98nnktCFN/T+07td2c83p6+/Q+hah8/VmB47qCtCA5HqODgwwRsx1oyIiAG51iMoTQWEoLLyG52JLqJyk5SUlOnJaO30hTpCdxWYXn2ZVM64a9qpVIbKp7I8sKx/wcec99MQIbG4iE3LzwyGb8MY2cQ+J9P7qC+xZJg1aD6xebxekk93ZeWnTf8licA+2deLJZ3LnlF52h6BS9cYEq5nVZIumZ7cioGY3w+E/yi4uUy7Gs7C2AJiqcLl5/y+P4OQqHVRt3m7yq26TqpT41hohNVHSRFh3kBJl6fQKvrTEkBZmU3DTpYdx2jk6llLjoolkfJ2hl6JOL0AUJE5FaY5ITzibCkpft2QFa9LGCGnhLAFMUMaW88Ux6HPTUldBZX0pfdRH9VSEGK4MkPY4ZA2sLU8ZXELCn/xaY7UF8DxlgUyzcWIU5IeS1gKovrH1Xsagjic38MPrcs7SBbmcZJwIbkLCpyEbYmejmmZLdKLbB/ZMXUC0Dv5nhuZJdfCD8DvXazbebLAbRtgvGNjgP751Hb80hWlZen2GJRVX3uhK2nB+iuX2c4w9uJOeQcS2i2meTN3R9ZSGO7W7h7c0NjBb5SLryhu/Z3/0rnHM0A3JSPvom2HmCkuvb4Wwg7nBxvHkrV0ur2TXUzd1dlxFsmwm3j+JMcqbgbiAU4m/uv5fPvPwKz+/YxudfOrZoex3kF5D+hMZwgbTd5xp3XZaQTQvvVHHiK7s3UBmOIc1ZOPV5Smn119Dmr6LdX0W/p4xJhy9/TUSZ/3v7x5DtxaMo5Zkon2t9ll2TnYu+b5OPAnxh+0eXHO+G2CB/dvrvV3Vu/9B8H0/WHaIqHeZLb/01McXDMzV7OV65A12U2D3RwWO9b1KdDiMAI64inqg7xMnSFixEdkY6eaz3TRqSY8vqVNjkeWF6POW8Xr6FC0WNhJ0BdFEkkEuzIT7InWOX2RrpxW1oSxb8FgqrNu7jjgDfrD2K08qxM9JFUE9imQLdSnF+ZSy6qDMmedJ7B/X6JD1KMfenrzEoB4mJLlrVcj6WeJtv+PZxSOui2ogCApqgoNomHvvWC2FsIO1SifucJLxOhssCtDaX095UTntjGaOlfkxpkcv5HjLYq4WmKjxxcFfBjre9u58tfYVXjloKsmEumYJb6W41amP8Tv8z87b9xuCLM6xYAAdibeyNdSDeAuHNUhDswjDFzUV/Y4jOjaXc+2xrQY/7swLJsJbVJK8cjPHG3c2kvSqaU0bRF4blLQEGS4J884F9fPu+3TMkNdMQpnTA5+KZQ1v5+kMHKY0l+e3vHWd75+C84474iviDB/8NYY8fTVGRbYu7eq4SdXr4H/f9Ep86fZwD/e0ABNMptvf1Y4gSmqJiSBLSErlugTzFcaFgTBl3G/jxwc14tBwv796AO5ultb6Ch964DMyO5b/f8XG6vBXXDWrWAUop89sH58Jl5FYUZsprl8h5R1Cc0n+cPr5tk1Bcy35+LjKSSlT1YIgSg+5ivttwF0/X7MMU8u5Kr6eMTm8l/+ny95Ftiy9s/yiXgvUznx/wlNDqr+E/Xn6ClvjAos+YiUCft4wf1B3mxao70KT5KdZRZxFt/mqeq97DHZNd/ErnT9ga7UVaoYbgVrBq456QXdSmx4kpHiKqB7+eJ4fZqfUTslKMSj66lFJ8lobfyuC1s4QlL4Yg4bR1kuRXhT4ry6FMFyr5/E2JmcRGoMaI3tQqxhIg6nfTWxOivzpER0MpnQ2ldNeWMBHyYC9C+fivAQmXkz/49OMFO97vfff5NTXu1+es08hcI8QIHnLL1Lc/RDdOFuYGF6vCvRE52BtBnsKysMa9d10xb9/Z8K4Yd10QafXXEsolKM9EVmSoi8suws4AoWxi0V7n2wHJWrrOQQDKh+Pc8U4/4+U+jr7Yhjs135mwgZFQgD/9+IO8uHcTxmLSy4vc0809w7TVlXNVrOTgxU429YygmrPj+M72g8iWxRef/wavNW4mpebnQbeepSwVu64VTiIrK1yuqaY+HEZdoYjNnywcX4YtCqRdKjlV5pHXLtJTWcwd1/oJJdKc2Na4QOHw0NgV1sVn88OGKPFK+TZMUUK0LO6IdFGiLU4GE8ylKNGWF9OpT43x6fYXScsO0rKDlOykzV9NW6Dmps9RFyV+WHuQE6Wb2BAfxGHqtAWqSctOzocaebliO0OeYtr81WyID+IxNDp9lcQVNx3+Kl4r30J9ahTXdQWDNnk52L9reYhzoWasKXrc2tR4XosFm6TsZMBTSlJxcbpkPcOuIv5t69McGrt627z3VRv3mnSYYWeItOzEbWapyEYYl5xzwhQC6/Qx+pQiTEEkaGTICAptSjnVRgTZtuZ5TtMIWWledq1HwWRXdvUN+zlF4tyWWk7vqKOtqZzBiiAjZQE0581z8f4sQTUM7j4/X8d+mgjClc1xqb6KnvISFMOkejJKWSSOK6djyBJhn5e+shAZVWH90BgffOs8Ry63r+n480IKeSRR+C4beJlaxgUXOtNUyAtxxB5Y1LivJaZ7bAuJtOfdK7ZMyU7+1+bHuWf0Ah/peXXFOoVWf1498sjYJR7rP7EmY5RMe0luARuIBV28dWcT66+N0tNcjCeVw3ldWP7vHr2LF/ZtxpREqsci3H2ujXUDY5xvruHJo4sTDNWORWgeGOdqYyXn19Xw2KvnUOcIzlwrreburktsHe3nTHXzjHHPL0YMNHn2vgbTaR49cwbRtqiMrFxfUUgKWgDNoZBVJdScQcPwxMz2Xdf6FyycPtF1HGvO1UvJTk6UbiQjSsi2yaN9J9kXXnwhKmAvy6cvkO8l/6Xe14H8/cuKCt9qPHJLxj0nKrxcsZUjoxd5aOAdVMvg1fKtfHX9+7BEiR/V7ScjOXhw8DSP9r+Fy8zyTvF6/nzzh7EEkctF9cQVDy5z/r2JK27+ruVhzoSaQRBoig/zi72v05wYxqfn5dHTkoNuXwVP1+7jfFETQ+5ivrr+QUq1GBvihSuim4tVG3fVNtge68ZnpKnQIgT0NLuNvnzJv21RbURx2gaPJC9iCSIeK4timzyePItim8jkw2AfS7yNMmfyrdcneMTK4rSXnzBs8kxK4ZCXn9y1iZeObGKoPEDC6ySnSD+TofVbgS+j8Wdf/d6C7QI2L29r4e0NDezu6OV3nnyJxtEJVMNAtPJqf7okMRwK8HcPHeFccy0No2E2996eB3ApzJW/vUgJL1FHMRqftS9QRmZJ7zHA2rWMLYW5ghTL4duf2cPJo02rOmbC72T9lXenPc0URHq95UxEffP4r5dDh6+SYC61dsbdslCX6cPuaS6maCKNJYqEK700t4fndSZcaazkqbt2YIoC+65085+++dKUhKuOYLGkcZdMi63dQ1xtrKSzuhTNocAc4+7JZYk75wu82IAmK0y6fXiHe2a2h30+vrdvL5XRKA9euEhpYnnSokCBjbsh5+uTbODM5jpKI0nqhic5treFI2facc8h7HFe571agjhvoeSwdNxmYTgHpm3MYoyINwIb8Boaj/adpHZKSvze4fMcr9xOh7+acUeA4myCX+18iZJs/tqHsgm+1XSUEVeIIVeIzHXhdhv4Ye0BzoXyv+O65Bj/4coTbI72LZijGpKjlGci/MWmD9Hur6LXU8YzNftouvrDWz63xbBq4z6p+viH+vupT49xsngjd49dZENyNr80zSpUbM2G4QSg3IzPvIa8pz4XCtbMPovBEvJ59KHyID/4wE6OH24h5svzvecP/HOjvhhEy6Y0vrB/u7+kiCcP7sSZ0/m7v/hHQonFeymrJqP8+d9+i4/8wW/ytQcOs7V3iKaRwhaeLQdxTmFUPz5ygsTv229TzbtHklJoTJR4MCWRysHYsouB6YXtu4Ub/WrXFI1oRL15ucobhbiM5w5QMRznzSPNTJR4EG0bJTd/35d2byLtVKmYjPOpZ0+wq61vtsd9mTSEZNk0jOS93OHiAFl5/pT6QPt5/urQ+2maHCXq8pBSnfQUlfHU5r3EnS42js3yglRGo/zaK69yurGB//LLH+UrX/nqshEgp1bYav/pTqDxkI9TWxoIJDUqamKc3lTH4XOLF7+9lyAA5ZkItVOFcwBeI8O6+DAd/jxB1B2TXRRnZxdVqmXQkBhlxBUiqnrJSfPvb0T18EztXiwEFNvkQ30n2BTrX7TwTsJmS7SXA+Fr9HrLyIkyl4L1dPoqaYkXVhoYVmnck5KTTk8FW+K97Ix2cdlft+QDfzNkMEtRhMZ9Tq6sr+TYXRt56a6NpNyOnxvzW0RrdTnXair4hTfPEEhpy3I5O3M67z99iS8/eBfnGmtpGgnPSFiaoohTz922is+5xs4GitDwsnTB1HsVDzx1hYe/d3FepOJ62MCLH97M2X21azewW0BKdmIhLBCguZ0QbXvJdi3NpeCLaTz49BVGqnw0dk7gTs2njb3cVIklClSPR9neuXjR1GIQsPGn8rnvlEudUs+cxf0dFxj0h/iH3fcQducjHyfqNlAfCfMrp1+ex1AX8Xj48bZtmKLI7z3zzIqpnUIy1AEzHUO+pEbNaJ6e25nT+eUfv4M3szTv+3sFom1RkZlf2yXbFsHcrEPalBhecJ4+Ix8hyYkypiDOY0s9G1pHXHaDIFCdDLMxPrCsFy4Cm6J9eAyNnMPHhMPPgLvk3TPuQ64QEdXLqCPIuUAjCcXFuuTCi1AoWILAOzvq+fHdm3lrVyNjJb6fG/UCIep1E/W6CSVSy3ok0wgm0yTcTib8HiDf1vPUlr3UxCYJpRM0T4zizS29SLhZzB1bBWkMRMI4F+WWf6+isS1MdX90Vfl5V7IwIc6bwWqrB2xgQvXxavlWcqJM2RIFVWuNi3dUEyl2M1niITiZ5uSdTTz85EXc6VnPN+p15/UvtBxF8ZVptudi+v6Ji7DYSbbNr555mcO9rfQFS9BkBW9OY314mOr45Lxn2ZXLsbu7m9rJCVy5lb3yGyVKWgnTnrsrq/PAyavEvU40VaF8Iv4zQaAkYOO/rsBTsO15LXtzvfb8Z/L6APk/hBka62m0BapnvPmSbHzFQkGAEi2OauW/M6m4ZloI5yLlV0kUO3GkDK4eqqT57BjlvTemLbEq416dmcCvp6lTw/nwoAbBAlN2TmM85OWbj+3l+OEWhsvy8q4/R+EgWRaSZdFRVYYlLK/NbQOt1RWIlj3jWbaWViHaNmnVQdzpoiwZx5u7vSp3OxhnvR3ly2znk1xhI5NLEta8l3DXi+35lqxVnMqmCyMUTdyY0blZZCQH3607zIWiRoAZnvC3izcQ3uZfsn3HFCQiqpc+TykIAgfC1xbdb63R1D7OJW81Nb0RmtrHOXmkCV2VsdOzkaDpanBLEPJ69qs0nJYgMOnL59S9GW3RCIxk27SEh2gJL1+3Evb5+Madh/joyVNcrqniIydPLdnnDoX33K0pPhCAyYCbp47uIJjMoKkKH3/uFA69wNKPawwBcBoLF01zF1jOReoElnOCRp1BzCmD3+av5k92fAzZWv7Z0SR1JmVlCCJp2YGFMC+UP1HtIRV0MlHlARsiFR5K+xJTC8jVYVXG3WNmyUgOLgXq0USFMUeQh0fewZfoL4gXZZMv5ji3pZb/71NHaW2uwJDFn3vrtwGVkzEqJ2M8vW87D56+xJ2XO1AME9G2Z7jYLUFAlyXONNfzxOFdlCSS1IbzAhIViRjPbNpNQEsjWyaHem5/a1YPfrJIXKKEP+ROvOgEyC66MPmfvE6Ad8/LXQ6WAkgCQjZfcOe9AW+8eCxJaHxt6g1sIKZ4uBRoQJNV9KnJa9RZxJgzuGhL2DQE8uI89w2f4+joxTUZ70ooDqeo6Y3w+j3rkEyLyWLvAjXJism8dxrzOBkJ+WmaUy2+HAxJ5MK6fAV37WgEx3Ue94C/GF82gz+7Mk+4L5OhYTyf+hr1B9BlGUlf2oNfjfLgjcCYYt0E6KgrY/+lHlp6R/je/bvJKjKqbry3o2Z2/tlcDjfSd24ioEnqzCwUVz1cUj03NiZBwBClGSn1aQTGM5z6QCOuhM62VwZI+x03fO1XXVAX1JO8b+QMNnAx2Ii0wkW6EcR9Lp69bytf/8hBJoNTlaU/N+y3BZv6hrnzSjvfvmsvv/1bH+f971zizsvt1IQjqIaJIYkMhwK8uamZZ/duI+1QuPf8VXa39wJQEwvz6beP01VczobwMCWpxG3/wQ/iZRgPNXMkTy2Eea047wVM/LKX5F6Vmj+JooyvfhIRyOsaaMr8Sl3RsnCuQtDjRuE2s3ym8zUeGzjBpUA9b5W08HzVHiozkzQnhxf/7dv5vHdQT3FHpIsD49cWCITcTizn0AhAY0cY2bAYqfLzgR9cwJuYn0o6fLGTpw9to688xJvbmqkZj67IpW4DlxqreHXHOgTbZmdb/wJZ1L889BCenMZ/O7awc+V6JJ1O+ouLsQUB2TJRlzHskL//hcQ0bTZA9WiUV3evJ+pzkVMkVOM9btinUMhzsIX5lOFuXaM4l7jhFEYgt5ALwh/WePSvzqNoJoYqYUnCDesWrdq4JyUXZ4NNmIJIWnbSkCpMW05/ZRHf+vBennxoB7ryM6ZjY+fD2bJhohh5jW9ZN1GmhEWmtymGyXixl5HSwAIGrEIjkNH49AtvoCkKz+7dxpOHdvLkoZ0Ilo1smlNUvfnVuyOn8/7Tl/ncM68QSuZDwiLQEB2nITpOW3ElmqzgNm6vp3w3/exlZFX7en9KvXaA1E6V1E4VMXnjKQVNUTm2aTMxlwvVMBFtC1cux8MXL9yGkeYnweJsgiNjl9ge7eHN0k3snmznc23PFqzFqaBYYe6TLJuGrgkauibo2FCKoUiocyrm77zQQd3oJF3VpXzz/r340hr3nm5dUsM8rSqcaanj//nY/eQUmbJIgkMXO3Fr869N3OFi51D3ose4HlWRCJ969TWibjcbh1euaVqUaOcWINpWvo0TaBgKI2AzXuTjkVcvFlxt72cBsm3hMrIzEc8tsT4+1fEirhucD4O55IKIQbLIga5KFGlpsi4Z3SGiZm7sHqzamrrMLBsTAyDk8xJ+/dbk6AA66kv58ifu5LUD6zF+mnPrto07kyMYz+BL5jWjXVoOV0bHreVwank1KmdOx5EzcGQNHDkDNWugGiaSmZcAlQ0L2bTykqBT4iLT8qBPPXQH3//AziV1kAuJ5pEwv/+957nrUjuvb1nH1boqxgI+Mg4FR9qgNJZkY/8wh652cvhyB1WRfGHUjzbtZsw7y2XdWlLFv3/9GdyJ2zvZezDw8N6fXGxFQB0yETM3btxly6QkkSDs9VI/MYFTz3GqqXlBOK/QEADZMqhL/XQqaU0j3y64+Ix0el8dgWiGM/vqEC2bjpYyPv//voKam+0TL4km+a0fvMIfffaDtNaV82f/5n38eP8WtnQN0lldOrWXwAt7NxH1ubncUMk7G+vpLwuh6iYPvH2Vne39C/QXNoSHGPQXreocFMuiaXz119mQCztnSuZsbY1s2TQPhGkeCPPaznWURJMrRjL+NaIiE0G2TXRhqqhOi1F+i4WkOYdE9/ZSEiEntVcnGGsIoGoGwfGBG/LeV2VJbKDXXQYCrE8Occ1Xk6ejzYRvysDbQF91iL/6tbs5taux4CvQ1Q1iVstXwAYbPJkclSNR6gcnqR2cpHokRslEguJIEpemzxjk6QIz0cq/nv5vmlN85rW1FEnqQgRj6VUVVhUKpbEkHzx1gXsutJJ2qOhyPu8j2DaKaeLO6ngz2jzDIVsmD7Sdn2GX8jVlltRuLjTsqf/iOBjAi4hNMzEcmDPvzfYl/3TC0aWT2eDKhz9uMKKqmiZbhgbpKC9j3OfDqedIORxrUsWsWCabYv1IP82JkCnBp8XQcmWUzg2lbDk/RPlIHDVnIC1iqN739hVGQ36+9PhRRkJ+RkJ+Xt2xbmbRYAvwPz71AQxZIqMqWJKIaFocvNzFr//odYLJhaQyHz/3Gn969MM8vXE393ReWlDQJdrWTQsy6Uph503JtBBNix/cewfl4TjP3rUN2TTpqC3ljmv9yxr3xZ7Dub/Jn1VsjPWjmjq6IDHsKmLEFaJMi93SeQu2je6QmKj2IFo2akanujVye8LySdnJNX8NKcnJuBqg11PGHdGumxk3NjBW4uPPP3svb+5tXtvcup0nunBpOu5MlsqxOC0dI2xqH2FjxwiVozFkI8+kJ0y1tcw1uD9LD2r+HG38GQ1/ZnXV7g+2nsvnf6eKQO5vv4A/W1iWrMVgA+O4+Gc28RJ1mIg0EOOPOUklKSI4eJomaklwF4NLVtJPqzYpU9Wsa30/i7+TInGnk4mPeAg+m0ZK2ywXkLh+fF5N46ELF3ireR2iDZ879pM1OQeHpfMb7c8BC+mjf1pgicKS0T9PMsvmC0MINsimxd0vtOJKzzeyAuDK6nzquTepnIjylQ/eSV95iLRDne3YEQRiPjfYNg7dwB9P8eCpy/zuv7yIR8stei++v/UA4x4/X7zncf7XkQ/hNHLzFsy/+dYLfPjK2zd1zoWOdkqmhWTZfOjl81xYV8PnvvsKleEYz965bcW2O6epI06FlnVRIi05/lUY910TnZRqMVJeJ4PuYk6WbmRdfAi3uXLL7ixx+3zIOYutrw2y7swovsksTDFe3paCOqepU6FFCDsCOK0ceyLt1KXHb+rGJbxOvvyJu3j9wBpp/05JMZaPxakZjrCuZ4wtrUNsah+hbDyxbKvJzzEfIpBWVM5UN5GTZLzZDNtH+pALXNhzPSI4+Fu2c4kSdjJODJWsIM0UbTsxiOPgBaGYvfYoXhZGEywE+j2lXAzWc+/IeUZcIRqTI2s6+VhOgdBTKUZ+y0/igAPP2RxyzILrKrcFE4LPLFw0WYJAxOMh4XTyyLmz9IdCtIzc/nMQYEWxmHcbliBgLhOmnizxMFBfRNFEmvKh+KKSvALg0E0efeMiBy93c2xXC+fW1zIa8qGpCqYoohoG/pTGuoEx7j3dytauoWXnkKbJUTx6lvs6F+8caBm/eVrnQtcoSZY1E43c0dY/Q1B179utC4RjroeIRWV6kvaAG1sQebtkPXsm2guSvv1phs/I8JGe1/k/mx8lJ6n8qHYfwVySe4YvUJpd3IM3BJGo6mXYVYTLzLEuMV+QSyCvchiu9jLQUoSsW5T0JykevjERplU9HYptsjXeS1Tx4rTy+VXxJrhwdVnkO4/u5tl7t97wZ28G3qTG3nM9bL86SEvHCOu7x/Anbp1wxQbCe70YHpHKl/OkBZYEEzs9lL6zNipYhYAlwHjAR09ZMXGPC0sQ2NfaTVFqaW+8o7iCvmAJLeODnKlupi4apjx5a2GolXCZEq5RzEdp5UpG5/IAACAASURBVAF6eY0anqR55n0XJpWkOEUFxhJBzqTipMNXyeVgPYfGr3KytIW61CjyGi7uxj/jI3avE6NIJPKYh8hji7fNiCmL4DMLGasSTidXq6o409DAfVcu8/LGTawbHbstvNTvNVhivn1zKXgTWUTLpq8xxDsH6vnAExfxpJauFSmNJvnosdP84vEzxD1OUi4HhpQnePGms7izi3vq1+PB9vM3cTarQ9pVWDGhuTn39royXFmdutHIkkWF1+Po6CXap4RdXivfisfIsn/8GqFcvoUzKyokFSc2Ahtj/QT11fE22EBWUsjNqUeyBJG0pOIw9Xd94Xn3yHlaA9X8qHY/ScXN19Y9wIWiRjbF+qnITOLVNWxBmJKe9TLqCjLoLqHPU8KR0UsLjDvARJWHzl1lhKu9VHbF0FWJ0HDqhubZVS/9MpKDU6ENZEWFiOrl/tFzrEvdmAToa/vX88RDdxS8EOR6BOIZjp5o48Hjl2non6AolkI2C/sAqBGD0cNFM8Y91uKaqXTXQjJyxkLKWuQCMpJmYSkC8WYHgXYNJfnuT8Zxl5Mvv/8uXtm2gajXTVaWUUyTv/7SP1OUyhuWq7UVPLtnG+uGx/jgWxdoK6mkp6iMCbePtOJAk+U1yfmO4EERTI7YA4t65QLgIUcWaclUtmRZ6KJETpI5X9RITpRvayHaYij5RpLgs6uY0JaIgMqWRU6SyUoyp5qayCnKTS2ybxWaqBBT3Bji8lRCqmVQll0bljpTEskt48mKpkXS66CnsZiiSBrJXN11k2ybomSGokXy6e824r6lNdNvBtM1RQBJt4PxIh91o5FVf/6BobO8XLGNDn81KcXFU3UHeL18Cw4zvxAyBAldlKjIRPjtqz9a1LjbwPPVu7kSqJuRe83IDnKCzLhztpi3z1PK7+75LLJt4jR13GYWt56lMjPBA0NnqdBWVtUrFFxmjk90HkOyLZ6t2YsmO3ijbDNvl2zAY2gzpDaGKJGVFDRJxRJEFFNfsorFFgVCQym8k1k0r4KhisyU5a8Sq+9zzyW5f/QsIHAh2IA9xduzmpWEDQxUBvnBwzsZL75NVLK2jTOrs+tCH5/55hus7xrDMdW+UXhqVHBO6PP5z0UYeiBA2akkqRqVVJ2D8jfi9D8czP/7gRC+Lo2h+4Ns/sthZO3dWW3awHAowH/+tV/grZYmLFFAMi10WUIxTHLSrPdTlEjztQcOs3FghG09g8QVN5JtURsLE3V5qI1O4NZX58Hc6pgl7CXzvTagISMvqfKe793eH25DtkwsQeTxvjdvupDpZuG6cmvFh55slqPXruUXKpLMr7z+2pqdg4lAW6CG79Yd5lyomZwwPXUsffc3JAb436e/sibjMySRrGPp6Wy8wocnkeUX/+UMDk0vOLvbUviv7/sYnz/5Y0KZJKJloZomIjaWIHC5rIbvbz1AVlZ4/PIpdg903lBNQ0GNuw2OrD7T8maJIl//4AF+dHQbkmXzf335WdzZpZ9fASjVovzupR/w1xs/wLVADYYgMeqc1asXsRGnuNyXUxc8WbqR18q2TO0zxQ8nzPwPBAFNdnClqH62MHqqPqo+Ncbuic41Ne4CeerZz7b9mI2xfr7VeJQhVzGGKBFVPTOzkkA+4i1bJrKdY118iF2TiwvylPbGCYym0Z0yHbvLqL16mwrqABKym3NTfe5jjgClN1Dun1Mkjh/eyNmttbfFsEumRe3gJJ/83kkeOnZ5nhb4WiHQrsGU5150Jc3gQ0X4OzRsSSAXlMkWy4i6AylrkwvIyAVWdFot0g6VLz1yDyc2NlM3PsHdF9rY2jvIVx68i87K0nn7lsXirBseY6AkSEdlGQ+cvULE5aG9pApLEMjJa+P9VpBCsyWuUcw+5keLLGAYD+cppcGOoS7h9uZEmawoc+fYFbKSQr+nlC3Rvts+9rm41WfSEgTSDgeVsRimINAfChEcLLzgxILvBc4XNfK/Nz3OoLsY1TJwmdkV0wHLaXYXGqYkoqnKog6HANT2RKjticz8vVYY8of40sH3E3F7KUvGePziSbaN9jHp9vGlgw8B+ev03x74Zf7s2X9k2+jqn8mYz1XAkebbfadZ6PZc6WXPld6Zd1dzzUSgJT7A/zzzdU4Xr+dcqJFhV4icpOAwdbx6hpp0mI2xfmqWaa2sTY2z9SZ/m+WZCC5zfhqhMhMhpCVwm1lKroskidiUZyJsn8xzESwglLFtatLhmfc9xuLFxwL5dvEHhs5yZPQS54qauFjUwJC7mLjiwhYEPHqWUi1KY3KULdFeGpNLc8XYoki41kfGp+CNZBFuoPNqGqs27g5LpyQXY1L1sSfaTkN6bPVee1URz923BV0tfA+3I6uz/0w3v/qdE2xpHV4TY6O7RSa3e0hVqSRrVRwTBtEtbtLlCvEmB97eLIHWDD2Ph2j83gRK0sTbk8XXnUWJG7jG3h3DDnCtpoJTGxopjSf44td+wP62HgCePLhzgXHHzjPatdaUMxb0zVTKa7KMIUqcq2qgZWwQt357xVw2M8FmJvkqWxnAyyBeMiicoxSbMt6mgg4hyGftiziXMO4CMOAp4XKwnrTsoDyz+nDjTwsSThcnm5txGAa+TAbHbWCnWwxZSeX5qj0MekpYFx/k8PgVqjKTi/Jwz4VPX8NQtiCQcSnkFAnHIrSs72ZRV3eonKNdVxj2B/nOjsPUvRomrahkFJX/8vITNE+O8Mf3f5QXNuy4IeMe9xbOuCuGiTeVneEuv9nrJZDXTD86evGm6IcF4LPtL9zkt+cjTNZ1ym0f7j9Jm6uaKm2CoyPzx6RaBo8MvM0jA4t3LEjYfKz7FT7W/cqqxg75AvQD4VYOhG+emnuy0sOVw1VUt+c9dku+8TuyamsrYHPNV4vb1Ljmq8Wra1RmV54gLVHg+OEWuupKbnhwK0HWTe55o5Vf/8ZrVI9Eb4hU/1YgmDauUZ36JyeRsjaCaaNGDZq+GUbWLLCh8lgM12gOX6eGYEHZiQSGV0JOv7v59uHiIMPFAR578yy7OpefSASgNJYgqyiknA4g3+vu1zLYQj4HvBaynsVofIIrfI8N/AsbMQURA5GvC1vQkAnYWX7JbmMvI4sW12QkldfKtpCRVS4UNeI0cpStYdhuLmwBMpsVcjVynmt+EQgmBJ/LLJhgJTsfkVIMg0Amgze7ukKnW0VWlDkfaqI0E+WT3cc4NH51Ziw/Tcg4FXKKvKhxXwk2oMvSikQtFjBW5OdqQwU5RaJuZJKW/rFlnAqbXznzCg+1nmHEG+SL9zxGWnHM/G7cehbFNNnX38Hx5tUXGmcViaTXser9V4JsWHjTa/M83U50eSoIO/zsinbisN6bxFexEifDzQE0j4wjbSDpFnLuxu3Gqo17l6eCCm2S7bEeLgYaSCguKrKRFX/gSbeDJx7eiS0WNjso2DZ3vdXB73z5GKHojVUR3irkrE3w2nyvJNCu5UPzU1DjJmUnkvPet6cKIt7NSTHtUEg7VOpHJxZtB7oe4lQua7pwTlMc9AeLEYD9/e2E0re/O0AA6kjwG1zgEbq4ZocYx4WITR0JNjNBMRnUJcrpJMukdioM2JQYQbBtvEuE124nLBUG/qiI2P1ObEXAlpifR5t6PqSkTfC5hV6vYhhURKMMB4NkFYVQKknLyI0Vtd7UuAWRSdXLzslO1seHfmqr89MuB1mHjO8mjJQN/MHnHiOYSPGRY6dZPzC+wGBbAvzo8Ha+9PjdRH1uLEHAldU5eraNf/f945RGFuos+LMZ3LksNvmFcU5SiDndM985/Q2+nIYhrp6UJhL0kFOkgqU5FT3vua8EazrtfZMe/vT5WgjYgjDVvz1L5GMizNSN2VOkWtN1CNPbZz+bf292u0iHt4phZxFb471ItjWPJGixzy533HnnjYCAnR+pcGvkQytB86poHoXQcJrJCg+CbeNK5AiO31gkbNXGvSIb4WTxRpyWzpgjQKW2urDmS0c2MR7y3tCgVoRts7F9hD/8i2cL0to2DYt86HPuyhryN78yvvJCZiUsx0BnzJFbvJ1w5nScOZ0JvzdfsLJMGsMWBDorSnHlcvjT+QerOBVnQ3iYUW+Q0mQc2TLXZLEiAF50NhBhPZF52+f+uxgU26QlPkhCcZKSXQjYM0Q2a4nIo24ij7pxtucIvKQRu9eFmLPxvaaRbVJI7FdxXzOo/sLiUQXVMNg4PIQpiezv7CThdK4JUYhg27jMHE5TR7F/er2hpMeB5lgiHLIC+stDPL9vM7oioakKv/8vLxBIzS4AbeDY7o384W9+mKwizxjVqNfme/fsIqvK/JdvPE/xdVrwewY6+fL++xn1BegKlTPp9vKFe34Bh6mjSzK9RWUEtDSXy2qoSK4+mjRW4stL0xYIimHiS6684D1+sIVrzRXcc6KVivE4vpSGbKw+imMhcC7YxJNVBxh1BHFYOkfCl/nw0EkU2+SLGz9CQ2qULk8Fg65iQrkEv97zYxpTo+iCxBvFm3iuYjcTqo/adJhHh99iW6wXUxT5Yssv0eqrQRMVXi3ZgtfQ+GTfMfZH2gEYcRTxhY0fYcBVQlEuyad7X2JDchATgfOBRn5UtZ8BZ4iKbJT3j55h32Qbsm1y1VfLN2uPsDvaycslW9FFmY8MvMbR8OVbuOJLo6wnTtFwCksUUDUDU5GwxBtn/Fu1cfcYGkfHLjLsKuKOaNeqSGxyisTz92y5geGsDpWjMf7j375YUMMO0Bsq5Z/2HGHEF8ScE2lQTJO//MHfF/CbFkJXJMw1MO7lkTjlkTgvb2/hE8dOUj259ITSUVXKm5vXUR6J0zQSBqA3WMbLzVvYODbIjzfs4BcunqQsFb+tBmYcFzZQQgaRpR9wC5jAhYaEF31KKDa/f0p2cLxiB2+VbKA5OYIuSHyu7bk1jaIk9zgQUxZNn59AGbPQKyXQofwrCTAhuVdl8I+KSO5RCfUuNKIph4NXN27k2OYtbBkc5Pt79vIbLx+/7URMim3SlBhhzBkkrngozt5+JcCbQcLrQHPenHE/saURUxJxazm2dA/jT803dCPFfr70+N1kFRlnTqelb5SSWJLuyhK6qko4vquFo2fbeOjk5Xn340OXTzHqDfDi+h24c1n+3RvPkpUVuovKcOk5vrX9MD/cvJe+YAn//vWnVz3e8ZC3oAx1as4gkFjZM8w4VX5033a++aG97LrYx4Gz3WzqGKZ+cJKi2MqENQI25VqExwdPUK1NMOgM8ddND7M11sum5ACmIPJy6VY+3/Uc5VqUf6m7myeqDvIf25/kqq+GF8p38vDIO7QkBnkr1MKTVQcpySWoyYT5g9bv8kTVQcKqn0/2HV9Q9HnZX8+/7XyaKm2S79bcyRNVB/n9tu8z6Crm+YrdHJ64wo5oNxf99TxXvpsyLcq61DA20OGtZHush99r+wGmIOK6jeJJliTQs62EVMBB48VxRhoDWJLAphPDt6cVLiM56PRWotgmDmt18n+d9aX01BYXtEJezRl8+PlztHSMFnyC+cn6beiizCffeRWXPnvzbneRnk1e4Wm59pBCYcPgGLs6+3jqwB188aMP8bGXT7GtZ3D2mREg7nLwzvoGvva+w8TdTu45f41N/fnwb1pVKU4laJocpaeobE3GfJJKEqg8TjsmAiI2TuZHDHKIHKeWN6gmiUIIjSMMsJ9hHFjkRInibJwD461UZiJcDtZhIawpAYYRFFEHTeSx/IQjZmz04vwIRBu87+Rwn80x8cseQt9f2ANsiSKKYeJPp2krr5i3AL2dcJo57h85x983P8Cxih18rDuyKnrNtUbc57ppYpcrjVVYokBpNMH6gfnFwjbw/L4tdFWWIFkWR8518B++8xJV41HOra/lP//W44wUB3hzWzNHznfgT88uDPw5jd954xkm3T5U0yCQSc+EfBOqk/JklP5gCR+88g77+9tXPd5wgT13l6ZTGk6suN/BM134kxpX1lVwelsdX/7YnfiTGi2dI2y/Nsjui300942jLlH3kK8qzzGh+rgUqEMTVUxBJKJOR3dtdsR62B3Nt4hti3Xzk7I7sBDo8lRQrkXZHenEa2ocmLzGmWAzQ84QNZkwimUg2XnFRMU2UK+THN4W62FPtANh6vWzFXvzBd+uEro85TgsnXZPFSnZSZ+7lDFHYIbLpTQbZ2+kjWpt8iav8OphKCKxUhejDX50h4gpizOFdTeCVRt3r5FhW7yHMUeAJ6sOINkWjwyfokqbXPRHbgOnt9ffdJhsKWxsH+bIifaZHvZCYtLt5WBvG/v7OtaU4MSUxDy71hoYymAqzadffIPL9VU8t2cr55pqKY0n6aooQZclvvDRh7ERGA4FGAv6aR4e49effw3fFEvVuokRRn1B3qjfyPrwMEWZ21/vEMPBSSrpIMg4LhQsDjPIQ/TgniJnbyXEd9mAikUjMdoJ8rdsp4QMm5jEr2fYEu3DEgROF69jS7R3zXnSRc3G9AmzufWIRXKvg5nknQXKmEnsgcX7l32axrb+fiY9HoaDQR45f25N6JMl2+LI2CWGXCFeqNhJl6eCI+OXaEiO4jUySxLpqJZBSXZlg1EoxG7BuPeXFWEJAkWJNNXj86NZ4aCXk1ubSLtUgok0n372TdYN5COXe6/18sCpq/zTQwe43FhJyqnOM+6QV3srTy5sHfblNO7rvDij7Lfa35ENDFQECye4ZU8Z94nkirsWR1McOdXO/rNdPPKTiwxUFnFiVxPHD27gxK5mSicTNPeOc/Stdo6cap8J9U+fW0ZS+ae6e5Ftg32RdhyWzvlA44yTkO8Zn71Wim1iCuIUI50Dp5Wbed4Uy8wrsomrW6KX5maPK9vmzOJYk1Q8ZpZN8X68Rj56sW+ylQ3J2TZTl5m77dLW01CyJs1nxyjti1M0mkEyLJzJG+cTWbVxNxG54qvlmr+GrbE+mlPDPFu5h890v7Dol1qiwOWNleQK2P6m5nT2ne2hoX+iYAZFk+SZB6txcoxLFTUc7GnDOcdzFwCHod82I5ZxKiQ9hat8XQ4CsLl3iL/9i3/ijz/5Id5e38BokR9TFLEEgbPNeVlM1TA4eK2TL/zDD6gdn81xi7bN4e5rSFNkDGu1CGoXihiwvRSRxUDkH9mCgsUH6EICOgiSEWT+xH6VEBkG8PGX7OTHNNAylaPPSgpJ2UltKjyjw7yWcHboxO92oldJqIMmrnad8U95mXzMQ9EzaSyX8P+T995xdp33eef3tNvb9N4xAwwaUQgQAAmCVSRlSRRVSKvLSqS1Ytlx7MT+2P7Eu3F2N/bazmrtSFasteXYliVLsihShV0sIEgCRC8zAKb3fns97d0/7mCAAaYCdwbU5vkHmHtPec+557y/99eeh/R2B2pk8YK16liUT7zzTr64R1o5kdStIKm6+V+3f4o+XwUxzcu0K8iJkg15/e8liIPa4iP83ye+ucaju2acXieRoAdbYtWdM3FvfkHlyeoUJa5GTQTQ0VBFZ0MlArjz4gC7Lg3OXbNq2ezsGuQfHtvHcFlRPh+/Ckiw6gWaqciMVIUKFpaXhKB8OoFrCZKa6+E0LGonotRMRNl1fpAvfvcwJ7fU8dx92zi+vYHDezfgTz7AQ0cu8oFXzrKhfxKHYZGTNS4GavnV3ufYGhsg5vCSVOe39C206FaFRameoN9bQUp14dQNIg4fOVmjSE/O/R6KEJiykl8QzBbBSUscFyBgpHGbOdqSIzSkJwEJW5LmmOXyuELbtvaQBRSNpwlNpG9KMOYKVk5io7mpyEW5r+fcXCjz7unORU88URpgsjQwR8l6yxCCiqkEB473FNRb+eb+h4i689WrlizzTmMbZ6sbaJ6ZRJsl4dBsi9995ZmCnfN6pN0OEr7CUkkuBRlomArzza/+D463NXBkcyt9FaUkPC48OZ26qTD3XOjmrku9uK4TjDhT1UBO1bi3r3PdxgsQIMfvc5QdTJFG5e/Zwos08Cj9KNgk0Aig40XHiU0zMbYwzTGqsAFLUhh1lzDqKSaueUhqbr60zjn3wBtZ4ve7SW/UcIxYeI/l8JzTGfrPRUx93oftljDKFCq/Fr9hXwEMFxVxoaaGLSMjOEyTH+/cxRfeeH3NvXdTkhnylKLaFiW5G8e2GFzrJAc8B0lipDKEoS7c674UrtxBxc5LHl9BTlM5u6GGsZIgii148ucn5t1vSQhKY/mOkZT7GgW5NcR0iY9YwF2wSJ9s29SNLhyBXQ62LJHwOhkvD9JXV8ZMsRfNtKgfCePSTV6/q5UXD7bz2Gvn+dSPjuGL6JTnYhwv2kBM9XA61ITHWll3w9Z4P6dCzTxds5/61CQX/bXUp6eozVx19qqyYd4tauXVsu2U6nFaE6OU6Us/sxtSo9Rnpni6ej/bYv3YsoyJwr7wxWX3XStILF2AvRKs2LiX6glK9fkhts2JoUW3nygLECugwZKA+pEwbT2TBTsmQGUiii93NYzWEJ6+YRt1jSur0y4H8XU07leg2Tb7L/ax/2LfivcJZtM8u/lO0poDWQgO9F/CZ6xtf6yEoJ0wrUSRAC8mO5nkLary7Sswmz+/WrUrARVkiONAIKEJi/rUJKW5GLqscqR887rn3D0XDOr+MIL7Ut7oqQlB9Z/G0CZtsk0q2oSg6MdpSv/pxvZCW5KY8ge4WFVN3O3BZRgUpVLrwu3vtXL824urX9z6zfXnYx+qKcK4iV730KxYkq4qpJ0agdl2uumQjyPbWhCyRMPINHs6+xc9hpglTxGA6ZKRLIFqCGaaPWgZm8DYytovdY/CdKuPqjMLCzJNlvhJeAs3X8i2oG50dblkU5Hpaizn/MZqTm2p48TWBhyGybaLIzx8uJM9Z/oJxTOc31jN4T0bePmediTgK996jU8NvsrxolYGvOUcnO5gb6SLytnuq/0zFynPXU2LNKQnuX/qLADV2QifHnyVY8VtjLhLaEuOsifSRdC8GmnZHuvHkFV6vJXkZJXGVJ4Fbk/48rzt6jLT3D95FglByEjziaHDHC/awJC7NF9AmhrHO7voKNET3Dt1/gbmu/c6Ck8ZN4upEh/JAhosxbK548LwHMmELiv8w4772DY5wNaJQXqKKzlV3cxnTr22qhXox8+8U7Ax3iySXifRQCGpJNcO1fEI9/RfRLPyIXllHXqenVjo80w3ZFDQUYjipIQsOVTs6wJv+b7U/D6mJDPqKWHCHUISgu2R/nXPuUtW3sBfC9flfOubUaYgmQJtykLK3RiKk4WgcXoaW5YJptO4DZ3iZHJduOUdtsnBqY51ONOto7+uFN2hwMoEx+ZQPxFBFoKo38NIaYjA4AS2JHFmQy3nm2sAePzwGdzXha6FBHFPfp7TTBNZCNIlDvoPlOCJ6NScjDLZHkA2BYZbITCSQdVtwk1ebFmiaDCNpcmYTplUmYPgcAbdpzKwr5iKjjjZgIYrZqCYV5/V8Ypg3nMvEGRb0Di0MuMe97k4vq2BN/duoLOlkvGyALVjUZ786Ql2dAzRNDQ9r2r+7uM9bO8cxlJkXrqnnX/z96+zMTnKxuTCUrcPT56e93drcozW5FUuh5psmCdGF5+zvVaOe6cvcO91bWr3T89npmtOTdCcukr/WmQkbzj3FVTmovzSxIlFz/lexZoYdwFEg27S7sIV06mmza5zeUa1jOrgeE0LXaVV7B7twWGatMyM871tB+bOPxIoZjBYStvMGG5DZ8xfhNfIMe4Nsm1ikKyq0V1SRUUyQuWsZOmpmkYsSWbnSN9c6K2zvJr+4jIevXhmTUK4+XvlYarUvwZHX/7cwLye/uUKe3y5LPsGuvLiF7M5rbXO+1aRYlR4+T6tHGKYGE5+RhNZFP6I/bQS4RTlZFDJouLHQJBXk/NizIa4bIpzibkojNMyGPCWU5aNza3Q1wvXdiYAyCmBK7V0gagEFKVS3DE4gGrbSEIg1inn/ouEgdpiss6FOeaXwl0XevnnB3bTX1nCC3dtoWYqynB5Ef/Pxx8g69Somo7y8LudCxDbyAxWFAMQTGZRLJtMSCPc7EXuFtiKhKVKjN4RIlHhpPxigqqzMSbb/aRKnMRr3ahZi66HyvOfbwpQ924YZInBu4oxXAqtr1yNVlqyxGB1cUF55Z26SWvv4jzn1+Kle9r52mfvI+dQ2Xeql1/9xzfY3D1GIJHBYdzIeSEBgVSO5sFpXr+rLU/k9T8pxNy/V7P3V0h41gIrMu7mbHHCvB3F9X7SVQgp35ZSyEp51bJoHsizjNlSvuDBn8tQlEkhC4HH1OdClLYkMRwoIe7y8qf3fJhPnXmDM5WNnK1sYPvEAGGPj6jLh6Eo/PO2u/mdw09TlE3zRnM75ckYO0euhqkV2+Yf7jzEI2tk3E1VZrg6VHBt5uVgSRITRQEu1FdzuqWOiVAAIUl85dmf0zQ5A0BG04h53TgNg1AqgyXJHGncRCCXZuPUKC+1bmff4GXKk2ubl9rJJLuZ4DmaeJpWJCBElt/nGG9TxUkqqCGJKWS+xk4eo49JPByhmp1MIiPIyQ6OlLcTc3hRbYshbxnbIv3snb6Edx1pN4UMRrlCrlElV6+iRiyCr2TBBlsDKyCjJO0bvHcBjIeCdFZVs2VkBKdp8Myu3XzmyJvrUjH/i4K028Hl5gqqJlYnNXvfqS6aR6e5VF/B1z9yiP/x2D4yTgemIqOaFh9+4zQ1UzcSWZmKzNkNec++KhzDYZj4J3IU9aepPBvDmTBRTEHT4Sm0jIXuUdE9KqZTJl7twh3VcQlBzckIlefjjN4RwnQpRBo8ZEIaB/9r17wWqGjQQ399SUEJrzZ1j+NeoZCVP5nliRdO86GXzlA7HpljuVxuNCXRFPtP9q5Yavf/bzCRmXIG6fJWcdlbTUTz4bRNnhw9TOVsFX9G1kgpLly2jrcAraYrMu5DnjJGXcVEHD4CRhpdVtkWG6AitzABiqEqpNyOgrZ2lYSTeNP5CnavkaM5PEF3dKJjtAAAIABJREFUSRUNsRvVhVKai4zmgCuVk7JMc2SCtMPB1olBzpfXkXa48OhZWsITJB1uirJpMpqDYGY+p7dPzxF3rl3IPOfQ6G4sX5c2uCvIaiqvbt/EN95/iLPNtXPdAk7d4JOvHp0z7r1VpfzBZ59ge/8w//5fXsRCoau0EkkIJnxBpryBdamW92DyrznHVqbpJYQLk91M0EaEPYwTwYUbkzG8/DXb+TPuRJIETSLGh8jLaOqyQnk2xq6ZHly2weHyLTwx+Pa6huaFBLH7XYz/RoD0NgfI4DuWI/B6FkkHvU5l/DcCBF/I3EA/a0kSo6EQ52prCXu9OC0T2V5ffncBJFUXEYcPXdaW5ThwWTp16RtrWNYUksTJbfUcevvyqnZz6Qb//jsv8Uef/yWGKopIePPvvGpaHDjfwwffOod3gWryhMfF8U2NAGwcnMCTyzsZEoKZVh/eaT3f9qiLOSM9uiNEqtRJaVcSZZYzXL3mewDPtE7p5QRDe4tpeGcGxch/GQ566K8tLHfIzvOL105dj4ff7OThNxcvpF4Mj7zRwSNv3FpqJ665uVBUT8LhQbUt2iNDVC0jAJVQXXQU1xNzeJGFzebIENXppVMQNpDU3Ix5iok4vWRUJ5YkIwmBwzbxGVmKcknKsjE8K+i6ycoah4s38/2qu+l3X53rQ0aSxyaPzxn3bm8V36h/lLuil/n42BHc9q0VpK7IuLssnbTiJKp5qUtP0+OtJKJ5Kc9FF7wwU1VumilqMVRNLO4dZlSNjvI6xn1FnK5qIphNc7GslvroFE5Lz78zQsx59j49S2k6QVpzUJ6KU5XIPyDlyTgn65o41NOB18hhSRIna5soW0MGtrRb4/ymmjU6+o2wgZd2bua/PPUY40VBGiemaRqf4WxTDQn3/BqJxokZxouDpNxOumrK2d47wo7RfnKqRnkyxh1jAxSnk+tiYLyYPMgQD3LjRFRCvlCphRhf4RS9BJEENBOjivz4vGaO0mycnkAVkhBsjA+ve849vVVj5PdDKGlB1VfjxA/Nv9/apIVRIhN53HODcVeEoC4cZufgIKF0CreuUxGPr1sroi6rnCzewOvlWxlzF5NVljfujckJfu/C99dlfNfi3R0N2LKEsgLthCuQgHvOdvOHf/dTXr5zE/1VJWimxfaeER49eoENwwsX8r6zpYmEx0kokWb3xQE8WT0vI3o+TrzKjVAkKjriqFkLyQZbk5CNfH+TlrEIjGZQDBtblnBHDCrPx/CEdTY+P05xX4pIo3f2PgtsKZ9vH64uKsyNmsWOVRj32xlVH/UU8832R+kK1eA1Mvz2maepGlnauE+5g/ztxofpKG7AYRn8h9P/sqhxF4AhK7xRtZW3KzbR569k2hUgqbmwZAVZCJyWTkBPU5aNU5ucpj06xMPDpxY18jYSr5TewbdqHyCluGjITFGVi3DW33DDtjXZMOOuIk4EN3BvuIPGzK0Vj6/IuFdlI5iSQq+3ElvK9wsvpQZmyRKmUiCChVn4r6ODLE/F+MDF40CeHrY+OsVvHXmW0nQCXy7L+y+dQEKwZXKI0lQcW5JoCY/j07M0RiaREITdPhTbniNFONR9gT965OP89uOfoWV6gqjHy7mqer701ksFvZYrEEBffSmDNcVrcvyFMFxaxPcP7masKMgXX3iTjx0+jj+T5d996SlOtdTP29ad09nWP8zx1kYGy4rZ3T3IjtF+hJSvtH8voo4kddxIxqEKi9b4CB4rhy6rlK6ipatQiD7qwQrKNP3bKVyXDPRqhVzz1VdQSQocoxapO2/kPJCAsnic4tkKeWkVpCe3CgF0BOv5b20fYNRTvGyWUBaCmvT0bbnHkJeYHqouonF4/iQugCwqQ1KAMpHCh0ECByGyZNAwLJn9p3so6opjeRQqrCS+ZI50TqVbKqaGBEk0SshgIZNBZffFAb79n/4W1bKon4igzi4oSnpSlPSmQIAzad4wjg0/n1xQRMo1u607lr93lReu3kNdUzm5tZ6Mq3ApvGAsTePwzIq3v7Jc6mko41JT+awTt/Dz8OjrF/Bm1of4pRAQSPy/mx7h+brdRJ2+G5b+tiRhqi5SqosxbwnnihvpLKrj3rHzeMyFU3uD7jKeK9tFWnHx6ZHXORi+gMfK8bvtnyNxXX+/18rSnBpn2F3KjOZfH+MOUJOZ4cHJM1z2V7M1NkBLcnzRV1zIErZS2KnHk57/kDgtk4pUPpyhCpuydIKy9NVWvfrY1XDglZEE9NkFwmzrVjA7v6S2KTzJf/nJt3lm2x76issJ5DL83stPs3u4t6DXcgVCgtcObMRS14dGFKC7upyLtVXcdamP33jmZXzZ/H1dTOqyeXya17ZvJOzP00Oqwl41DeKtwkTCQMYxWw9/M09WRnHwfM1uws4AHjNHaTbG+8ZOFXysSyHXpOIcMnFfMBbtYVXDNmZg4SvMaRrf37uX57dtpyyeoDwR5/d//OyaG/ms4uDFqp2MekqozIR5Yuht2uLDnCzewPcaDvJbnT+iNBel11/Ni1U7SStOfu/C92hJrL1i3UIwVIU39rXS+IOj8z7PoXBYrsNG4idKK580z3NRLqXVnuGoUsNue4wuqZiBTJDJrJePmBeJSU6+r7az2x6jSiTpkYoYlfyYyDiExdbwFNXhGxcxEiz6niz13WIQ5MmujuxtKaj7vOXyGK7sygm6dE3lb546wA/evwtLyTPHWYqEZtqzxk9GtgWejM7dx7t/YYy7AJ6r283TTfvJKg4026IlPsaeqS5qU9NolklSczHsLaWjuJ5+XwW6orJruhu3uTh73GVvXqVuT7SLJ8cOo806kgt1GUlCUJ2L0OGvI6XeOqnZio27LUnkFI1SPUFdZhpTVnBYC3PMS7ZAtgprAbRldJZvGMNNbCMD1YkoX14jT/16RAMeXt/fui7nuoKIz0PE5+HuC9249KVzOhLgzeYwZQWjwJGY1eA4FbwkNfBp0Ukjq/cGLSQm3CH8Rpad4V6CemrNuQsWgpwR2I681Ku0QHG8rYFRqaDOLBwVSbhcVEZjPHLuHAe6u3h1U/ucROVaQpdVzoUaKdYT/Hbn0+wK5/m5x93FSEBzcoyW5Dg7I33smb7En2/+CF/d9GH++PS3cOhrLwl8PWxZ4sjeDTz57Alc19BUx3EyKvnRsPCio0sKPqHTKxcRw0WxyPKWXEsGDQ2LqOTCg0GLHeF+ewABbBIzvCC3UCQytDO9rjwJl1sq6K8rLdjxJFuw70QvjlVQeZ/eXMsLh7ZQNxrh0dcvMFhTzLmNNTz22gXSbo0T2/KU47/2969RFl6ezva9AlNSeKV2B1nFgSpsHht8ly91Pk/QuJGrwQbGPMW8XbGJO2b6cC0hIjPj8JNWnOyJdS0rlSwBLlvHlPIMe7eKFRv3UVcxJ0Ib8icVgqCRnse9ey0UW6Cu0hgvh9XQ2E7KXhzCIiSyCCCHioKNtoje9xWcrm6gt6Tihs8V2+ZDF44X3EN64f4thIu8BT7q0hBSXrNYs5aXahVA3ONGsyycxjqzjV2DYfycpxQnN/dM6YrG5UANU64QCc2Fz8wS0pPsnV65UEch4D2VI37IRfxBN4Gfz580bCfED7pI7nJQ9LOFyV/82Sz1M9NUxGOcaGwimFkfkhhbkplyBbkj0kd1+iobmCKs2TylOtd6Vp+e5qGx0/x162O8UrmDjw0eWZcxzoMkMVxVxIVN1ew+Ozj3sQeDOhFHxcZn69SIfLruJamJzWIaBxaNIkYMJ6V2hkYRZUryzNP99mDgxCSNRolYP5IeIcGzj9xRUK+9KJZi8+Ux1FVUsI9Uhsg6VH7nv7/I5q4xXr6nnckSPw+92Un5TIIPvnKO//bZ+3jtrjY2Xx7Dm/3F8NxTqpOYwwOShMfIsTk6tKBhh7wTWJMO87G+t1Z0bIGEIpafuwSQlp04bLMgzseKLaYiBJYkk1EcjLhLCRiDi24r2/aqPe3lkF4kz2QhMSn7yEga9VYUA5lROUi1HUMICEseTmnVhESWLcY4KclBkcjnzLKSik/ocy/vcKiE0zWNQD5SEXb76Cmt4KHL5xY8960gEnDz/P1blqxdWAsE0hkC6QwdDdXYkoyyhMG0ZZljbY0EUxnKY+snAHLDOJAoIoeHlakRXg+HbXJHuA8x25cPoNnrr0seei5D7AE3w38Qwv+gi/RWB5ZXYuKLfnJNGqk7HahRm5LvLeztugyDtvFxAMriCRRhr0tBnQAsScZl6TiuuW8O20QTFlHH/AVqe2wIRdicDTXeHuMORANu3rqzhTvOD83lwb0YbLcnmZHcKLPpnWqR4F57kCqRxIHFFnuKUcmPRL4HuURk2COuEq5o2FSJJFlUvBhc2l6BK2XQ0LO2amHdjeUc3dVU0GNu7J6gfHp18r0ZlwNfOkdpOF+oqpoWuqZiqEqeFXIqzt7TffzTh+8i5XH8whh3p23k5wQhyCgO+v0VpBUHnluUdg0aadyWTpe3mkenlk4DmrJCp7+WIiNFoADsjis27pXZMAdmOrkQrKcpNU5jenLRh0Kx7LzknxAFa9mYKvHd8JkNXFAruKiWU24nqbViCCQGlRAaFmV2ioTsZFgNoZphLEmiRy3FLQxkBIaksM0YnWP5evDyOQ72zHKmS/kJ7edtW+ksry3INVyBAF6+tz1fSLfOxr1xIkzjxAwv72jnrfZm7rlwowKeIB/a/Od77+RiXRVbBkZoH7w9+VPI67ibQsLg5khbZGHjtA0GveU0J8Zw2CavVW7jodHT6xpW1WZs6n8vzNhvBok+5sb2yCDBxJcDIIH33RzVfx7D2X/jIkYApizPfR5Kp3lh2zaeOLn2zFmyEHjNLCnVRU652gXjMXN4zBy9vir2TF+eu5eqbSKAiPPGd3a9YGgKZzbXMlhbQtPgdJ7ICCgjTam4WmvjwGajuGqYA+j4xcy87z2zBZoCeFuqYUzy8ZDVhwTMlHuxVAXdpVLTF8GdMRirDZL2O6jrCaO71HwlfNog5XMSiGaYqvKjOxVAorY3vKzIjQCeefQOkp7CiUtJtk171xglkdWFzh2GSdapYSp5l8iT0Yn53SS9zrk3yZfWSXqdBZWkXWu4LIMd071cDNVhyAo/bthLSnXxyz1vUJ2amcuRr3buacpMUGwkeL1kKw9Nn2FTcuQGAZor9NnPVuxl2FXKnmgX1blbXyyu2LinVBdluRgfGj3KuKuIrKzhWqQPTzNtvOncHItWITBeHsSWpHmEHQYKUdnNdnOMjWa+313Dokhk5kh3Su0UNVaMjeYUXmHQbk7wI9dW6qwoNVZ83uTuNg3c5vxr2t9/mX/cdbBgTGAC6Ksr4fn7t5DyrC9xDcCGsUkePXGerppyfu3XPsVTr7/Lg2cuknI5EZLESGkIXVP58d7tPLN/J07D4ANHz9I0sc79ytdgFxMco5Kf0cwH6cGPcR0Z7Xxc/7ktyQx6yjhd3MykK4Rqm4y6169D4VpoEzb1fxCh6i/ipLdqmGUKUkbg7jRw9RhIxsLP2YzPxx89/gTBTN4wWZKMrirrYtxVYdGYnGTUnee6qMrkRUaK9QTFuQSvVmznfWMnKMklsCSZy8E6TFnBZd6+VA6SxOWWCk5sr6d+JDwv9Lzce7zU93eLYYR1dRtLkenbWELK7yARchGIZDhxTz1VQ3Emq/w4cyaWKtN4aYaLd1Rw5+EB/vHX72LfK33UDERWJA7S3VTOO7ubC0pcUxZO0t49jmquruulcjKOkCR6GsqonoxRPpPAYZr84P27+MzTR5FswenNdaiG9QtFWCMBv9zzBueKm+goriOluvhxw15eqNvFgfFOHh4+TWtshJCexGWtvACxLTnC/shFnqm4i/+t9RM8NnWCO6Pd6HKeLnvcVURc8/Jy6XbeKN5KyEhxb/gCpQUQrFlFQZ3M4dItVGYjTDmD7I52EzIXJnCWgEAiiztrkC7QajPudzFV6qNy6mp4WEHgFgZDSgiBRLM5Q1ZSGZf9pCQHzdbMXMi9XymmyM7gEiY+kWNG9rLDGJn3I034AkTdV0OMtiRzuqaBQDZTEMMOkHVq/PzgJi63VKy71w55L+wTrx8j4vPy/YO7+ZtHD/I3j9wz9/3vfOHj5DQFIcsUx5M8efg4n3j92G3tb82hsJNJ/pmNdBFiG9ME0GcXZvNnx0OM3JCbl4RNSS5BRSaCz8zgNnNsH+pbV699biwANjjGLBxjK09d+bJZvvzzV2gfy4eIDVnhcFvb2gzyOjhskzsivXy76X56fZVsjA+jCpvKTISm5DgvVO/ma20fZN90J2nVxU9r9qDLKu3xlfdPrwWyLo2XD7Zz4HgP1eMLi7CsBtJ1/0I+Srn96DDulEEy4KSnvYxdR4bwJnL0biol51JRLIGQmeuK8cVzHHi5Z47dbSmkXRo/fXArE2WBws0XQtAwNMOmrsU7nhZDa/8k+071Eg3klTSrJ6Lc/W4P33ryAEd3NqEZFlPFPh5/6Qy+dWR+LASCepr/ePI7/NOG+zhW3sa4pwhd0XitZjuHq7bQmJhg3+RFdk/10BYdIWgsL2CgIHhq9E1Sios3irfwnZpDfKf63rnv/6TloxhSXna8RI/z+PhR7p25UJD5dkXGvddbwairBFNW+FnVHtoTQ/NybwshmMgU1LhbssSllsrrjLvNBnOaXrWEjKTOhjdkGqwICjY2Eh5h0GZOMS15sZFwYFJvRZmQ/QTF/N751zZs4XBz+9zfkhDYsswnT71ZkGuwJTi9pZafPLSdXAGpeVcLfybHrz/7c7YOjPDatjbONdUyVhwk5XSi2hY14xG29Q9z/9lLvO9kB+5lqurXGu9QzXeljSRwMISfo1Qvuu1u8ZMbjLsMVGXCFOsJnLZx2+ha4wddKFELzwUDaZVOjWZZVEcjWJKELASqbbGvp3ttBnr9uW2TPTNdDHtKKctepXX1WDoPjZ/mdHEzr1Vs442KrfnlliTTkJzg4XVuNVwI59ureW1/G089c3wu915oKNd0BjVfnKJjVzXeRI5gOEPpeJK3Hm4hHnSS9eTf+ZV2Egmgo62Kw/taV1VQvBxcOZO9p/opm159HU35TJwvfudNiuJ5w+Y0LD708lm8GZ2TW+qwZYn3v3qex167sGJK27WEQFpx9FgCKjJRvtzxMw6NneN4WSvvlrdxOViNJSv0BKvpCVTxUs0udk938djgCbaG++fa2xaD38ryr4ZeYnNyiJPBFrq8VcxoAXKyhmabVOthNiZH2Be9xJ5oF44VFN+tBCtkqDMImGkCyTQN6UlU21q2mi8Yz+AuYDGFpSqc3F7Pve90zVtBF4ksu4yrVfsukWGvMd9jqLei1JOnyj2p1dCrlHBI77lBUeu+7g52jPTP/S0BbkOnIr4wze5qkBeI8fJXnz/EaGXwlo93q/DoOo+9e44DHd1E/F7SjjyPtmIL3LpOUTJNKJl+T/CW72OMJrEyrvAACz9zSc1Nd6CaDfHRFa241wIzH/eQ3u7Addmg6LkMgdeyKJGV5fIyDgevbN5M89QUGyYm8OZyuNepg0EC2hIjfOXyj/EbmXk9uluj/fx2x9N8t/FeLgdq0GyL7dE+PjZwhNr1pp5dAIam8u2P3sWB4700Ds2syCOaryy4+GcA246NoFg2kg2WKuFKG/hjeY/VH83izJrc95PLyJaNM2fizJg8/venkZZZaAjy0cqfPbiVkcrQCka9QghBIJHhwcMXb0pRUBJQOXU1ZCzI88Y/8cIpHnyzEyQJbzqHU7+54tdCw5JlTGnlbbwS4LZ0dk33sDkyxPsHj9MVrOKVmh2cKGslpTqZdAd5sXYXF4oa+NWOn3FgonPZedJn5Xhw+ix7ol3EVQ85WcOSZBRh47J1gkYav1W4CDGsmKEuPKe3C/n8u7qM5145GScYzzK8uJO1KliyxJnNtcT8bkKJ+ZWEK82hCeAOY5RtxtiCbXGhTIopX4ChUAm6cg1zWLXNL3WcvOmxC/KCFn/65Ye5tKHytoTjF4IMFKUyFKXWX3d7NSgnTflqNTyvgyIsEqqbsMM3W2Vu33Il7GpR9GwaKyCT3Oskud+FnLAJvpqh5AdpXN0GUlYsmnPPqRovb9mKSzeojMVwGgZVsShPvntsXcbusM0F+6E1YbMj0sOW2MBcb64q7NvSjbAYpot9/NmXH+aP//en8a8gVNxHiBHJxy4xjhsTGeiQSnlWbuU/WO+gIrABCxlvPE87mk8QScgIyoYTXCs3XD2Qdw6u/K4Vo8t7zLYs8co9m3j+ga3YBS5Mu/edbmrGb85heeXuTRy5s4X2rjG2dI1RPRFFMy1U0yaYyKKsod7BDUWmKziRLqtk1dXXNl0x8rWpaWpSM9w93smIt4Snm/bzcu1OkqqLQV8ZX9/yAdpiI5Rnlk/7yAhCZnrRdHahsSLjHtPy+ZVpRxABdPmraU2OztPZvR4V0/F8JWahKuYlifHyAMd2NfLw66sXLoD8D6YukKe9gu/v2M9z7TuoiYXRrKuRCc0yb8m4R4Ievvnpg7y+v61gBYb/M6Eg+Sdhk1ad/LjuLpyWQWUmwuPDR5ffsYAIvZwl+HKWXItK9H1uEvtdxA+6CX/Ei/uiQeiFDN7jORxDVl7X/Zr1Z3EqyX985kd0VldTNzODV9dxmO8NAyoDzveQMb8BksSprfX89WcP8sV/OEwgtbSBtyWJKcnLKSopIcNGMcMWMc0rNAL59tsh/IxJPvxCp44EU3iISU5aRZguqQgvJjEcWMjsFuOr4mgQwIWN1Xz98/dhqoUlj/JkdD78/KmbfqcE0F9Twvm2ar7ldaJYgurJKE1D07T2TVE/GiaQyOBPZamciq+K3385yOIqVbhAwpCWNl8CiDp8hG+hayPfZSFw2ibNiQl+6+yP2Dd5iT/Z8TEiTj/DvlKOl7Xy/sHjC+5vSjKyEDelY3GlBfWKo+m0VpdSXKFx9zLtCHDZV0ORkWTEXUpD6kY1tmvhMCwah2d422zG0AqTL4oGPLx1Zwt7T/bf4L0XBEJwf9cFPnT+OF49e80LcHMPqADCIQ/f/uhdPH//Fkzt9rG8LQQBZDWNjFPDlpf2DjzZHJ7bnHsXgI5CFCcgKCE7u1hbHi7L4P0jx5dY2q0PJMDVY1L5VwlKv50ks8VBcreD9DYH0096mfq0D895He/xHN5TOu5OAyWd78kuj8cJZDI4DQP1PZAu+UWCqSk8f98WKifjPPGzU3iWyQdnUCkXKd6U66gRiXnpngguOqVSGkSMY3I1hlDokvIsdxoWnVIZHgzKRJpuqYhGcVXAaDkIoLO1kq9+8UHiftey268W9x+5RHP/0nP3UnjwrYvsOdvPRGmAsfIgE6UBJkoDTJX4eHV/G0mvC0uW8Key/B9/+gzFscJ5qU7bmFtEmrLChGfpdIUuq3SFqkmphbuPErBrqoe9k5d5oW43AEPeskW3Px1oYtRZzKbkME2ZSbRV5NMF8HzLrtkomMSe0S6KsytvXVyR1W1IT1GsJ9mYGMFnZZl0BnFay0/0Wy6O4dTNghl3W5E5uquJM1tqOXi0a9n+0NViw8wEX7/7fRxp2pin1p09vmaZ/MXT31rVsa4Y9m989hCvHNxEylu4HtVC4HRTLc/uu4O+ilJ0bWkhIIBffuNdHn/n9DqN7kYYSPyceo5QQwIHFaT415yjlCwJNE5SQSUpNhBhoSWUISm8U7aJU8XNeKwcJdk4Twy9s+7XcS3UuMD/dg7vyRzRRz2EPyqT3OMg+qib2P1unEMm3pM5yv4uibsjv2r36utPCnLlNftFjznFAm7+6Ym9eX32507jNBafaIvJUE0SDwY5VLjGuGdRcGBRR4IukSWDShqNRqIcl6poExEykkqDiBGXHGQkdcUryu6mMv7qc4fobKsqePquYirOR3968pa8aVlAKJEllMiysW+SnEOlr7aEs5tqObO5lvHyIGNlQQxNIVvgouFQLkVAzy8WDFmlM1RH2OGleAGKYwGMeYt5qWbniu6jNZtSWekCTJev2rTFRGMALvjq+WHVfqqyEZrT4+yJdrEr3kvATK/sXBJsCOeJq9xLnGchrNjq+q9hzLm+qGYx7Do3gC+Zy5MvFOhBnSny8p0n9tDeNU7ZzOrYlZbDsfoNNM9M8KELJ3BdM4mulgXMlvJ5vv/02x/g9Na6gi1uCgEB/GTvdv78I+9jpDSEqSgr4iY/eGF1+tiFhIHEP9LOT2khgwoIUpKKLvJmXCBxgnJsJH6N07gXCIHGNQ+mJBPUU9w7cZ5jpRux4aaKim4Fc3daAbNYJvo+NzMf96LX5p+RwKvZPEOdBOGPeIg9lA/fN/8v03g6bk/kxJRkXqncQXegmseH3qY2vXhhWkzz8M8N9zLuLuLLl39K2W1ShlsQksRUiY+/+twhMi6NT//LsXwx3PWbCcGI5OdHcv4Z8WDwplRLP0Gel5u5xx5GIPFDeSM5FO61B5nESwUpXpfreczuoYMyJMSswVj+/RLAYE0xX/v8fRy/o6HgBDCSLXjk1Qs0D0zf0pyZcjvoaizn/MZqzm+sobuhjJTHiSVLVE3FuPNMP5u7xmjvGad8prCslj4jQ0t8jGPlbeiKxrmSJn7cuI+P9h7BbV5T+yBJDPjL+ZMdH2PYV7Ki1PDJsg28VLOTvVOXuHOqG6+Ryc8Ns+qLAkCSsJB4qXYnb1dsAvJsrDtnehY9bnUujMfKMeAuY8BTxttFmygxEtwVuczD06eozc4gC3uOZGk+JM6VN+I1cii2RUkmfgMPy1JYWVhedZNWnHMnv+ivoy4zRUtqfMn9Aokse0/15TmRCwVJ4uS2er756Xv49b95lUAyu/w+K0R9ZBpJCLy5LA7rag5xpcIcAkh5HBzd1cTXP3eIodvAQLcc+ipL+YcH9tFfUULNTJQdvUM0TkwvKyKz91L/+gxwAZynlNepYyeT/CvOcYIKnqFl7ns/OtWkeEHy9s1RAAAgAElEQVRqQBfKgsbdaRsEjDTNyXFOlbSQVp1c89quC2wHmEUKuSaV8OMeYg+6ERpoExal/5Sk+EdpnP0mV2o9A69lSdztZPCPixn7zSAtX7o91ee6rHG4fCvHStvYHB2kJh1e1GD5jQxj7mLeKmtnR7iXD42sb13DspAk0h4n//2zhxiqLuZz33ub6onYPJKbZqI021FMpLm0z91imAPW8Nwc+KDoxxTyHJnS4yK/+N1qTSEBlSLvTdbaeQO31CxgqDIdbVX8yVceobuxvPBzhhC0DExx31uXbrk97YeP7uSrX3gQd1anvWec+9++xI6OIbZcHiWQzM0Zw7WADDwwcoY3KzfTE6gipTr5200Pc6Synd1T3QT1FAnNQ1ewmlOlzeiyyqboMDYyl4qWZhnNKg7eLW/l+frduK0cG2JjNMfHKcnGcJs6OUVjwh3iXEkTA/68I6HaJg+OnKE5vrgdfGj6DAciFzkWauXV4m30eSuYcgT4QdV+flR5F1sSAzw4fZbNySGKjCRe66ouvIRg38ilPDPlTaThVmTc3y1uY9RVPMdIN+oqpmQFDDoS8PAbnfz0oe1YagF/ckni2ffdgTed45M/PEb5TGHUh7x6lovl1VyorMNt6Fd5yC2LP/7Jt5fc90of/ov3tfOzB7cRDXoKMqZCo6eqjJ7qciojMf7rN7/H3kt97/lw6wABdEnhV8R5KrgxnCWRN/BZVBaTw/GaWbZH+klqLkbdJRTnEutOYhN9zMPMk15SOxw4Ri0CR7L438wSfDWLGl4gEiYg8GYOz1mdbOvt40UwZYV+XwUV2ShVmfCSxUEygtbECO+UbqIzWPfeM+6zsBSZnz60jf66Up589jh7T/URimfmeVDaddd5/ZO1UMfNat4lG5gsC/DWnc186xMHmChbmxZZV87k0FuXae2duuV3vWoyxpbLo2imRTCeIePSGK4swlQUysIJSsNJiuJptFUy3y0ES5YwHQqO7NW2uubEBL9y6SX+dtP76PNXYksynUX1dBbVz9tXtS32Tl7mM12v8nLtDi6FapY8l9vM4bL0fGeT6uJsSRNnSxbh8hcCv5HhwEQnn7n8czxmbtH7KgFeK8f9M+c5GO5g0F3GiWALHb46+t0VnPc3cjrQTHU2zO54DztifTRkJqnMRdCEhSQELsOgOTqBIuxVMaWuyLjvjPZyj9kxz7gvJXN3LbZ2jrKpe4wLm5a+uauFkOBffmkX0YCHT/7wKBv6p245B797qJeNk6M3fL7UqikfTivixfs288a+Nnoaywpe4VpIxDxuYh43T71xjF3dA+95ww6QRcWDgXsJ4Zi8D7741ViSzLmiRo6XtPLLfa/TEaqnMhtZ1+uPPegCU1D95zE8Z3Tclwzk5PKejjpjoflvH0+3jUTU4WVDYgyfuXykrDQXR8Zmxulfh9HdPGxF5lx7DaOVQXacH+KhNzo5cLwXd25t0x+CvC76O7ub+MlD23h3ZyNp99pQUUtCsOXSCB986SxO49Y7Gvaf7KVxeIaJUj/jZUEGq4t5Z1czUb8bW5YIxdNUTcWpG43w4RdPz+m555wKhkMhWuqhdCyJM2sSK3YTK3FTPpKY5d53kHOpZL0a5SMJRhpDjDaGqO6PUt8dnntPDo5doCiX4s3KzZwtaWTEW0pSdSEjCOopGhMT7Jrq4b7Rc9SkZzhV2oy2DC/LhvgoX+x8gTMlTfT5K5h0h4g5veQUDVOS0WwLn5GlLBulLTrKzukeds70UpKNr9zYCpvm9ATN6Qniqps+dwVd3io6/HV0+ur4Sfkefl6ynab0BL8y/ApbEgP0B8vRbBNdUWmKTuDTVx6pXpFxL9Hn504qVjEpunMGH/nZKTo2Vhe2DUySyDk1XjrUTndTGZ/40bs88mrHHJ/xzZypOJOiOLO0/vQVMy8kGKou5mcPbePInhYGaovJOdSbC6kJgSTymvW6pqxpKF8SAkkI6icj66IoVggE0EkLjQQOQswvKhHkC5wG8FMsFq+eT2huUqoLQ1awZJk+XwV3ziyvsVxIVP9ZDEkHbdqCa+bZ5VbjFd9IIAoZ+VolhCRhyCqaba6o1sZt5if0jPreKiJdEBLMFPt45eAmTm+to7V3kvuPXOLQ25cpuqbS+1bu/rVPpKEpvLOriR89toOLGyoJh7wF72O/Fu6Mzpf+8TBVEysjgVoO3oxOy8AULQNT+XfPpZH0OIn53fQ2lPLWrhZev6uNpNfJA29dnDPu4/VBjh9qoK47wrm7annoXzoYaQqR8js4+mAzH//GcY48ugFLkSgbS1IykWS8LsBoQxBfbL5Bk4Ft4X42xEaJOr2kVSemlFel02wTn5GXdNbsfBzvAwPvctfEJZAkKtMLC7KE9DT3j55l7+QlkpqbrOLIzxWSjLjCCiksXKaO38jgu5KTv0kEzAzbE/1sSQ7ywMw5oqqXS74aXizbwXl/PZOOIFuQmPYGUGybrOqgOTqxqnPcVKXXasKZkhDsOjdI++UxOtagAtTQVC43V/B//sZj/OiRHXzmB+9wx4VhXDkjr0zHrVf5CvJhPF1TSLsddLZV8fx9W3jrzmayLi1f/HKz1yXybYNPPXOc973ewb/540+S8BW+BeYK8qQ1aUZKQ3khntvaGLYytDNDCRm+zh18hdNkZzOdaVTCuHiFel6lno/QhYuFvRO3qZNRHMw4Axwp2zyrsby+whbOwfzzeEUD/bub7yGtOflA17vUJMP5nliY6+U1ZAVZCBwj+f3M2YkmTxSydrnN6yELgdfIkpid9JZbjEScPmwkvCvw8gsByRb4U1n0KxXaN/MuShIzRV7Cu5o4cUc93/jcvew73sv+E71svjxGKJ5BNfNiKIotFnUirrxNtixhKjKWImNoCiOVIY7s3cBrB9oYrClG11REAYVgFoJi2Xzue29zx4Xhgj0rWYdKLOBmqKqIy00VdDWW01dfynhZEEuWkET+t7ijc3hefl93qJSNJtl0aoy3Ht1AtNSDpclMVQcYrQ8iJNCdCjuPDFE5FEO2BCUTKYQs0XZu8oZxSOTpjz3p5SPIJbkEJbm8gyrIFzxL4kq0L++oSeQ/85k5PLaOrUioxtrND4I8CVJO1oipHo6G2nijZAt9nnJkbGTy7/j7u09wrqyBTTMj1MZXVwy55mXcElA+neCXXj5Hf11Jwbjm559EwtBUzmyt4/ymapoHp7n7WA/bO4Ypn0ngS+bwpnO4swaauVhWNg9Llsg5VbIOjYxbI+l1EfO76a8r4Wx7DSe31zNT7CtYFEK1LO47comnnj1O+XSC+uEZLmysXjPvvXVkgq39I7y8czNPvfEu7UNjBW8pLDQaifNRuvgOm/h1HsAhWWTQ+Kq0mxlcpIXGfka5nyEcC+RBAVy2wT2TFyjLxrAlmUdHT9y2lIQlyfQUVTDpDbF5eoiyTJxpd4DLxdUowuaOiT4MReVseSOWJHPneDeSELxTsxGnabAxPEJJprCVyEtBFRZNyXEuBuvo9VXSkJxYdFGYVhycLmrCkhWakksX3BYK3nSOv/yD73Jmcy1/84m7iQU8N7eilySElF94RUIqzz20jece2oYzZ9A4NEP98AxVEzEqJ+OUhZO4ckbe2Fs2ksgb9CtOQLjIy1B1EcPVRfQ0lNFfV7rmxnzepdg2+4738NGfnFyR8txK8dMHtvG1z96HUzfxJ7MEkhlqxyLce7SLDf2TbBiYonw6viCPf6zYzVRNAFORCZd5GdhQwuaTY4w2BAEJxbRn1eTy+zqzJim/k6kqHzV9K2PUs1SJrFdDyOBKmcimTc6roRg2WtYi69OYbvDhieoUjaUxnDKTTX5cSZPi0RSyJZho8ZMJOGg+MY2pydiKhOmQcaZMZEuge1R0l4KmWzjSJvIq1gA2EnHVzaQjyGVfDUeK2rngryOrOPGaGZrSk2xIjdGUnkAA3UVVeAydgUAZtfFpSlfx3q9Lj5Zm2tx9rIcje1p4+86WNX3ILVWhq7mCruYKVMOidixC5WSMspkkRbE0vlQOT0afW4UjgakqGIqCqSmk3Roxv5uY30005GGiLE/WsCbhciHYfWaAL3z3COWzIg57TvfnjfsaoWY6wudefpv/6+OP8IefeZyPHz7B5sFRSuNJHMbiOW1PTr9tAjIScIhhysjwDpUMCz9JHMgIaknQTpj9jFLG0sRGHktnZ6R3fQa9BGQh8OtZijIJauPT2JLMkdp2+oPlTHqDBHJpytMxZtx+jla30RSbwG3oHKlt56MX35oLe68XnJbB3plLnClu5of1BwjpSXaGe+YZeAEkVRcvVu3iTFEzqm1xYKpz3caomhaP/fw8w1VFPP3YDowCCq3knBqXNlTmqaOvgWzZaIaFNjuXmKpCzqHka25uZ5eMEGzsmeAL330LX3rxYq+bQWv/JF/6zmGqJmP5hc5UHH8qu6JzmJpCuMxDS8cUtb0REiEXsSIX298ZRhKClo4pvImrz3bpWIKJGj8zFSsz7rYsMdIeIlLlYabex53PDJDxa0y2BNCdClteGyVc56XzUBXlvXF84SyxSjcd91dTOpDEG8nhzJhEarw40vkI4OjGIBMbgmhZk7KBJCWDSc4+UouWtXBkLDYeGUfWl7fuhqTQ66ngkreGTn8tHb46xpzFKMKmKhdhU7KTLYlBtiYHqcnMoM4Kn10oq6dtZuSm+nrWrQG7cirGB188y8XWKsJF3uV3KABMTaG/vpT++tL8B7P5Zses1vAVyUVLkbFmV923FGJfJTZfGuM3v/kKzYMzc5/tPdXP3z11YM3OeaGhhpd2bibm9dBRF6K7upyqcAxfJodqLR7VeOr1d/nwbSSxkYDNzNBGhDgOsqhICLwY+NHXvV/9ViAjCGZThHJpKlP5SUtXVLZN9VM2EKcsHaezpI5QNjX3vSxsKlJRNs8Mr/t4VWGxb/oSb5Vt5lyokb/Y9CG2RgdoTYwS0pPYSEy6i7gQrKcjWE/E4eP9o++yKbY+Y7VliUjAQ2vfFJ/+wTuMVIU4sqdlzd9jW5HJKTI5bl8nw0KomErw6R8cZVP3eMGjctsujrDt4siqFwyybVPbG2bnkaE5DfldhwfmogoS3BB+d2dMdr85uGKjZqkSM3U+FMNmusGHbNhcuqeSSLWHaJWHslkDHpjKUNkVR8tZuBImgeksld1xnGkT2Ra4YzrTDX5gimiVB8shE5zQCdd4cSYNxtqCtLw7RTqYjwgshbDm40SwhXeDrfR7ypl0hEgrDjRhsifWxV2Ry7SlRqnOhfGbNwrH1MWm2Tw9NJsaWyMSm1uFLODg0W4626r49kf2Yt2OinIpL/+Xc95mUyAErb2T/P5fPkfzwPze5Q19UxTF0kRCa7MAulhXyfcP7sZU8pzHSZeLrurlc/z3nrt9JDaQbxu6QjcSQJ9HB2ohz3W2q6ydcEWhkc/5C5ymQVt4hNcatuGwDH654zBZTeNkRQtJR/63yRcLFUYKcrWQgLrUFJ/veYmvbfwg/b4KRtwlvFqxfa5uwZQVDFlFEoL7Js/yhe6XUAskXbkcbFkiGsq3nlZMJ/jdv3yef/dHT9LTWPae45lYUwiBJ6PzqR8e5b63LhekHe16XHs3BYCS/49k3/j3tWjoClPbG0G5ZkwrWXhcyY2vBKpu44npDG8u4q4f9KGaNq6EwfYXhikeSeGJ6WT8Gs60iTOVl13WshbOtJkPr8+mA7SchZBnc/NAaDSNJ6qj/3/svXeUHdd17vmrcKtuTn07Z3QDaORIMIFRlCgGWVaiJMtK1kiy/ZzGz35vzRuvmefl92yP7Rk/a8mWZUu2km1liaQokmIEmEAixwbQOeebU+X5oxrdaKAb3YiEbH1r9QLq3qpzT6XznX323t/2y4RmNBxJxJRF1r42saw5/VxiC/9WdzdlSUFwHBJGlkcmD3DfzDGqtQw+W0Nylh6zJgMRnm7bjug4fKjzNfy5xQMCF70eK97zGkAxLT77rVfobq3ijR2rbqgP6maBaNms6xrj97/8PKt7Lw4UUXWDzaeG2XPH2uvy++sHR/ntJ1687ONuOdt/7TuzQpSQeZMaXqOOMYLoSEt41uGv2EN0ibKvNxqmKGILAoYk4TVc+VhDFNE8Hvy6xnvPvokhSegeiTWpETZMD2GKIoYscdfwKe4dOI4jCOiyjFfX+cTxFzFEEd0jI1sWyjLxI9cSEg7bUr382eGv8ZP6Xeyr6iAvebFECWYjiavLaR4YO8I9E8fxW9d2OfhScARhTldCwCX4P/7LJ/mT33+Ertaq6xqNfjPBX9L5lR++xWNPHLhqi90RwJEAAQQLHBkEA2xVQLAdBB3sgMD4p8P4ug3izxRxFJj4RBgpb1P17YXaI7JpI1/n2kKWLJKPqWSqfZx4Rz3xkQJ1Z9J03luHYDns/lYX3rzbiUPvbeaOf+1xyb1gcvjRJm77bi+2LNB5Tx3FqML4mghqwSV+j2bhzRuUQh7yMYXRdVFE22H9y2OXDLxzEAhaZTbn+nlg+ig7Mj34rXkL/FLviIDDo137GQ/GqC6kqSpcntrjDddFVQyL//3Lz1P4vYc5tr7hPxTBC7bDllPDfP6be9l4enTRG+sxbbaevH7kvmFwjA2DS1fzuxmxjxq+LGyh1ilQRx55SWpnxYVkbgR6qyrpq6pEcBwakklaJ6fZt7oNUxKpyuTYNDTMM1s3ES6WCWoaWwYGOdDeSlFRqE+mWD0+wbGmBlKBAF7DYOPQCMebGtBkmZKi8Mjho8j2jYv4F4AqLcOne5/jY/0vMeGNkfd43RgCo0h1OX1DUwvPwRYF0uF50SgBaOuf4vf+4Xm++Gv30bm69t89wUcyRT70k4N8+jtvXJOleNsvUNisYqsCvrM6xbUK/tM6hY0KiALhN8pIWZvgQQ0r5I5kog7h18sUNlyfnP3lUAp7cCSBzc8OM746TCGq0nAqTd3pDOAguCFWbP3pEI4oIFhuxsn6l0ZZt2dsbvudXzp1XqvzKYTVvTlO3lfHpudG8OYNykEPpipektxvSXexI9PNquLSQahLwUHglab1rJ8eYl/9GrymTnNm5UJEN5zcBaBhNMVv/dNLfPEz93F0fcN/mKWzOw708Plv7mVt98SSN0iybNb0TOAr6ZSuk7DF5WIiHGY8GqFjdAz1bSgzmsSHB4tf5ygtZC5J4FdSWvF6YaiiguF4jA3DI5xsaCDtD/Dkjm00zCQxW0Sap6d5eX0Hv/bSKyRyOQTH4XBLE6GSRvv4BCPxGL1VVWzvH+CZLZtQTJOzNTXc0tvLU9u2cntXN1XZGxc1fw4CbonXpuKVVxe7lrAFgXTYt+AzyXEn0r/91Zf420/fy4mO+n+3hkRFMs9HfryfDzx1GNm8Nq4QWxXI7lLBdq324moPeo1EaZ2CGRWRMzahfTcm1XGl8GV1KtJ58m0q9WfSVPa774Z4QeS+4MBpbx0tpUl8tuEu/VsL95nxBCmJKrVacgEptxyeZqolhGA7VPblUAuXHg/bSpeXm74Qrp+9PjvDcChx2Rlab8t0VnIcNpwZ5Te+toe2/ilX2P/fKxwH0bJ55Lnj/MGXfnZJYgd34Ewk8zSMpq5vt4CheIy+ygTd1VUkAwEsUeR4QwMnGuopyy6F9icqOLCqhf7KBIYkMhkKcbi5iaKi3DAa3cg0FU6ZPiJuHvhsQY7F/m42xPMFKrM5RMcmHfCxpX+QD+3bz8defYNAWSOgaaweH6cmk0WybT7y2pt0jIzy5M5tlD0yfl2jJpUmoGmUFYWKfI76ZIp4Pk/Zc3MFcr1dsEWBTNh30d2XLYetJ4f5o//1FNuPD7rjzM33iFw5HIdYusjnvvkKH3ryIIErjIx3FvkTLHBEAWXcIr9Vxdfjkljk5RL1f53Gf+p815ewsJSywMJtQItL5JsVklt92CKYfoHUZh/lhDvOmH6R1BYfmQ4vjgC2BPlWhdQWH5YqoFXImH4RLSZhBEVsEfItCrl2FVt23QilJg9y3KJyJkfT8SSSufjNnvEEKYhuLIuFQEoO0OOrZkitwAZykkq/twpdlHEQSMsBiqJCVvJBTqDxSBL5jEMyFSYj+RlXIhiCiC5IpOQA5hK0uth1vvDvfAhAQ3aaNxo6qM/NUJ1P31x57kv+sO2w9eQQ//PPH+d//N7DdK6pveaVkN52OA7RbIkPPXmQxx4/QDi/spSRaLZEe/8UXW3V17V737zrTiqzOforE2zrH+QdJ0/SVVvNUEWc6VCIpulpvnnXnbROTlHwqoxHonxr9x1sGBnhWFMjH31jH8o1shQuhVYyPEovX2Yz/8gmYpSXzGf/M14hcpP43AFeX7ua3uoqNg8Osvt0F1988AG+f9tO1o6O8+DR44SL5bmI4bxX5bnNG0gGAwRLZdompjje1MhX3nEPHsti9dg4x5oaER0Hv6Yj3cAl+ZsZjiCQCbmkcGFOt+g4tAwl+bM//SF/++n7ePbeDZTVK1SSvIkgWjZ142n+2xeeZtuJoasq46r7Jb7w4wcoh93J4ud/5WUSIwV8vQZyyia/QyV4VMOMiUy/P0h+u0r113MY7R5S7/LjKAK+bgPBdJj5pQBGQiJwQsffqc8x1uCH4q7ATZeG5RMZeziCI0L/xypY/2djpLb7KTQrBPs0Qt2Q2hFgZmeAUHeZYI/G1J1Bgj0apQYP6pSJYDqMvyuMqDskXs8jF21GfimKJ2thJy99byXH5nB4FS3lSWx0flaxFa+t0++r4uNje/BaOjnZx6QaoUZLcTC8iobyDClPkIBVprE8zVfr38nu1CmiZoFhb4Iefy1xI0daDrAj23MRUztARvZzJlDP/thqhr0JCpJKyCzRXJpiV/os7YUxgtY8RzgIdCYaqSxmiJYLly269bbWIhUdWDU4zV/8yQ/46889wL6dq66rOtuNhGjZrO2Z4MOPH+CBvZ2XRYKhXIlVA1NIln1dJzwOcHtXN35dw8FhJhDEFEUKqoouy5xqqOeu02eoTWc4VV/HycZ6LFEk7fcRLGuUPJ4bQu77qeafhI2EHZ0KSsiXKKN5M1nvomPz0JFj3N15Zs43/gc/eRpDkvDMph3+4U+ents/VNb4yOtvYooiyuz3H37jLQxRRLbdiNr6lJsa98lXXrvxJ3SzQhDQFQ9l1YN/kapnAhDNlvndf3yRVYPT/ODhbQzVxXDEn0NjwnEI5cvccnSA3/jaHppGkldkrS+EgOGT0P0uHTiCgKA5JB53pbije139CGXCovHPU3NBdkxZtPzxwujtpj9ffMVRKttU7c0R7NMp1nooNniQyg7qjIERlfCPGBSbVWxFxJYFtKhEYl+e+MEitgSO6P7ZHgFHEkhv8mErAqLmYPlE8qtUql7OIdgOevxiWtO9EoZXwlO2iJaLBK15l0LMyNNRHEF0HDKyn7iRp7k0RZ/fNa4cBDcwVpCwEREcqNeS3J86Abjl0L9Rdy8dhRGiZgHlgiwRC4EToSa+2XAfJ0LN2MLC5+5AdDU/qrmNHZkefnXkJdbkR5Fmx7itE330xGr48drb+JUTe2jNXByEvRRuikLjFakCf/D3z/HT+zfyxIOb6W+s+Pl88WYRTxW4+42zfOCpQ6xdJCJ+Oci2Q+NoimimyEw8eB16OA9xVtceQeB0fS3DFXFi+QKyZVGXSvHihvVMRiIYksSWgUFGYjHaxyepTWcIlW6Mz22CAALwWxxhPTMoS9Z+u7nQOun6pM8PehMAxVp6QiQ5DtIF33t+YaEvC90jUfCri5L7OQSLGh988iCr+qd4/MEtvHLbajT158e1Idg2bQPTPPrcMd790kli6YsrJF7/Plzhcda8S8STtwj26mA7hHO2S+yNHrxjOuMPhKncm0PJWKQ3+jAiEvH9BTx5i9QOP1qFjDqTJ9xZBgEC/TqRUyXkgs3MrQGkso2SvNgPfvyhega3VrB6zwTR18qMqjE6Aw2sLwy5Lr1Z1/BipkHILHEq0EhB8rK+MAQ4C6xo2bFoKCfp91XxyNSBi44/HGnji80PM6FGaSpNsao4QczIIzsWuigz4wnR7a/lYKSNlCfIbw78lI25QRwEjla1ENJLbB/vIapduu7JhbgpyF0AYpkiH3zqEJs7h/nee3bw3N3r3p5c+KuAbFpsPTHEB39yiJ1HBwhdRa35htEUiZncdSX3x/btpyaTJlIsIjigmCZ1qTSKaVKdyRAtFF39edsmVC5Tk87gsSxMUSRcunEDyyamOehUk0W5oZrqV4uW6Znld/oFrglMWaLkW56oPabNrYf7WTUwza2H+vnee7bTtar6pg+285V1Hv3ZMd7zs2O0Dk6jGm+P5sGVovql3BzpyjmbxBt5ylUel/RtBzlvExjUWfXP00iaQ+RECU/OAgdEzSF2sIg6aSI44BvWkYs2ctF2Y5pm9xcNty01aSGe52/XvRLHHmpgYHsFkZEi9XszvHv6MAGrjN/S2JLrI2iV2ZXtImS6qxS24OrkizisKwwTNQsIQKWewW9rvGPm6Fz7Ig5N5UnKkoeouZCAp5QQP6i5nUk1yiOTB3hw6jBVega/5ea3m4JEXlaZVKL8sOZ2XqlYz9OVO2goTRM2i0iOzYwvBAgXWfzL4aYg93NQdZONp0dpHZzmoRdP8LUP38GJjjqMcyR/M/rJHLeIRMNYmk999w3ufKubUL58VT4wgIaxFNXTec60O9flvAVg9YQbyRkqu3mXDlCRz899D7BpaHjBdsfo2II2bgQKeFCw+SJb+Qc2EUJHXcJ6/2Nev6l87r/AjYEhi5S8K88uSSTzPPLCcXYd6eOZ+zbwvffsIBUNYInCzTHOzK6oqZrBriP9fOJ7+2jvm8RXNn5uJrfnwz8yv6IiAN5JE3Vy3sJWz9sWAE/eJnyyPLct6hZKqjS3DRA5UVrQZuT4wu/PYbolSLbKzaYQcYiZBWLnkbBvtpS5T89gIvJ6ZA3Hg83cnjmLYpuoGISKC9tu1JZDe6UAACAASURBVOYn7ofCqzgZaOTh6UMXhdKdDjTQ56+mvTDKrw09j8/WF/RPcUzihknMKPDrg8/QG6ihM9jAiLeCSL7IBztfR8DBY1ko1uXJf99U5A6zxe2LOrcd7GPLyWFe3dXOEw9uob+xglQ0gOG5Cax5x0HRLaLZIi1D0zz48ike2Hsa72wd6Gvx8vnKBm39k+zb0Yp+DXWyL4XF+n3hZ2/HwDJOgGl8VDNfgtP6uRKc/QWuN0xZouRd+RK7AMiWTc1klk995w3e+8xRHn/3Fl64ax0TiRC5oPftyY13HLyaSSxdYOOZUd7/1GE2dw7PSbb+PBL7UljuXIQl/r/UZ4vt4wBjHREK8ZUVLJOwuT1zltsz84qcy/Vze7aX7dnFa1aMeeNkZR8fHHv9ImI/HwIQMovcmj7DE1W7SHqCCEBYv3S9jEvhpiN3mL+Y/rLBu/Z2cs++Lo6uq+f1W9o421bNcG2MqUToxs6yHQdf2aB6Kkv9WJqOnnG2Hxtk/dmxuZrF1xICcMeBXgp+1S1jeRmons6+rZHUD+w9PacYdqVY2zOBMLv68W76eTf916Bny6NleIaHXzhBWfXgCDDSHMOSBBr7U3PylMuhfjxFsFDmyC0NNPcmic0Ul9x3pCnKRF2Y7fsGr6rfR3Y1oBZM7t7X5aaXXiEEx6Gja76a27Ed9UTSJZp7Vi57+XbAlMXLfk9gfqyJZ4p86jtv8J6fHeOtbS0c3NxMb3OCofo42aD3+o4zjoNs2dSOp2keSbLhzBi3Hupj/dmxOV/wL3BlsBSRidVhyiF5Re/vFRUTvMR3hiBhCyIhc+kxYK4dBwJmGVOUsISrN2JvSnK/EKpusuvoANtPDDGRCDFcF2OwIU5ney0DjXGGamNunus1DMITHAd/Uad2IkPTSJLm4WlahpPUjWeoH0tRkSpc91n0llPDbDl14wqFWILAiZomqvIZanNLV2FKe/2Mh6J0TI3iAK83r8WQJO7tdZWdfutrL5PyBhiIVbJ2agTVNNjXtIY7Bq9On94ByshM42ZU1FDEcwm1uivB9uNDbD8+BLja0s++bwOGR+LdPzqJoq/cz2nKIv/0h7v55W8duSS5n9pay6sPtF81uT/1oc1se3OQT3/n9atq50I8+74NrD41edOTuyWJaFe5wiUAiVSBh188yX2vnWGkNkZ/QwX9TRV0t1TR25xgpCZ6TSrOCbZDNFukeWiG9oEp2nsnaRmeoWk4SSxdRLpGpO4IMLYmzNDWCtJ1fgyvhC+jU9mXo/nQDOHJ8lx/Lge6T2Joc5zxNWHylV4MVUIpmEQmSjQcS1HVk8WzTLW0/u1xTryrHtGGdS+M0npwBgfIVnvp35FguiVIKexxSS+lUd2VpelIEn9qaQtY90mk6v0kGwIkGwPMNAfpua0SRxKxcei8r5ZMzdKGx4P/34k5idprAb+l4bFNRrwVy+5rCwJjahyvpaPaV1+B87Kf0vMfgasht3PtXFZSvmVTP5GhbiLDzqMD5AKnKPpUij4PU4kQfU0JxivDTFeEmKoIkowGKKkedFVGU2QMj4xgu/WXZcvGq5kEihqBgkYkV6IiVSCRzFMzmaVhLEU8VcBX1vGX3D+vtnRJ1OsFG9jbvo7abJqs14fPMKjKZfjh1lvYNdBDy8wUpijx+OYd1GVSxAoFbh3ovqLBwUGgP1bJZCCMLsnUZ5JkvX4Uy8ASRAQc/LpOT7yGsDa/XFSVz/Bi+0bu7T2FAwxH4nQl6sioflYlJ+iPVc1FjBc8KkWPwlA0QVN6mopijml/iL54NV5TpzU5SUgvX9AvMBB5gjZepIkiMvXk+V0OUUWJNAqvU08jOdYzc9kyj0tBcOCOF3twBAH5MgOYJNPm03/zGrHk8jP2X+DqYc1q8l8r+DST9v4p2vqn0D0S+YBKwa+SD6oM1lcwWBdjrDrCeFWYdMRPNuij4FcwZQlTFhEcB9mwUQ2TUK5MNFeiIpmnZjJD/XiaVQPTJJJ5/EWdYFEjUNSvqZXuAFpQ5tVPrebYQ/WUoiq6V8KRBCTDRimYVAwWuOufztJycBqlaFKMXXrp2sFNR+vdVclrn2xnss21iE1FwpYEJNPBU7bwZXTa9k1yz1fOEhm7uNLZOUy3hjj83mZsScCf1mk+nOT4Q/W88qnV5Cq96AEZyyO6Kxu6jZo3SQzkecffdtJ8aOYiTQMHOPWOOl769bXoPhnd76a/ObPuFUcSGN0QY3RDbIkTdLj/705fU3KvLyeJmEVertjEeyb2kzBcffiLCvAAw94Eb8TWUqOlSeiXpyO/GC6b3LWEzNStQeJHCvjG3Mo6iLgyheK8iIQjuQpHouGW9XFEQBAQDHeH3GoVPSqT2F9wcxiF2f1NN2XCkd2bIVggmAsjpGd2BNAqZGr2ZInMRqSvGp5h1+F+bFnAcQRsWaD3oxWs+tcZsCG12U/3pyrZ9kdDSCXbrRAnQ+/HEtS+lCU4UGbinhDx40V8E4a7CicAOAiz99r2CCDO9scCRxZwRBAN56IH7dpBIOPzEykVyXp9WKJIU9Ig4/MzFo6yenKcV9rXcba6lrFwjLbpCbaODOA3rsxVkFe9TAYjnKpu5NHOg5ypqqMuk6LoUZBsh22jvWS8froTNayeGXMDWcpFhNmBaTgS51vb7mHd5AjDUXe2KtsWT3dsY9dwN33xKn60YRd393fyze138/FDe/nBptvomBzlyXU7+G8v/fCiPmlI/CObeIEmFCwsRKYEH6bjvrQSDqeJcZAq2kjjZ/7l1FSJkl/BFgUky8FX1PHoFpYkUgp48BV0ZMvBFqAYUPAYNqpmYguQjvtBAG/RuOj+WqJAMaBgKi6hKGUTX1FHdNzfLARVVM28aCnQwf0d3SsjWjb2BatNhkekGFCwJBGPYeEr6EjWyjIEbEEgG/FiySKKZuLPz1s4ukeiFPBc1K6uSBiqjCUJyKZbBrnslfEVDbxlc67P59r1aCb+gj5XrcuU3f6asohs2u71nPUPmx4RzSsjGzYlv7tk7i/oqNq1j/R2BAFbuj6Bp6phoaaLxNPuRK2jaxxbEHBEt8qkM/v755TZzj+Wc1XNZgPkxNmy06K9/D11RHAUEC8z6cYBijGFn/6XTZx8Zx2WR5yrgCYZrsC6qUqMbIzyvT/fycN/cRxf1iBdf+k2TVXk4Pua2fPZteQrVBCYrWtuulXhRDBUiXKDn4Pvb6H31ioe+y/7qe1MX1Lv3lREZhoDvPaJdl75tdWUwh48Zcv9K1nYkoDhlchXesknVP7lb27j47/5Bo3HLs7xNxUR3efSmlK08JRttKCMqboFjpSiNVerfTFc7grGctiYG2BNfoQ34h3897Uf5ePDL7GmMIrHdg1FGwFdlDkaaeVr9fdTkhR2ZrppLl1+CvWFuHzLXQA9JjOzI4Casqg4UKDQpOAbN8g3qwRGNGxZILXJj2BBYn+eXKuKlvBgywKJ/a6aUKFBxTfhLj1k270U6xQcGcJnyvgmDSbuCmF5RdQZk4qDBST9vIsuQqFRYXJ3CN+YgTpjokckgv0a2Q1eImfKlOo9BGd0VM1Nn6g9kGXqXSG8molk2GTbvRQaFKwqEcU0ccICouQ+rILgkN7oR4vL2B6BqtdyWD6RmW0BN+/ySBHfmM7UnSHEkk38aBElu/yA5UoMCpcptOIQ0DUG4wnS/gCrpicRsHnkxGH2tq+jqypNbSbFjoFetg/1k8jn8F0hsQMEtTIbx4eYDEaYCoSxBBFbFDAlCbARHYfm9BRTwfCix3cl6tg11MXq6TEMySW+1uQk54Y9RxDYMj7AtpE+hiIJih4FxTJRbJM102P4zIuXo46TYD813M8QH6OTN6nlx7TN9xmDBvL8VGhFd0TOLbqVvTJPfHQrvWsT6KqMWjZ56AfH2XholKFVMb7zmVv4yD++RUtPkkJQ5Rv/6XbWHRvj/p+eQffKfPX372KwNc6uV3r50D8fnCM7SxI4fGsTLz7SQSmgINgO646N8eh3juErGXSvq+bJj2zh5LZafudPXuCWVwfm+jpdHeQb/+l2sjEf0ekiouPMaUaXvTJ7H1zDkV0NlH0eQhmNO17sZvu+QTzL1I12gO71VfR0VDJTGSSULfOrX9pHzUiWkt/DSw93cOyWBjRVJpQpc9fzXWzbN8j+u1vZ8+AaPLqJZNpUj+foWlfFhsOjvP+bhwHoX5Pg6613MF0dIJDT+MhX9tPYl0JXJN68p5W37mqlEFIJ5DR2vDbAbXt6UcsmnZtreeKjW9lweIQzG2uwJZEHf3SCHW9cnQtiyWtwneNvzrV+jqS5DCJwcKuoOaqAmHEZ3wq6swGx4OB4BbcKGyDobtU1vVlCb5YJvaxhewBZwPaClHfcCm1+ASvgCrmIBdfgADC9Em891krn/bVYHhFP2WLVm1Osf3GMyt4ctigw0xSga3c13bdX8dP/uglP+dLjlyMKdN5fy0u/2UExqiLpFk1HkqzdM07tmQxq3qQYVRjcGufM3TVMtodINvr5/p/u4IP/50HqTl5COlUU6NpdRdfuKixFZM2rE6zdM05VTw5Zs8hVejlzdw1n76khV+mlHJJ5+g838onfeB3feVa2AGx9YpCNz47MfVaoUHnmP2/k9H21SKbD7d/qYffXupY8z+W04i8XflvnY6N7SSlBTgUb+e9rPkJCy1KlZ1Btk6KkMK7GSHsCyI7FncnTPDqx/yIhnCvBFTmPbEXAO2kyfm8Y74RBZp0PqeyQ2uxHKtuYAZHURj/1z2YQTEhv9CMaDkZQojChEj3p5lVnV3uJniqRXeOlVOPBN2qQXesDASZ3h4gfLlKq8SzIWTwH0yeiTplM3hEieqqIVuHBN6aT3BogMKTjmzTo+mwVjU+mLlIm0MMSya1+ggM6pte1nOScRaFZIXJaRi7apDb7UZIm+RaVYJ+GEZHIr1IxAxKxowXMgMjE7hA1L2fnrNblYAgSj1fuYl1hmLbSOF57+bQWAdg4OsTBplaqsxka0jPokofJUIR14yM0pGaIFoukfX6G4gm8pkGsdHliB+dDkz2crG4kr/i4t/cEtiBypLYVQ5JYOzVKTvVxrLaZvngVI+EYIa3M4bpWhqIJBiMJGjIzPN++mbzipeBRKcseziZqmQxGOV1ZhyFKyLMiLQ7gM3TKHoVXWzq4dbALUxAvqiw2RAhbEHjMOUOci80YAQiiU76gmvtkbYhX39nO5/5iL019SVIJP9GZ4opm596SyX/+o2f58ce2UQgtTLEabo7xg09u576nTrP9jUF3oJZcaxlgw5FR1p4Y5//48vsuavfZX96AKYt8/i/2YIsCX/udO7FnSenMxmqO7Grkvf92hOqRLIdub+KlRzpYdXaayon8RW1deBHSMR+f+etXkU2bb3/2Fp7+wCY+/YXXOLmtjmM763nPt49SP5DiyK1NvPhwB/X9KSxJpOzz8P5vHOJffv1W2k9P8cGvH+KZ928kE3XTh6argnz+L/egaibf/+QOnvnAJj77V3sZXBVn/50t3P/UaVadmeL05lpefHgtzb0ztHS5qUJjDWHufKGbz/2/e7EkEbV84wsPvd1wALNOpHCbilh0CLymYcVFitsVBMvB/5ZO4XYVUXcwExLqWQPfQR29WYbZR09b46G8xYOUsZGnbdQTBtlHve7EwCMQfrrkriYCU61BTryrHsMnI1gOm54Z4f6/7SQyOf/uNB9J0vHyOG9+dBV7/7c1yy7JZ6u8vPy5tRSjKoLtsPXJIe77+9NEJha+j237puh4eZznfmcdvbdWkmwK8PqvtvHuvzpBaEZbonUoVHiRyxa3/0sPu7/WjT+z0EBpe3OKiqECez+zmnJYYaYlyMD2Cjr2LizMIpsO8nnFrQyvNM8fjoNsWHivMYEvh7biOL/d9xRPVe/kcLiVMTXOpHfeNaBaOqsLY+xKn+XdU4eougZL8nCF5O7JWgSGdNSkiRUQ3WV0wRX/BxBNh+CgTrSzhC250oOhXo1SjQdbERBsUFImetidqgo2RM6UkXMWWoWMknGfUqlsEzlVWlQVyT9uEBjWSW2x3aV4ARAErNk+KBkLUV98ELcVd/k/MKDhm3QtRbnsCimcmwjIeYtwd9mVQ/QKqEkTMyASO1rAP2ogaTbN30+S3OonMODBk1v6wT0HU5B4ovIWnklso6k8xc5sD7enzxA3L03GYslGGHZo1Sao1jJYgkhyKkB7eRyPbiHaDoPTFWwqDeHTdHKilz3hddTrKRxgU3GI0746DEGiIHoJ2iUMQeKEr5F1pVFWaZPUmBlEx+bO/tPkFHeC1ZSaojabIqwVER2bqnwWr6mzcXyI9ulxArqGYpm0z4xTk0sT1EvU5pLc13MCEQefoRPSSlTnM3zmrReIlQrU5NI0ZGbwGxr395xgOhCmPpOkJTnJyZpGRsMxmjILxV8MRLyYS+a2AwtI/RwSkwU2HBrhh5/czqYDw9y2p5dwZmX6/uew2L59qxMomsnuF7rxFxYPfBGcxZddT+yo5/6nTlM35JaSXHdklKO7GgHoX52gb02C7/zaTiTLoRBUGGmOUQwqOBPL1H52YN2xcRoH3EDIrW8O8dRjW7BFgaHWOHVDGdpOT+IrmWw4PMIb961istZdfYnOFGjtmsZbNFhzcoJwqoRo2eiq+352HBujqc+VFd2yf5gffHw7AKONUc5urCEb8+HRLco+DyPNUdedgXsPKyfyrDs6Rnz6P3DsgeSSs6A7+A7rCGWHwu0qyoCJFRXROlziDj1XRmuXMRokvMcFPGMWhXtUQMOsEbFVAXXMQm+VkccttPUegs+X0To8CLPDjy0LDG6NM93qil8Fkhr3/sOZBcR+Dv6swW3/2svZ3dUMb45dMiPg2MMNTK0KAVB/IsVd/9xFeOLiNkXboe5Umnu+cpaxdVGKMZXeXZUMbY6z7qWxSz7DbfsmufXbffgyF688KiWLjpfGOH1PDYPbKzC8EiMbYxeR+3LIqj7217fRPjNOrHzlRtBiGIhWkvIFWD8xhGIvtLxbSxN8euh5HvTGmVCjJD0hNNGD39JI6FmqtTR15SSKc+0mHpdN7oIDWoVM/2Nx9KhM+FtlivUKfb9SQTnhcS1Z2/VDA67PyXTcWrmW69MoNCoMPRrFiMqEz5Zdf5Q5W2/Xco/JtXldgi/Z+Ed1xPOvlQ3ZNe5Svpo0iZ0scfazVeRbFIywjOUTGX4oQq5dpf9DcWpfyJLa7CfT4aXvw3GaHk8hmA69v5qgWOfWAJ7Z7mf61gClag+t35lBnOuPAzZYHoH0ej+mVyTUUwZBYPzesDuZWKG6lc/W+Yuub3Aw3MaL8c18o/Zevlt9JzuyPbwzeZTW0iQe20S8QIVNdBwGlEoqzDwtTNOlVhMxi6wrjyI7FvZs5aItxUE8jsmbgXaO+xoZUBLEzQJt2iQpKUCvtwrJsak0sqwvjbA/0IbHMUnJAZfcgfpsCgd3IBcA1TLZMDE0tw2wemZhPfi25MIXrGNqZMH+jZkZGjMXq7XVZ5MIoRgZ1c+Bhja3b4WLZ61RNAqOhxm8RFk4iXJwRW56iFLlLIye9xV0PvF3++hvq2DPu9fwv/7vB/joP7zF5gPnZyC4vXREAUNdWUCWqUiomrXi1LjzYXikBdarb9af7wBln4e201M88r1jKOf5pWuGM4u0dDE850XzezQbwyNiSwKaV8ZjWIizl0ayHIRZn6/gOMimMzcZ8ejWnK/43A1UjPPbtbA8ouueUyXqB1K899+OEjhvclt7Xn8VzUTRb6ylZM2+j+cLSTngBn0tcs9swTUOVuIHv1I4EohFBylpz8UpiXkb2ztrmDggT9pIaduNohVA0BzXrz8blqGMWkhpG8ECKWWDLWBWSoSeKs8ZJYYq0b8jgS27B615dZzoyNITK1/OYMNzoy65L9V3AY68p2l2w2HD86OXDJQTgOZDSeqPp+i6q5p8hcrgtjhtb06iFhdfbpY0i7V7xglNLd1ubKRAdKzIoBPHkgVylZdfh8SQZIbDcbJeP2umR6nJpUj6QoS0Emmfn7BWQnAcRsIV5FQfHVMj5FQfaa8fQ5JZPT2GbFv0xavQJZm2mXE8tkV3RS2dlQ0EDA1bHIELyF0AgpbGmsIY7YUxbMEd5QXHQVrUNLl6XDa5q9Mma748iSPN6gw7UP+TNHVPp+e2ASInZxV9bGh80rUmoucpCm3+n6Nz/4+cXpioP/jLMdZ+aWLWN69g+iWU3PzFqjhYoOJgAUdmLtht4/8z6vZpdts/pFP/9PwgU/NilpoX54mj8Yk0jrSwz7sOzftG/cPu7DHUVcaRoP/DFaz7wjilag+FRpWq13Ks/bsJVw51hWOXANToaR6ZPshD04fo9tfycnwDR4KtvBlZTa2W5q70Kbbm+ogbeQKzFYKCdpmHM4d5MbyBGiODzzYoSioZyU/IKiE7Fl7bwOu4VmTQLrO90M/2Yj8+W6NPrWLSE2ZXvodTvnpAQHVMFMe6aAn8XD8X+/9KcTnH1OZSfPzwnkvus4EZ6snzBbbzWY6RQsVEZAofJWReopHXqOdjdOI9L5hOV2VMWWTV2Slaemb44a9u48itjaw7NoZsuKs02agXTZGYqQww1hBh8/6FqYfOImdUPZIhWeFnpDlGQ787ETJlcS6g7txxi5VxbOpL0r2uklte7cNBYKC9Ant28K4ZztC1oZpwukzleA5LEt1ANWP5l98RoL+9gnxQQbQdetZV0tSbRDZsakYyHL6tiWTCT3y6wGhjBEcUiM4UyEaWHyD72xPkQyqSZdO7NkF9fwrBgYrJAoID3qJOQ38SRxAwPBLKeZMXYbGLcJ3x1t2tyIbFxkOjKJqJaDtk4j6Ob6/nrue7AdAVCV2RkE2bvjWVTFUH2bZvEF/JmAsIvGawQD1jkvmgj/ImD9FvFwns1Ug/5kfQHaI/LKK3yQhlB2nGRiw5WDGR3EM+9CaJ0hYFsTgbsaOBlLGxgiJWVEBbLYME4SdLCOX5vO5zaNu3vOZBy8HpS34/0xQgW+3m+vvSOtVns25g3iUg2g5rXp2ga3c1iAITa8KUQ54lyT0yUSIxWLikdr1kOG4wnOPGV5wrdHO5mAxGUbIzfGXnO/jc/uc5VNvKpolB3mpczdbRPjTZw4/X72J3fycOAi+3bnCDmQU3C8ISRZ5Zs5WArpPyBoloBZ5v30JlIbuiMswiblGp640VXx1NlRBtB49hYwuQC3vx5/W5gUdYwv9/JeSQeKtAcrsfPQLh7jKe/MWzIGAhqTqsmGTn2llpzIIN8UNFcm0qggPRg66rQLyCyMpzfZdwWFscpb04Rpe/lmcS23g90sFX6x+gQs+xNdfHnelONucHyEh+utVqmvVpIlYRn63Rr1Zy1N/EhtIw1UaGDaV5UlpfGmHCE+WYr5FmfZo6PcWgkmDCE6FZn8Zv6/htnY7SCBVWHs81CN64Uqzk+agnz2Oc4Tt08H9xJ5LgUEbii8I20qg4DtzHELsZwXMekww3R3nrnlaXyIHJ2jC37u1DNm0iqRK1wxlefmgtZzdWk434Fli+3R2V9K1O0LWhCk318NLDa6kdzrD1rWHWnJxkx2sDfP+T22npnkG0HSKpEnc/exbBcejcXMtEfZhs1MeRW5sohLy0d07S2J/iXT86yb99/la+96md+PM6+ZA6F1G9+cAwPeuqePxXtlA5lseWRCKpInc910Ugv3yQpOmReOrDm9FUmd61CT74tYMAbDw0Sk9HFT/58BYiqSKTtWE2Hh6laizH4Krl828dAX7y2GYMRaJ3bYL3fNvV1W4/PUl75yTPvm8D1SNZEAT8BY07X+gheonc/usNR3AnTYYiUT+QpqkvyVBLbO7+aorEgd0tsxkQDoGcznhDmFPb6lA0ky37h6/o3V4KAqAMWyT+Zlbaebbpyr/OzW1XfMVdIlaG5p/BxJcWj7NQRizyd6sE95SRp2yMGgnbJyKWbSxZJFPjm9s3Prj80nNstLhgpeZCzDQF5zIRAkkNb25lEriVffP9T9f6MbxL040vY+DLLJ2/zmz3RNvNOnCrxF2ZvVuTT7N1rJ8ZX4jJQHg+o0RySwHLtsWGySHu6zuJJQj4DY2OqRFS/iBlj0JPvJqgrlFZyOK1DHrjNdw2eBafoTMeil5Rn87BQqAsKSi2gecqJwArJveJ+jCH7myiejRH5ViOkZYokmVz57M9V9WBxRAY1QmM3jz64IID0dMloqevXArwQpiCSLevlrci7XQGGkjLAbbm+tie68UUJA6E2/hq/QM8NvEa96ROcm+uEwFnTnT1ndnjc++jANyXOzXXtozNA9nj2LOR+QLwcOYIFxZduTt/5pqdz/WEANzCBJWUOEYl/U6YDCoCDjspso4km5i+aMm+ajzH2uMTZGI+BNth/ZEx2k5PIlkOoUyZh79/nJ6OSkp+D+uPjHH3z87iL8w+d447kOx43Y3sFuz5dEdVM3nvvx3h7MYaujdVYvgl6o6nsTwinbfW0LumEjVj8sGvH2S6Psh4a5jmXtctUTue4cEnTzBRG6ZmMEOsUGCkNcZ0XYCK0QIf+PpButdVkZn1Y9cNplcUhPbId49TNZZlrClKPqSy87UB1p5wleYqJ/K871uH6e6oohBSWH1ygvbTU/iKBqtPTRCfLiAbNg9/7zhV4zkkw+ZdT5wikirx4I9OkpjMM9YQIRfxsv2NQdbMthvKajzy3WP0dlQyUxlwVwlGs/hnl+gbBlI89IPjK5qYXGuoJRNFsxhtilI3lCYxkefoLY3ctrePQkjl1JZadr4+wGBbnGhyEsF2V04O3tHMmpMTi/bZAU4kmohqRRpz89ZuWvUz4Y+yNuWKOpVkBQeBgLnwebwwnfJq0md9R3W0VbJbSOWgjpS2cQDLIy5wL/nTy197WbddN45vcbdUKeyZy0RQyibyCsWc/Kn58y8HPViepcnYo1nI2o1R1Ux5A7zW06mETgAAIABJREFU3MFYOMa7zx5mLBRjb+t6euPV7BzumUtVnIezwNLeMDHEqy0d+A2N2mySkFbimTXbCGplKkq5q+rbhBrlm/X38vDUQTblri6rZMXknot6qR7JEchrHL29kY4jY3Svr7qqH7+WWOw9WfCZsPiOiz1u183vBpRFhQPhNl6Ib6LP59YLvit1ijszp6nVUoTNEg5wR+YM36y5m8crb+Xu1KmLhFnm0vCXgAALjrkWdc4vbOGiFhe5xtfq+gpAC1mayFFCxkB0LSJMfEsE2oWyGtuWUH4TcElvqSj09jNTtJ9ZeknTVzRY1TPFwC0VrDs8RkN/CsMrMdwRp24kw4bXRgglNY7c10gprJBq8uOcgSd+cwurD06y7cAgVQM5XntfO3XjacoBdzkvki5fUarYtrfcuIj6ocX98xVTBSqm+i76vHYkS+1IdkEbAJsOunETm2f/rVui3VBOY8v+xVUU49PFtyWQTnAc6obSRJNFpmpC2KJAMKu5aYdAOF1GNm3yIZVb9/SRjvmpHs0SnyrgLRmYs/5qBzBEib5INaYo0Z4eYzBcyZhlUZRVWrITrv81UkNYc8+z4PHybOt2vKbOzrEuqkqZ6zKeSBkH/+GFwZwOYIvCghdMWomLwQFJt5Yk9/M19gV75ZMScVafwcHVLLlUqqJgOVcUw3K5iJQK3HPmLJoks7u/k0Qxy50Dp8mpPu7rOTEb8+PMKXSKjsM9fadQLJPW1CSSY6NYJrFyAQeIl/JU5zN84MQbCEBAL6Msks67EjhARg6wp2IjOzM9wA0i96rRHG/d14pkWORiPuITeQxFYu7uvQ1wAEOWKCseiqrCcFWU3rpKeuoT9NUlmI4EKXgVCl6VolfBEgW8homqmwTKGolMnpqZLLXTGdpGpugYnCCeLeDVDby6gWxd20CHsujhtzo+Q07yU6cl+dDE69yTOoHfMhAvCKqIGXk25wc4EGm/hj24PDiAKYmUFQ+aR2YyFqK/toLeugQDNRWMJKJkA16KXpWCV0FTZCTLxq/p+MoG4WKZmpkMTZMpmiZmaB+epmEyhU/XUXUT1bg8xb9zE5YgK395HAnKdR6MiPusBnp15MK1sRCUkoklC5SCCorminlUDudo7EwSnSxheCVMRWSyMUh83B0gLVlk19N9cysBidE83rxBfffScr+/wOVB0SzXwjRsFM0iF/by6rtWM9wS49jOBladnWK6OogjCliySPVIFm/ZQHQcvCVjQapkX6SaF5q3smWyF0OU0CQPfZEaRkIVlGUPG6YHySo+eqI1tGXG0USZKV+Ehtz0shNqxwP4RFAFOMerNmA6UHagePkBfh7dDfS0ZffIc4Iul+zHrKjNUvDljbl0X1MRseSVyXxrgXknmVI0VzbRuM6QHZtEMbdAITWsleYUN89db681X6EuMqfGOT/u1GXn5ZgF3ODg84+/UuiijHkNdOXhMsg9MZ7nvV8/gukRCWY1etcmaD0zs4DYcz6Vg2ubLlLdWgyBksa6gXHCxcuveW6KIuMVYfpqE3S21HBgbROH1zSRCvmXLfBwvgeqq7F6wXeSZVGTzLKpZ4Rdnf2sHZigcTJFdTLrKktddk8XQnAc1hdG2J3qZHO+f67U4KL7AjGjwG3ps9dtJWEpGJLIZCzEUFWczuYajrXXc6ytnuGqGJa0/IOXCZ2v3dw49z/Bdkhk8nQMjLOpd4RtZ4domkhSN53Bq1+fUpaFNoWRx2KETrnPmXfcnCP3skems6XWfW5WAMm2WTUyRePULBE7sP25AU7dUUdsIkLrsWmUssl0Y4jEaJ6B9RWkq/y0Hp+mNJsrLxv2nEUD4M3rzNQFSVf6iU3OW7m2IDBQHaevLnHZ5yzZNlu7hogULv/duhIcbm8gFQ6saF/FMFkzNEFVepm8/avAjjfmA2Obe91B95f/5cjcZwdvb2L3c91UzAYX1g2liRx3r9X9P13oqqoo5YiXcuQVHyDgNQ1uHz2NLspokgcBaMxNM+2LABDT8tTnZ2hLj1FZWjxf2RGB9Qrc4cfZ7YMOFSokdzTO2TBkwuEywrN5nGMaQmplpOjGPjn4MjqFCjdQMlPtI9F/6WutBTyY6tJjdni8NDfhKYUVtMDKaCPVMP9e+VM60jJa8zcS1zpo+FqMXQ5QlJRlOWylWDG5FwMejt3WiDyb1rLuyDgVkwuDNYaqY/zW7390RdWZWsam+R//8AS3nbp4qXAp6LLEmaZq3ti4ijfXt3JoTSO5gG/5A1cIS5IYqYwxUhnj2V0bqMjk2dw7wi2n+rnjRC8dg+NXVadddUx+d/Ani0aoL4Yt+X625Puv+PcuF5pH4uDaZg6taeTI6kaOtjesaMK0UjiiwFQsxFQsxCtb2gmUdZfoZydTt5/oJVRaXi/gsiAKeMcMQp1lBBukwry/sKR6+Lv33cNLO9auqCmvZvD5x/fy2z94GXCtnWyFj8bTSRLDeQQHmk/OMNQRp+z3UNedJh9TkQ2bhjOzeeIvDy1os6kzST7qJR9TF5B70evhX9+1i39+5I7LPmWPYfLlv/wX7jrafd0nhros8aeffIjDqxtX9Jw0TiT5k688cV3JfTm0d07StzpByeehpWuGUGbpSZApSlSVMhyuWsWmqX4EHLfi4iwX5hQfxxPN9EeqGA3EqC6mCepFjla1EtUK1BZSCxuUgF8K4vx6DDaocKFkboUIFTJsUXHeE4Tv5+CraYThlUULS4ZNxUBhjtxHN0Rpe/PSEfOTq0OX/L6yL4cvY6D7ZQpxhXStH1sUlg06HNwWn2O9RH/umqu/XT6u39uQlv3sqdh4VW04QI+/9tp0iMsg97GmKIJls6rTDSQJrEC05VLIBHyMVSwuYboYJmIhvvronbyypZ2hqhhlxXPNSGcxOKLAdCzEizs6eH3jKp64awv3Hj7LR144QN3UJaQULwEBLknsJdHD6UADW3N9y/rUwVW8K0oqQbN0VcVSTFHkwLpmvnP/To60NzARD1//GvKCQMGncrCjmUNrmnj6tg1s6B/jPa8e497DZ68ZySszFo4AufVuVLqSNJFn03EiBddtIFk21gpqd2semaHqGJYoINkOgbTGxldHZvPF3X0qh/JzRA+w89mBeX1xYPPekQVtBjI6O5/tv8iPmfN76WqovKJzNjwyZ5qquf1ELx7r+lpL4/EQqeDKJ4DxbIHa6ZXl7V8vRNJlti4RJ3Ah/KZGQ3aaxuwUFaUcO8e7US0DW3ALmqiWwcbpQVanRgkYGoLjsG2yl9FAHJ9x8TPsvDOA818T0CRf+poJAiRk+HgEx3TgCymE3PL3UtZtGk6kGNzuZkGcvq+W3V/vdvU6lsDZO6uX/A5cAZk1r06w/0MtWIpEz22VrHtpDF9u6ZXHXMIVrwE3h73uZBrvJfa/XhAtZ84d4AhcMmL/apBUQny18YGrbudalHo9hxWfaXyqwEu/tBZVc30nvqKBt3TlM7FswMt4RYRzdWcWgwOUFQ/P3Lqev33/vYxURjE8N75KbVlV6Gyuoae+kuduWcfvfecF7j90Bo+5tGLalWBcjfH3DQ/y951/T1FSyMp+xrwxGkvTBM0yedlLxCgypUaIGAW6gnWcDjWwqjDOzlT3ZQXNnQu+6a2r5Eu/fDevbm4nHfK5UqjXWaP7or6IAhPxMJOxMG+ua2Hn6QF+77sv0jE4ftVxD8qMSc0zOfLtKr4hHU9y3nIXHYdVo1N4dYOC79Lym+AODlMRl8wS2cICUj+Hiz5b5pYs1oYDZP1euhuuPGD1WFs9piQtIHdn9hzO/e45ZUkEYU7UxpRFXn+kjbsf73K7LrgugqUEXgZrKih5lUW+WQSOQzxboG765yO+QABCeomQPu+PXWypvS0zvmA7ohWJaBcHEjo1Es4nItC4DLGfj4AIH4/8/+y9d3Rd53Xm/Tvl9gpc9E4QBDspkqIpqosqlqXITZYdj9u4pKzEXyZeWZkkTjIz32TieLIm+eKxE8uJ7XjsJO62ZFtykaxeKFEkRRJs6B24uL2XU97vj4sqgMC9ACjJs/KsxUVe3tPue85597v3fvaz4bEM4uW11RXndORPvrudgtvC1HYfPXc1sednE8uatwgg3OHm/F1Na17GkW8O8up9rWgOlQt3NLH7l5PseHJ6mfc+x9g//v5O4k2lsHzdQIr2V6OvC2HutbDkDWwZDYTAVCRmtnrQVQl1BUnzjcBEIq/YsG2wXasuC7RN6sRe9lGsBZ1rnxnBlEvdkDZq1QxZZrLGR8ZhW9FLMyUYq6vmH99+Ew/dvJ+CpYIX4mpAkihaVPqba/m9338fH/7ZcT7+k+dpiCY3zcCnZTsZpWRkgjY/T9TuZ19ymH9tvZX7pl5m0NXA4VgfT9fu5qbwRfJKaVJ16XmkCg172OfmscM7+cL9tzJT5Zn/jW8YZlXCMg4bTx/o5sTODj76yAu874mTNMxyHtaDQp3K8H+sxteTI7HfS8MjSVz9C/W028ZDOArlGXckiZjHSajKQ01yc6UrF0MAk7V+gtXlR7Zei57OZjRVwVFcNNFIcO76FqwFnapghtqJFP3XNJB1W2jpj1E7mWZ4d2A+PZAM2Ak3eUj7bPjCOdovL2+zOdwQIGsrz7hbDJOO6Sj211mxbiPYVBGnax2wx1a2ouU8vDLi3V6k03nK4ZI2XE6w7fkg5+9oRrfJ/OI/7QYB7aciWHOlxa1mk4m2u3n8kzvJ+q0omolhvbLXGBhNc9NX+3jm493oNpkf/PcD3PPXPWx5JYwtrSEbAkOVyVRZ6bm7mRMPbEG3ytjSOvsfHafx0tWpGlgLlrxB7VAaa86g6FAY31vFqXe1s+Opaay5kiCOqUiYqoyhSnhD+Q2VKH5w4inePX0cdR3aIQKJ56p38j+2vW/9F7AIZRt3d6rIgRdL+cKZRjfyRsN9ksRUwEfKaV9m3AVwqruNL73jJp65ZltZJK7XDbOtJb/x1iOM11Xxe999gh2jwSsan2mrn5OezrIOPeqoQZcW4hgt+TA7kmP0uxpJWEqEJUGJUakKnZpCkoKssjWzumbzYmiKzKvbWvmXtx7hZ0d2vbnGFuYXGFmHjQffdQvnulr4jYef5WDvKDa98hemUKvivZin+oXsbIvOpd93ToZxFIogRFmLm5jXScjvZufImpuuG6Ys8+rW1g11ORuv9ROs8uDJLnh7pizx1Hu287avn8Oe1UGSiNY7CTe5iTS6uP3bl/DP5Hj+o13sfXGCcKOH07e0cuDpMc7c1EL9aALHonaZJjBaX02uDI4NgLWo0T0WfEMm+TccKtBlKRHnKoUkwQ0OUCXQVrc8EuCdyXPoeyPMdHqY2eol0eTkx3+2n5ZzMarGZxdu9XYmdleRqbay5+eTxJudjO2vvvLla4JD3x8h3uig560tFDxWfvRfrqHxYpzawRTWrF6KFOzwEer0YKoylrzOwYdGOfjDkU0VBaoEErD1xRnOvbWZ8b1VpGvsPP7/7KL35gb8EyWFRc2ukPdY0G0K7/3DE1jX6JB3RQiB0yigiPVGdAUuY/M4R2UZ96HuUv7mwsEmhFQStLnhF/34Yhtj404FfCRddpoiS3NwL+3q4C8/cg8X2xveWG9yFRiqwtMHtpG1WfmTb/yUHaMrT1oDjnr+seUuPEZuTe+zIKsYi4z7hD3AczW7yah2mvIRhlz1PFezi2lb6SV06znGHQH0wC6Ohc6u+UAVVIXHD+/kwXfezOXW+iX1q29GGIrMc3u3MhXw8tFHXuAdz56p2MC7BgpktlgJ3+pCTZvYQktfvLpYkrpYirG6K2trL0bM4yTkX52AtFEYisTp7pYNHUPIMme2tdI1sYhMJcAfztF9utQremxbVakUbDRJ2l+KXPhDmSXh0+b+OA3DCYZ21aDZ1CXGPeOwMV3tRS/zObJpOtvGNt6n+lcRwiUjmi2Ve+1zaLOUQvS5tZ9/CdhyMsxdn7vAU7+5nfG9VeS9Vvpfk1u3pTUOf2+YG77Wz/Mf6VrVuAN4wnlue/AyvmCOE+/ZQrrGxvi+asb3vWY/AYGRNNd+f5hDPxhZNTf/eqB2MMWtX7rMLz61m5mtHnJ+K5dvaVi2nS2lzSvxrQdzMuAbmVXtRnHTpGnLMu71E0nGOqtpGonTNBJnYHct9tzGb9hUwEfSaZ9XWhPAma4WPvPht72pDfscNFXlpd1b+OsPvJW//fz38KezKxrY/elh3j/9LB59dYW7C64WvtZ0bP5zYz7GrtQoh2O9VBfT3BY6R0FWuTbWT6CYQkJwZ/BMWdeqyxLfuuNaHnznzYT9nqve+3qzYCoyfS11/PUH7iJnt/Khnx6v6OVRsoL6X6TQXbOtfV9T466agp3DU5zqblvm1a+ErM1KsNqLpshXjaymqSo9W5s3fJxXtrdx/1OnlvyftOiadVWm92A9/lCWqpksWY+Vk7e3M9Hl5+wNLVhz2rzHtVJb47DPTdjvLu89FQJHUadrfG2t8/8rYZXAs4Fp3wJUyRAub3Gr6IJtz81QN5Diwu2NXLqlgdBWD0WHijtcoKUnyt6fTtB+KoIto5fK5dbwriXAN53jhv/Tz46npjl/exOD19USbXFRdCo4kho1w2m2PTdD1wtBakbSq6vOidnnal79cfn553gf8x/m95FK0bYyIJuw7YUZqscynL+rid4b64m0uym4LVjyOo5YkcBYhpZzMdR1luu15sL8fc+XqNlAu9Y5wnUpZ7/xSEdZxt2W0+m8VHopJQHOTBFLYeN65POkOklCFoLLbfV89kN3c7Gj8U1v2OdgKDLP7eviv3/0Xj774A+xaUvziaowaSjEaM1HcBurRzpSqgNltquZzSyF3evzcWyzfd/92kL50Nzo1BbXZh7nrSpfved6/vcDx0pe1q/I2M5Dkoh7XPzP/3AXeYuFD//sRRxl5G0NhwQmJcMuIHbEiedCHtfI0oXp/v5x/u3Ot1CWkoEkMVZXRcZuw5/ZPDnixbjQ3kDSWXnHq9fiVHcbpiShzE6CAol3f+40uiyjmiatvTHu//9O4cgXS966gEM/G+EtPx3CqhkYkoSmqsimyc0/6F3WUCXsdxPyucu+nq0TM7jym1zquE5odhlJULHkqZBAcyio+YUue2VBoWTgNwJHZYsD2RT4J7Ic/ZcBrvu3wVJaitIcLpklidW8aiFud3LkW0Nc+71hChYLBfnKZkECbFmDhr4E9QNJbvvSpRJfZvY7yZztOGiuzVc4+NAo+38y23FSCJSVUg4ynHq6BWvQYPf7prjnr89x9//qmf99ZY+FIagZTnPzl3u58at9IC9c89yCQTJZImBUCWxCZ0tu41GpjmyQL/T8I9XaxmRsoUzjPtXmI9TkIeWz40oV0KwK23pmsJeheb0qJIn+llo0VSFrt/K1e45ysntjucYlmNUIFlKpmcTVMmpClvjpdbt5y8Vh3vvEK0tq4bdlJ6kvxnGUkUvxaxl2pUslOk35KE356JLvV7r6tX5R0mnjW3cc5ovvugVdvUr5dbHQPtScG+OrMNZFq4UH33kTimny/sdO4Cqsrpud3OPAcEiEbnFjSRpk2604xjRey0raOziJLATlztUTtX4yDutVM+4ndnZsynFmqj1M1Pppm4lhSBLPb99GUVXpDM7QHI1xtr2VhMNBQyJBfSJJb2M9eYsFi2FwtHeAiy1NTPu8hD0e7jl9lip9gQE+R8qM+Ms37vv7xt80+fbBW2pRNZOmU3GsKR3ZEOh2Gd2uYEtqyCbkPSqGRcae1FB0gWaXSdfbGbkhQPfPgjijFWjmG0Bhg95YvHKHSpdl4i4nzkIRZ7FI2m4nb7Hgz2ZRhMGlpka+ctstfP5r38BSFPTWBvjincf4u6//G6p55TcidsyJmjbxvZBnvV6mbArk4tr7mjYJ01oicSu6WHfZb2nxMbcoeINkVdeATei05lfv0lcuyjLuimGSqHIw0+Rh+9kgk20+aqbTVM9kNvyyDjTXUrCq/OT6vfz0uj2IMtTtXgtVLynLtQWjBOJpqlNZXLkCVl3HohulVpSqQsZuI+5xEKzyMlpfxWTN5pXW6YrM1+++jl1Dk+wfWKhlrtYzVOvlMavrigk+OfboplwPQMZu5Qe3HOArv3ZD+eVKa8CXytISitEUTlKVyuBP53AUiqiGiWyW6sU1VSXltBF3l8hn47VVTNb4NmWsUy4HX7/7OmpjKd720nmsq+Tgq05kyTeoeM7nsUUM0l1WrNHl27cGY1SlsoSqysulj9f6ydjLYNevE6/saN+U4+iKTE9nE20zMUxZ5pd7drFvdAzVNOlvqOdyYwOHhoZ5Zsd29oxPcGpLO8d6LvKjaw+wNRjiROcWuqaDvNzVyQPHTyw5tiHLTFd7STnKizBIAvYNTKy94esFGYK7vRSdCtWDGfyjOQaO1WLJG7in89RdSDG130fBa8ERK9J4JkHfnXWoRZN0/TrufUGsyzjPIyMgUVmUQZdlfnToAK5CkdZIhO1T05xrbSHmcmIxDO46d54DI6M4iwuOx+6JSXzZ1RetpgUmf8OH76X8rHH/d7wZUdZs2ziWxJQlplt98x2qyslPloOeLU1859gh/vme68srR5qDELTMxLn95CUOXxqmMZygKp3FmSviLBSxakZJSYrSis2kJO6Rs1pIOW0k3E7G6vwc39PJz47sJlJBeHFFSBLDjdU8fNN+to3P4CxUzklQEHjWCN2XC0OSeHFPJ1++70YiZUqDXgnubJ4jF4a44ewA20eDVKWyeLN5HIUi9oKGxTCW9DE3JYmCtTTWGYeNhMtBsNrLS7s6ePpAN8ONgbIkiq+EyRoff3//rbSGYhzoHVt1gWmfXoguOYeKK/aLVg2DncNTZRv36WovSdcCV2QzkXDZ6WvdnIZMuqJwrrOZe46fRzUMPvL0c/Q11vNSVyct0Sj1ySStkSjOYpGs1UprJEpbOEJ1urQY9eTzjAeqeN8LL2E1lhqmnM3CWH1V2W03FdNkz9DkpvyuTYEA93QeR0wj1eggXW9j6JYanJEirnCBquEsRa+FSLcLpeDAkjUwbAqNZxPEOip/n6SMCeM6whDLVenKQU+hpDdfAUYDAaaq/LzvxZex6SV9+JTdzmBdHVG3mzvOnV8X+SvXaaHQrF5Nwbd/xyagbFeqcSzJ4aeGGOkOsPVimLb+2Kbc28kaP5974Hby5SiiCYFqmHSPzfDRR5/nltN9s4ZcL+WRVtlVpsTWtWk6/kyO1lCcXcNT3Hq6j0/8+Hm+dfu1fP/Wg0R8rlkSR+W/TlMUnt3fxd3Hz3P40kjF45NS7Lzs3cax2LkNja2gJAX82Q++lelqb+W/RQhkIXDnCrzniVPc//RpmkMxbJq+pqiMBChC4CxoOAsagVQWQYzdQ5PceLaf3/zRczxzzTb+8e03MtIQWB8HQJIYbKrhrz50N1/9zNevqGYngPghB5Iu8PbkiV7nwjGu4RxZ2jdaQrB/YJxnDnSXdfqiRWW4IcCB3rH5fPZm4WJ7Ixl7GfrSc+ddZTtdkbm4pRFdlknZ7Ty+dxdZm43GWJx9I+N87ZYbuNDShC+TozUaZaCuDgmBxTCQhCDicdNfX0fY4+GakVGci2rmszYrIw1r94KfQ8d0BF/66qQx1gUB3qk89oRG2i5jWGQ6ng3T8UwERTMJ7fCQ96p0PhFi4qAfoUgoBQNL1kBeTwMUA7hcgGkdmssrHZyHKeCRdKmhTCW7yRKKaeLJ5VBMk3OtLQT9Xt569hw/OnigFNGUZUxJRlcUFMPAkGVMWUZTFBRz4V0XCkTvcDLzgIfMTivFeoWpj3iZefeCU9T0lQRNX14glAkJ8h0qUx/xknyLA9Mh4bxUpPH/JPGeyCMtWi8KQPfLTH/QS+ReF8ICnlcKNH0lsWS7uW21GpnQ/R7iNznIt6pIpsDZp1H3nRT+p3PIs49q/AY7g39RQ/ODceq+n156TgnGP+knepeTjr+K/l8XhSjbuAupFFqrmUrTOhAtidmw8cWbkKWytOgl02TrZJgHnjjJ/U+dmm+KsZHzy0JgL2q0zMT4g289zrGTl/nce4/x8s6O9eWnJYmhxhqe3d/FnsHJJZNhOZi2VfG1pts4FjtX+bkXIeF28OefeDvDDYF1GXZ/Osetpy/zWw8/S+d4mDRWLBhYZrPSOhIRyYlbFHGVoaoxJ6Vr13RsiTTvfvo0d718ga/+2g18/9YDTFd7K/fkJYnT21r5h3fdwqe+88sVw/OaXya5145hlyk0WEhvtWILL+eJSAL2902UXeuOJHG5rR5DkVHWUXu/Gs52NVNcI30hCcHegQnOlsGoD/ncjNf6aQ9G+fiTz2LIMla9NAa/+4sn0BUZdZZBv2NiCoCPPfkMfQ31NEei3P3qOZ7bsY2xQIDtUwtKbFm7leHG8o37vv7xN6zWeSXYExrWnIElZ2BP6tRdSHLyI+2cf3cTjWcTuKcLDN8UoOhWcYUK1J1PMnGoisv3NKAWzfUZ+FN5OJ2HBrV8790UcLqA9It0aYFQAbYGZ3h65w6+fOwWdk5MsW16mrDHw1M7d1CbSpGx2XhuRzdFReGxvbu5rm+AUx3t5CwWfnrNPu4+cw5ncYFXoOQEnlfzqAmDyN0uXD0F/M8tLNjcZ5YustP7rPR+vg7JBPfZAnJWkNltpe9vamn9Qpza76eQZw+vV8sMf7qa6J1O3D1F7MMa+TYLA5+pwXQsH6vY7U6m/4MH26SO76U8QoXUATsDn62l88/DVP88i2SCfUzHNqETut9N4NEManrhGTTtEqF3urBNGtiH3thyvauBso37TJOXE7d2ACVFH08iP68zf7Wh6gbHTl7i4z95gQN9Y+tWK1sJc4+NJASHekf5n//wA/7uvbfz4xv3rTnJrnxAiacPdPO+J07iqFCDPitbN0wm1BSZr73tKC/v7FiXYe+aCPGRR1/k3hd78Gbz5FB5WW2m0UzTbsZxU6SAyjNCfaDHAAAgAElEQVRqB11GhAPmNCaQwlbqlU2BDFbs6BRRkBDYMealceeuyJMr8Ns/fJrdg5M8+M6bObu1eV119989dojDF4c5drp32XdqRuAaKFKsUrBGdOoGCzhHVu4+1zEdxp0rkC6Tpd7fWoex3prlK8CQJS50NKKtsbD0p7L80b/+nA/++UdXf14kiaTLznBjgI5gFNU0l5GkrlTO1xSLE/J66G2spzGWoDO4wAQWlBaQ0xUo6M2RFt8saD++iKw6UEpDXPcPgwhFQprt2nfD5wbm/w1w+J+GEDKVseQXQYqa8M8JxFYr7CgjOmMK6CsifT6KNFE5eVkCPvbUM/O9ECTgP/30F0hCzIfj73n1LPe8enZ+nzvOX+CO8xeWH8sA/zM5/M/kiN3mIH6LE+/JAi1fWLlaR3dLjP7nahCw9dNhvMfzYILWoDD4XwNMfdiD62wB1/kiSJB8i53YrU5qfpyh4zNRlKxA98iMfaqK9P7l6dqaRzI4L2k4LxdR8iXSdOKonf7/VUviOge+Z3OoaYF1UsdzMs/Ux33ktllxny7M38/UQRvFRpXAz7JYp0srJwGYVRKFa60YtTIYYDunYb1QGv/8QQvFXSrIYBkysL1cRH6dCkBMj0T2bju240UsY2uv9Mr33AFFN8k7LQSbvbgTr88vsheKvPeJU3zix8/SGNk8qdcroSGa5A++9RiyEHz/1gPryg33ttbT09lMUyhOv6OBH9W+paz9IlY3mrQx0tmJHR08fNP+shqhvBa7hqb49Dd+xrWXR+a9OSgZ7qxs4bJcw1v1PlxoNJvJeW2CGcnNaaUREOwzgryqNLDFjDEjuXGgsduYWbEHu003uO3UZWoSaT7z4bdxqru8zmLzmDVe37rzMHuHJql9TacxWRNUvZRFKKDmBEKBlSjxEuDMF9kyGeZcV3niMb0t9Rgb4A2shLDPzUStf6Hi4AroHpuhbTpKXTy9pkRt0uVguLEGXu2r6FrchQI39Pav+J2QJIaaaihYywsvK4bJzuGpFWvl30yQYEmDlddqoc+xrTd0juM5+C8hxO9Xw5FZ1bmVYAp4Jov0YLy0zzoDRBIlLYc5bHYa6UrI7LWR3W6l6rEM3uP5+XGzBA3qvpem729rSO+34bxcRKgS6d1WhBXqvpdGzpauUU2Z1DycZubdSzkOEqCkBZ5FkQJJgGNYw9FfpFinICyl2UnWwfdintC73cy8x4379MI+4Xe4UVIm/qdzC5KzVki/z0nmbhuWQQPJEChxc964I4NZJZO7zYY6YZT+v8JyyvXCCMjE/shN4NPJzTXu9ZNJ9r08Qd+eOppG47QMb07O/YoQAotu8MCTp/idHz5FILFxZn45kIDaeJqP/+R5Jmr9vLCns2IPWFcVfnloO3ecuMiM1cdzVTsIaGmUNZSHsrJ1zYn9ShBAxOviu8cOMlHjr+yahaBrfIbPfumH7BieXkaycaDRbibol6uZkLx0i8iS708oTVyQ6zAkCTsG47IPCSigksGyav24IgT7+8f5y398iE/80YeYqK3s2k1Z5nR3K8/u6+Jdz7y67EzZLVYKtSo1z2WIHHXhHC3iGF3uvduLOtvGQ2Ub91CVm6jXhTtfQTnUGhiprybqca35+7tHgzgKGjuHp9Y07lm7ldH6KoqqsmplQSUwJYlL7csVvq6E+miSmsQb1+J1vZgrmNrUJZwJPJ9DGgzCYTviLjfss0GdAhYJQgacySP9JA0v5ZBCxooL0jc7sl0WTAu4e5aSWCUB1qCOJWKS22ZBWCRMm0SxScUaNFBjSxUkHYPF+fz5HAQlDkBmj5XonS5yXRa0agXDLVNoVfG9mFuSr3W/msfZpxG73Yn2t3GsEYNCo0LysB37qI7n1YVcu7BJ5G634ni2iPerpYiOtCgVbzupYb2sY3okjMY3mXT3a1C2cZdMgSeeo/NCiPa+CLIhrgpbeA6KKTh28jK//dAzr5thn4MEbJ0I8euPn6CvpbYkN1qh0X1+79b58OrhRD+/OfEYPn15p6jFOONu5+/a71vXNYtZdvxTB7orC28LQWswyn/76iPsGJlecZwFYCBhIKFKJrqQyaMiSQIdGY8ocrMxzHYzjEtopFULEclJtciiILCu4XZIwNbxEH/xTz/iU7/3AAmPs5KfTszt5KdH93Dd+aFlUsaaR0b3ymhuGa1KwQitPDb2osa28WDZeXdTKoXQ22Zia25bDgQw3FRDzLvGbxeCHaPT2IsaewYneerg6r3ohSQxXldFxOuiMbp+9azFMGWJ8x3lG/euiRmc+eLr+g5vBh6qPsxlRyN/PPGjTT2uJIBJHfGjNNIj6dmQwCIZNhMw2FADkzcahlMGWUJJL1+ZSDrIBYHhlksrJ6WU/5bzYhl5TsmKZUp0QoGJT/qZ/LgX27iB51Qez+kCQoHw25dXPckFCPwkTfJQgMi9Thq/niJ+swO9SqbxawnkIpg2ML0yZkDGqJGR8gLhlECAvKjVriRAFAXSKil6IYHpkxCzXAEpK5CTYsn9LGub2SiBsAEaiApFkMo27uEGD4+9axeqblJwWEqG/tJVyrkLwdaJGX7ve09QH9u4Us96IAE3v9rHkQvDPHrdnoo1h0N+DxfbG3BNFWjLh3EYBSxrdAryGJV1d5uDAKJeJ49cv5eUy1HRvt5Mno/87DgHe0eXtYQEUDDxiQJDckl7vdlMEZRchGUnKiZJ08Y1xhRPqJ2EFBfXGePsMWcYlvx4KOAQ+prGHUrv+MHeUd77xEm+ds/RymriJYkTO9q52NFAQzSx5Hf4evIM/UaAor8UqlMTKzd1sOgG7dNR7EWNfJldzi52NHL3y8vzk+uBrsiM1FevWTduL2q0T0ex6ga7hqeQTHNNbYiJGj9hv3vTjLuuyPS2rt4DfDE6J8I4NzHC8XrBQN7U/tqvhSSA+VT6r7AlXwFqwgRToPuXP5umFQynhBI3S+kGA+SswHBKpdTZIswtEhYju93KxG/58J7I0/npCPZZPkK+RSV+4/L5TwKqfplj/PdMYre7qPlJhsR1DjCh+uclh6u4y0L6/Q60TgW9SSH9Pge5YzYkTVDzqQTqVHnhEyFD/qiV5MecmAG5JLwzZeL95wy2l0oRQyFD/vrZbapnt5k08X41g+3E7DYS5G61kfhdF8ImoURMbK9qiCulcVZA2TNo0a7gj+bIeKxE6ly4E1evbMCm6fzOD59m+xvcZMKdL/L2Z8/wzDXbSFZoNJHg1PY2fn3wFZoKURxl9Pit0ZLcHT5d8XXOafK/sKe87nNzkA2T63sGuPeFc1cM2VoxucEYw2SB9d4iUrxXO79ku3fqFwEJCUGdgC1U3rPblS9yz/Eent+3lQtb1u4xvRhpp51fXruToz2DSzQGJF3Q/IMExRoFS9zAFlmZmCQBgWSG+liq7BKvCxV4r2sh5nEy0lC9Zt14YziBL5NDEoKmcJyqdI7oGjoGkzU+wj73pkXapgI+Yp7yar1l02TLVGRp69l/x+uGqMPFqK+Wa6aHN+2YklHyps1VKBfuCwXkoiB10EbDv6TmPXIhQ6FFRa9WcPYWkbWSt2of14m8zYVWq2Af0eef01yXBfM1Viq7w1rSKThdmDfsAtB9MoUWFfXS8oWkJWlS80iG0H0uwve5yXVZ8L2QxzpTujDrJR3/36Yx/TKhL/pxfTeH++EcCFBC5edFtG0qsU97sB8v4vz7DMiQvt9B7E891P5WHHXKROtWif2pB/tzRVxfyCAUSL/HQezPPNT+dmkbvV0h/p/dWM9qeL6Vw7RLpD/gQFRQRVl+zn0sScflMEM7aqgKZ2gcu3r9eW9/5RK3v3LpKh29Mhw9P0TbdJSezqaKQvMCuNDRiNso4C6zjV+VluY9wRcrHldjVh0vW4kIEODJ5fngz1+mNrG2gt5agf7S9xvzPiRK+eQbzg3Q31JXcbXCY9fu4Dcffob26ej8GBZrVGZud5fyejUqzd+O4x5c2YusTmZoiCTLNu69rfXzOu0bRcTrYrR+9a5cAC2hON5MDgnwZAu0T0fWNO5pp53R+mr0TWp2c35LU9kiVr50rhRNeROVwa0EARQklR9XH+KCo4VGLYZlEUcmJdt52reTc852CrJKd26Kt8VexWtkOeHeypC9nimLnz3ZMTRJ4aKzmXtjp9iZm2Ta4uPn/v2M2GpRMLk2PcBd8YUujiddW3jCt4e46sRAxip03h9+nh25SWKqi8d9e7ngbMGvZ7g90cOe7OrCTYsRt7s4X9/KtsgUBdWCL58h4vQQs7vpiM9gM3QiDjdBt58tsSCmJJNXLUx5qtgSC+LSVjCUIRPJgPReG7pLQs0sv7fOS0W8L+VJvsVB9E4n1b/IggmFZpXpD3lxDGm4zxVK6QdD4OopoGRMpj7ixXkxjJoR6B6JmQfcmK6ls48lYoAMxXoF0yohFwWGR2b6A14KTSquFYw7QO0P0gTf6yb8dhfFBpXWv4vN8xnknEDOCYycAK1EolMnKn9XcjdbMd0Svi9mUMKl/dUpg6mHAuTusOP5RpbcrTZMh4Tvi2mUaGnslCmD6YcC5I7Z8PxrjvxbrOgNCnW/EUedJc8Jt0TuhvLn+LKTs6pusvvkJHd/p4fus0Esxc2t752DL5XlY4+8gH2VxiBCmi1ZkBdIL0IqleiJRduYculP0aVw6W31pBps8/uWC0dR463rDL1ebqu/IkFOrPBHAqyi8pKXy231vLRrS2U7CcGtp3o5fGm44vNdTdh0g2MnL1MbT5Xd9WkOMY+TJ18jRGMN6TQ9lKDlO3G8Z/MI+cr3vzqRoTGcKPu8Kae9RADcIAQQ8bnXNu5C0BKK4Z3VePBk83RMRVbfh1Le/XJbPfky2e1roaez/KhKfTRJbSz1K5Fvf9y/j+Oebu5MnGNLPsQz3p3zz4qMSVGycGf8LG+PvkK/vYEnfLvRJJURWy2XHE20F0J8t+Y6glYf9VqCFz3d5CQLsjBxmkXeGT3BTcmLfDdwlB5nKwD99nq+WXMDd8df5ePBJ7ELjcPpATrzM+RkK49UHWTG4uPdkZfZnRvnX2tu4ryjsnbAKaudJzt3M+32EXL5+PqBWznX0M4TnXuZdvv52sHbGKhu4Cfbr+VCXQtfPXSMKY+f7+y9AX2FlI+zt4jnRJ70Phvnv9PIpS/Vcf7rDYTesbDIlIvQ8ZdRrEGDwb8IcP6bDVz8ah09326k2KTS9OUEzj5tPhrofSlP4KcZ4jc7OPdwExf/qY7z32ykUK9gG186J3pfzuPs1Yi+zcWlL9fR+7laer7TQGavdVUhGuuUjvflPKn9NiwhHdeFzU8VGS0KyoyJnFxYGCgTpc/F7lLOQW+d22ZhnlHHTaSUoLit5NAYjQpSTqBML9hZdUhfUs2xFsoy7vFqB5PtfqZbfYQbPJy6sZ3J9o1PassgBG976TxbJlfP5Ye3ucn7LQzeVkc2YCXe7mBmp4ehmwJEO10ICab3eJk86GfsSDWmImHYFApulalr/GjOyvJoN57pq7xGV5KIux1EFxGkBKBJCuO2ai47m7ngal3257KzuaLFhwC+d+vBsntqz8Fe1PmPP31xSZObNwv2DYzTXobRWgZJ4rG37FpS+635FcI3uwjd4ibbaV3VyHizeZrDsSVlgKudS1fkTZGK1RWZweYa0mtEXhTTpCUUxzWbv3bnCnRMRcoqMbvU3kCuDLGotSAoSUaXi4Zokrr4rwZT/gnfbm6Ln+doqpc7EufYl11QmXSYRW5NnCegp3AbeZqKMaasVeiSjISgvRDirvhZBBK3x3vYnpukIFkoyirVeoabkhdxGXkainGq9TRDttJzM2PxYhcaHfkQLcUIbYUwBdmCjCCmuOhxtNKVn8YidGqLSWxC47yzpew5wpQkzte3kbI56IpMc7pxCxmrnRmXl6zFxsnmraRsDia9VRRVlZTVwc7QBNdODOLQCsTty6NCkgFb/zhM84NxLGED3SejJg2sU0uNsHXKYMfHpmn8ShI5IzDtMtU/z9D9yRkCP8osIc8paUH7X8fw/1Me26iOaZep+mWWrZ+OUPOjNM5LxXmymZwT7PhEkIZvpECAViVT/ViWHZ8IUvvDFPZBDWkFH0kuCKofy4ICVU/nUOOrq22uCwalDoBLTgxCkRbKGQ2x4jYoLIyJuYLkaoW6GmXFPc8faiJS58KeK43YVJuX+snNIecshi+d4/pzA3hy+VUHfXqvD0e8WGr0kNOJtzkxVZnGV+P031HL3u8W6b+zjvpzSaqGs0gCNKdC/x111F5OY8lWFnXYOhEmkMiUrT0+B0ORiXjd1MyGvU0kTni7+LeGm4hZ3MQsLtx6HlOSyCh23EaeXekx/tvgt8s+x4zfw7P7t1V0XQCHLw2zfTQ4/3nK60dTFFpjkVXH3pAkLtc30REJ4VwhZDeHnGph2uenLplYMbS3Gmyawc1n+nl515aKlQIvdDQyFfDSHC6x5uWiwDGugQD/qznskyuL2EDpXWoNxnDnCsTLYOzrikJ/cy13bDCFVLSonO9YO+3jzeRpiCTmF5pW3aA5HMdVhvjOQFMtSaed2nh6QxNa2mFjrL6qrG0lU1AfTVKVLK9x0huNhOokoC9EGeqKSYbttQAM2Bv4uX8/FmGgCIMBez01emkOlDGxmjqKMJEQuMw8Et7ZiJzES54ujnu24THyKMIkorrRpdJivKMQxmZqPFx9LW6zQEJxcm16EEUYaLLCmC3AS+4uzjtL3rrXyNJcLL9CQxKiFF4vFrhc20xrIkzQ7WP3zBiNqRg5i5WQ08PumTFqsikiDjeXa5s519COIStLmsoshpoRtH4+AVy55bQEWKMmLQ8m4MHVW1NLlOaMSL+HI18cREgSuipjzRs0/EMKIUkYkkyq1o4tq+EIabT9zfJxqPthBlj5eRMyaLUKaszEcyKPnNt8x8YyqJO9w4Zep2AZL9kZrUtFuCVsPSX7aenXyd1mw6iTUSdLjoS2TUW4JKznS9wUdcxAWCW0DhVrX2m/4nZ18wl1e16ZwJHVsM72cJ9pdM83kNlMbB8NsnN4ekXW9mLYkxrBPV62PBMmuMeHI1rENZOjejDD9F4vRbeKJWvQdCaBe6ZAwaVgWCUSLQ62PBOueHJTTZOuiZnKjbssL/HcNUnh4drD1GkJPjD9DF9quYu3h07QWIhx0dXCcV8394VOrHLE5Xhh71YivrVro5dACO594RzKIg91sKaOqMtNXrXQGovg0DXiDifTXh9bwiFshk7M4WTCX83xLduwaRp1qQTuYoG0zY4sBK5FE0HGZuNMSzsNiTjdM1NUZTNkrDbGqgJUZ9PUpZIUFZXxqmryFgsNyQTVmQXDc+TCEIphVmzcCxaV47u3cP/Tr5Y+11uQDPCdK48A2haM4snkyzPuqkx/c21F17cSChaV82V4w1WpbCltsAj1kSSBRGZN455x2uhvqWPrGlGxtTDUWFMK75fxvNmLGq0zMWybLNF7teA2CiTUhfueVBZItE97d5KXLXwg+BxeI8t3a44yaV1Y5Cyucnltxct3A0e5IXWJe2OnsZkaF50LssENxTgeI48uKQS0FLuy47QXZuZ7NDRoCe6LnWRXdnx+H1WU73HWZZK888LLeAp5iqpKXTpBUSlN+4pp0hkNkrXYMCQZ1SzdJ1cxh6eQ5ehoGKd+5YW5rkgEG31knVaqIxmsRR1VN0GUZMVzDgu+eK6kMe+w4MwWUVYJK2e8NoItHoQkkfHaGNtaxZZLYXqONNNxMUyixkGkzk2ixsGtD12edzbLhbBLhO914TpfmE8JVLS/DKZbwqyWMf0ypktCb5RLvTgSpTI+x5MFMvfZif+JG9cP8iVC3XsdqMMGjidL86Pzl6VtYn/iwfVQHhRIvc+BZcDA8VRpG/vxIuqEQeyP3Xj+NYuwS2Tvs1dEqCsrluuL5bEUDHIOC6Nbq7AWDDzxzWXLy4ZJ91iQltDaq9Kq4Sx5r4WavhSpRhtbngkR3eri9Ifa0O0Knul8SV1q0XNkS+hc881x+u+oJ1Vhy0ZZiDVTBStBV+SSKMksTEkmaPVzc+wC1yV68WlZ2vJhrkv08v7pZ9mXGuFF/+p1y4thAid2tpMrs3RrDg3RJPv6J5alGk62ddJb38gjew4w7fHxzWuvZ7Sqhu8dPELY5ebbh44yXhXgcn0jUZebFzu7SdvsPLT/WtK25WM64atitDrAF2++k7xqIez2MFJdyxdvuous1UpPUyvPdu3gJ3sOkbQvrUbYPhLEt45+6YYic7ZrIWwpEKS7ZrkWZfAt2mZieLP5svLupiQxWesnu4Fw95yU60BzzRobCvzpLI2RpRGzhgoEYk53t264m+NAc03ZREd3rsCW9aRX3iAcSfXxtHcXFxzNvOLawml3x6KcuyAj29AkmTOudk64t7JyUeVySAgysp2MbOPRqgNMW/zMxVyzso0RWy1HU71cn7o8G4IvGVm/nmFLPsiTvj2M2moJWvyccXUwZS0/Jeop5umMzVCbTdKcjGIxDQ5MDXHN1BAN6TiKEOybHuGaqSFaEhEspkFbIsK+6RE6Y1euVhLAaHuAZ2/dxkvXbyFU7+bcNc0MdQYY7gzQ113LQHct565pZqytmou7G9dUzQxMZ0hWOxASOFMFEjUOwg1ueq+px1owOHekmd4D9fQcaWaqzLTwHB9LyBB8wEOu04L/2Ry2ySsvDCSTFScKrUsh9EU/wa9XkbnPTv46K6EH/Ux/q5r8kdIcrEyYBP4kiVSA2J96iP2xBzllEvjjBHLEXNjmj5NgzG7zRx6UuEn1nySQ5wh20yaBP0si7BLR/+Yl9UEnru/nUCJm2aSxsunI8YCD7/zmtbQMxYjVuLj+8QG2XKrcC74SfJkc+/vHy8p3BvrS3PaZy0iG4Lb/Ufr72q8MY1hllKKJbMK1XxmeV0ayZgx2PTyJZELtpeQyoYS1IAlBQ7TyenshScvynDICq1kq9fAaOYKWkpqbw9TYlRnjmw03lX386YCPwaaaijXOdw9NUpXKLrt3B0cHOTg6xPNd23lhaze7p8bZPz7Cz3fu56WObTTHYxwd7GWwpo6mRIyLDc1M+fxM+KupSy1P0zQm4hweGSBrtTFcU4timMSdToZratBkdVY1TcdTyOHLLRX4sRgG28eCzFSgXw6z6mltDfPMcKUgCN7tIXrUiaxD+1ciuAeu7I3URZPUxZJc2NK49jskSSRdDiZrfHRNrN8jPr2tdU09eYBAIkN9bLlxr4ulyhLfeWV7e4mPsAEJ0r6WOoplRlPcuQKdk6F1n+v1xtujr5BUHPxDw110FEK8J3KcPnsjAPfGTvFvNTfy/7Y9QEc+xA3JyxiSjIzAZRSwSKX+CTVaCkUIbKaG18ghC5PfCj7O12tv5pT7fg6nB7g3dgq3mS+Vb0ky9VqMz7a8AwmwmTpH0n28N/wibjPP+8Mv8GjVNXyh8a0IYFd2gpbixhZMcyS2lT7vnR5hrqR1LegWGdk0caeKeJIFphp9aFYVARStKrvOTfHNDx1m/+lxFMNclYQtpNLxdFUm67biiefxxPKcO9pMZ08IS1GnfizJNc+NEQhm8MbKcy5Th2yMf9JPsUGh0GzBdzxP7Q8yYCyQmAUL/5ATgsb7IqXow2vGydJnUP+R2Mr1pLNrBQlQBw0Cf5hYcJ1nRYnmIAGWAYOaP1i+jbRoG+sZjfoPx0rblJTEcD5eKFuKuCzjnnFbGdkWYMfZILtfmaB3X/2may740zn2DE6WtViQWNCAnv/bBDm/sDBY3NxBYkETulLDDoCAwDrkM4UkoVkWJkIZQZWeZtrmRwCduWmO+7u5Jj2E0ygyYQ8gV6A1OdhUU5IfrTAkv2NkGndu+csx4/ExHKhDINEajTASqKW/th5dkWlKROlpamWgtp6MzY5imnTPTPG9A0c4dvn8ivcta7MxHKgj5nRj1XVOtG+lKpuhMRFHESb+bJaLe5o5OtiHTVteB71tbKZyPsEskXGixk9HMIpzTOPgb42vvd8sZErpoWf3bysrJZBy2JgK+Ddk3E/sbF/zHlp0g47pyDI9Asds6NuiG2sK//S11pF22OY7KlYKTZEZaQyUlyoRAl86S8tM5XoHbwQkwG0W+O3g40v+/y5KHRrrtSSfmnp0xX3fFj8z/++/Gv0mAP5sln3ZMQC8uUk+M/qtZfsVJYXvB47QVojw8eCTWIXOhLWaBxvu5LZ4D23FCDV6ig+HnuXDoWc342euiUpKWqtiWcbaq9h/chxPMo8nVWC81Y8kwJPIYSnq1AVTDGyr4dgvLq86t+sWmeEdAZzpIpcPNLDvxXE6LoUZ3hFg94lJXKkizYNxBvbVMbSrllsevlxWeaWaMrGP66gpk8CjWeq+k8IaNsg2WbCHdCRNoHtlUtvteC7nsSRNdItMcqcd13BxiS7GUtGhK0OCeXGeTd+mgkxEWcY92OIlUe0g1ODm/KEm0l4b7b3RTfPaEYJAMk3rJkl5bjYkBI5C5WUTQoKiujDEijA4lBwkL1vRJZmbYhd4onov/7vt1/DoWS66WrkzcmaVIy46NjDUGCDkXy63uBpsRX1WiW3pU9IeDaHLMoYss39ihM5QEEOWyVpt7B8fYXuwVCebsdo4MtSHp5Bne3CS7x04wv7xkWXncRaL7JwaJ2O1c2BsmLZopETecXu5qf8SsjAZqa5h9+Q4IbeX/roG9k2MLjlGWzC67LjlIGu3MllbMu7ZVguSIXBM6mS2WFETBtbo6gHVncNTqIZRnnF32pms8a3rOqEk5Xpye/ua29k0na3jK3vBXRMzOAramsa9YLVwfksj1/cMretaIz5XSQynjMWkLASdk+F/F69ZBYYkk1CcBPQ0DrOIQCKpOFDFQhfFNy0kGOiqZdvlGfJOC0OdAbb2zZBzWrAUDRonEqi6ybbeGc4eaKYmtLpzZCma7DgdZMfpBZKvPafz9q8tdKzbcq8EqkQAACAASURBVDlCe1+0VB1S5vA4ezU6/3xppKPolUntsmM5nkHWSh3lwte7sMQNLMkCSBA/UEoTXkn06lcBZRn3htEEVaFZxrckE7V6cMaK6JI8yxDdGGQh2DY6g1178w6kbR3XJpCWlKipwuTe8Ek0SUERgi25GX5n7Gc8Ub2XvGzhHaGXuTNypqzxzFstTNRVVVy7XJ3KUpNYzphuicdojseWNMo4Mty/5PN1w/3zSnVBr49f7NjHr/WcwqYvn8CdWpFDY8NLlO12Ty140JqiUFRVclYrWYsNT365N1m3TunhokUl4nWheWSiR10ISeDpLRJ9i5PAcxms0dXDN7tGpssrhwMyDitTAR8m62swMhnwMxVYe3Fg1XS6JlY27lvHQzgKRZLu1VUUTQle3da2buM+HfCRWOMcc5BNwZ7BqXWd52oh47MSbXFRN5jEllv6DAzvC5RK2s6ub0G5HthMjdsS5/lF1T5ecndhShJuI89d8bPUaVe/A+ZGIADNqlCwqhiKjCOn4YvnuPGpUidBCTi/t5Fz+5q54emBTSu53QwxJNMuk9phw386C2mwJkxsIX1+waBmTBxT2pu+i+FaKMu4O7MazqyGiUSPr43v19/AA+nnOFO7hV8ffQZlg6tM2RTsGnlzTQRLIFjCLK9w13lIsKR5jACOJPq4JjWEQMIidNQ1OsfNIeW0Vd79jZIKW+AKinSr5eLmMGfAalNJ3nP6JayGvuoktNjgLclfGQZ3XziDLpe8Y/sKC4T1dhIrWFSiXheSIRAq5OssyDp4evM4R9eOwLQFo/jTuVLd+Rrja8gy0wEvGYcNT67yNsinu8vItwuBvaiz9UrGfTKEs1BcM+8uJIlT3a3rlqGdDPiIlynDrJgmu4cm1nGWzYcAdKvM+O4q7GkNyQRDkQhu9aLZFBr7EmSqrGT9NgxVpqEvgSNVOZu6UsjAvuwI2/OTaLMa9oowsZsaSpnpudfOvK/XgkAScN1zg+gWBYTAWjSQXlOa3X0xSGdfGFvhzRW9sYVmJWsrKCv7VURF+p6GJJNR7GxPjqNLCnnZgiHJKGs0RFkLsmmWpbT1RmKzHoM5IZu46qIgW/DrGTzGArlGlElmWa86mjeTK7HBNwhFiFXr3NeCBFgNA6tx5Wdnvc1GChaViM+FkhU0PpTAtMpY4gvnWeteyqZg38AE4+WMryQRrPISdzvWZdxf2d6GWQYhsmMqgvsKx3flimydCDPUuAbjnhJPI+F24E9XVokgKGnKrxUdmIOtqLNjkY7CGwmJUm92WTcZ31WFfypLtMXF4KFafDM50jV2TFliqtuHWjSItLg4+Ogoin71PTcFgdNc/3uUc1i4tLMB2RRsvxzEkavMkKbcNqaafCiGSeNkAmeZ+0uARTex6FdehKz1/RsBIYHmldHcCvl6FWvUwLBLaH6FQp2Ka6SIYZUoVqtIusCYlbf9VVwGVGTcFWHQlIvS62mm391Icy66ZqezciALQVtw7Xy7qUqkOmwktjko+lVMy8pD7r+UpfZEelnLRMMmEdvpJN1mQ3crpVzsjIb/Yhbn9NVfXQogpdh5onofJ7xdRC1u3hd8nltj59EkhV9W76MtH2JXZnzNhyntsJUVzn0t3LnCFY3Emw3rSYVASVwm7bAhAaZVQneV2r6mu2xUncxiSaw+4UjAvv5xHr1ud1nnC1Z7SbidtIYqI4/lLSoXOxqvKFG8GHuGJlYV39k3MMETh7avfixJImez0ttax1suLudJrHqtVgtTNb6yy+C6xmfeVJ3gFENQN5wmWVtanEx3+eh+IYiqmQQ7PZiyTPuZCI19CcZ3VaHZFBT9zZsmnMNks58/+pv7sRY0Pv8736ZzsDJiZ193HZ/907txZor87uef4vCJyp6LNwLCIkGrCm0WqFbAKYFVAk1AVkDUgGkdhjWk1wjVCBkKdRYcUxr5RguuoSKFGhVLwqBYpcwbeqkoMOwyhlNGXoHlL5wStFugWZ29hlJ3NzQBRQExE8I6jOtIEYMKeNIrwpAlEl4H/kQp8ruWFgxUaNwB3EaO22bOogiT2sLqqkPlwqoba+ZXTQUG3xNg+N01pNptaB4VobI0DCkElqTB9n8OUvPKUuOebrHS/6E6pm/wkm20ojtLxt0W1fFfyrL12yEan0ogX0W9DQOZJ6r38UjNIbZnJ+h1NZJSS5ONLATn3G30uNvYlVmd3S0oee7RtXp/r4DzW5r4rx/7tbJzym8ksvbK6vfnICTQVBVTgnS3DcMpU6hRKdSrOMeKqIm1e4vv7y+fYR+s8hAv06NdjLH66rIEiCRgz+DkqtvsHRgvq/93zmaht7W+YuOecNmZDJQfKapk/N4ItJ2JcOHWJmxZnfqBJEWHytjeKpK1dryhPJb8r4bwzkbhTeYpWhSKVU4mm9ZPDH1d4JQQRx2Ie9yw1QqNKvhksMtgocQkz5uQMCFswIiGeDGH9HgGabauXTbA01fA07fg4FhSRdxDxUWfTdq+vbKzKapkuNuNuNlZMu51CvgUsEulF9WgZNyTBsRNCOqIy0WkJ7PwYg5JqzwapCkyJw52cPxQJ5/4l+d4dU8LNx3vX3MOq9C4S0SsXp6u3UNdIcGN4fNUFzeuG+1L51BXCc8CTN/o4+wftoCAzu+GaX4sjpBh4q4qBt9Tg+6Wueaz4zT9Mo41rs+Xvgkg12Dh3O83M353FUrBpPWRGL6BHJpTZuo2P9M3+UhtsQMSTU/Ey5ok1wNdVniqajc3xy7wztBLXHItKFWpmGzJzfBk9Z41j2NKElGvC0OpvNf0ZK2fyU1odvKmhiShKTK6rGANGQz/th/P+TzuviLlJlg6/3/23jtMjvM68/19FTqH6ZmenAc5B4IkmIMYRFKRyoG2pWvLQba1ttfrdJ+79vqu7+7KYeXdvbZXa8sKK8mSaVkiJYoUIyhmgMgZmJxnejrHCt/9o2YADCZ0N2YAUld6nwfPg56pqa7qrqrznXPe874jU/gLJbIVOO2l/F6maoJYQqBWQcI5095Apox/Ozg6C+WC+9beUVTLKisU4mTujVX33efm+SuClG+r4D73jdiKQEin7dJ8LkUgUQQE3lQJqQg6jsaQisCTNlAtiXQL5LdaIXL1PN0rQr+B+KNJxNDqVxJCqTyKlBTcGsma6pOFpSDv8SF/uw4Cy1+P4v+ehqeyy16LUgBrdOS/q4O9Xuf7UFm4KHYBLhVCKrTrsMMNd/uQnwwhv5REPJpGlKp/uEtwFg/3+pG/GoF1Lggu9JkHnONyCee8W4BNbrjFh3w4CEeL8Ndx2J+/EJ8qQTLkxdBUNNPC1BSGWmuxVKVsglZVcBdIerLjRItJjoc7+fP1D/Mfj31lxf2IcjPkUsCpzzRh+lTWfGOKrX89ipp3TixyKo/lUTj/4SiTNwRZ/w8T847HdgmGHqxl6MEI3imDW3/lPDWncxfu+I3/MMHR32rlzM81cvKXGgmfzeMfLF6VHouNIK4HWJcfI2gVFiwiglaerFr+YW8rouoRuJ82WKqCpSr4+0ps+LNJlIKN7RYoxcpubo9hsmFwgjc3dJTdViqCvpYoJV3Fu4yb4by/AU63N5Y1iwFnaqBpZnkvh1CuwJrRaU52NS+7naUIRupriAd91KZzy2576bEmA5UHd82y2VZmMXKt0XddlLM3NtJ2Io4vUUKxJaFJh3syd6+7cua81yhAjw7RqgucqwsJLNGCXClcRRMhHSnlvHd1XAMBJ/it0Z1AuwzkDV7EU0t7D0i3cILqn0ahQXMCeqUfhSKc99+qwOcbkHf74E+mESPmQibiUu8P0KwiPxOBT4bBP/vmlRKZBU6wr1edhcZuD+KLceRXUhCrTOMwkC2S97qYrg2w76Z1aKZdEcG7qqtWAqOeWg7V9GALhV/se2pVgmCkzEPG9Ckk13rQ0xaRE1m0/MUT0zMW4VM59KxFYpPvEskhB6UajaEHIkhFsPGL44TP5OatmtSiZN1XJpm8IUhsZ4DpXQH8w8UV90gWgwC8VomU6l1wbRlCZcBTT32pvCGPrShM/yy4LwtLVbAUge0SJHZ6QIHA6SKatCsK8Jpls6l/rKLgDtDbXEdJ0yoO7lmPi4HmOkp6+axwc99Y2bEcIR03vXLBHSGYrAkwGg1XHtwFTNUEiFegtw/QOR674Dn/doAAeg5M03NgesHPl3v904BU2DtL6KwiaK4mti69uJUBBfmBIPxhnbNYqHIy6AKEcDLqBwPIOg3+ZApxtMJn/Bod+ft18GDwymZdLz0GgIiK/K06aNcRn5+BsfLPC3fJ5LZXztA0kUQKwbaTS/NvLkXVh+uzStwcO8m7R1+jO7s6bNhyxBupChACISXKIgzWOR150yvmaWdLoBDViG/04Y6b1B3KLqpQ50qY1L+RBlUwfmsIW706V7kmLXZk+tkX2cThQBeGopJV3PR76vl+9DpertnIXfGjZfdjC0EstNCK8We4iLmrJNvjIrXVQ2adm9QWD4WWyrIT1bLY3F/5eKaTuVe+Vh6vCzFeobrglv6xsvO9Qkq2n61s9Gy6JshYFf1zQ1U531qPXMTbezFsHJxA/wkxi/lphgT2X99JyaWhGxaB9FtAtN3oWrwq4RHwcBB+u3Zlgf1SKAKu9yD/IIpcV57PI9fpyH9fDw8FVhbYL4cu4CMh5O/XIRvLL+6zPjcjzRHWn59g17GhivlSVdebptwhHm++HpftWB/+2vnvr3jBd7mk5uXQMxa+8RKpbg/pTg+2ygXim6ULMh0eTJ9C3dHcgl5GvtGF7VHw9hbQsouXQdSijX/UWWCku93I1fwiL4EmLR6cPsCXW+7iCx0PMa2H+H79dTxbu42k5uPWxClui58o+3nawsn8foby0PI2RkQDWyIVgW+gsqkI1ZZ0jcXwFkrkK/isBxtrKbj0invZ47XhiqcdtvSOViSosbVvFMWyscv03WdCPkbqw9hCLDAPWgympnK2rXLf+g0D42Xv6Z+hekhYUDovuOce4YKCWyNXQWldCkEi4uON6zv57vt3UnKpROI5WkfeAqngsOIw33svuS8FyD0e5K/WOOXs1Qjsc9AE3OR19v3vpxFLTM7IxtkM+07f6r7/HATw7gBy0kT8+QwsU020FcFoUw2RRA5fofKprirn3FWm3SHWZUZZnx5lf+1aFtTBrwCuMiNPwoa1X5vkwB93MPDeOvzDRVqeS4KAkXtqGHxXLcKG9V+6rJIgwAjMikMU7SV15YXNhR6+EZi7mFafVSeA9sI0/2bgcQ4Huzgc6CKleak1MlyfOsem7DDeSmZehaCkr2J/7P+nkIrAM1Ki5TtJkju9BE8U8PdVxqcQQCSTo3U6wbkKAlverdPXXFeRq6ENjEbDTFZgIVyTztE1HsNWBeOtISfDShTwFJy+4aXnUpvK0jExQ3/L8vPulqLQ1xwl43URyi2frUnA0FTOtVcW3B13x0m0nwX3VcdMnZ9f/NIj8342p5FQcqn8weffX7GCmz3re2G4nOdj82iSHYeGVveAK4EiYLP7QnCXAPUq8g+iDhv9agRWl4CHQ8jXC/CN1ALuk/QKp7/+gN9ZDFwNCAEe4AMh5JEiPL5wdHsOqmXT2xnlh3dvxp8z6ByO8XPferXsW1QV3JXZ8Tfdthj01ZPU/Y4z0gpn3SsROuj6TozUWi+9H47yxp91oRgSbKck750y2PpfR2l5JjH/wS1BTzvHZnkU5BIVEKmA5XWyHT1jrcg1qxwUHOLcrYmT3Jo4ueD3lVxKEii63mKSz08AsmtdlEIaZlDBO1SiVKtiBtWy8rNzCGYLdEzMVBTcEYKTXU3cduRc2U3zHhfnWhvmSRMvhXVDkwTyRQbX1fLEx7ax66Uh4lEfdz52Gl/64kJQAN6iwcbB8bLBHSHobYmS9nnLBneY5Qc01pbdDqApnqI+ma6oihnMFthxfHVY9aFMgUiiMg7BTyoUy6Y2liUV9pIOup3JiLngJwQ5f3VW1gDCtmkbTvDIV16jJl69xfKKoYDc4kY8nrn4+tM1sKu8OuSKoAv43Tp4Ljev7y0FTmb/SBjcV5mHIAQ0qcj3BxFv5GF88eeSL1/iU994+cLrchMxc6guuCNZlx4lo3mZdofYkhpEWwURm0ogbEnNqRze8RKh3gL+oSKK4QT2ukNZak7kUBeZIfSOl1BzFvkGHTOgLlo2tTwK2Rbnxgj2F6saU6gGEsiqbrxWaUWSvVIIihUQsX7aITVBrlMn1+6i5mCe9EY33mEDvYxxzBxCuUJV5jUnO5srKstnvG7OtjdU9PBaPzSBr1BkxuenaTBJ0a1S8GqLmrd4SgYbBib44d7y45R9zVFSPjetZbeE8631FCv0MGidTBCuUP1uy5lxvvi7X6to25/BmUn/P//DD+jridLXE2WoPUJfdx2jbRGELamfSqMblTy8JLphEUoV6OiPcedzZ9h+uDKS1qpDATZd0vba4Xay5qsZ2OfQoCJ/qQbxp9MXC7W1KvJXItB4jZInIeBWL1znQT6RXTT2XPpJmKrCqbVNbD1dfhql6jMY9NXzo6Zd3D15hDPBFlrzsYrkUleKfJOLY59rITBYZNtfjBA6Xyh7MQrAEzOpPZZjao/DhK85kZtXnpdAKawxfX0AbEnjyymHoHcVYAiNL7fcxdb0ILcnyvfWl4N5BTPuP20Iniggs4JirYp7wiDXUV0rw1ss0TYZRzfMso5rAKc6G52btUzl50JwLwPFtlk3q/TWcW6GqeYAmZCbNSem8OQWap+7DJOesWlcJYNSmWA8FQkyHg2zYXCibJZ9vLul7LHOoW0qXnFw/xmqg2pLuvtidPfFsJ47QyLi49Cudv7L79+Hatn8/JdeoWWkMmExzbTwZ0tEpzJ4Cleuoy+5GBdnOffVQQBtukOaK0rkL0cgcpVIT4u9931+5NeTiHOGcx7vC8JN1QtSrQhBFfnuoCN0k5v/7Mh6XRza2kbO62K0OYKlKsTD3tUP7qZQGfdGCBo5bATD3iimUHHJqy/TmGtxkWt24Z42KdZpWMNiPnNezvrtXtaHdMdN2n8wQ2yHn9P/RyPR/RkiJy+W72xNcPaRBuIbfUSO54i+mbkqY3DgaMe/EVrLhuzKDTWuiI38E+5ydCXw9xapfV0hs8FN6HgR72B5dbo5KBKaY0lqMnmmKuiPDzdEyHpcy2rMS2Ckwn57bSpL61QC1ZaU3CobDk8QSBbx5A3URVpZCtA4k6IxnmaoTBndUhVOdDVz6+FzKGXYt8e7y4zXzULYNq1TcYKr4F3wtkBJwv8Tg5ZZidFaFerUi/+vUS72ZC+9qASL/Gx1M1HVltTGsqw7PUFdLEsy7GHt2amq5WevFBLoEzV8VdmKio0C/Lb1Gp7lDMoXg3AEX2SHBjUq7PEsXCHMPbckF5/NRelEL1U4F76g+s9YCGjSkPf4EecSEFWRv1yzfJ99zm527lgM6RyDypUfB8A9fkdx7/x8wpynaNA+Eme0uYY7X3I87F++vqeiXVapLW8TKuUZrqtHSDAUDe0aBHaA0Jk8ofMF4lt9PP/lDRe+cAEoBUlgqED7E3F6vj2NO3bRqUyYkrYn48S3+Bh4dx3Pf3UDHY/HCJ/JY/pVRu+qIbbTj3eixMYvjhO4SgI2c3DZJj5rZe8hpLwi3XVFSurjGepSK1cVfLujbSqBIiXCgMhreSKv56+II9kynaQula0ouJuqytm2enafXbqPbCmCw+vasSsYK2uZTtIUc3QPTE3hxQfWops2m94cY+2xSdRFKkzRRIa2yXjZ4A5waG0blqqgLxPcpRCc6Koscw9lC7RPJlbN3vMthwXiG7O6E5cHbIHzMA9dEvDr5v6vOOXdSxcCQcUJAlEVvKuTmQrAWzBoHkmQDDetyj4rhQTG8LPXHqGRLGdEZZyMReEXsM2D3OVxyuHzJMWBjA3nDMQPMvDjHJwzHJlZl3BkaB8MIN8XhDategKcb5Y9/60U/EIYmpYIibZ0juO8gXgsA6/kHBJgznaug24X8m4/vC8AG9yOWl41Qd6nIO8LwN/E58UGxZa0TCRpG4sjpMMJuO+5ExXtsuqe+/rMMN6RIgndz5rM+DXp0xh+ham9QQIDRUohFfeM6RDqAARYboVsq5tjn2slscnH7v8wiCd2UW3KO22y9QujuGcsxu4MMfjuWkyvirAl7rhJ4yspur81TdvTiavWbwdnFG5zdoh+bwN7UudxXSFfQeAIG1QLd8nk53/4Cp/53o+v6H2Xg6Eo5DUXPqNUsW3tNcUVxpumWIraJSxyF7yFEJxtb1w2uNtC4eiaCjrdUtIcS9IYd4JL7VSOm586T/+GKPseXE/7uRl82YVjMXWpLC3TibL2rwAnupopqSoelr6WpsP+itUQI5kcrVOJip4J48Eww5Farhvse9uKxwi4eN0sdf1MW86/pf5+DjpQr2H/RQPcuXoaFZ6CQctoklObr21wF0CbTDMjvCTwkMbNFXunhVXkr0WcsbhLg7Mt4XQJ8ZUkfDuFyF72JZjSkXQ9WoRHU8jfq4N3+KtbPAkB3S64P4B8ILB4RDQkHCsivpaE76URmUUuhtMlxOkS8rE08rO18NGgM6tf8XEA9/vhb+PzrrWiW+NsTwONkykiyRy6aRNJVdb2qrrnrkubnlnxmjcia9kTL88OXglsBfo+GOXErzQTOp9n958O4Rstopac8sgc0z25zsvh32tnak+AqT0B2p+cP7PpmzDY+oURWp+Ok+72YARUFEvimTSInMhdmHO/mtCkxTtiR3kiupvv1t/AuvwYPnN+CVNF0pOfWPY2udLM3VYEhQqJUdWiP9LAi12bedep/TRky6vs/aQgks7SOJNCse2y2bYUcKa9cdltLFXhWAU9bM2yaZu82L+ebgowsKEOX7bEu752GE9u8XlXf75I+2Qcl2FRKjNRMRkJMhYNExqaXHKbc20NFVnSAtSk87ROVzYrXdQ1vn7DzVw32AdA1uViLFRDazKOZlnkXW5i/gChQp7abIa87mI8XIMlBE2pJMHiT1jp38BxKyusblXDkze4/YWz1M5kqbnG0wIdpGiXaeK46ZYJ3MssEpeFLhy99kthSzhQQPznGLyaX3KMeQ7inAF/PI0sAe8JVJfBt2nIz0acsvjlC2Jbwst5xH+aRhwpr2onhk34T9OOqdknqzTi2ehyKhfjFz/HrNfFY/dtpz6WoWNkBk/BIJzKs7OCKZOKg7sEvthzP/YlYedcoIXr4uXdaVaCXIuLgXfVYvpVNvz9BC3PJRd9v9rDWc5/pJ50t4d84+KiI6ohiR7KEj1UWSa22igqOv+z7V4mXWFeqVlP0MwvsMz1W0W+cPrvl92PkOArVK8mZSnKVQnuNnC2rpl9XZu5+3x5hb2fJGi2pGd0GnfJLCtmYwvB2faGZRnzI/U1FUkH+/NF1oxMoUqJBGpiOXa8PIwvW6LgW/o7VIDusRihXJ5p1/KtBCkEh9a1s2GZ4H62rQGrEmU6KalLZWgso4E/h86ZGO5ZS9WU28NTW7YTLOR5dtNWHjx6iB9u2U5TKsnx5lY++/yPeK17LWmPl0NtnXzu2R8ufyiXvV7qu8j43Tx1zybWn51k6wmHoDTWGCIR9rLpzIRTEfa7EVLizzmL/6zPhVQEgYzTWvvxzWsYaa7hI48eWPDe16IioZk2u94cZPvhEfQrtEheCQSSWgqcEFFaZAZltQhLAwbiv85UFNgvHMuoCV+MI9fpsK28R8cFeBWnvH85pIQzJcQfTiJ6q7AET9qI/xFHXudxjGMqhVs40wKXBPdgtsgdL59hoL0Of66Et1DCW6GQTVWZe3d2glumL9b7v998/VW/gIu1OkZIBSkJDizeq5aA5VIo1WgohkTPvj0FNDRp846Zo8tWiCshJyq2TX2i+r65pShkvO6y41oSx8GupGoXHuyKtNEsG5dlXGDFWkJQVHUybg+Hm7so6DoZl4ek22GbqraN1yzNc0qTgKGolDQde1ZS2GWZuCxz3jE5nBVBxuXGaxqo0qag6hemBDTLwmOWUHBaAiVVx2UZFDUdicBjGijSJq+5sIXz2mVf2XWxbngCT8moSKluIhIk5fcQzi6eWR5e21aRf3soV2DdsBN0DZeKZthIISh4dV55Rw83P3XeEbJZBF1jMULZAtM15XkCB9e385FnDyz5+3Nt9RUdr25arBmZuiKi52BdFH+xyE29Z5kKhjnT2Ew4n+O6gV4GI3VMhsJk3B78xQIICBaWL0uWWlUGP19LqVkl+rUM0a9nUfML7zpfrkjjRJrM7Hx4SVcZaa2hccKxn06GvTz+4HbCiRx7X+/DUzR47KHteHMGN77RR/N4kh1Hhjm4w/EfsAVMNIQYbo2wtneKmkTuqj8fBaBZEs1a3cAucdqdSFAN26mSqo68t2pKfqCsoU7meV7pRMNmkBDXWePoKw3uEijYiG+l4YVcxYH9wt8eKSIeyyB7XOBfIbchZSP+cHK+el4FEIAcNhBfTyH/JLq4e9xiUAXs9sCTF5NP3bDYdXSINf1ThNIFdNOq2LW04uAugHsmDs27WO+dOFjpn18x3DEDV9Ii2ybof28ta741jZ6yUCyJVMD0qhTqNHo/Uk++USd8Ok/kyFuTmZeDLi3eO/V62e3KXQqqLcu6hC0GqQhSfg95t46vuPgFK4HJQJjvb7iOp9dsJ+YPYioqoUKObRODfPaVJ2jIOuM2AzX1fHXXnRxu7mQ4XIclVD773l9CtZ0bfPPkML/z4+/RHXeClI1gOFzHdzbfwHNrtpFye/EZRW4cOsuHjr7MhunReec+HgjziY/+Fr/58veJZlN8c/ttnI02Yagau0f7+L0X/oWGbIpXOzbwP6+/lwfOHOTxDdcR8wf59P5nqc2n+cfddzEajPCJwz/mk4dewGNWd6MCrBuexFMyyvexhSDv1hmJ1hDOji+6ycH17eXfUErCmTw9ow7zCkAuawAAIABJREFU+cDtnbgLJgdvbsebMxhYV8sNz/XBEpXpzvEYtaksvS3Rsn33Q2vbMRWBtggJzpoVu6mkLO8yLTYMLN9OmoOpKMT8AbJuD5PBEN5SiaKmMxqOkNNdBIp50m43LtNEtSW6aeEvFdnf2cP9x4+UHb2NfTRA+mYnYE9/MkD4uQJq78Lgp0jmuWsptk064CFWG6B9JI6lKpR0lUC2iKtkYgtB0aVRP5XBM3v/6MZF0auJxjDf+PD1rD0/ycEd7XzyG69VJRf6doEEimGN3nvrUUybzudjuFMmEzuC5CM63c/GeNA+z1ki/KJ1kDoKvCjaVx7Y5959wISvJqsL7LMQNsjvpuFDIVi7AoU7S8I3UnD4ysjPwgS5Pw/9BvRUKBWugtzgmvd+RbfG0c1tFNwaqmXPVm1L7DpaXk2wSsvX+agxrn6Pxz9WovN7MTIdbs78QiMzOwIE+gtoBRvLJShEdZIbvGQ6PAQGCqz/ygThc2+/fpzdpiJ1gdJnLnuxmDtdqIeccS27TcXa7EL7UX7eak2RNg3xdEWkqcuR8brJeD1LBve87uLRLXv51803cGfvcXriExiKyngwwoQ/jHIJWS5UzHN733F2jPXxvU3XMxkI89HDL1GfcxYe4UKW6CX999FQhC/c/BCn61u4+/wxWlMxYr4gz/ds5XR9K3/yo2+yJj5fQjivuXihewu6ZbJxaoTb+08Q9/pxmwauS7KVkVAtL3ds4J7zR9jXtZmv7LqDtTPj3NF3nKNNXTyxfhc3Dp1h28RgVZ8XOKz12lS2Ii34oq4zGq1h88DC4G4JUVG/XUinFTCX/e99upeJ9hDrjk4QTBY5sbsZ9xJZO0CgUKJ7LMab6zuWN0ESwum714Vpn1rYK48HfY4TXAXXmMsw2TBUmZGUoagMR+rYPjzIcKSOzWPDxAJB+qL1rJ8cY/3EOL5iCY9psH1kEIEk5fHSFo9xsKOLluQMnTOxJfdvzxGZhJg1narosNAsSeNkmuFWx1QnnMzTOJGiY2iGcLqABBonU7QPz1AbX/jsG20O48sW8WeLrO2drFhJbLUhgXTQzXhzmHitj7xHRyoCV9EkmC7SNJ6ibjqz7FRDvMdH4+EkllslH3VhuRRSrR6EDalWD6HhAuuIU0QlhYv1Moa6SiV58e0UInbl1VcxaCJfyTuWs1daOhk2ED/MIHIVpsmLYcyEk8XKg7sQjq1tQHGY+TjfZSrgprerntaxBLppcXJ9EzuODpXVp3jba5gKG7ofjeGdMBh6MMLMVj/Tu/1YbgW1KHElTIL9zhhc04sp6g4trdH7VsJeoyPrFfAKlCETkZbY7SoyoqCcMxE5id2pzncJKkrMh7xoT88f4xISajJ5PCWDgrs6A5mU30PS76EhkV709xmXh9P1rfTMTPCL+5+mPptCACVFJeXxUZvPXLhfork0d/UdJ+3ysL91DUVN59aBk3Qlphbs11AUXu3YwBtta/mNV37Ae0+8jsu2MIXCjvEB/ujej/OP193FHz/zT/PK+AVdpz/SwOdeepzb+k+iStvRxVY1XObFABf3Briz9xjvOfkGjekEf3T/J3j4+Kt88tA+3mzp4Y/u+zhjwcgVBXfdstk0MFGRmEvRpS3pez5VE2CiNlR2H6ot2Xlu/sr8UqnZdUcn0MookW3tHeF7t26nWCbAmJrCya7mRYP7WDRMvpLrS0qCuSKd40sH3EvhNQ2uG+y7QKYDuKH//IU2jQBq8k7wvKnvLDM+P5ai4DZNJMxb1C2Gmh/mSN7jwYwoRH6QQ5taGCgkTjA+tMOppPT0TSOkzet7OolH/Gw+OUbTRApfvsSLt6zDmzdoHk9Sk8jz41vW4i6aRKczvHpjN4MddRzf1EzbSJwjW1vJeV00TqXxVyDtu5qQQDrk4bm7N/DmdR1MNIVIBT2U3CpSCHTDwpcrURvLsvbsJHc9e5o156bQLh+pFJBq9xIaylMMaWhFlcmtIdxJA8WSzGzw458sMVIK8LLSRgDn2nyn3bvynntOwmOLP5uqgfh+Bvnx0JW5uUng9QKcWRnJWsxYyF7DIeVVUpoXgE8gm1TEOedz9BZNbnizn6zfTSSRQzctBltrK1qzVEWoK9WoFKMagb6iE0Dta0Qaydu0Pp2g4bU0pk/B1gVSEQgbhCVRizZa1kYt2NfkeK4U5s0eVAusXS70R3PYm3WkX8G8xYP7b9JIl8C624P2pNNTVKbsRdm1Aodw1TiTZqC5rqpjiIUCzIT9MLIwAAN4TIP6TJJn1m7nxa5NvOvUAXTLQrctorkrv+lyupsDLT00p+PsGT6HPtv/1qTNjYNn6I5PcKi5m9FgLe2pi0FCCoVNk8PsHOu/MGKnSol6WXndYxq0pWLotkVrKobHKNKWjOGyTBozCQxVo6BdOZlw27lhHr1j56Kyr5diLrgvxms4195AUV+EkXsZVNtm55n5wT0bcLHvXevxZ0rseb6fxpHUsuN928+PoFk25cKLoaqc6mjkvjcW+hyM1YXJV0jA3DA4jucKxjMvxVIOdTX5HO88fhhjlm9Rk1u+7eY/VGLdx6eQKmgJG2Wx0SWgfjrDh2aJcKG0c8+95/tHsIUgmC6gWjY3vtHH5pNjhJPO73cfGmRN7xTBdAF3yeSGN/rZfnSEYLqAp2jw/u8doqhr+PKla5pkSKC/u44v/Nbd9PbUk/ctXpKOAUPtEY5vbeHlW9fwoX86wP1PnMBlzF8ACUtiawLTq6KYEtOjEOkrIRVBLupCChgXfjbLaTql06bTViNzP1SAiVXgD+zPO9lvzRUoeaYtOFhAJFZ4PgYwZEJWQrDCyORRHC2Ecxefb8FsgT2HBji2sRUp4F1PHVnd4F5s0Dj5b5rQ0xb1L2VwzZjUHLt2MpNCgpqyyWZceKWB5xqJ56wmtBcLqG8UsW52I8MCqQnsllmxCwHKWRN5ubfxEt9iOJOnY2Km6uA+HfYzFQ4sSaoLFvO89+Qb9Eca+Ktb3s2Xd9/Fe06+zp29x2mfDZZXsoAqqTojoVqa0gn8xvw+liZtuuKTDIWjjITnB3eA+myKQGn5a81lGejWxQWD2zJx2bNaB1IihcAWV14m3XluGEVSVn/LUFUmakMYmrrA9vR0eyOlCjwBArkimwbnl7gbRtM88M1j9G+I8s3P3sAv/dk+/OmlM4v1gxP480XHGniZxYSpqpzuaMIWTg/6UozWhcm7KwvuO88OXbWFtSJlVaNvwgLX2PLflABchkVdfP5C4fJyuzdv4M0bF/7GXTRpmEpfeB1OFwinLx7b3CJg7vfXApYiOL61hT//d/cy1uJUjdwFE1+uhKvk8BaQEltRMHWFvNdFzqsz1hzmv//m3WT8bj7wzwcvBngJdWeznPhQM4ph07lvhu5npjnySCtawabn6Wm0ok1EFHhW6WLa9qFic6scQlupFPmL+fI3WSUoSHizAHdfgabApIU4VFwVY1AxbiJTliNiVAncAiLznxGKhNaxBC3jCWwhiNVWpjtRcXAv1OtEjuYpRDWMGhVX4uoGVwkUhUYRDQ0LnzTIC50nfZvZYEywqziMDWSFQ5zxSydHSSpeFCQBu4iCJCtcFIVO0C6gY1EQGoXZ1yu+EKvF7Gw+gL1Rx96gox4oYV0vkDrYrRqyRkF2azBmIRsVZETBXquh9JqISz7yUC5fkb3o5ch63Qw3RBYNPuA8kLZNDPJnT/1vnlmzjR93beLbW2/hu5tu5ENHX+bDx17GbVavRS2FI+CiSntRb3LNshyG7iIBWLdNNHv5VfRclnSpgNhqZk49o9ME8gWSAd/yGwpBIuD0qhvjFysdEod5bmjlb7md54YWWKYmoj6O7G3FcKnOnPsiAjaXwm1abO0d5dnrNiy7nVQEE7Uh4kEfdamLgU0CE7Whikcnd5xbHXe3txsuv87LXfdvReVwtCXMV3/+RsZawmimzZZjo+w+MMjm42O0jCTw50oIW1Lw6EzXBzi3roEDezo4sqONRI2Xrz9yI60jCW7bd+7CJEztmSy3/pmjYTJ3H+39qz6kAMUhz1MrC9xkj8y751YEWyKOFC4Iic3dvkvt10AhLVy4pcV5pZawLFyoIiCBY8Xqg7uUMGkizq2S7sn0bOZeKdxiXrWh4NI4tqn1wjPT0FVOrG/m05e4xC2FioN74HyRzBq30/C3wTt25WYDlcBA5U13O0WhkxEu7s6fwS9LdJozTiYGTCsBDrvbEMCW0igRO8cJVzMpxcPm0hhNVpqnfZtoNRN0GjH8ssSPvWvw2yVCdp6tpTGUaxTglXMGUheIpI1yqOSUfDQBpkR9oTArpSfQv59DegRCBekR6E/kF7Ue9OcdUxPNtDC1KkpPQnC8u4W0z0NdaunyZkM2xUePvMQ7zxziUHMX39myl/91/T1siI1yw9DZqs9fsy0i+Qwpt5eSuvCymw6E0W2LSOFtOulgmmwcmOC1Ld1lt00GvAuCe9rnZjRag1VB723PyYEFP1Msm22vj1I7kVmQYS+F604PlA3u4PAwhusj84J73q0zXROoiBTmLRqsX2RWXiqQucFN8l4vUoA+ZdH4d+mKVSBTt3lI3e5GqgJPr0Hk+3m0+Pw/NsMKE78axHYt/rl6zhtEvpdDS1d/n1tBQW6Li2KXhtGgYgUEQoKSk+iTFp5zBr4jJUTxrQnq4Izvvb63m5ObHf3/e588wYe/eYDWkYVKgb68QW08x7ozk9y67xzP3LuRb3z8euK1Ph5773a2Hh29ULmYp843CyHnL5hTwk2DzNLMKt2zcRtGTQq6xnN7NnC8p5l3vnKCjX1jnO5qoi6RoWnm4j2VwcV5pZY8GjHhw0KhzUw5jpsSxJlS9U93C0ffPb1K8/oJ25HKrRS6gNDFe67g0Tm6uZXuAWdyxtBUjAodQSsO7mrBpv7FDL7hEmrORstcXYlRSyikFC9tZpykXs+oFmatMb9PfMDTwTFXCzYKOibXFwZIKR5OupqJWhmazSR9eh0hO49XGpxyNXLA3UHEyhOQBXqMaQLy6ivTASgjFzMx9YyTgivPFubdQOoJA/XExYxMPW2inl68QqJKScfEDJF0riLd80txrKeFlH/x4H65CEekkOWOvuOYisqJhjaONnYsGtw128ZQtSVnon1GiU1Tw3x7283E/EFa0o6GsgSmfUHO1jVTm8/QEV+cC/BWQ0iHpFZJcE/5PSQuy/DHa8OVMc+lZM/phaS/ybYQ7pxBdPEJu0Vx3SL7WQxpn4eR+hp2nL9oaJT0eytmyq8ZncK/mFmODVZYYeY9Psx6BX3CIvRcAd+p8uNhUoXpj/lJPOBoJtR9M0vtvyxkqFsBwdQjAWz/Jcd5yTEH9+UJP1Nw+qgVQAKlDpWZ9/lJ3e7BaFSxQgLbqyB1ZwNhgpK3UZMS71mDpi8k8R25usnOUkgH3Rzc1U7Ro9PVN83PfflVotPLB1sBhNIF3vmDY/R31fHkA5sZbotwZkMje1/tW/ZvL0UWnYRw0ySzF/a7IowYkJc8fts2Hrt9Oxmfm/WDE6wfmODQhnY0y+IjT13UZNCxGBdBziq1fMI4wj6tc945yv4KxlcvhyERJ1ePCCnSNrJYxRJDA+kTFz7LYKbABx57k1DGaf2YimDNQGWt2IqbkPkWnRO/30zs+gATd4QoXCO/WwWJkBIFiYXAQMUQGhYCtzS5vXCOz6ReZG+hnwPuDtzSZGtxBE3aqEgeSb1GUegccztM5z3FQT6WeYOHM4fwXaPAviRWWDRYPzRJy3Syare34foIJzqbsRe55hMeP0+t3cFATT0pl5eMy82ML8iphlbyuov25ELXKZdl0pqaYSgc5WRDGym3l4zLQ053Yc3eWG7T4ObB09TmM/zNDe9kJFRHxuVhMhDm7268n4THx0eOvITfuLYM40qhSOmUniv4rJN+L4nAfNvI0WiYZKC8lWTjTJr2RTzkVcNmoi1Ezq+T9+kVXTprh6cIZ8vzYlJ+D0MNkXn7TAa8JIJlWhCz2HZ+ZFEynAC8x0t4zzjB3AwrpN7hqejYc5t08hucUSZtxib4ehFlkbEkNSupeSJP8KUi3pMG+oSNkrWv2AHRDggmPxVk/NdDZPe4KLWp2D4FYUmUnETJS7AlVlCh1KGRvNvDwF/WkdnrvtZNPgByfje9a+sBuOO5M0RmKh9P9hZM7nj+DO6iSTroob+rOv6OjeBRZSOfV/fyF+qNFFmhDfWYiSxIvnfHdn7jm89x68G5toCkJp2jr2X+8fkxuMvs5RdLB4jKHHeY/fMrsTHLIbVVAxM4u4pxIWc77dhKoQnwXQzLqi0vBHZw+u/BdGX8k4oitOlTSK/xEH0pTf3LzqiZVuFKeCXIChf73Z24pEmLmWREq2FAr3UIWEaMXcVhnvRtYkiNcGvhHM1mipe8PejSosucIaV42Oddi4okamXoNmJ8x7+TZ70bWGNMsa20cuvVtxKdEzN0j01zdE1LRSpic5CK4Lu37eCe/SdxX9bbzbncPLr1Js5EW/CXCgRLBdJuL4ai8o7zR7i1/9SC/bksk5sHT/NS50b+8x3v5x9334VuW6ydHuOXX3+K1tksfev4IJ9+4xm+uvsOfuGDv040lyI9q2b3sSM/5l2n9r+tpx3WDU1WNH6Y8blJBH0XSIsSh3me8JcPllv7RnAbC0mLmmFzdG8b57Y1oFiS9//DQTz55XkvLtNka98oL21bu+x2RZczm19waXhnGe8XMvcKsO38CMoSM9OuEQvfoSKZ691IjyC9103069kF5fVLIQVkbnRjNDnBwjVsEnhtcTERLWHT9W/nL4ZiD/sY+IvaK0ol1Ywk8HqR5N1ehCVx95v4DxbxnjbQp2ykgGK3RvIdXlK3ebCDgsJajelP+PGcNtCXOa+rAUNXSIUcqdWu/plF+SzLoW04jmrZFDw6mVDlUqkC2C4n2WhNU0LFh7nyFueECUWbZMBLw0x6Hs8m73YtmMuXQEa4OaVG2WsOY16+uDAkJCynlVwpLFmd1Gw55OUs16rCCoKK04adhakqxCJ+TE0h53VTcqm8ub2DR779WtldVXTW6R43hQaNzBoPUhMIS1L3ehbf2NVUX5LU2VluKA7QZjrBodOM80h6vsLbhzMHkAgUJI2k6U5PO9k+zpf/3uwRwDFkAfhg9k3s2e3fzoGkEmiWzZ5TA/zo+k1kvVVoGAOvbummvzm6QHgkmk3xK689yan6VuK+AKai4isVWDMzwZ6R8/iMhQ9ZAWycGub3X/gO+9vWMOMN4LJM1k2PEi5ezCQ0afPAmYN0xyd4s3UNca+fQLHA5qlhdo30Lsj+AqUiv3DgWXaP9C57Lm3JGB898mOaMokL5/DRwz++UGUIF3J8/NA+1sXGqvqMLj/HQL5I5/gMpzuXd+AyNI2ZkB9DVXFZFqaqMF4XJleBfO3W3lFci2iEt/bH+cjfvHHhtWsZEZs5aJbN1t7ywR0cE5mU34u35PQ0ndZC+UqDq2SwbnhyyTE2ISH0QoGZD/gxmjWKXTq57S5CLyydfZgRhdxON7ZfAVMSOFBCH608mRAr5PoGXivS8udJ3IMm3hMllMsec4E3S4SfyTP2uTDTnwwgXYLcNhfFNRr6/mtdDXQkYcFR2Kv6ry/52qpKEIAkbt4UjUgEPgyul2MrIimLmJPl7j41xHfu2sl4NIxmWjy/Zz0vXLeOd70437cih4vjaj371Va2W+O8qHXyQeP4hWc9Nk7vvKHSk5IOy35qFcniEsd7vlII4fTdFcAGQ1Po7YoyWRfE0DWElCRC5e9LqDC4+4dKuOMmtQdzFJp0XHETV/zqsuVd0mJrcRS/LC0bhJ0CxsUPT73k/+Ky10v97CcZtxw5RzibLzvydDlyHhdffnAv//Hvvjvv83VbJrvH+tg9VnnvDRwOwMbpETZOL18NUZBsnhph89Ty2wkgVMrzWy89Xva9u+OT/MrrT1143ZRJzntdW8jya689WXY/5eApGawdmSob3AEmawIUXRquvEXG62asLoQsQ6bTTIuNA+OL6rNPNQc5fn0Lhq6SqvXyvi+Vz9zngnslWcNEbYhEwEtjPI0tIBH0kfaWN99onU4QSS+voe4/YuA5b2I0qRhNKtmdLgIvFxYEzTkUuzVy25ySvFKU1Dx+9TXaL4U+Y1P72PLlbS0hafiHNPH3+TAjKqVGFSO6wrL0FUAzLYLpIjG3znBbxNFhqCJ7H20OY6kKmmERyFTXEusVNbix6ZBJ9ikd7JCTaFfqDAdOll2S/Nzjr/JP9+3h2NoWDmzuoDGW4p7XTnHzofPzNhezxDlFSk4rUQyU+dLEtnQy52owYVZfyi+HgnQWGpVeHgoXgru3aLL78CB5r4tA1vl+ejujFe2mouDuSlog4OwvN+AbKWEEVVofSxA6XbhqN52KJGq/PZnT1WIqGMBXKtEfjdIzOUXa60E3LU61NLNufJxAsUjc52O4rpZoKk37TJxDnR0c6Whn7cQkt54+s2Rm1Dqd4I6DZ/nGPXuqPq7ndm3g6JpWtp0f+YmvYlwLeIsGa0amKgqWU5EgBZdOMF8k5fcyGq0pu/+mWJLGmfSi30X9WIYbn3FGkX78wDpMTSlrACSkpGUqQV0qSyy8/GzsZCRIMuB1jH00jcmaIHYFTPme0Wn8heWzVaUgqflBjvReN9IlyF7nwmhQcY8sXMTYGg5DvcN5NHlPG/iPvMXcmCXgGrbQRyzMiIL0CmyfKPudrDa8OYP2wRli0QD77ljHe757GLWCqg44mfrLt66h6NIIJ/O0D1Y3WuvBpFdEqJEFsugrK8tb0jFqsaBrNMZnv/U8P/f4q5iqgsu0qEnl8FxW0fJhsN2ewBAqOVw8aJyZTyKzqS5rBhhd+NnN7cGpbIh5EtxzWPY7N6o8BkXMKto5f2dqKm/s7CId9LDu/ATdg5UpQVZMqMt0uml4MU3nN2aIHM5hBhUs/7UaJPvJxoHuLl5du4a/euh+BqJ1PLNlM//71psYra3hv7znIfoa6vnq7bcwFQry3995Lymvl5TPS96t4zHKPDglfPxHr1fv7y4EsZCfv3/oZtK+6kr6PwkoqSppj/sCoW8x2DgmJpVew27DpGd0uiLns8lIkMKsn3rS71lSkvZSdI3HqE1nF31QxKNeju9p4dieVqabAhWNwwkcmeLu0fIPg6maADNBZya4qGuMV6CjD9AzMo1vMab8ZYg8nkdLOQed2eWm2Ll4AdcKKaTu9jgOWVJS960sK0kGryaEBDU5+6AXjlbFtV4lB9MFth8eQTVtTm9s5GuP3Egq5FmULDsHCeQ9Gs/dvZ59d6zD0hSaxlJsPj5a8ftaCNbKOOvkDIMixLvts7hXoj6Tl5C3Ha2LWWtqWxEIJIamMFUbIB6cX44WQFTmuN88x51WHyVxWWosAbO6CCWm55+DxPFCONzQzf+76yH+6PZPcrTeYeVP+sIMhipwTbRkdeTpucx9Fp6iwc5jQ9RPp/n+vdv44iO3VrSbipkG/uESE+8IYQRU8s06esqiVKPR9HT17mQ/bfAYBsc62rj7+Ele3rAWf7FEzu0i73KxdmKCjNvNmokpbjjfR3+0nqzbTXM8Qd7lYuvQ4kzkS7F2eJJbjpzn2T0bqzouW1V4dWsPP7xxC+/fdwjdWhkZKBHwkvO6aJpOLXvMEjjXXk/LVLJs1lcpDFXBUFXSXi812SzDdbUc6upgZ/8gTYkk/mKJkqoS9/sI5/K4TJPxmjCDDXW0xOK0x+JYikLC58VrGPiKC9tBAogmMtQnMozWL5+JT9UEKbocVnsi6GMyUkZTXko6x2eoSS9eDnYXLSLTOYQt2XBkAk92+XbVHELZPF3j0+zf1LnsdpaqMtToiBsVdZXx2vLjlZpp0Tkxc4GEtxzUlE34yRyxjwWwQwrpWz0E9hcRl3z9Eii1qWSudxab2pRN6Llrp4J5OSTOrLtZr2LWKNg+ge0WoAmk6ozrmW9BKf5SeIomN7zWz8u39HBuXQOPfmg3E40hbn3xHM2jScKpPHrJQkiJqatk/S6m6oMc3tnGM/duZLo+iD9T5N3fO0JdFUz7kyJKg8zSI+PsYKHGQdUo2Rey7Bd3r+XRd+zCVNV5joR3HDjDJ55weCcGCmMiiCkUkrixhMJJpZ5HjMPz97uMOc6iiC0M7i+2beEftt9LuJjlVG0b9/Q77qgHmtbwQvs2/vCVbxMuLfPZVftYvezGLrh1XrpxDVN1QW58s4+20coqLJXPuWctmn+YJNemE32thJ40UapcFf20oimRZN/mjbzz0FH++oF7+dwPnsJQVfyFIk2JJA3JNMO1tfOYroFCgeHaCM9v3sg7Dx1ZtsSiWjYf/9Hr7N/YSaoCEtSliIX8fP2+G1g/NMn288MVC6QshqGmCH2tUdYPTNAxNoO3aDAT9jFdE6B1MkEgXyLjdTEVCfL03k089OJR/OOrE9zHIjW8tq6HcC5P2uth28AwA/V1uEyTFzet5wOv7udgdycZj5ukz8t73zjEmdYmDnZ3sG1gmKZEihPtLfTXR8m7dN61/zCB4kLyYF0qQ9NMsmxwj4X8FHUNS1EYqa+hUEbG1W2YdE7M4FtisVNyKdRM54g1+jlxXTM3/eg8ShnzGIBgvkjX+AyqZZcVpOlvjlLSVEq6Vn4xAtSmszTOLL+QuwAJtf+aY+ZhP9ItSNzvpfGLaUTpEj8IAcn7fNgB5zhrns6jJd4av4hil0byLg+57S5KLSpmnSNiY3sFUhdIDaQ2S3y6UlvRVUJ37zQPP3qQ//WZ25ip9fH83evZf0MnTeMpIjM53EUDIR3Bm3TIw0RTiFidH4RAL5m8/9GD3LbvXFXv+ZLShkta1MoCUZzAtleOXDmhrsiFkbF/fPdN3HbwHLtPDc6z5K1LXmzTSiAhPEwJHyrOqLS52FOy2sNJzA/ulqLyLxtu4paREzx85hV+985PXfhdWzrGjCewqCjXakJ7LyvyAAAgAElEQVS1bJomUiTCPs6saUQC6/rK64FUflRCMHZfGM9kiWyHm6ZnUvj7K8sergTLfSeizDZimd9Vs91qnVv31BSffm4fDckUv/nEj+iamiZYKJDyenAbJvWpNHecPE0wn+fhN/YTyeZQbJt3HjqCFKLscQhg59lh7n3jBI/eubu6h40QnOhq5i8/eg9/+d/+mbpkZkXnfXxNM+6SwUs71/DI46/R21pPIujl8du38avf2seP9m5GN01O9DTz0GXs15Ug7fFQ1DW2DI3wnRt20xxPUpvJsGNgiG/ftIdTrc1MhoPceLaPJ3duob8hSm06S/v0DNsGhym4dE61NrOjf5DDnR2camtmz/n+Be9Tl8zSOFPebjfv1kkEvBia6viql0EknaNjYmbJRdzg2josTVD06cSaAhR8OlqyvNe0atm0TsYJZfPEQ8tLcQ421lLSNUq6VpF7XX0841gPVwjPWUfRLXu9m2K3Rma3yxGYmYXUYeY9zvidyEvCT+XnZfZXGxJAg8R9XiZ/KUh+nY4duOQTtkFNSbSYjZK3UfKSwlod289bGuBdhsXtz5/FlzX4m1+/nYnGEJmAm3PrlqCJzy7GIrEsH/v6G9z75AlcVZr+3G/18qbShB+D0Kwr3Io+gaKEoqM8Gg/5uP+VE7RPLJ2hurDZbE9SQsM927fpshc6G1aN7PwFswTG/RE+dfRpmrIJ1Ev67W7TwFC1smZSK4VuWqztm6R9NI67ZODPVkZ8rDi4Jzd5iBzJEX01w/SNASz/1fMqlkDa6+Ffr9/FY9ftYiIcxF8scfvJ03z2yWcJ5QvYwIm2Fv72vrs40dqCt2Tw0MHD/PwLL+EvlhiN1PCZz/w8n33yWf7H/Xdz66mz7BgY4m/vu4uO6Ri/990f0B5zLp6+xih/e89dvNndiWbb3H30BJ955gXCufyqBHi3adE247zX2gmnhNWQSlOfumhA4Zm1L21KXmxzdMQWipksBgGEs3ne/dJRXt3Sw0h9TVUPG1tReGVLN7/9Gx/kL//bt6lLLt73rQQb+ie44Vg/Y/U1TNQGMTWFkYYazrc3MFEXREibvUf7ON9e6XzK8pA4s6CW6gw3qrPjQEJKanJ5wtk83pKBpSiOm5xtO9tZFpploVq2M6MtHTc21ZYOS3sph7JMnpbpBIoty3qlj9WFKemj9LbUlz2PumR2WcvUmliOpx/ezN6neyl4dZTLbTqXOgygbSpBNJmpILg7Zfmsx81MqMyMu5Q0xlMVB3cBqGmbmh/lyV7nAlUw80H/vOCevsVDqdUpc/uPFPGcN1bFvKNiCOcYhv+vGmfGXoI+YhF5IkfouQLe0wbKHPMZQMLZr9eT21Wd7fLVgLtkcdPLvaw7M8Fj793BvjvXkajxOvfF7LNASIliS7x5gxtf7eU9/3qErr4YqlV9daSLJGG7iIZNgNKFceQrhinBdMaTd50e4rWtXURSWRT74siyatnz/DAUJDoWYnasuUFmVk5ovEwoSQChYo4pb3gef8cSgqFQFI9ZQrOvruaLakui8erJ5ZVry/cV6f9YACmg0KjjHyyfNVwpLEXhCw/cyzPbNvP+1w+wbXCYhN9HyuvFX3BWLf0N9fzGpz/Jzv5BfvexHzLj9/HVO25mMhzi9/71B0ghSPh9PLd1E+/df5BH9+7heHsrn3ruRb5yxy08vW0zn3r+JaZCQf7tJz9C+/QMv/P4D8m63XzpztuIBQP8+3/+Hr7S1UsdKsnIq9nXzrND3PvGSb5+7w2UXNWViqSi8MqWHn7n1z/I7379KdYPTuKyqr9oUwEPo/U1lDSV6UiAU91N7Dk+wOnuRjwlE1NTGa2vIbNCEp8lBDNhP/1NdZzoasadtxiujfD47h3Up9J0Tk1jC4FmOwur7slpirrGE7u3odsW68cmSHs9vLG2myd2b+PjL77GpuExnt+ywfEqGFqcXKTMyv76C0XS/uVbIM7noJXP3KWkPpGmdXLxzEMAPaem+dTnX+L/I++9w+S4zjPfX1V1V+cw3ZNzQs45ECTAKIoSRSUqW5Zsy7LC9dq+z11715bD7sp3dy2vg7yWbWkVbFmyKJGSmKMYABA5DcIkTI49oXPuqjr7Rw0GGMwAM4MZkLT9Pg8e9HR3pe7q853zfe/3vta8TqrVhmOBNXeAqvEIJdEkndWlN530jQa8JJx2Bkv8GPLNJ++yEJRPxgkkFj7oyHlwncqhDuvkqy0k9tkoVCioIzpCgsj7nGaaWxO4j+awht7alLzulhj7ZTeFCvO34zyfo/43w9h7brKqvX1rnEVDFoLS8SS/+q3DfPyfj9PbEGSouoiEx4YhSbjSeUpHE6zsCOFZZNvb9ZCAAFcnZv14qCJx623GupgmTvoTaf720f28snM1peHENDt9+6V+3nPoAjBlbCS5uSwHaTLCWNF5zVLPBwutS2t1Ts9cucvC4O7+8zy5YicAKaudPm8pzzds5anmnWwLmdof70QsOAKoUZ2aJyKk6m2UHErgGrx9Qa+/OMiT2zfz+Zdf47OvHkKZYxX1g327cObz/P4TT1OcTAJQkkjw1Q8+zHtOn6MiEkMWgvtaLrJmaJiDa1Zyz/lW7m+5xImmBiY9pu3pM1s3Enc4+N2fP0tlNIbA7A/+mwfvpb2ynC29C9PnfifAnc3zsVdO0tJczZmVNYtOFwlZ4si6Rr7yuUf49HNHuP9EK67swoNISSRJzWiEkRIfGy4P0TwwTtjnYizg4Y4z3ZREkjQOThAKelnfOYQ3tXAbT5hi+dqstNeWcaGhkpNr6jm6rgG1oPHRF06xqXeAAxfbceTNc66MmA5R7ztxFoCyWJx9rZ3IwpzpezNZPn7o2JTTlWD9wBBrBoenX78R6kfDeNILCO5BU3J2JHjzFLdV12kaGseZv3GDrSRAzZuTLXd8cYNJIJ6iYtL8PdyM2asrCr0VQQbKiubdpyNXoG50cpZq2Hyw9Wm4zubJV1kw7DLRdzko/W5yuv8dCawhHdeZPPJi25iWAAEYTnmazAdQ/vUEtpsEdsMKWuDtJdTdCM5MgbWXRll7aRFmBItAqxTELfJ0SUVICE7L5XxBP41yq4x5DTPAY44jn37m6Ky3BK6puRtIhCUH3bKfJFZUdHwiN7PP/VZwXduaLATvu3yMvGLhpyv3kFWsPN20A28+zc6RTt7XeQybfjvF3G4dCw7uiSYbBa9CydHk7TwfAC5WVyILwa7O7jkDO8C5uhqaR8emAzvAtu4+chYL3aUlVEwN7MWJJKqmoxY0ShLmLNCmaRRkc353qrGeca+HP3n0kekafNTlJGm3Me5dnCHLOwFNQ+N8+fFX+c3f+ihJ5/wiJNfDUGTON1byZ594gEObVvDLzx1hY9fCZHrLJ+OUT8ZneIPffaJ9hh3r1rYBjKnHC5005C0KXVXFnFhTz5F1jfRWBBkq8ZO2mcI9FRNRqicjbOvuw1a4uYHH9feTLGa2qdzofrsWtaOTuDPzT0xGin1cripBm8fmVc1rrOtZeBvSYmExBM0DY9jyGpl5VPI6q0vnnYwAOLN5Godn+wzMey6TBq5TOWIH7BhOifgBO8U/SJLcYUMrkgEJe5eG8/xb39tuOCWMK77bApwXbn4O6Y0qmt88539vcIs8o5IbGUGADC6xROOcqZW7BHzqWVOFNKuahFTnHAsMGUGVEUcg4RE5bGj4RG7piZTrOKoS4Mul+eTF17inr4WozYUhSXjzaSqTYVxzKHbeLpjtjebRFkJiXXBwl3VBstFGUUvadEYyltcv+1pkVSuyEDhuspLJWa2o2sxZtWWqdlRQrs6mp2uw1z6+5rwzqkpxIsG+tplOZ85cntVDty5X+nZBAvae7+ZzTx7iLz9yD2Ke9OrcO5EYC3h5eu8Gjq2t566znXz85ROs6g+Zn/E8K9trGfdzse/nOiNx5Z8kISSJvFXhYn0Fb25o5vDGJgZKi0jZVTJ2dc6MhCeTpSz21rRlVkzGzVXEPKS64WIfFxsqb74zIbAVNNb13N57bXV/CEcuP29w76gpI+qZp+NCCFzZHA0jiw/ukgGukznUUZ1sk4VcnYXMapXkThu6R0bKGrhP5LBMvrUa7WCq4aEL07wD0ILKnLK3AjC8EuOfmXKjextju6bIRP1ObDmTaHWr3S4CM3N3NYCY9XlJzH151SQoFWkUDCwIikUG61L63A3zJASmeuZ3H97Di3vWkLWplE3G+PUnDrH3XNf09UmAjxweY3y61p9AZXG9QnPgug6wK3/Z9QKNsdCcr91u4aKcauHpBzZwaFczvliGilCMX/+ng/Mec+HBPSsY3+chvNWFrAnqfziJr3VxadWFom58Ek2W6S4roSk0NudFNIbG6A8GyFmUafOT4SI/AhY1yDeFxhgIBnjfybN4M7evp1ZgKhzNl/K99taS5nh+ITeRYhh85tkjjAa9PH5gK3nrrbVq6IrMaMDLY/ds48l9G6cY+a3sutSDL5nBkStgzxdQNX1h7VBT16ErstlyZbGYfdWqhbRNpaeymNa6cs6uqKGluYqkw2YG+ysbv80tR1egGAar+0OcWFuPrtw4LdtbUcy337t33v0F4ynqRxamOnWrWN03ijObJ+y9+YTk+V1rF1TOKUqkqb4BR2A+OC8WsLcVyDZY0IIKkfc4yK6wgGL2tvteztzywkHAzB/JVIpovgFYAuSMwNFaILPBLA+MfsFD9R9HsMQFUkEgZDAcEoUKC2OfcRO9z4GUFwg7b9u92dsQ5It//3EOvNrBF7/+Ov7Y4scwXZborQ/yxoEVXFxfSdqpEgin2HRmgJ3Heqkais0i3UkwQ7RmyelwwTSL/3sP7+bsqmq+8OM3CMTTtNeV8reP7seQJO6acorTkUiimtLtkgUNmaOWaj5UaF3aeVw3pxRAW7CGslSEYHZm1jrk9BO1uWiOjmCZQ7VuuRDz2AmGk+w83cu+Y5c5vLNpQROKBY/6zpEC23/rrak/b+gfpHl0jB/esQtfOk1ZLE7OYiFts7F2cBibpvGhY6f4g49+kMd3bWfn5R4yqpVv3rufNUMjbOwboDBPKvQK3nvqHK+uW813D9zBA+cuYC9oJBxmW9Wm3oFbIpVdDwM4X16HLsusHh/COY/q3LjLS0FRqIpfbQWJ2Z3oskwwPX9ZRMLMPPzGzw4Sczl4eceaWw7wVwatrE3l6PpGjq5rwJXN0Tg0QdPQOHWjYUqjCbypLPZ8AYuuY9ENZCHQJQlDljFkibzFQk61kFWtJJx2xv1uxvwexovcjAT9DJb6yVkt75gAfjNIwNqekane8RsHdyFJ5NSb97cDbG0fWFA5YCkojiWpHo8yWHrzenphAfeJLASre0dnuHYtBpIO/uczxO+2o7sl4vc50Kf82B0XCzgW4PcOkC9TyNVZzN5zGwirKTKT3GmbHvkK5RYi73FiHdeR8gIpL5DzAikH7tO5GSYzcsog8LMUI40WDJdM9EEHWkDGfSyHJWpg2CSzhW+nqbDnPJfH3lkg/JGbdyHcbghZZqLYTV5VMCRIeOyknao5DqTzeOI3lwk/va2Wv/nNAwxXz7w3ju1u4JUH1vBL3zvK7jd7ZvArzkvFFK4RSz8mVfJZowX7ra7eBWCY/72yczV/8ndPsb7LzGZta+1DMQQn1tVPB/c8CueVMiKSnaRkQ0IwIS3MxXAx0GWFv972MPf3nuFDHUdmvNYeqOKJVXv5f0/+FF9uJrHUm12eTisAdzpH6UQCbyLLiS31Zu//AjwEFjzip2rM2ayrP0+yTkXJGthDs60plwP2QoHf/fmzfPfAHXzt4Qex5wvIQtAUGqdpdAybprGno4tf+8UbPLtlI89u2Yguy/hTaf6f516mLJ5gMDA/KQhg3cAQX37+FZ7atpkTTQ1T7VAGG/oHWd8/yFIyTWAy/wf8xfxs/Q7WhQZpDIfQZBlHIU9KteMs5FAMgzG3j4xFpTo+SW9RCWnVhieXRRYGqqbRHSzDn0kTSCdJ2uxkLSoZi5XyZAyrrjHq8ROzO/FnUhSnEliEQeVElC898Toph41DG5vnFTFZECSJlMPO+eZqzjdXm88JgT1fwJktYJ1qMZMNgS6bwV2XZfJWhYxqpWC9/X2hC4Hukyg0WVAvaWZ70yKxrnd40WSyG2F7W++y7OdmkIAtnQMcW9sw7SJ2q5ANwYbupXEEvAezWEM6uQYL2YapYUhA8PEU0gLnDIl9NkKf95ra7jYzsAuVKV1uE9kVVob+wAe6mXaX8gI5J1ASgpUfGsMSu3owKQ9FT6XJ1VoIf8iF4ZZJ7rGT3DOTuyLlBN7XspR/PQ4yb3twB5gsdnNhQxUDdUX01wZIumxIgDuZo7kzxL43uqgais4aryeCLr79uTsYrvKDAHcyiyeeJeZzkHbb6Gou5Zufv5PykTiNXRPT21+QSqkTsekVe+F66dfF4ppAJSRploWwLMSM+9aBxg59iKSk4hU5ZAQd8uL86Bd0WsCEw0NtfLZoTFk6SkJ18GrTeixCI+Ywx9+43cmvHnsF6y1Ofq+HI1NgdecoCCgOJ7EWFpYpXVBwL7hlJne60O0y3rYMkc0uis6msYeWV/T5SqpBAjb2D/L7TzzNuNdD3mLBYhgUJVO4ciZTWDEMPnLkOHe0d5Kw21GEoDieoCwWRwAl8Th/981/pH5sAquu859++gzBZBJ7QeNXXj04XTeWhOCh0y1s7e4j6nJiyBK2gkZxIrl4vfY5oMsyEy4PIFGeiKJLMmcqG1gbGuBI3Sq2DXUTdro5W1FPaSo2bVs66fTwcvMG9vdcwmIYDPiCZCwqDeEQx2pW0F5SiS+bZv3oAJXxCD/euIeyZIxgKsGBrosw9TmuGAzxle88w3/97Ht4ffOK27MyliSyNnVen/O3AgLQqhUMt4TarZlmHg4JrVRB7daQEgKtSiG/1oIRlLF2aXAL1aWGkUk8mey8Nez5IAnBtvbFZ8SuL98spGyzuWMACbFks2NZCNZ1XyVZLqZkdAWWqIH/xQyhz3tABoGEOqLhPTjzy7hRmQpMHfp8pTItNCMwtcjnXNEoJmEOp4QOaH5TPvZaSIA1ZFDxF3E8R3NEHnaS2qyiBRWQQAkbOC/l8b+UwftaFuuoju6RUaIGuv/t7YkbK/Xwzd/YR8znoHBtG6wQHNvdwOltdfzKtw6zqi0043M8vquegZoiJAHbTvbx0R+ewB9Jk/TYePY9G3j13lUMVRfx2Ee38Xt/etVZ8SHj8rR4DUCFkcK6aJ3VazBFupGAvee6+cajd/Hhl8/gT6S5XFvCs3es57NPvjljExs6NpGZ3rzauH2cm4JsmZEKF4AmK2iyzJahbvqCJdTEJmmcDPHkuh1osrIswX36tzX1oGwszps7m6kYm/9aFxTcJc1kOuSCFtI1NjyXs3g6l6/erikyT23bTMPYOGuGRlAMg4zVymN7dlA7McnDp1umDQWu1K3BDPA1k+FpIp0mS/SWBDlXX8uBi22sGRxGwqx1N4yNTxO5rgjKXIHVMKgJR6gJL84VaSFQNY26yDjV0Qk2DfeSVywkbXY0WSHqcJK02Qm5fWwa6WNdqB9ZCLJWldcb1/KJMwcpSqeQEJQlYqRUs00nZ7GyYbQfVddI2OykVZWCYsGTzeAo5GeoKMkC6kJh/tfXf8wffO59vLx9DQWL8q8i/X0rKDQqJD7lRBnV0aoVkCFztw31XIFCk4L9dIHYZ13YLhQo1N+6bKRV01nbO7ogmdaboX50clEqb2D2+GdUlR/v2sFDZ1sIJpP8y55dfPzNoyhCmIY4imJmoa4JdFs6B7BoOnl1aYHIlcmxcshcyWiyzGO7d/Lo0eNYDANDktBlGVmYgkBXfq+GJCEJgZAk8zFQ8bUYZV9PoMkyL25cT3kkipQ0TX8kIab3N+rzYTV0ihNJpKl9GbJM8fdTBB9LoSsyfcVBzteY/g2eBXQyAMip2ZMACbBGDPzPZ/C9nDElZq8Z0SVdIBUAY4qkGzNYv3cEJHNF/3Y5aWlWhYliN5IQ2DOFadlWXZHI2Syc3VLD9z67h9/5s5conrgqUtW2ppyCVcEXy3DvS5fYdHZwerJYGkqQdVg5eFczx3c1MFbmoSxk3qte8uhIpKdCuk9klyZicw1+/YmDfOsDd/A/PvMAWZtKIJ7iV396mLtOXyU9G0AKdUbm4LBSy3u1jmU5hyuQhaAmMcHh6jWsn+jDoeVBCPKKlTNljbjzWapjk+StFo7XNpOxqky63DPG4KUg4bbx6r7VxLwOemuD5gRWGNx9uH3ebRc0ulmygsrnYhg2CTV6e9R4UjaV87XVTHrcbOvuxZ3NsetyNyNFpoZ30m7nVEMdEoI1gyO4cjlONZp17FUjowSSKY6saGLS4yZlt9FbUkxRKk1pLEZ3WSm1E2E82dtDAJwPshBT/wwsho5AYtAfJOT2IwuBTdMIeXx4ciVUxyax6jrNk6MM+otpmgzhLOSYcHlIqXbSVhsgsBcKmLJqFkqTMcZcXupUG9uGumcdXwK8qSz/5VtPUzkR4+d3bmLS535HpMeXG9kdKsIKYqoOiyxhfz2H2q2TX2Uht8mK40gO9aKGsC3t+jd1DvDalpVL2seWjoHpLo6FoKAonK2rJeJymloNEnSXluDJmnVVXZLoLSmmt6QYdzbH9u6e6QDvTWVpHprgUkPFks55Xc8IlikSa29JMa4pDf6UzUZ7ZTlRhxMk2HW5m6FAERXRKD0lJZRHo4z6/YTdLjRZNrNusoPzNdV0eUopGU2QVlVONDWgyTI7unvIqCo/2ruL8miM+y5cwptOc6qxnrzFwsqRUYJTv/sJj5usbkVOCpTryiw6ErokowqdtKKSlVX8hZurMEqGmaa/ond+w/cBSvptiujXwZbV2H6yjzsOXjZT8AKGqvy8fvdKzm2u5uSOOs5tqeGel9umV4LhoAtDlnCm81QNxaY/EwkoHU+y+81uzm2qJuuwcnZzDe964dL08S5JxTwur8ZJgSwW/qN+BMdS65iYhkS/8/1X+NKPXidtV/GmMjOU6gDi2PiWuo3A1Mpdl2SSLH/mUBYGH2p/k7/a/jAhl4+NY31YDJ2OQBUXi2v51KXX8OSzbBzuw6prjHgD/PKJ17AuA1cLwJvMcfehdsaKPXzkZyeRheDIjsYFbbvgpYthkyj4FAoeheh6B8ETKewTy5eW1xUZe75AZ3kZ/lSaTX0DM14/srKJmNNUqDvbUIskBGNeL5WRKCcbGyiJJ2itrmTdwDDjXg9WXefZLRt45MQZjq1opDE0v9D+7YK9UGDjaD9WXcemaTSGQwz5AqweH6IkGceXSdFSUUdXsIzyRJSa2ARlySgxu5O8xYoEWAwDdz5LSrXRNBnClTcFGzRZoa2kivrIOIYkcbGsGl8mhV2f+d1IQFEyzZcff41V/SG++9AeWusr/s0FeLVNwyiSUdsKWDs08uutMxiwln6dzF0qhl2aqRt+C9h0eXCJZ2sG9+vrizfDQKCIM/W17LrcTcxhNv6UxuP8w70HeM+Zc0RcTp7eshmLodNSW0PN5CSVUVPzQRKCzZ39Sw7uG7uuXndJPMGbK5t58Nx54g47x5sa2dzXzwsbN1AzGeZiVSXOXI5TDfVs6+nldEMdJfEE7RUV1E9M0lFRRkE28+O6bDr7jXs9HG9qpCwWpyoSxZnLURMO48pmaamt4Ymd2ymNxRgq8tM0Nk5bZQWrh0foKC+fda6aJNPuriJpddCUHCFmddHlrmBFYoiSXAy3niMnW5hUPRTlUziMPFnZSlaxklZslGfNUlpYdePUc7i1LJqkEFbdSEBxLr5sK9al4r6XW/nc3x3Clb6aLl/dOsrWU/188/P7eOX+NbyxfwUHftGBPLWyv6JkadEMHJnZRN8VnWMURdMMuv301QWmnxdAFgt3GIN4yHNRKl5yuecK/vYj+1nbPcKGziEqJ2JzclucFPhY4QLVwkxPa0icl8uW5fjXQgZ2jHbypdPPcrh6DScqVmBIEsWZOL984RX291/AYugM+QL0FpWi6joXy2qoioWXjSTrTWZxpXOknCqaorCmfWHCRIsSsdHcCtkyK3mfgmsgj21i+Qh1toJG/fgk9kKBjKrO+rmMeT1sGBhEEjAQDDDhcbN2aJiiZJoLNVWM+n3Uj01QOzHJQLCIqnCEgqJwsqmeynAUR+HtUxFyavkZK+p1oQHWhgZmCLmUdrZMP3bFZiuQPdhxdvpxaWpmvWXAF8RRyFOQFSyGcVOyhSeT432HWlg1EOJ7D+7mmb0bFsTo/tcC28XCVO1OIGcFtrPm33JGIMcNlLCBpAszrXq+MGdqdqFYMTCGIzt/7/gNzzVXYE3f6ILbCAESDgcl8QQN4+MEkyZD15/OoE7xQzKqiieXZc3QMLsud+POXr2XJCHY0jHADx7YdUvnewXXTmp8mcwUN8W8hrJYnMbQGOXRGGmbydgWkmQKDgHedIaVI6Mk7XZSNhtJm521Q0MULBYUw+BcXQ2eTJaaybBJCMtmCSaTpm1vPs+Yz8uG/kG29vTizuVoryinftz83Y/4Zzv15WQrl3y1aJJMIJ9AEoJuVxk2Pc+xwCoeHjnOqaJmJATD9gAfGDrCicAKxlUvwXyC4lyCS94aJmxexmw+Hhk+xqAjSIenmrJshKJ8ElncXm3xhcCdyPLgsxdxpmcGaAkoiqTZfaSHE7vqudxcMoOYJiRgqmSizOFXUDqWwJHOY8gS42WeGftdKcKksTIsuakUSSxLqblfg+Joimf3refH929jdc8od5/sYGtr/4zuDBVjOrCboliCNcbtWcDZdI27Bi6wbfQycZsTIUm481k8uTQKZldQd7AMVyFLfdg8B2mZAjuYYj4HdzczXuzBWtApmUxS+ub8pbwFB3d7SKPjwwE8l3N427PLXlvKWq28sGk9eUXh1159g96SYp7euomEw86qoRHuau3g23ffiTOf43JLE9YAACAASURBVJ4LbWzv6uUbD9yNJ5tle3cv6/oH+fOHH+RCbRWeTBZZCN599jz/9UMP8/Vv//PynuwSMZc621KqoLsGOtk42m+y1rUC1nmMDCyGwdqeEf7ge8/xwPFW/veH9nOpvgL9iuDNv4bV/JQJTHk4wSdePDYdcCQNbC3mRO4KzeoK5KT5WG3Rrnn91iCBKeQyPMGlxnmEam6A+tFJAvHFmfQ0jo3xyvo1fKf4TsJuFxlV5flNG+isKOex3Tt517nz+FNpDq9agS2vsbnvKllPAtb2jqLmtUV7D1yBLVdgVb8p5pGxWjm0eiXtleX8aPcudl/uMjslpr4bVzaHkCS+v28vI34/d7W1o0xNPmXDwKprNIVCPL5zOxmbjYpIhIJi4Y3Vq9Cm6u6KYeDJ5vjB3t380sE32dvRyV8+9C4iLid3dHSyo6ubP3/vuzlfU40/nZ7Vb+3Q8xTn4tj1PPWpEP3OUqoyk2yI9zHoLOayu4Jz/gYUYTBiL2LvZBtJi4PGVIhViUHSFhuni5qIW52E7H62RrtAwKAzSF06NK17/rZCCALh1A1b3iSgKJrGk8gxVnq1HLeQIdyZyqPmdYQkkXLNnMS6yeMhT4lIk8ayNELdNfjEc8d5+PVzXK4t5cXda/mj33gvgViKh99o4Z4T7QRjpqFMAZmXLE0cVmrxiRwlpPhs/sxt6eCSAU8hi6dwjYshZslHEgJXPsfphtVIAhRh0BAem7dVbaFIumxYCwZ3HunEk8xhWWDKf+F97oN51v+3YSwZA80pL6uXu0U3+OTBI9NmFVdqkH/4kyeBq4PwVx5/EkOWsEyllK7/+6v/8gS6LKHoBlnVSs5q5c7WTrwLJNj8a4Wq66h6elHbSIAnneXeU21s7ejnJ3dv5Wd3bma42EfSYXtnBvipljt/MkPT0DjvPXyeh45cwJmbvVq5GZbryqyawdq+kVsO7isHQrgzi9OJd2dz/NazL5oud4ZJJK0fn+CBlgvTq4UPnDhF3qKg6Mas1KA3laFpeJzW+ltLzTcPjU17ztsLBe47f5F7L1yaPvaKUTPw/9qrrwOmSJRJsDN5Jxv6B6efv4LNvf2miIwwLT/3t7ZNTwIAHmi5wD0XzK4RSQj+6Cc/Q1NkrJqOBHz1R4+jS9KcffcyAqeeI6k4yCpmcHLqeWx6AUUYKMKgOTnCmvgAbi1DMJ+gzVONU8/hMAposoWa9ASV2UlKcjGCuQQ5xcq7Rk/zRNVeGpMh3PrbN77IhoEtZxoyCVm6obiJJsvo8sxlhSFL85blJK56Glw/ITwrlWFDZ4MY5zm5iUeMjiXX3CVA1XQC8TQbO4eoDkXZcHmI7z+0k28/cgePPbCdAyc7+OSzxyAh8IssO/RhduqDHFVqlnTsxaIgKzzTtIP7+s5RloixYaQfCTFdZlouuNI5JgMuRsqbCYZTFEXT7DjXN+92C6+5qxLhbS4Mm4xUEPgvprFnl6/mfqWufP1z10IRphxi3q5gzekIGSLlLtzRLM5UwbQ0nEothXxeLlVX8ujREzcdzCfKXci6IDBuBkcBaFbTJvHKTf1vFVc+l0Aiza8/eYhHDp7jhV1rObixmba6ckIB77zuYLcdQuDIFaiciFIbCrOpa4i9LV2s7R3BvgytikuBRddZ3Rea/41zQQhWDI7jWmRwvzI8y9f/Vq4L4jZt9r0rAY5cnpX9oVsO7qv6Q6iFmeW4m6Ug5zrX62G5ZnsJZglHSULMeE4RAuWa65OmJg43QnNyhNdKNtDjKqc8G6EiG8YidBpSIWrS4+RlC63eamy6xv7x81Rkw3gL5njg1jI0pkZp91TR7yzhnrEWRu1FXHZXsj7Wh1W8vfegNW/gj2aYKHYzUumjanC2aqAuS4xW+Ij6nUhCMFhdRGPPBAmvnZzNDAFCYir4z962YJWnjjXze1EwGJHcOEWBDNZlqbkLTD/3tvpyzq2qpr22DMUw+NSzx9nUMchwsY8n7tvC3z26n9/5Py9RIlJ4RY6zSgUWjKvtkG8BMhYbL9ZvYe9QK9WxSapjpsrkhbLlnWQousHW8/3EPA4zY7jcfu6JJhvRTU4KHgXbRAHbZAHb+O0RsbkRDAlGa70MNhdROpigdCBOx+ZS/ONpKntjFI2n0RWJSKmLQDbFh4+eIOtSydkUDEVCNsCa1wiXuUm7rXgiWbrXm17btR1hAiGzhnlhVyWyIWi4NIE3/G971X8tyiIJPvXCMR48eolL9eW01ldwcnUdLU1V83qBLyuEIBBP0Tw4zur+UVb1h2gcnqBheILgItPYtxMW3aBxeBw1XyC/SN6CO5OjNjQ5w5/6rYAzV2DlwNj8b7wBVg6MveXnvFQE8gk+MHQEEMhAac4kGO4Km21TW6Pd6Jh+5BKwPj6zlNGUGqUhFZoOGg2pEHWpsen3v51wpnM0dE8Q2tvIjz62neKJJA3dV6WMBdDVXMIr968i41QBwd9/4U4e+fk5Rit8jJWZrZx5q4WEd7bRVNznIONQkQ2B7zpp2yYRJYadPsnHejE+Q452Kfjvn32QnFVhRf8YH3npFOsvD+FJmx0ZDcOTFMdS/P6XH8GORpNhcjMCIoNb5JeF3Ji2qAy7g5SnIjgLWVqDcwfrqN1FwubgcP1qVk0OcaaqAVkYnKxp4v9/5gco+vJM/HI2K1mblRXdY7jmID3eCAsvvEkSzsEcQpFJV1kx1Nm3dcPwJD/842/d1FZyLhTHFuYJbSgy41Uehhv82DIaJYNxYkEnKZ+NSzsredcPLjJa62NgRRE5p5XdL3TTtzJAcDRF2quiZnTcsSxHHmzEltGo7QhjyBKdm8tIeVSCoynq2ifpWVuMfyJDZY85C5aA7a19PP6f/25R1yULQXn4rTEzWS7IAkqjCUrPJth7oZsPvHGGsSIv7TWltDRV09JcRUdNmdkrv2zHFFROxFjVP8rK/jHW9Y5QG5rEl8zgTWVxZfOLIp1dj6sGDxLDniIulNfRXVTGsCdAxO4iY1XRZRlHoYAnl6YiEaE+Osam0T7qomMzelZnrFiBbe39/OgPv4Uxx6rnZrDoBtVjkRkCNDfCXHu+dpu8YqGtuJqLZTX0+UoYc/lIqXayFisWQ8dVyBFIJ6hMRFg5McxdJ7rYdfHvb2mFUz0ewbpMwf3K0Q1JYsRdxIWyOroCZQx7gkQd5vdiIGHXCnhzacqSUeqi42wI9dEQCc3glty0rY35tc/n8wC/PmgsyTN8GeGLZ9lyqp9T22s5v7GK//aHD7Hm4ig1A2FkQzBU5efslhpGK3xYNJ0VHWOc21JNb0MQzaqQ8NhBCGJ+B21ryll3YXjaLEYAHStLiRQ5kQ1B9TVZAQnwkOcOMUgBGRv6snUN3Hu8jTU9I5SEk9jzs93mqsaifO6JQySwMSx7qDWilIoUR5Vq7tD7l+wM1+sr5Ws7P8CXTj/DhvE+vnLnJ+csX+iSTNjhoXlylIjDzYrxEUqTMcJOz5LGq+shJIlQiZfiySTOqeC+kNFmwcHd15bB3ZOl4FOIrXXg7p6dGnDkC0uWpbwZFM0gEEqRdVpZcS6EZIA3kqGubZLOjaVMlrsZaC5i/dFhRup9DDQXkXGrFGxZsg4rCMhnzdqUM1nAkSqQ9qisOTFCIJQiXObEnipQMpSgsjdG8bCp4y4BvnSWTQu0Pv23AlXTKQ8nKAsnWNc9zCOHWtBlmZTDRs+U7/dgSRH9ZUWM+T2k7TYyNisZm0rGZsWQJKyajlXTsecLFCXSFMeS+JMZysJxasbC1IxFqB6P4srkUAyTaKUYxg3dqBYDgRn4JpxeXmncwAsrttDrL6UgKxiyue66fiIqIabrwxZDpzIe5oHLZ3nX5TOUJaPY9JnZKlc2z4Yl2rWGHW5++YP/gYhjZnakPBnlP73xONuHu2Ztk1MsDHuDPLVqOy81bWLC5UWXZAxJnmm2w1RwEwIJk6Bm1/KsD/XzgdZj7Bi6jDufue3a9tfiyvcSsbt4pWkjzzdvpTtQNv29GMyuBV/7vSiGQVkqyt3d53mo4xRV8TAObbYt6L8HKJrBvkNdtGyu4cjeBvprAwxWF5nftzDr6oYsoWgG977cxie+f5w////up3VtuclbKOjsf7WDE7vqeO6h9VQOxVh/fghFNxgv9fDUI5uYKHHjyOTZcmZme7IEWDGWjUh3BfvOXMaWL2C5QYuoK5PjnqNttMvFtChlhCQXCoJOOchefYClsr1r4+P8zvGfU5sYByTyipUvnnmWYGbmQi1mc/HtDfdRloiycnIYTZax6jqPnjuCMg+peTFQdIPemiAv3L0WZyZP3WCYTz822+/+eiw4uBfc5krNMaqhZNKmat1bDAmw5jQyLgvjlR5KhxLImoE1r0+LHLjjOcaqPUSLHdRcjqBbzNV+wmejoi9O0Xia8UoPJUNJikeSTJa7ULPaVK+xqZxlTxeYqHBT3h/HPUdb2jsdAgit96IUDIKdSeQ5fnuGBAO7A4hrVpxVJyNYcsacg2SszkG86qqhYuPIBNtf77vhgBp1OTjbWEPSYcOeL5i9ofEkW7sGbrDF8iKrWOgOlPNS02Z+tmYnUYd7QdsJJHTJtBQoKBa6ghV8I1jB9zcf4P2tR3mw8wyN4dFZOgJLgZAk0lYbaXVmWnTYE2DC6Z1BktIkmX5/CU+t2s6P191ByjY7lTr7mpgOlrqskLdYebNuDW/WrmbzSA+fbHmDHUOdeHPLZ3ZxI+RlhT5/KS83beTxdXuYdHgWRN6c+b1An7+U7269l8fW38EDl8/ygdZjNE+O4NTeeh/4txMSZrvar/7DQRyZPC2bqogUOaclaBXNoCyUYOfRHj76w5OUjiX4yh8/wzMPb2CgpoiG7gne/exFvPEdPPGhLXz1D99NzUAEZyrPUJWfSNCccG5oGWL1pbfGAvuvPnEPH33xJE2Dc1sKS5iMdB/ZaV15u9B4WGtfloyKu5Bj/aRZmsnLCtWJCfYOteG/zhwmanPxxMo9CCR6i0o4Xd2AvVAgkElSnIovG1velcnz6ceOkHKaCqXu5ay56zaJiT1uDIuE/2Ka8FYX7q4cRecWx9BeDhSNpwmMpYmUOqnojVHdFcUdy1HXPklRKI1vPE33+hK84SzVlyMUhVJ0bC3HF85SPJKkf1WQyt4oOYeF/pUBSgcT2DIatkwBa05H0QwaL07QsaWMpFd9RwR3Q4asX8Wa1rBm558l66rM8/99PZ6RDA/+7kVc4dkDnlAkzn6qlqzPQqzaSarUxq/cdxDv8BwcAwmGt/lpfbiCnNvC+BoPG/9lkLu/2nZjkw8BCaed9uoy1vaPkLVaGQn62dw1sOS02c0ggIjdxTOrtvPz1TvpKSpbFlJgwubg+5v2c7h2De9vPcZ720/iz97e+n/WamXU7acgK6iGTk6xcKRmFd/Zeg8XSuuWfl2SxNmKBnqLSnl/6zE+dv4QpanYbbkmAaStNp5dsZWfrd1Fe3EV+jKwitOqnZ+v2cXJqmbe33qM97WdoDgd/3e3iq8ZjPKbf/ELzm2upqchSMzvwFBkvLEMK9vHWHtxGHfKHAf8sQyf/P7x6W0F8MDzl2hfXc7F9RV0N5Vc3bEQrGwf49PfOboosaVbhQCObGw02fA3gQxUigRBLY26jCWB62ExDH6l5SXc+dlWug4tx+7hdux6HkXYkQRkrCo9jjL29bQtW3DXZYm2FeUMVBWBgMrRGBtb588iLyi4GxaJfEAhU6Fi2CQkA+zjb48ojC2rs+HNIYRkiunXdYQB8ESvBuGiV802AQmwZTR2P39VQGa43oslr1NQZSwFnequyLR4Q9GE+QUGR1Psfr77lj2llxuZgMqJX6tn9dOjlF+Yv4YvGQJLVseaMVC0uaOvrAkOfLUNzaFw+tO1XHi0+sY7FND80hhVJyOkim38+B+3z3sO/nSG3W3dhPxePOksst0g6nbc9kF33OXlL/Y8zOG6tSRsjvk3WASEJNMdKOeb2++nrbiK3zryFCXpxenCLw4Sg74gOYsVuSB4o24df7P7IQZ8QYS0TFMkSSLqcPPY+juI2Z389ptP4c7f3CL0VpBU7fz17vfwctOmBWdRFgohSQz6ivnulnvoDFby228+edsmKe9kOLIFdh/tYdfRHjSLjJClaYb7fJyE2v4IX/jfr/Piu9ZyZlsNca+DokiaLaf6ueeVdpovj8/ah6nvbp1iykMJ6SVP3CVgVV+I8SIPdaM39/owkDitVHJWKccpCgRFmke0+TXXFwMZwc7Ry3O+puoaH25/E4eWI1mwo+oacbsDfZnbiOMeB6OlXur7J7FoBie21LOufX5XyoVpy6cNyl+Ko3lk0+ZVn/JEXpZTXzzMGuLNX5/rsQA2HhkiP9X+oWa16da5xez/rYQAEmV2evcFaXxt7jTV9ZA1wQd//QySIbDF504fS4BvOIshg2OOlf3177XHNexx7Yb7mwvBeIp7zrZxurkGe0Hj4aMtt+2eEUDI5ef37/sk5yrq518VTs2qpanHppIaV/nPN/mBJmxOXmjeQshTxJ++9E+UpG7fSnHAV0zGotIVKOdr+x5hzOWb+9yuvZ5rfNqmb+N5Bpy0aufpVTsoSSX4/MkXlk1h60om5asHHuVg3VoKyjxDzhKuI2lz8HLTRiacbv7ni/9EUSZ5w+/FLFWY3/cV8lPSaePwpiYePGLqpxtXT2R6PLjWuOrav99JEwkJU4NhMbDoBs2dY9T0R8h+3zqtF2LPFrDntFnjoQDapSAvyQ2UCjNd/UGjHXUZ6u+PvnSKf35oF/0VQconYtMZg7JwnIbhq50AccmGgUSRkWWf3sdJpeqGff63C86Cuagcc/voCZTiymdpnhhdVoU6Na+RdNmYCHqwaDo51bKgLMrCXOEEqBEN3amSqbCaK/dQAUv6HaDOtAhImH2h1vzbV5cTgOaQSRepaA6T3CcXDNS0jj2aR5maNOkWiUyRlZzHSseDZRScFqI1DhwRc9UjFwSe0Sxq+ipxI+uxkKiwT9/d1rSOLV5YsD/2ckMAzlyOjT1DKIYxZx/tch1n0BvkT/d/mDOVDTdd1SqGjjebxpvLEMgkqEhGcecySAhSVjsht58Jp4eEzUHM7rphMNIVhdMVDfzRPR/nP7/+E6rjk7dlUOn3ldBbVMqf7fvAnIFdNnQ8uSzefJqiTIryRARfLo0kBAmbg0mnh5DLT3zqegxJumGALCgWvrvlbrYPXWbH8NyrlcVAABNOD3+552HeqFuHdpMygqoV8OYyePIZguk4gUwSXzaNIUnE7C7CDjeTTg8J1UHM7rzh5E2XFU5XNvPHd3+UP3r1RwQzyTnfpykyJ9bVU1AU1nUPE4yluNBUSXHk6vuPbmzEqum4MjlW9I/RVV3CSNBHw/AE7kyO4WIfk343xZEk67qHb2u56a2ALMzVvyM7f1ZWAAlU9hv9NAqTRb9cxLo//6X7yaoWuqqLZ/ymHjx8kS/85I3pv+1CwyNyNBDhrFJBWrJiIL1lnQyarPBK3SbuHLzIlqEetgz1ADDkCyzr5FjNa9x59DKnNtWCsPDRn51cXrZ8qtZGz2eKyXsUbBGNqqeiFJ1N39KAJjDdrcaKPIz7PUTcTtJ2lbxVQZdMNSuLrmPPa7iyOXypDIF4iuJ4alk81t8uCCBVonLhw1X07C8h47NiqDJywcA/kGbvX3dRdiEOAtLFKuc+VsPwNj+TTS7yLgvHvtCIJWsGc08ox56/7qLqzNX2lNAGL0e+1ETebSHc4KTyTIz3ffks9sTb85mFPS5+uncLp1bUsb53GFcux689f2jZg2DE4eZ7W+7mVEXjDQO7bBhUxSfZPtzF3v421o/1U5qMzarVian9tZZUc7R6JSeqVphM7jmCvJBkTlU28e2t9/KlY88RzCSW/drGXV6+dsf76QqUzwzKQlCRjLB96DK7BjvZEOqjOjYxK8DoksSE08v5sjper1/H0ZqVTLh8NzxeXrHwna33snpiEE9+aRoPKauNH6/by6sNG9CUuYOxqhVoDo+wa7CTrcNdrJoYojg9+3PUJYmQ2097cRWHatdwsqqZQW9wTt6BkCSOV6/gHzcf4HOnXsY9x3VkbFaevHMj955omyYb+hMZnt+zlu1tJpnqe+/ZzWeeOoI7nWOgrIgfPLiDpsEJjq+v567TnRze1Mw9J9p4fu866kcm8SxSkOhfMyQgSIaDUg3jkhNJwH7Rh2UZAus/fuU7c+7lihIpmFkVGcEGI4QMdMoBikTmLW1RTFts/Lx5JwWbTCB7tTz3avM6/uOrP58lxnQryNostDeXUz4WY//hzkVpTCw4uGsehcCpFIZVIu+9dSJMRrXwxsaVHF3bQHdFCaGAh4jbRcqhkrNaMGQZ2bgS3Au4Mzn8yQzBeJKKyRibuga5/2QrwcTCeuMXCgF0VZbw+qaVDJb4UTWdNX0jHDjbgT81m0xxS8dQJC7fX8b5R6tZ9ewoFWdjSEKQDqpEa51YM/p0JtIW11jxUoi6Nye58MFKBnYF2Pq9PkovmTeRJadT1DuT0FjeEufeP2ml4FR48uubl+Wcl4KcaqFuLIwnk6M4nmSkyIuYMqlYLhRkhVcaN/Ji02YKlrmFZCy6xp19l3j0wptsHu3FfpO2KQkIZJLc0d/GjsFOOooreWrVDp5avYOMVeX6pF9BsfBy0yaawiEevXgY2zKy6MFciXYWz5a33TPQzida3mDTaO9Ng7AiBGWpGGXdLeweaOeN+nV8Z+u9XA7eQKFOkmgvruRkVTN391y45fM2kDhWvZIn1u4ho9rmfI+9kONj5w/xYOcZGiIh1Ju0DylCUJmIUJGIsHugnTMVjfxw450crV6JNsfEK2dReWHFVjaE+ri758KsVj9XJs/DB1sYLvEzFvBQEk1SPhmfUcf0JzLsvtCDBBxbV0/5ZIJVvaOs7jXTrisGxmgeHOPsqhoyNuu/q+AOUCLS3CP6posmy0FqkwB73vwNGcztwwGQw8IZpYJKI0G5SLDKmJzjXbeGjEVlxF007/siNjcp1c6Zqnr291ycft5YLj4MkHbYePqBjRSHk1QPR7BnCxTF0mxr6Z932wUHd29HBsdo3rR83ejEMbpwQt2Vr/x8YxV/8/4DnGuuIexxot9gNq8rEroik1OtxNxOhkrMD1oyBCNBH3sudi9rcBdAW205/+Pj7+LUyjrSNiuyEHhTWVrryvkPj/8CZ27pBEJDhvFVbpwTOVY/PUpJmxmohQyaXZnRhqamdcouJhAS9O0JYMkZlLQnqTlxY5KJLalR0pFEAEr+7S+ZlEYS7Mz1ICSJ1zauZFPP0LIGdgGMuv384+YDJOzOG7xJ8O7O03z52LMUp+KLSp2qhs66sQFqYhNUxyf5qz0Pz1laSNocfH/zfnYOdbBicmTWYHT9FS9pdT91PV84/vyiSwGuQo4HLp/FruX5X3vfx5A3OGeKPmZ3cqRmFfv6LmFdhNf8tUja7PzDjgeYdHnnvAaHluf33niCe7pbcBcWHhQlwKEV2D3YQU1sgq/d8QiH6tbOuYIPuXw8vWoHm0b7ZjHoJ/xuLjRVEnc7qAmFiXgc/PTAZlqaKzmxto7N7QMzZHNX945yZEMDLSuqqRqPUjMaQRbXtI2+k4rubxF6JD9u8jSJ2ZK3twoBpOwq33/PLg5vbuJzTxxiT0s3Rzc2UhWK0DhVc89i4SVLEz6RpVSksAmdoEhzQO9d8jl0+iv4k30fn/d9uiQz4fDy+0cfY+WkqXUhAVWx8Iwsw1LgSWZ54LVLdNcVE4ikcGbzy9sKB4ABloSObVzDNlaYrg0vaFNZ4vWNK/nTT72bnvLiaZvBRUOCyokoZZHlVX0rKAqHNjRzcEMzYmqQ0IGI18U/3b+bj7x6msaR2WzRxUI2BCXtCToeLOPChyrZ8k8DuMZzWLP6jNr5vxVYddPYo626jIePtZC33JoT2Y1gSBLf23w3g97gnK9LQrCvr5XfO/gEjsKtiZxIgC+X4WPnD5FS7fyfrffNmWIOuXx8Y8eD/M8XvzcjIArMFPBIsY/M1KRxTc/orZ2LMLijr40vHn+eqluo8UuARRjc2ddKZ7CS7265h6x1tl2tLsn0FJUx6i6iJr74FZEA/nnjXXQG5zbUUXWNLx17jnddPoN6i5kOWQiq45P87sGf0hWoYMg3xz0gSRyuXc258nru6T4/Q6WuJJrgl549jgDs+QKyEHzquWN84vnjWHVTN+NP/uHp6fd7U1m++JM3yFkt2PLatLOiYhj80rNHl20wX25EPQ7zfDXT2c2QJdSCRk61ohgGWdWCLa+Ttyq4pyReFworBkN4KMHMIHpYHiGh775vD2dX1aDoBlGPAyFJ9FYEaasvo/GnhwHwkeWz+dN0ykGqRRyH0LCxTHKvFisl6RjvvXwcx010E+I2Jz9Yux9/JkXGqmIxDFRdo2nyFv0m5oBV09l8YYCmnnE8qez097gQLHi0ja+2k65UqXo+RmSrC9ukhu/C/KIXArhUV8HXPno/3ZUl87z75nCns6ztW37DkIzNak465pj9Z20qndWlNI4s3StY1mHFi2OEG930HCih48FyGl8bp/nlMcouxHGG8+8Ylv5yIOpy8LM9mzmyppHmkXF+sWk1H3v9+LJdY09RGc+v3HLDiWJddIzfPPLULQf2a2E1dD5y/hBtxVW8Xr9u9jEliUN1azhd0cSuoc7ppzVF5vCmJtrqy6mYiKHoBmt6Rhd/AkJQGQ/z8fMHbymwX38t93Wd4+WmTVy+vpY/dS2jbj/DnsAtBfdhT4Cfr945Z5JWNgzu6zrHg52nl1zCkICKZIQvH3uWP7jvE3OS7DTFwk/X7mJfXyt2/Wr2TRbMchO8vp557d/S1N/XPneFNW99hwZ2gP/82+/nI8+dpG44TMTnJBT0svtsNy/cuY6akQiHtzZy4HgHb2xfyRd++Dqu7GK0y3VapFI6pSJkBJ80Li6ZLS+AbHHyYQAAIABJREFUg1tX8F++8SQv7lkLmJP04liSsyuvtute6XMP6BnsaMtS678Cq6GzdnKAe/tacOg3zthGbC5eaNiKIcmcrG7Ek8uyZbh3WZM4EqAWdIJRM1OtyRL91QEa++f/XS44uFvjBsk7bUxuc5Gqs6GGF/bDLCgK33n3Xtrq5q7xSYZB7ViY2rEIRYk0tnwBJChYFJJ2O1G3g3G/h5GAl0AizfolSn3eCMZN8rViGb8t10SeO/6ik8bXxunfFWBwZxE9+4tpfG2CXd/oxjOy/D3GbxfyVgvetMmX6C4vxpDgqmr10iCAx9fuIWOZu55r0/J84NIxapaRxe7PpvnApaNcKqlm3O2f9bouK/zLxjvZMtozY0Vq0QyaBidoGhybtzf1RrAYOvv6W9k60r0s11MfHWPdWD/dRWVzekFEptjpi4UAnlm5jYjDPeekqyQd497uFgI3YLHfCvYMtLNicpi2krkNPk5WNjPoC9Icnj2pChV5SDrt1I1Ozlp9X2ioYNzv4e4zHct2rm81iuIp9p/oRIKpVbBJOMxbzaF/Tdcoa7pGaW2qIO1QFxzcJWCFiFApWpAQOFg+EzFdlmZ0+AhJYqzIg/M6Fr8FgZvl11upiY/zQM+ZeSefDi3PttEu7FoeVdd5rWkdg/4gkhC8u+3Mskg6Z1ULnY2l5GwWJoIeNEWmvzrAF7/z+rzbLtzPvT9H8YkUuWILno4s7t6FpXBa68p5Yfu6OV+rHovw8V8cZ/elbopjZj3hihG9rsjkrBbSNpWkw0bY4yLptLNycPlSHldgK2hUTUTNHtvrBiR7vkDj8ML6yxd8vJRO/aFJqk5EiNY5aX93Oec/UkVxR4L/y957R8l13Xeen/veq5yrOueIHAiAIEAwU6QkSpRlU9ly0lgj2Zrd2XBmbM+Z2eMZz5nj2bOeXXnXx2NLHkfJtmxLFmWKIsUoRoBEzkA3Oufuqq6cXrj7x2s00OjqiG5QlPU9p4GuflXv3brvvvu79xe+3z1/MzJPqvN+RyydY9vwBFPhIFOhAPdfvLZhggrjgQjHGzuXFCnqSExycLRn3W7fSlCQ7J4a5K6JAV7o3LtorEjgUlUjl6oa2TtpEykpliSayXGmNkLJ0YBmWXSOrN0LFCgX+MjVk7iNjZnMVCnZN97HD7r3U67gsco7XCTdPizEmhKlEp4A7zZ1Ua6Q5CakZNv0KHeP9m6YIRDYtcaP955Z0riXVY3XW3csMO4SGK6JcK6zEdWyaJyeJRHwkgj6aBuP49YNAvkS3354P4+cukrO7UTXbNd1MuAhnC2Q9HsYqo3iKZZpmZpFCuirr6JtIk4w/+ORXHczj0dtPMOLR7YzG/IRD/voHJpe92JTAuP4+J6yhYNynAxOjsiR2965C+ChEz388ScfoOh0kHc5mYiFeGN/F1/51soGbSMQK2aJFVdefLpMnU9dfoOAXqB7ZhyHaaBZps17sEFTuGpZOAyTmZif6pkMiiWZia5OoXPVxl1YEH3XzpYXpkRZRYhYAt+7by8F1y1ZzFJSn0jxb7/1PB989yJOw1xENgP2jbaYI5pAYioK6hoSfAyhoEhrxSQql25w6GI/3SNTXGusnjcYmmnx6VeO0zS9PFPSaiGxk+eEtH8cJYuqq1ksxxQDD8SId/iQqoBbjLtWsjAdAsOl3HGShvUi73IwWB1FAIcv9yGkRNtAqdAzdW3MeIMVd4eKZbFzapiO2ckN76tIIcf+sWu83byV7K0MeEKQdPs41rSFPZM2774iJVuG7B170u9h+8D6XPLd8XF2TC3Nyy+BuCuIy9TxGkVKqoNrgXq6MuN4zMq7se3TowsU7xacTwjSLi+6qq7JfX6puomRYFXF++IydO4a77vtErtboVkmO6aHcZjGkrwEb7Vs41dOvTw/HvrrY3zrsbuJpPP4imUmIwH+5rGDdIzN0NNUwxNHL1CbSCPmjF9PUw2TkQCHLg7wyv6tHLrQz3cf2MuOgQnevncXX/7u6zx3eCdVqSzP3LeH/+1bL+Epvzcsnjfj3/zZC/O/182k+eLfv4EUAs00cZUNLCFw6QaffO4k7jUkDUtgWASpxzaC08JLWaobQmLzi98/xnce3cfTD+/h7JZG2sbj/IvvvsWBS4O3fe71QsKcMJM9gmwRI4tw2c43iOWzxPJ2XyQ8PjZKU95hWHRfm6JjYAbNtOPtdVOpVX129cYdQIJaXn2jjblEtVuhWhYPne7hw+9cWBSvshBMuULkNRcNhQRD3mqGPNXcnezFZxTmBCQEqrQwhIoiLSwh0BUNzTJxShMTQUl18Ez9PRxKXKE5P0NJ0VClhUOaSAS6YtfUu0wdBcnBK4P816//I996+ABXm2sJFEp88N2LPPn2WbyljUkUsRyCc59qIjhaIDRSQNUtDKfC4JEo2VoXVT3ZxYx5EqquZCmFHAweiREcLaKWTYQFvpnSPNe8pUApoGE5FCxVYKkC0ynI1bgwXQqKKXEn9fkVZcmvYrhUm1rYbw+DXLULVbdQTIkza6LNZdybmqAU0JCqIBezE7B0r0q2xoWqS1Tdwn0Lc13a4+G13Vu41FJP+8QMZU0jlC/wL3/w+m33pSkEF6pbyDgr08sGSgV2Tw5u6K79OgSwd2KAaCGz2LhjJ+Ncrm4k4/QQLBewhOD0liZOb2mmejbDa/u6+K2/+OGaqwYe7j+PtoQhvo4zkXaOVm3jYPwq454otYUkPYEGnhp+u+L7a7LJZduRdzgx11DWYyHoidWTWIJe1quXODh6bVMWp5FCjvpMgqFwTcXj/ZEask73/MLiTHcTD57uxVMqMx4LcXxbG7MBL8M1EWpnMxRcjgXStoaqUHZoWIog77IlggEC+QK1sxkSIR+DdVGSAS+GqjAd9tMytTGbgttBKHtjIaVISThjl/Xeeg/WEmu//vk6chyjkbRw2WQrG6Tn7inpfOF7b/GF772FNVc6u1RJ3GZDYpPVXIw182LbXVyJNGApCu3JSR4dOsNdk/3kXC6cpknSY1fsPLPjAP/y6EsLJIlvB6qUqHNj0VxDMvrGpi/fgpGqMImAb1FjvKUyj5y+UjERpaA6eb1qJ6ZQOYCdmDThjnAu2IoqLbZkR4k7g7TlJzkfbKWuOMuIpwpDUSkqTh6YucCAr4ZJV5gxdxRLKIx5ogx6azCFwj2Jq6QdXs6G2vCYZe5K9uE17RDDvt5h9vVunmqZFILJHQHe/WIbSHDmDSyHAhK6X5ii8+XpRcZdAI2nk2x5bpLex2vof6gKR8EkOFbk3j+4Rs0lu5yuGHbyzpfbSTW60T0qxbAD063w8v+xDWfWwJUx+MDvXMJRsPv87GeamdgdRPeqJNptN89rv7kFd0rHkTe565vDNJ60S1wSnT5O/HIrxbCDUsAeMkOHo+SqduIoGISHCjz4ez0L2l2XTPOxY2eJZnLs6R8l73JysaV+XhPgdpB0+xgJxZYspQyW8myNj27aZNCenKIql2EoVF0xGW08EGE0GCU4M4qpKhScDvZdGaZ9bIbB+ijmLRSmK0GRkgOji2Vfb4VfL7AvcY2s5qbfX8sjE2c5EVu8uL4On15ccucO9uJ8qbBHJWRcbgbCNZQr8Q1Iia9cpDOx8cpiAvDoZarymcrGXQjKisZQqIqd0yOArQl+rrORQL6IZlm0TiRI+j3s7R2hLpFGMy0utdYxE/YzVBPBny8xti3M+Y4GZgNegvkiJaeDN/Z089CZqwRzRbpHpmmamiWQL9I0vXHlYRuJjQyHNMkMH5M9jAs/bTK5YQx1f/Gxw+zoG2fL4CRVydx7yvwngXfqt/Anex4nWszSnRxHSMm0N8Qf3fUEn7zyJl5hL6Cu1DQQKBa4Wt2wpudmORRcDk7uaZmfL3SHytXOWr74jTdW/OymGvfR6jCGuvjWuMoGu/orq9o4LAOfWcRrlKgrzTLujqIgaS7M8KPq3QSNPBOuCI2FOFcDjTikyZC3mj2pfl6t3kNHbpxrvnpaCtOo0qKkODgR7mLaFWTcHaW+mMBCYcwd45HpMzitlXd3UkCy1d6lBUeLlIIarpRBIeLAWTBRdEmuyonuUwkNFTCdClrZQi1ZlP0aatlCK1mousU9X+un+4UpikEHlkOg6hb+yRJVV7O4k5XdYu5ZnSO/30vXC1MUok6EJfFNlQiO3CDXUUsm9aeSxHrsW7r1BwtzExRTLpDprbqSsa9XaQxKiTd+YyXvTuo0H0vYCxFgx9MLJ2hXpnK7Y+ks9YkUZ9qbUKXFrsExlA3wVsW9AdslXwlSEiwVaE5ubJ7EzXAbOp2zE5yub8MSixcYcW+QKV+I7TOjaIZJ5+gMJ7a3MFoTZv/lYbQ11o4HS3maVpG13pGdIKe5iZYyVBfTXAw105BPLPl+zbJwLMOidV0bfrVIuX2MBZYm/6jPzm6oVO7N0CwTt7707tNSFKZ9IZgz7nuujVJ2aGimRXUyQ/PULIYqMNQbGQYO0+TJt84B0DKVYE/vCJpp8dDpq0xEgzTMJGmaSnKptZ6GmRT7rg4xG/Ctqc/er7Bj7n5OKnV40Tkjaqm2BjbEwJccGt/64AEUCbt6R7nv9DW2DUxuWL7OWmAKhX/YcoR9k318/uKrROdi8Wmnh6e7D/NC+z7+0+t/je5QOTzYg08vsWd8CG2Ddu0ll0ZfWxVd/TfydFbbC5tq3ONB37z7ah5S4i2VqZmtrKbllCZ+wzbufsNeEdUUk1SVUniNIsbcZCqFIKfaWtZhPUdNKUXQyJPXXChIqkppgnredr1bOvfGL1NVThMtZxhzx6gup2gurK7Mx3QqTO8IkK9yUnUlS7zbR93pNP2PVtH4ThLfVInZTi8lv8bUriDVlzJk6l00HE/R/0iMtlfjNkGNhMhQgcjQ2hjvBOCfLuP/0dIGy5Uz2frc6pMN299YfYlTYLLEzqfXvuNyGSaHrvST9bhRLGvDGLwSHv+Srl8B1GVnl61P3Qi0z06hSllxKpt1+5jx2TrsQ/VRDFWhY3QGXVMJZ9Yuk9yUiq/KxRcrZdiT7KeguhBS4jHL+I3lx9pGkgqlXR6m/EtT27YkpzeNHFSTJu5l7rkpxIIx49YN7jtne0Oum+JDFwfmPUsC2NU/zq7+G+P+8IX++ff31VeRdzvpa6wCBMFsgYaZFNZNn/9JhgRG53bsrdLmHVE3aOf+q999k3jIR09rLa/t7+I//KuP0zIe53PPH+fQ+YENucZqIYVgOFTN5y69tiDJLlQucPdEDy+17kVXVGK5DCm3l7gvQKBU2LDnKpAt8nPfP4U/X0Zic8a0DK9u7t5U415wObEqMHr5CuVlywTac5P8XdP9lFQH9YUEbquMAnjMMtWlFJcDTXyn8V7ymguHZeA2yyhS4jVLRMtZEo4Az9QfJOH0E9DztOUmORrbiiolnx1+zZ4IlkgyqgTTITA8Ksl2L/6pEoZbZWp3EFfaINPoxj9ZwnQqpFo8qLpk6z9NcPWjtUT68hSjTtyp9y8f/u3CYVpEsms3aMsh4/SQdbkrHhPS2jQRl5vRlJ5BWcKlbSgqCY8fXVF55cBWRmvC5NxOahMZhuoi/N5Xv72m9jVmEquaLC6EW/hh/X4aCvZu/amhN3GtwjO1EZBAzukm7lm6fO4fdxzmma0HN+f6ws6jWe544RbCnlvvgWD1apNt4zP8yrN2PwvJfKLvRnim3g84LuoZEwEmhJ+05UZBcp8cvm0KWgF4SzqeqSR18TR3XR5muD7CH3zmEX50oPuOG3cAj14i4QnM5XrZ389CkHT50Cw7Gfx4cxc/3LKHxpQ9Jn4pObOA4XC9UC2JL18m53FScDsQQDi9us3hphr3Si55ALe+fFZmU2GG/7XnewgkCpKOvL0jfXLiXQA+M/L6fInOzQ/cx8eOAVBfnOW++CXE9eOFGXakh+deS9ryU7Tlp1b9PUbviVAKaFRfyKCULMIDeUYPhmk+Okuiw0eqxUOq2UPDiRSJLh+KIak5n+biJxvoem7qJ34VfychgbzTRd5Rub5dkRDbVI11G9F8dmmDKwRJtx9d1fj8c+/wxt5O/IUyreNxvvnEPXP6Cat/8MOF7Kqyb4uKk3unL7ElMwZS4rhDhv068g4X+SV45BECU6grS/FuEiRiZbnZNUABlB9j8prNhg+dVpme37XDxngrJLb07khNhMtttZzc3sJgfYz2sRkeO3Z5A66wNqjS4sjoJb7bfQjNMqnLziKQzHiDPNN5kB3xIfJOJ3Gvn4evXWDL9NiCxd5GIOd18f3Hd/PWwQ46BmdQLYt/9T9eXbG/N9W4SyGoSFK7wjwlWN7FYx9f/iS3riBvx2VUfSmD7lXRiibhgQKurIFatoj053EldULDRco+O/O89qyt6lZ/MkXfY9VUXd44so6fwqaczTtcyxgJSbC4sZ6CSgiVcsvu0rJON7qi4imXaJxOcnpLM6M1YdrG4mt+8P366krHqsppXq/eScbhQQCPTJxFkXeG1tgSgpTLu6zc7nuNjRT0+OeOHXKGUfz40QlT4qqIbpjc6u9/7lEmqoJE03l2947xy88cpWNk5j2JuStS8omrb1PUnPzlzkfmk0Wdps6u6SE+dflNrtTVk3J7GA7HmPX4EFLykQ0isQE77l47neLBoz20DCc4v6NxVQJcm2rcf1LgHy+x5fuTcFMszTdtu19Cw/bE2/2Dqfnjs20e+j9Qza6/GUU1/pn46e4QLCEqEqRch4AFNKObBdv7tPS9LWkalhCMVYdx6QZbBycoO7V5Gsm1YLmkt5tRn0/wxNhxwJajXCpssBmwKri9f4o5CMAnIKqCZx0LDAdQo0HchLwE/b2fU1K4eFdpICKL1JLjZaWNXzHPbkhCXffwFB//0RmaJpOEsvn3NNQhsFkVv3jmh4wGYqRdXiQ2qVRDNkGolMdtlblc20jS48NhmnTPjG9oLkswU2DnlXFMReHE3la29E6s6vzrMu7XTyuFwJqjM7z+YwmbhnM24GEsFq4Yc7drRSvLcy4FxZK49PVRHJqKoKRp6/IbzV93JW/DTceDI0V2/t0Yjtzq3KISKGsq5i1hDKduoFVgkJKAqSjkPE5OdTZzbEc7l1rrmYwEyLld9lxSKFGXSNExHmd33yh3Xxkkls6imSaqtXrRn5XabSqCksNBb2M1R3d0cKm1noG6GGmfm7Kq4dINItkcbRMJdvWPcvhiH62TCZy6sa6VrURgrODa1VYwhq+27SRUzLNncmB+N5B0+zhZ38EH+u3saAsxvzqu5H9yWMuPRUOx5YufP7yDiViQjNdN9WyGyViQ//P/+8eVvuYCrMZIm0LBYRkEddtr8W6sm7rCLOodc81vrNv7TkMq2L726zdViJt+n/tRBfgVCCkQvPlHXfBahtQbfw8o9o86d/71yGU3O5B/Wo/Upc3qVZKQNiFtQdpCpC3I2L+TMSFl2cev/+36T8G6zgp20ySOzcx5/XdzdaWqCpICGkXhpyxVDlsjuDdIuOWpl07ZORSaiqkoCNPCVASK5D3ZvQts71nX7Ph8JYSQEnVOFbA6l8Y/XKQqm+Gl7t38w957+cNvf33D5J9VUxJKFdBMkw+8fmmeOnglrPlpLGkqI9URJqNBpsIBJqJBpsJBJiMBpiJBJqJBZkJ+Sks1QAjOdDax609/e03X3dU/xtP/4Q/X2lwAXt27hX/z658k5atMerIcDl4e5I/+n28Qya4+w101JGpm9Tc263HxO7/0Ub794P75vzlMk6/+wd/x4XcuLDAiphAM1UZ59tAuvvXI3YxWR5bcP15uqePVOVl3xZLcdW2YT716kiePnl0zacWtKGkqY1VhXjiwnX98YB9Xm2qX5ODvJ8bJ7ha+88BdaKbFrv5RPvfSu9x34Ro1s5k1l4atFH9eSQugoDmZioUoag62xseIFLJcqWokVrBj9YZQONHQScrlJVTKs3diYJE3wA45LdsKhJT8wrPHeHNvJ9F0nrbxOH/50cOYqrImFbHVLMRertuLzyjyZvUO3KZOX6CWhybPb5hS1kqw7cIyLZUSt1HeMGKPtcJpGDiXo+79iA+533PDGAfmjPXNr70K6y66vp3yOE1AZG5RIK//c2N+XbW5M4Dc3EIgY0JGQnbO8M/9L17MwbsFVuKjCVDmY1YvAolvg/ndR2si/PdPPchbezv4jT//IR96+yLP3r+LxukkBy/cWZY6CZQVjZ5oA8fruhgI2fNcUzrOgcletsVHcBs6x5s6eX7bXu7rv8xHLp3EsYEln8mQhzM7m9h7YQRPQefpJ/by+X84tuIibM3GfawqzG996ec4saV1/QP2PagDlUKsW2b2juCW9umaRtLvxVTE/O4953Ly6l1b+NqTD3Cus2mpM1WEpQpOdbfQOJ3kiWPn1t1MCcRDfr5/aBd///ABLrfUVdTSXgqGpnK6u4WzHU3cc3mAz794jIfO9OAvrq5MTpFyRTe1oSw/rKWAnMPNjDdIWdW4b/gygVKBN1u2cdfEAAmPnxc693JopIe+SC0HxhYTyOgrXMNhmjblrmlRncxypruJgYYY4UwedRMSsR4fP8Wwt5ovJZ8jrOc4Hu1aFYfDRkEAzhUM94d6T7NrcujONOgWqNJi+/Qy9L1PBuBjfqjgafyxgpj/Z+1wAGHV/mEJz2nWglPFRRTYlZDBiQtjw8Vb/vTj9xJJ5zh4YXB+oV5yahzd3f4eGHfBqy27+ctdjxAoF6gqpBHAmZp2XmnZzc/1HOVneo/RGZ9g/0iQ3qp6+mJ1fPntH25ItjyAaloU3A6utVbjmJd8Fay0rFuzcbeEsPWs/xkQNbzXmA75MVQVzTLIeFz83cMH+JOP3s9kZAkClxXgKZXZPjh+W/XmY1Vh/tunH+PlfdtIr8MTch2WqnB0RzsDdTF6mo7zxe+/ga+4Ms2vIi1chl5R5AfmsumXyKS/cQ5Jx+wEoVKehMePKRRqcql597e/XMRhGSjSYt94X0Xa17zDWTlZdA4uU7dpI6Vk66Bd7SGkpG18aVKZ24EAmvPT5DUXM64g9YXEHY25CylxLpPrIIA9EwM8delYxeOmInjmg7upmcmy69IogdziMXpuewOd/dOL1MGWw5XOGlxlk5aR+D+bMrU7hWnhxUIQkxunFSCBc92NfPX/+nu+9/AewB5bwWyRa023Jxm+HpiKwt9vu4+7J3r5xJW3iBUyCCSzbj/Pt+/nuY79PDp4Bs0yaUtMsX1qhGguu2Zv5HIIZIt0900x2BQDAYeP960qPLH+INlaYh83T8LvUcxESDmfhLBsC36MFi0zc8a9LCXP3Lub//7xh0kEvGvrz+vvlZJQrsC+a8PrWvdLYCIa5N9++RMc39qKod0SO7ylHeKmz1Vsz9zvE7EQf/KR+0l7Pfz7bzy7YhxeYHOUO02jIs2pFIKU27vsOSyh8E7TFlyGzpHhy2Sdbr639SBna9s4VdfOjulhBkM1SBQcpkFXYhztlnalXd5lmcgCpQKaZWAJuNhRz19/6B5+8dmjfOeRffzCD45taMIN2P3cG2jg+Yb9xEp2eOGpoTfvWMxdkRbBUmHZRVd6iftS1lSudNVyYk8rj712CWfZYDboIet3kwx52XFljETYR87rmt+vlJwag80xgpkCNTMZkiEvs0EPQkLrSBypKAw2RTm3o4nW4TgtIwk2Qmr4p7gBE8F3lK28RjMKki+Zp3HdJr+8ALzFMomgd37xbCoKPS01xFJrT0a9XUhg1u3n/pGLtGRukIh5swkOj13mR807MRWF2mya2mx66RPdBhQJXf3TNI3ZWgXu0uqe6TUbd3dZZ9vQxKpci9ORAKNV4UU8u75iia1Da5Nu7Ryz6fcs7EQLcZN7VldVFClRLQtDUSg6HGiWhUvXEcADZ3t4/jd+n0TATzxoyzomAr75/+NBH4N1MXoaaxYbrfcQ02E/ZU3lxJYWfu8zH2TWP2fYpcRhmPiKJaLpPN2jk1SnsoSyBSxFkAj46GmqYaQ6QsHpoOByYqgKVcksu/oq0/4uBwlMRoL8uy/+HO9sb1vohp9bNAXyJVon4xy+1M+u/lGi6TwuXSft8zAeDfL2zk5OdzUzE/RTdiz0/OQ9Lr75+D1UJzN88dk3KmoOXIfA3ln7ysXKxh3BlG9pljSAD/We4rG+s1gInKaBQPJLZ15BnrEFiV7s2Msvn3mFWC7DlaoG0i7fPO3kdUz5Q0sbdykJF3M4TRNdVUkEvGwZmkAKwWzAg64puPSNjz3HnQEenDxPR3YCkHfcLR8oFXAbOsUlsuYn/JWpaS1FUHZquEs61fEsmmnx3Ad2UXJqdA5M2clLpsU7+9roHJhGM0zePtjBeG2IgeYYv/R3R3n9cBeGqhKP+njq+6eZqgpwdkcjqaCH6pnN5z3454i9cpLt5gwWAg1rQxThAD71wkn+7198DEtRGGiI8dr+biaqQvzHP3pmQ86/FihSsjUxyoQvQllR53NGDEVlyhumITu7KQJVt0IAnlUa9etYs3FvjKf43T/57qre+1ePH+K/fu7DFNwLH/Ytw1P8w3/62lovjQQGq6q4WltHOJ/n7oE+JkIhrtbWo0jJwf5rDFTV8MOdu+mYnuKxS+fxlMs4TIuaZJaa5NI15y/u38ZvfvkpZgOr08q9E5gJBehrqOI//9JH59ulGSZbRiZ58Gwvjx+/yLahiYrSkhJIBH2c7G7hrZ0dnG9v4MiFPnxrkHW8jozHxZ99+AjHdrQviq+7dIP7z/Xy2VeOc+T8tSVlLn/+pXcZqY7wt4/czdP338VYLLQox+DPnjjClpFJHjl9ZVkXaqiYJ1TKM+tdzIYmhWA0GMNi6dwnVUrUWx5Ix01utF1TQxxr2sKEL0xLaoZQafGOYSQYw1zCuDssk2ghi2aZSAk1sxmO7W7n+PZWgrkizk0w7ADVpRQv1t/FhCeCIiUPTZ7DcYfq3K97VMLFLBOOaMX3DISrK0oWu8sGrSPXeAaIAAAgAElEQVQJaqczbLk2iWpJXGWdu08P0D3HqR1N5QnOKZwVXQ4GmmNYikJ0Nk8q6CGYKdLVP8211ipSATcj9WEOH+9jrD6MZw1u/J9i9Sji4JKIUUQjRoEtMs7tbo0E8NHXzxHOFHjl4BYSIR8NUxm+/O3X6RqZXvHzGw0hJbumB/n21iPEPQFa0tMoUjIaiPFq8262JkY4XdMxL8DUPTtGXe7HQzTofVW7IoXgWnUNV2rrOXKth5Lm4NWtOxiJRBmJxKjOZHAZOuF8jtb4NA7j/U37Olgb5fc+80H66qsAu7ztybfP8vkX32HH4PiycRcBxNI5Hj9xiQ+cvMRwdaRiWd1KMIXg2I4OvndkDyXnwp2yt1ji5198hy898zpV6eVdZgJonp7lX3/nZbYPjfO7n3+C8Vh4wXsSQR9/++hBtg9N0BBfWrM4WsgQy2cZiNQuOiaBKV+QzFym+3rQmEnw1KWjy76nL1K3JClKpJAlls/YLmQJTVNJPvvD4+RdDponk4xVh6iLp1HXcT+WQ6yU4QMTZ+ZDJHcy5g72oqs2m2IiUNm4jwRjZJ0eAuWVK09UU85nA5uKYKQhwnBDhDM7m7j79CDtgzOkQl7CqTwtIwkGm6ILEpjqptOc3t1M1udi1+WxDfl+P8VCXBNhZvDSLNMcUxpolSm8G1Cd4TAtHj5xlYdPXJ0PpLxXwVIpBCfqOhFIXmrdsyCNTZGSC1UtnKtunU/8+xdnX/ypcV8PhJQcGOwnlC/w6rYddE9NIBEc7O/jI+fOUJ+aRVc1IvkcdanUgt3Y+xEzIT8zIVvswlco8cvPv82vPPcWkUyeK6216JrKnt4bbvb++himqixY4V5qq0PXVHb0ja3LuGe8bv72kbuZjC5M4lMsi6deO8Wv/dNrRNYghuIyTJ44doGsx81/+NWPL/AEWIrCu9vaOLa9nY+/eWbJxUtVPk11PlU5visEWZeHgXA1eyc3J7M263AxEKleUtYxls9QnbMXJ4am8NaeDs51NvLE2+f5wZGd9kJnMsG95/o3tF0j3iq8ZpGO7NpCXhuFUDFHQybBmfr2xQeFIOd0c6WqgbsrVB8EMwU+9MoFlLkxeu/xPvw5e6cupCSYKfLU90/hy5Xw5svsOz/MZHUQYUncJZ2DpwbxFsqEUwXcJR3FkkRnc0igJr4MVfBPcVswUHBjYKBsigF+rzOgVGnxlZPPrroh1fnNibuvB+8r466rKucaWzjT3ELO6cJbLtM9NcEr23bgME2+9KOXcRolci4337j3fr742iuEimtTYFsLJDfqna8bIkuI+QS+66+56bgUgrNNzaQ9Hg5f68W5XFnX9c+aFg+dvsIXfvAW0UwOSxEk/R6KLgej1SE8RR13WWe8KkTrRByJzet/pbWWc12N1MXTc/HhtU1wEjjV3czre7oXJfHtuTbCZ195l0gmv+YHUJGSJ46d54UD23ll/7YFx9JeN6/t6ebBM1eJLbFo8OpluuPjvNq2k2KFzPiUy8ul6mb2TA5uyuTQG6u3BVIqGXcpqc/O0pixs+KlEGimxYOnexiqi3G+q4FfffotLrcu9jrcLgSScU+UxrytGuVcgWhnoxEu5ulITKKZBkYFQpuCw8Xxhk4OjF1b1C6nbtI0fmPHU3NTnFyREEnliaRujAdnxiSYuZGlXZ2wQ26+wg3+hu4+Wz/ivTYQP6nolrPkhIOzooaHraE7xqlwJyGAztR7s1i+XdyWcTeFoKw65qUWN/shcpgm9/de4VB/Lw7Dnrju6b/G/sH+OcYg23h95p23sYTY0HKESpjxB7hSX4+uqtzTd42Sw8HFhkachsGuEVs3+t32dqQQ7BscxGGZHG9rZ6C6mqpMZnW6z1LSGE/y6997jWgmN9/HliK42lLLWHWIj712FtWyGK8OUXA7aJhOcWZLE2/v7kC1JN51EtZYiuAbjx1alGTo1g0eOtPD1uHJdd1zAfgLJT729lle39O98PxCcHRHO1ORINElFg7Xy6pCpfxc8tbCd2Wdbs7XtvDkFRd+fWNkZq/DQnC6rp3EEupnLlOne2aMQMleVGqmRSBX5EcHuvGUdMoOjd7m6g0VlrgORVq8WruHd2PdKFLypZ7ncFt3Lt4skGybGaUqn2Gigq57SdU40djJzMXghuxwVhp7yzIICoFUlBucCQM6nC6tn6SmAtIeD25dx1khPHidZfLmeasSLE0gLBAVvG6GU6EUcOCLb+AYnzRYbV6chSBCCR86eaEh5U0biFkLzpXAt8oO7SlTDimUd7lstsAKUHSJe8RAy6yygYaE3rJNPrQMdLeKVMCZN+06f6DkU5GKwJVdmZ3UUkAqYmmq8WEdzhRthsNVQAwbG1LYsW7jXlZVjrZsYcYX4P7+y5hCoS6b3FQDf31H7DKWToYC5uuLNxv91VW83dnFBy5dQEjJm13dACT8fhymSfv0FFPBIKda2wjn8+SdTq7W1eErlSg4V8fDLYD7z/WyfXB8Qd+WHRoXO+r5xMsn8RbLaJakbiY1T004WBfj8Ll+Cm4H+jorAMaiId7ZvtjFWjOb5r7zvbcVM1alpHtkisaZJIN1sQXHJiNBehur7cXDEvdxx/QwrclpJn3hCtqdgos1zfTG6tk7MbChY3LaF+RUffuS9y9UzHNopGf+mqolufvSIFuGJvEXyggpOb69lS1rrBZZDTqz4/zPl783/9p1Bw072Ldh2/QIzakZJvzhiiGTvkgdR5u28JGeE3fkGa0EXVU519RE2aGxdWyccD5P6b+lSfyxSXU6g9M0ybpc6KpKWdOoSacpOJ2UNY2Cw0FtKoWuaST8PsK5PG5dt485neSdTupSKUqaxunODraNxalLpzEUhYTPhxSCqkyGrNvNhcZGQoU8XZNTOI2FXhYJFIMOJrcFcad0qvoyKKYkV+VCmBLvbJlMi4/eB2o48LejWA6Bo2BS9mqUfSreRHnTdS2uiChXRZTmOWW4m68mXs0jXl1bzsvEv44y9OsRLHdlY+ycMtj6m5NElzmv7lbIR5240wbOpIH4rWkKYQemU8E3XUJIKAY1dI+Kb6YMEuK7gmhFk5orWaSAfMTB6P4IimHR9lYc3a1iuBS8szqmJrA0Qdmv4ZnVEZZkcmcQw6lQfTWLJ7X4mVO+OgtfnV1TX2wE1m3ch0MxxoMRRkNROuOTFBxOajfZuP+4wWFadE9Nsn9wEF1RSHq9HOntYTBWRdLr5WJDE95ymbqUHX+dDIXYPjaGalnM+vyruoZiWXzm5eOLNhRO3eDwuT6GayMM1sfwlHSO72jDUBW2Dk7SNTLNK3dvQUhJ99Dq5W1vxpu7uyg6bxkiUlKbyLBtA4xTKFegZSqxyLjbFMXNfPidCyhLMGV59TIf6jnNyfpOm1TpFgwHq3izZTtbZsbwGrdHtXsdphCcqWvjbF0bFfeFUtIdH2f79Mj8nywBvc01PH94Bw7DRLMsvvSd1zflOZlwRzgT6UBXNJJOHz/f/wruOyz7GitkODJ8mbN1rZS0xQughMfPc937ODB+jYbMnZ/wAHJOJ29t6cZTLhPLZHGYJj/Yu4dwPk+hzcETZ87y7YMHCRYKVGUzVGcyPLdnN6aiEM3lCBYKHOvqpOB0MhPw8/NvHeVkWytX6+upymT44LnzFJwORmIx2qdn7CqfWIyj3V00JRIcvFZgKhjkrS3ddE5N0ZSYrbi7z9S4GTgUI9afIzKax7Iko3siZKtctB2bQUgwXArDByJ4Z8sEJ4pceKIBT7KML16i5UQCZROdl3b5m0mE0iL57fVAFCWiJMFVmSthJUgBmXo38U4/M91+Dn2tn3inj7G9Idwpg44fTZOtcTF4XwxPQqftrTiKbpFq9OCfLNpcDPVueh+roezVCA3nyUec9H6gBlfGIDycRwoYPRAhMFHEldapO5dm8HAUR9HElyjPG3cJ9FXXcKW+jtpUml0jwxSdTvqqq+mrqeGhy5cZisWI5HKUNQ1FSqrTaV7YtZvdw0PUpewF0yvbt1OXSuLWdfYOD6+JW3/dxt1fLjEeCDPr8XG+rpn2xObqlksAAWMPhJnZ5aPlhQThnrXF0ze6fULK+Qxd1bLYOTbKt+++BykEnzl2lN66Wt7u6sZQFASSfYMDfP3hR/CWy2wfW10Gb1Uqy9bhiQV/UyzJXVeH2dszQlnTcJftBKJPvXgCgGgqR3UySyyZteuPc8U1cZmD3d/HtrctyghXLMmWkQl8q6SLXQ7eYpnqZOUa5J7GajvZbpl2f7D3NN/c+yB90bpFxwxV5bvb7+HBgQvsmhq67XsvsQ3TP+w8QtJduVxSsyw+eeGtBTz0uqoyVBvBWyqzq3eUd3a1YwmxKbvWmmKKI9OXsITghfr9mIqKvMNxdwF89MoJvr3jXkaCsYq79+ONXfzj9sN88cQLGyausRb4SyVqUmnqUklaZmY439zE211dxLJZcm4X9/T1MRKL8vk336Iqk0FIyXg4zEOXL9M6E2c6GCDp9XK49xrP795FT20tsz4ftakU9/b0opkmobw9N1338JiqSm9tLe3T0zgsi2g2S2Nilp0jI/iLlRnevLNlQmMF6i6mcOQNdLdKPuxg9K4IVX1Z/NMlxneGcOUMGk+PMHA4xuDBGN5kmcBUkbpLadzZzevfMCUcSEaxQ1Rd3B77Yv3fp4i9nMMIKBgRFT2qknjQy8yH/EjnakKYNvNlps7NwJEY+/9qiNlWL+HhArUX0yimJNHhwz9Zoun4LI687XJ3p3SKYfs+je2PEB7MY2kKUsDQPVGGD0ZwZQ3yMSe+KduoR/tyzHT7UQ2Jf7qMf7JIaOiGR0FXVa7V1OAvluianMRhWaRVlYsNjby2dRt3DQ0xFKvCEgp5pxNVWtSk05xqbcVbLtE0O8ubXd28vH0nwWKe1nic7slJ/KXVz7vrNu7VuTSfPvM27zZ30ZyaYefE+pjP1oJMi4upfX5cKRPvZBnTpRDf6cNyCsI9eZxpk9ntPgyXIHo5j5Y3md3moxxQiVzO40pu7ES3Z3iI3SM2X7UA7hocZOfoKEiJ0zRpTsS57+pVVMtCM21Zjd/+7j/OJVmZq1qF3dUzvMgwC8A3X7t7Y1faMLOwfKxxOjn//rVCV1WuNi8Wg1GlRdtEfEP60WGaS+YDjMUWkx/dCp9e5FdPvMhvP/rZiglcM94gv/vgJ/jqs/+D6nz6ttpccDj50/2Pcbyxa8lEuv1j13hg8OKCPytSEs3kMScVRmsiTIcDWKqCamx8/fmsy8/lYLNNOuQOv2detKp8ml85+RK/+9AnMcVir0pZc/Bn+x8hXMzy1MWjeJYTdFknJHaMX1c1vHppwWJKsyx8pSJlzVbuc5gWB/v7efTCRZyGQShvEzCF87n5yfT63wPFIimPByEluqpiaBpO08Bh2qRSoUJhLqFVRVdVci4XpqLQEp/hF958k28eOULH1BS+UhmnoaOrWkVtbgGoZRNVtzAdCghB70O1OAsmTScTqLo9J3iSOlJAtsaFqku6Xpui840pHEUL5ypVKdeLZpmmUd6U+HibgWLHrIU2OzcfXCfWVCH+qG9Vxr0QcTC6L0zd2RQTO4OARFh23oJUBAg7d8FwqnOvBZYCpkNB96hYDoFStmzDPrencRRNul6epvXtOI6CydA9EbSyhaNgIiybwMuZM7AcAqmKeTlegR2S01UVSxHoisprW7dTm06zc3QUdY5kLe3xkHZ7CBVyuHSdr7z0Ij/ctZuzzc24DINHL13g/qtX8Og6vjUYdrgN4552ecg7nXz08gn6orVM+YM0pzaHN/s6FB2caRPPVBm1LClFNK58rpau706j6DCzx0/fz1ThTJuUIg68EyWufrYO/0iRbJOLjqdnUDdQC1mB+Zri61T+N+cDqFLi0RdOXMtmx1dA62SCKW8In14iuER9sMRmTJLY4h23vl4PJqJBcm7XIkMmEUxFgrxwYPu6znszyprKRLQyT34y4F1R3U0ADw5e5OH+87zUuQd5a925EFyqbuI/P/xp/pejz9A+O7nmHbMEpr1B/mHXEf5u131LJkHG8hl+7d3nFi3YNNNi57VxpsM+LrXX88jxy2ibYNjBlqH1GCUMReXTg6/hNkrviYEXwEd6TvJqx27eaNlecTFkKhq/f++TJDwBfubyOzSmExU5/NcKXVGIe4MMhGs4Vd9BwuPjV0+8SF1u4cJ39/AITx/YTySXZ+/QEFfravnh7l3UpVI8euGivdu66T61zsTnCZrqkklq02le37YVf7FI5+QUWZcb99yzbgnBQFUVaY+HC02NtMzMMOv3cbyjg/bpabylMu5ymdp0mnc6O6hLpYhlFxNsOQsmvpkSQ3dHCU4WiA1kuXZfNYopcad1nDmDrh9NEh4rkKlx03wywelPNHP58XrqLqVoPJvcNMbd+ZySDb7A/Ei5fto1DAlHwcSbKDOxO0hwtIBWtKjqzdL3QBWzrV52Pj1G9dUsVx+v5dKTdWx9bhLDqTDb6sVyCDK1bprfneXcJxpQdUnNxTR1F9Kc/VQTVz9YS8OpJL54GSkEjrxJYKKEMCS1F9JcerIed1Kn8bQ9zjTTpHtynJe372A0GuVjp07RkpjhdEsrmmXiKZfZMTrK9/fehdM0aI3PMOvz8aNt2/Hodii3KZHgG/fex3O797J3eJBdIyNrep7Xbdw1y+JSTRPn61qwhOCeod71nmpVEIA7ruMbL+EfLqEVLcqWJNRfoOFNu0PTbW78oyWCA0V84yVSHR78oyUiV/O4Esb7siYmmC9yor6LSCHLlsQY0aLtJpx1+zEUlVghg65onKjvxGXobE2M4jJ0TtR1oiDZGh8lXMqRc7hJuP3U5pK4LIOky4uuaLZoSj65KKYfD/kqJuIZmsrXn3yArz/5wKZ+76JDW9W04S0X+fzZ1xiM1NATa1j8BiF4u2UrWaebT1x8mwcHLhIoF1Y1FEqqxtnaNr67/RAvde7GWkIxzK2X+fSFN9k+PbrovKaicLGjjuGaCP7CJhtbCTOuIJo0uRRspqaYQrlDDHW3wm3ofOn4D5n0hyvfF0BXHXxj70Ocq23l0b5zHBq5Qmtyek0J6xIoag4GwjX0Revoi9TSG63jfG0rCW+APRP9mMricdycSPCVF1+yQ2vAJ949jqkoqJat0f2zJ04ueP8Hz5+f/12TkiM9vRzqvYYi7VjzXUM31O5UKdk6McHWiRvhtKpslo6paRTLmv9+913t4d6e3iU9eKoh6XhrhvajMwjL3qXXXEnb2fNz7wlOLXTpH/zGAFIR9q5ylX34kwJH0aLr5bn8Imn3UWQwz76hoblkbHBlTfZ/075X17PgD/7FQj6Me/5kwN7lzx0/8BeDFfs0MGXvpEMjBQ59rW9BVr0AWhIJfunNN+Zf393fz/6BgfkxI4Ffe+UlwN4oSuCzR98GmN+EfOGN1zCFMq8dvxasy7hfrm7gXF0LabeHU40ddMYn2DO+OYQhy0KCuGllV3MyQ8+nasnVWYSuFag9niGxzUeuwYV3ooxSfv8RWQTyRYqqg0uxZgZDNdw3fAmnZfBK6248Rplt8RFihQxnatqpzyZoSU9jCoWzNe1ECxma0nFUafFPXfdQl5vlcqyJRwfP8o2dD9OWmqIhm6Amn+LWJX7a66mYqHanYKoKq1mNKcCO6RE+d/Z1/uDQR0hUoKQ1FZUz9e0Mh6t4rnsfDw1c4J6Rnrm+WTwm4h4/Z+vaeLl9F2fq2hkPRJfsC8WyeHDgAj9z+R1cFVXRJMmAl3d2tVEXT+MwTB5998qK32s96PfXEdZztOUmebV2D2XVgWYsq7K+aRDA1ulRvnDyJf7fw09WLI0D20X/bmMXl6sb+e72e+hMTLB7cpC25BR12eScAI+JrmoUNCcFh5Os082EP8JIKMZIMMaEP0LC4yfhDZB2eSoa80rtu/neC1hz6exavECVzn9rG5b63M1znFihifb733/z3EahUtnarVTWK5W2CVgwHa7Up7feowXXvuX1rWPu5mez0ngQsG6P1rqMe3U2zZ6JQSSCe4avoVomsdzmizOoJYvml2cRc5sRT1xn+1+Mzx8PDhTZ/UcjSEXgyJkoumTPH44gNfv1+xGaaeEyLfZMDZB3uBgPRBkNRLkUa8ZhmSiWxeOp0zRkE2yNj8zXD9fnZmlLTlKfS/BO/RYuVTUz6QvjsAzunugl6fbx4PkLeIwyosIeueTUVox5byrWcGmnafCRqyeY8Qb4830fIO9wLg4nCEHcG+Stlu2caOjCZeiESnka0gmCpTyqtMg53Uz6w0x7gxQ1B0XNgbWUoZASgWT35CBfOv5DarOpik2WQuAtlPnYa+foHp6aZ2DbaJyOdDDmjTLpiVBUHSSdPpR1rPY3Ek7L5APXzpJy+fjjgx+0ExErjSkhyLi8XKny0BNr4KWOPWiWiSqtBbscmzBKYAmBqSiYQsFQVHuc3uZYlQqYHvscig6ifOO6qHbcFsXOyBYSsEAxJJjrdwhKYceU5fVzM3cuC4QpEcbKhmjJc1dq99wBYYIwpH2d1ZxLAdM9F7M2QCltft9sNqQAqc31vZjr+7m2z/f9e93I28S6jHu0kCVczNEfrWUsGCRcyN4Rqlc7yeSmlY8FjsKN6wppx+Rvfr8zs/D1+xGWEKTcPoqag1A6R1Uuzf0jF9k1PUikmMVl6riNMgl3gKLmwGOU8ehFEh4/Bc1FoFxg/8Q1Do1dwWOUiBRzOE0Dt1GeVzm6Fbqqro5k58cETtPgV0+8hETw13seJOWpnNEuhaDocFLUHKTcXoZCVZVPuMJ3V6XF7slBfvuVb9GanF5ybClS4i7rvLmnk+PbW9BMi1//9mtr+Garg0BSU0xRU7RDVAfivTjucBlcJTgsk8+cfwOvXuKP7vkQ4/7I0n0rbMNdUpxsLPXQykjd7eHsXzYgnYLab2fo/C/TaGmLfIeDxIM+Eo/5yHc40cMKjqSFt6dM9JUcsVdzuAd1lDXsHSRghBWyO10kHvSSPOSl1KhhBhTUnMQ5rhM6WST2Ug7/+RKOhLlqIy8ByyPIdzhJ3uth9gEv+U4nekRFmBJHwiR4pkT01RzBd4u4xldue3qvm3N/2YjpE1Q9m6X7t6dxJEwKbQ5mH/Ay87iffJcTPaLgSFl4r5WJvpIn+moOT195TX2z2ZAqlGs0UvvdJB70kj7gQa9SsZwCLWPhHtSJvJkn8noe39UyanZ1C2RTEUxUBYmH/Qgp6RqaxrMOka6NxLqMuwAmAmF+1LGDnRPDXKuqw1BVDg4vppW807j1+u91e24XAmhOzzDhC6NaFq3padxGmZda93C2po29U/3UZWfZGh/lbE0bs24/nmyCHTMjvNPQzbQ3yPb4CGP+KOeq22jIxgkXc+yb7Fs2W181rYpa8appceR8L8F85fKdjYLTMNbkJrVdXJIvnHqZSDHH3+6+n/5IzeIku/kPrH9k+MpFHhy4wK+9+zwtqZkV3x9N5fj8c+8QXQMH/1px1+zG8tRvFK738seuvEu0kOGv9j7M6fr2inK9PxYQgnKNihFUSR72MvRrEbJ73AsmknKtQrlWI3m/l6mPB2j+2ixVL2RRVkmnkN3lYvQXQ8Q/6McIL/QMWS7Qoyq5HS4mngpS9UKWhr9KEThTXNFISkCvVhn7XIipnw1QaHXALXkiZkCl2Opk+sN+gieLNHwzSdXzWYS+wlwp7L7RqzT0iEpmr5uhr0RI7/cs7JsahXKNRvJeL5M/N9c3P8iilt77UIFUIf6Ij7FfCJE65FlEllP22G1P3+1m/DNBap7O0PCNFO7R5ausJDBaE+Z3vvJRju3tQEjJb339OT7z7PH3jKgJbiOh7jp1Ym02yazXj7nUJPoTBKtawWrSUC/riMIcd3yLiv6IB+ff5xD51d9Ic58TUhZK3wq7Kym5a7JvfnBd///Ja8eRCMQceUR7apK21NT868ZsnJ+9mph//fjAaSwh5t2cjw+cWfaybt2oaPwdhskXn32T3evQhV8LhJQ49bXvPB2Wyccvv0Pb7BTf2n0fr7fu2DBDIqSkJTnNpy68xYd6ThErrByKsoRgpDZCMFvYVOP+4w4BHBm6QlMqzvPd+/ju9kOVWew2ClJSk0txz2gv/vLaFqLlWo2pn/Ez/ukQpSYNpWDhGdRxJEwst6DQ4kCv0kBAdreL/t+sQqpQ80x2xZh48rCH/v89Rvput/3dLXDMGHiGdZSCxPQrFFodGGEFy6cw9WSAQruTrv84TeBMcdkdvB5Tufbvq5j5kH/ecAlD4r1WRps1QRGU6jWKTRrSKUgd9lBo1SjVazT9yeoy68tVKtMf9TP1sQCFDgdK0cI9pOOMm1hOQbHZQbnG7pvcNid9v1GF5YC672Tmw6nvFSaeCjL0P0UptszNB6bEPaLjmjAQJughhUKHE8ujUK53MPqFMIVOJ1v+3RTO+NKNLztU/uznjnBmaxP7LwzR3xTjz3/2XprHZ3ng5NKJ5hLI+Nyc3tbM5Y46ZsI+Sk6NWDLHvaf7uOvKMA5j/R7xdRv3htQse8cHeKN9O/WZWXZMLs4U3kwsoDpcw2ckt0Ef7VewujQIK6hny4i4hZiykM0qOAUyL+H/Z++9w+zKyjPf39rh5Fg5qaRSqZRzaqnVOUDT3TTtbhhyAwYDNuAx2IyNx8O9XDPXd8AzGMOMScZ2Aw0d6Rzo3OpWt1o5h5KqVDnXOXXy2WnNH7uiVFKVUgO+8z6PHuno7L32Pmuvvb61vu/93i8osJs0lB4b0eeAT+A0aYhhB9Ftg0fgLNCw1nlQDxjQMrtLT+eRmBwrn83n81lFRrJ59GkEZKQAS1GIZS9fQZ6Lhce2WNd9ksbhHrbPWci/rLmBE6U1E70xW4MyKc0xYBa5+9Cb3HF0B3WpQTz27DQThIRUyM99t2ygNJnFa1r8558+83vvUboQKEjmjgxwz96Xub7lAE8tWsfjizcy4v9gqeUAACAASURBVAue/7OZjMnpqFLSONzLHUffZnPHMSqyIwTP07jnGjx0fNaVQS17NkPdvyTxdlsIU4IiMOMKfXdF6PlgBDukUKjV6P54jOAxg+AxY3ruBZCfp4/udn0AeDtN6n6aIL41j5axwQGpCuyQwsBtIdo/H0d6FNIrvJz4ehkrP9mNlpp+snc0aPuzEgZuDbtxZFsS35pjzg8T+DtM994FOF6F1CofHV+Ik13owajWaf98Cb5Oi7JnMjOOy0K9TuenY9h+hZKXs8z5URJf52j7CphRlf47wvR8NIoVUTCqVLo/FiN41CB84LeTmimBkY0+Tn2lFKNKAymJ7Ckw54cJgkcNlKID0o2/GxUqXZ+MMXBbGOlRGLo+SOtXS2n6m36UafYaEvjnu6/imWuWc83OZr7wy1c5UV/ON/7kdn7wwWsoSWVZeqLnrL+7eW4F3/7DmxkJ+4mnciChvzTME9ev5Ov/9BRX7TpxwX12wcbdUlUGAxE0x2YgEGEgFLng+tmzhSsOoWArCrpl88Ta1dQkkqxqb0cfZQVbqoIjBJ7RHFVHCAxNQ7dt+qIRnli3hs++8DLOpLSX84HTpCPSBoW/iuL7etLdwY+FVgIC8xY/AOZdQbz/lMJe60XGFJx5GvrPMshGHXuJjjNHQzlx/jGZvKLjcaxLnl96OspHMujT7JylEAzGZied+9uEgqSkkOXdzXu4tvUQ2+cs5MlF6zlSXkdW92KqGqai4ghlSuU+RUo0x0Z3bLyWQW1qmBtb9vOuE3spy6XGPR+zheo4vHvbIW7afuSsOvmK41CVSeCz3BxaRxFuWGQUoeK5U/dyHg9e0zxj8SaBnNeDbtnj+gqSscqFbr30qkwSz1lU4qKF3HnJXc4GAvBbJk3DPfzpW09xz95XeG3eMl6cv5KTJZXkdS+momIpKrbiPpuxHlekHE1dG31Gto3m2Hgtk6ahbjZ1HmdTx3Fq0sPojo04z2c1BukR2LpCxWNpFv1l/zi5bgyeXmj41iBKQdLxRzGkR5Ba62PoxiD+U+a0LmjHJ+i7K0xyUwAU8HW6OunR7fkziG0SqP/+ML4Ok+PfrED6FNKrfHR/OEL9D8+sFS6BgdvCDLwnhNTc3Xrlr1PM+/shPAP2eL+7x9r4OkyCx4uc+EY5Ixv8WHGFti+VEDxWJNBy7jlJ6gJbUyh7Nsvir/SiFE7vG5t5/zCEmnNo/5MSHL8gs9zL4LtDBE8YqPl33k1tlqqc+nIpRqUKtqT0lRyNfzuAr90cT5kbg7fbYsH/PYA24tD9kSiMiuiUXhuk7MXsGW1vX9nAKxsX8t6X9/HF+14hkikwr2uQkmSWv//UzTxz1TLqehNEM2cuMAXQ0DnI1//XU8zvGCCcK2IrgievX8l37rmRh961liv3nLzgGh4XbNy7I3EOV9WxtK9ztJjL5SUPSCAZCLCnYa6bV3+yhaKuc6iulqFQkHWn2vAXDXY2uvKeS7u6KUul2bZoAUVNZ1F3D+qoqlR7WSl5j4cFvX3nTQRUTpio24s4jRqyUkGcmnDXyLAATaC9lAcF7OUeZFxBe6mAvdaD06QjK1W0VwvYS3RE4fwemoPg4eoruX5wP7XFy6vLXZlIuXH102qm24rCkfrqy3rtSwkFCFgG17ce5LrWgwwEIhwtr6UtVkFfKEbKG6Aw6rb3WiYho0BFdoSa1BCLBrupSw1dtLjKcDRI85yK0cWDw4ZDU9NGSwpZ7nvoO9hC0FpRxnA4SFNPP7FsDkNT3bTE0cWooankvF68pkmwaFDQdV5dtpBNx04Sz+UxFYVUwI+tKngNk+dXL2PO4DCLu3oJ5wtu8aKaSsL5AnMHhvjZw//wW0uVE0isSsHV6YPc+dR2hv1hjpdW0xEtoz8YJeEPkdc9mKqG4jj4LQOfZRAwDUpzaWrSw9SkhqlND5+3630m6AM28747fIZhH793E+p+nGDoxiDZxR5QBEM3Bql6OIXaN9WFK3F3vEM3hpC6u6uu+VmS2JvTe78ELiu95JUcJVtzDN0cAgG9H4xS+/MR1OzUecOKKQzdFMSMu/H7QLNBzb0jeAfOdCWPMeZDxwxqfjZCrsGDWaGRr9cZvCVE3Y8S0+5Qp/TNoM3cfxw6w7BP7pvae0cYujFIerUPFEHi6gDVD4zgb39nSZ4SGL42QK7JFeTy9plU/zyJv216eyUAPelQ/mSaxJYA+UYPZonK0E1BSrZmz+BVbNrfygNf+fHUNiSsP9zOr776zzPeXySTp753mK6qOJmAF1NTSQV9BPMG/SVhJuUlnDcu2LgHjCKa46DbYykrl5ct7wjB4boaemJR1re0ols2Ugi8lsVwKERreTkZn49nVy0nWDRI+30s6epmR+N83rdzD6GCW4mtOxbjpWVLuf7QkQsqCSvLVeyrvci4gkhJ7EU6cq6KdYUXdWcRkXewrvHhzNXw7MhhqwLrGi9OlYb+ZM71Pmz2IitUlN7zC0INecK8EV/ChkQzcHmNu8eyWdLWw6F5NVOU4mxFcLS+ioKu4buAmPhvEwKoyKWoaEtB25HzOtcBnvIvY1AJoiB5T/4IFc6ZqmKnw1IV9i6sY+vqBTT0DNFaU8b6w23Txk4LHp23FjWSCAbQbYdFXT1sXzifoqZR1HU2HzvJW4vmo9k2/dEIH31tOyMBP4ORMI6iYAvBsdoqmqsr6SqNc9uufWR9Xg7U19JcXcG79x4iGQjw0oolNPQPUJFKo+fOHV6RwDZPA8f1cgCuLrbQaA2e94JAAo4qSFQHUS2HyEAeISW9jVFSZT6uvv8EXm+Rjb3NrMi2oRVtNNMhXeLD9KnEe7KYHhW9aKNaDkZAQzMcTK9Kco4fhhxCyUtTIAggvi2HPnDu0Iuecqh4Ik3rYrfwUWapF6Ncw9N3mraAApnFXnILXf1yNeNQ8euZ+RpayiayqzBq3AVGiUpmqZfojqkLmdwCD9nFXlAEwpREd+YJHZ053yD+Ro7gCYNkuYoTUBhZ66eyLI2399zvdXRH3o1Tn+ve0w4VT2RIr3LJiNlFHopVGr72d7bWgR0ULiM+roCU+FvNM/pvOgSPGfi6TfKNHlAF+Xk6xSodf7u7KBiKBnjhyiX0lka564U91PUmzvhdg7EgL2xeQl9phLue30Nd39RjLFXhrZUNPPTutXRVxPAZJqotyfk9jIT8lCbP9BScDy7YuMtRctaJ0koEUJ4ZoTZ1+QyOxC3CUJNIsqinDweXUT1naIi0z4+tKAyFgmxoaWVlWweRfJ7eWJTykTRNvW4Fs06PB1NTyfh9eGcZM50M0WOjP5VHekA5ZELSQQjw/CCNGJGIEQf19SKyXEHdZyDaLLSEjVOlojoGSqsFvTaiy00AFT2zM45ZzceTFRvYG2mg01fK9xpuJ2C7L+/Hul5haaaD7zTcwe39O1mS6eSN+GJeKFvFR7peoynXw6slSzkZqOZTnS9iCYW9kfm8WrKMhCfE3NwANw/uZV7+zMI/Ww6e5JFr1mAztd56T2mUo/VVrD7ZyTsJ83MRnBvcsIf6dA71gcw4sfG82vlCFOcaN+6pPp5FfSjrVqM6BySC532LaNYrUKTDxmL7rIy7FAKPabPyRBe+oknznApsRUWZRobYa1qUZLJEczmauvvoKCshEQqy8XgLT61fybHaKhKhIJuOneTQnFoGw0FK01kSwQAFXcdRBN0lcSxVpTcWIVQools29YkhOktL6IlFqUqmqEqOsKSzZ9bFf3Z65/CsfykA9VaS+dbQtNoIM8EIaHQsiTNcG2T5q11UtqapbEmRKfECsOdd9cw9OETzhkrqjiYo7crStTBGIaTTtTBGWWeGTNxL/cFhjl9RSePufnbdOo+KUykMv0Yoeenkr0MHi7MSvYq9kYOvusZd+hSyi7yEDhWnbLakJkiv8bmxcCDQauIZnnlhL0zwDFhgS1AFUhfkGjxTjJMECrUahRp3KleKDpFd+VmR17SkQ+hwkZENfqQO+bk6hRptRuMeOmqgzOK9i22bCNNKr0J2oZfozsJ5ScpeLIxyjfw8N2tgnFyYnfkGtLSDmnLGPZdmXMUoVceNezIS4LEbVnFgYR2b9rdQ13um7UtEAjx64yoONdWyaV8LdX1Tj0kHvPzbnZtpmVPGn/zyVZY3d+ExbfYtquP7H7nuon/7hReOyaT4+K5X3fQLVbvkbrHToUrJ/P4BfrFlMzsaG/joG2+5MTfbcQUvHIdrjh7nn26+no7SUt697wBLu3p4Yu0avnX7e7jm6HHqhodZ0NvP1UeP8cbCJm7du49gcfarfVGQiFNTB74YcmBoUq79sAPDkwbPiEQdmXRORqIePb8Qhu5YLE+3IaSkNVDJdUMHqC0MAVBbGEICQ3qYk4EqFmc62RlbwLFgLUdCdSzI9XAwPBdVSmwUtpYs5Re117F6pJX1yRPsjTTwvXm38cVTT9OQ75ti4K/Z30ywYJAK+qfcT29JhFdWL2JFa9dF1XQ/X6iPZlGOGlifjuDM17lQAT314QzKYQPrj8I48y68ndlAt2yuONiKPbqDv/X1A2hnqS+gOQ6BQpG81zNaAlSg2Q7+0ap/UrhyxCXpLKFCcbxUpCIlpqaiOpLSdIZ98+Zw+859BAtFvKZFxUiakUAAS1XxmiZ+w8BjWe+4yI2jCPJhna5FMebvHjhjeZCLeuhcEnfllasDlHVmMH0KfY0RgsNFlr3Wze4vrKCsI0Oq3E8oUcTWBL0Loqx9pn3aa14QJPi6TcQs1t6BUybCAOmuT8g36JzuSZUaZJZMlL/NLPPy5tsNs7oVx6NMMIAVME9LnZM6FKt0nFEBHmFAYJZcHgEEjhsIUyJ1gVGpYZbO8DJI8PSaKLOo0eFvM1GKcvzeCnN1pDKzyt6lhFmiUqx2zZxUofcDEQZumx1nyApNUK8dv4ITuLRvi60qDJSEKElmWdzSy5zeBP0lYd5aNZ+hWIiy39bOvS1eRme0DM2xeWbxGu44vJMNHRfO7JsNGvoH+OtHnwBcck3N8Jmrpf/r4cewhRh3uX/9kcfGNaMBPvfiywAs7ZpdydXfBXgci+XpdhwUvLbJilQbC3MT9+8gmJsfoMcbxxAarf5Krkoc5mSwClNo9HjjXDd8iIQnxPPla7h+8AAf6HkDj7TYkjjC9+fdxstlK6jtHMIrJ2a0SDbPna/v5d53b55yP0Vd47VVTdy06zDLTp2dCXqpIfpsRF66WQcX006vjSgWEb3TC91canhNC0y4cr+bGnGu/lrU3ce9112JFIKb9x2iubqChzavx28YLO7s5UidQLdtKpMpVEeyb94c+mIRXli1lPdt38NgJMRIwM9LK5dQMZKmNJMhUDQoyWQJFov4DZNoNs8za1fywdffpiJ1+ZUlx3B0SzWqJZlzOIFqOVgehVS5j1zEQyGoUd6epmNZCfP2D5GoDNDdFGOkPMDCN3sZrA+jGzbVzUl2vHcey17tRgrB+qfa6G6KsvVDTdzxD+dO75w1pHTj2rNZt9oSLW1jet2p1IyemYsjFdfIjH/WBWbp+U+9UoD0Th09UhNYYWWcFyMciZ6c/fuhD9vjxtbxCxyfmFDKm/4u0LJyVrtvYUu0lIPhd/vEjKjvuPCI4xPYgdFnIoT7G/3nny81riR4CeEvmGze08JjN67my3/1ASLZAraisP5QG8tOdl+0/PeFs+UVlY5YKRmvj1uP7gGYYVBcHMbanVGLeXQnMxkXElv/XcO5xHkEMDffz75IA6cCLnFrU+I4v6q5ig5/KTnVx/xsL1nVy4Anyvxc77gRLzPSNGZ7ORCZi6FoU+prC+CurXt4+orlDMYmabYLwYGGWn5540b+4v7niWVy78g7OxtqifQJnCu8yDoNbFCaTcSe4lR97lm0c6lwvv1SnRzhLx5/FqQ71m/ftR9LEaijRSsqDrvG+K7tbmGTuuEE61pcgl7a52Uk4OeGA0doLy8h5/Vw7cFjLiu3f0Js54YDR7j+0NHLJoV7NtQcS3BqZRn+lEEoUaQY1MiU+PCnTTJxL3MPDKPakvJTaYKJIqWdGdKlXnIxL1UnR0BC045+jmyppu5YAktX6G6KYfg1Fm/rmfkGZosxGdJZHism7WJPF0YBQAjsSQZFS9h4u8+fgKwYEn1wqjtBKqcZfMl51dCYfKzUBM5MpVVH84lnO66FMblvxOgi5J0bd1N+kyPx9FnoswiJnA7PgI2aubR2JFAw+MRjbzK3Z5iuyhjeosmaox2sOdzBb7YsJREJnDXDZja4YOM+NznIcCBELJ8jns9gKer/L3N3fxcgkDTk+tkWX8KhcD3VxQQ1hWEEcDhUj88xiJlZknoQCwXPpEphChKPY5JXPJzOfRXAgs5+7tq6h5/cdhWOMjFB2arC41euRLdsvvLA80TyFycYKnE9Au2VJVQk0heWR6+D+WdRnC0+lJMmeAT2R8Oo96bQHrw4F9c7idNDHdosjXCgaLDpWAvJUICydIamnr5p30kF4B027AKoOpmisiWFmGQc1j471Z1e2j31OW16pGX8+IH6EEe2VLP54ZOolkS1bJZs63H5P5fy9ygCqc9ysyJA6hPvhTJdBoxkihs7fKDAvL8fOn/deGc0Bj/58o6r9T7lMH32M/EUY27LKQuVaSGEq8nO7Az85DrsSlFOq3p5WeFIhD2qg2BD+bNZKh9JnXczwpD4ui8tgVgAVUNpPvTMzjO+u/Oli/dCXbBx91kmV7SfQCBdoyB+/6Vef9cxnvE7TUdHrBw+x+BoqI752T78jkGZkeJIqI4yI4UuLTRpo0t71JC7zVgo5FQfYSuPMo2vzW+Y3Pn6XnYtrGfXwrlT0uKyfh/3X7+errIYX3rkJZad6hn3mpxrLMhJf0sh6ItHeHnNIh6/ciU+w+I//eq5CzLu1g1+7PeH8HxtCGVnERSwPhLG+mIUZY9xQboC0+F3dZyrUlI/NEz90KUjll0qjD3zyQbNECqvly3jhoH9ALxZsoiD0Xn8Uetz48dMrugV7cuz8oVOQonCxN5PMr67kcC20iWsTrYQtIv0+OI8W7mOW/p2UV1IMOQJ8UzVejYNHWNB9hw7fQF2UJmVi0eqAisyMSK09DS7QinRkpPeLUUQPlS8JLFnYUnU9ATxS6oCK6bALB0ZVlRBjq5N1IJEnSa9beoFcd3cCjO65t2+meSxSNvv5KYdALUoUfNyXIpEKTiED77TlQt+O7hg4w7urg9GldDeeW2CGeFWj3KLUTiKcAVLlIl/T/5uOByYtgqaoWn0xyMT5CXHzVV2/5YojoM6/m/3u8k7k0uJoF1AlxadvjLm5F03q8ex0HDwOwZhK8++yDxuGNyP1zGpLCZ5tXQ5NwwewONYRK0cC3I97Ik2sDDbRcgu0O4r40i4jk2J43inKTQigEUdfXzmqTfoi0foLJ9a+KOoa7y4djE7Fs/jur3Hec/2gyxt68FrmKP94riVvEb72lJVsn4vp6pK2bNgDtuWN9JcWzFehW5JWy+WemEags6NfkSfjfJCftyIaP+awvpsBOdq3yU07nJ0UTu6OBIeikLDRrieEGwCjoGH2ZVbdT2dgoLQKAoNExVnlI+uINFx8EoTvzTPW11x7B7zQscYvUcH4SoW4qDh4JUWXnnxwkgSyAudlPCNSyOHZJGgNLCEyslQFTnVS2Oml4iV42h4DqXGxC5qQ6KZbaVLztp2nxKn3SmnypugWg7T740x6IlQYmRoyPXS542hSAd1NC23qpAgZmawRqWxS4wMVYUE5ixKwhbqdFcQZoYhU6jX3dz1UfhbzTPmQmFB4KRBeo2boZFZ4sHxCNTz1LmYDsLELf6SlzgB1+OQm+8hdGRmorAEck2e8ViyPuBm98yEQo02q/svzNFxJoUM/G3mO0qmAzcjQB+wKMzRkSpkF3mRKu+MFK5w1Rx+W7go4/67jpdXL6S7LEbW5yHn9ZDzecj6vKN/Jj7nfB6SIT+pgO+MNg7Pq+LTX72HUL5IoGAQKBoECkWCBYNgwSBQHP3/gkmwUGRhZx9rm9svC4u8pjDMqtQpHqnazLFgLZq0uXlwLw35fryOic8xKSgeKotJdMeipjhMvydGVTGBJm0iVo7b+nbyi9pr+OHcdxMzs/R4Syg3UlyVOIwupx/xArhp1xGGIwG+/wfX01MSnTDwo3+ngn4e37KKx7esIpbOUTeQIJwrECwWsRSVvFcn6/MyFA3RHwthXwaKuixTEQOn7Q4yEpGXyJlYwOcBbdT0DihBtnvnsd0zl1a9lJTw4ZcmNfYI64wONhVP0WgNntMgGyh0qTFatVIO65Wc1MvoVaJkR70rfmlQYWdosgbYUGxnmdlLRBZmNWVYKLRopRzSqzioV3NKK2VE8VEQGqqUhGWBCifDfHOIVWYXa41OgvLCcsUlMCJ8PBRczVO+pRSEzlx7mE9kdnCF0cbRcB2HIvXU5YfYUdLE1YOHiFpZnqi+glUjpwDOqYZnC4WtZcuImllKjTRJPcgbpUtYnO7kiZqNfOrU8wTtAvui81mU7sLnmJw+tQqYdQwzs9SL4xEzssKTV0xkkghDEjpWPMO4K5YksitP311hV742ppLc5Kf0lYtX9BS4ane+LpNck9eVl13ro/yZmXXu7ZBbmU6OSq37OqwZ0+DANZCOb2bjPnJa3wSPG+9oGhyAp98icMp0F1aKoDBHI9vkIXT0EmgizDCUTFWh6Pntmdh/18b9e3ffwP6G2gvTqx6FqWl0l8Vmffxdr+1mRUsX6mUot+l3TD7a9Sp7Ig2M6EG8jknQdlMQfbbBklQ70pbEzCwKkuXpdj7V+QJLMx0ooy79pZkOPtf+HEdCdeRUHytTbSzLtFNmnDsOpUrJB17ZTbBg8P07r+dEXcVZj02GAyTDgUv502cFkZbIutOMuFeAB5hFbutsoUubfjXMg4E1vO2txxQTr1FaqBxTfBzXytnhqeee7A7WGx1nbatVK+WHoS2c0MsxxZkLkLTwk1b8nNTL2OZt4N35o9yV20dMnjv1NC80nvMt4Te+xbRr8TMKO9kChggxpIY4olfRWwzTZA5ckHEf27E/FFzNs74lFBQPc6wEn8m8xVqjEwVJl7+U5ak2YmaWo+E6DEWjspAc32XPBFU6bBg+Tre/lJZgFauTJyk10szP9nIsXEdKD1CXHxp/Hy4WI1f4KVZrqCfMsy6krJCYklYVOlrE0z+Nt8aG8IEi/jaTfIMHFOi+J0Z4f3FW+e4zIdBiEDxikJvvQeowssFPboHHNabnQPIKP9kFHhACUXQIHyzg7Zv5flLrfBTqdLTk2bXiraCg/9bQ+OoqcMLA2/POCtgAaCmHyO4CQzcGsaIqhRpXic/flrgoKVzhMM7zKHinL0yV93lIRvzTfvdO4N+1cf/3iFIzzU1D+6f8nwT69CjScliWPIXfLtKvRTAdhS2Dh6mwUiTUIINamLnGIDEjw02D++nTIpRZaTKKjxElQK8epdIcIe5Mv6PQHIdb3zpI9dAIP7r9al5btRBTu7Q78DHN8wuBsrOAeWUMOV9DjFbbc671ISUo+y5dnM0UCr8MrqNdjVNtp9hSbKHWdqtqtWplvOFroF8N06yV85PQJkpHsjTY08fBy50MfWoYE4UyO8Mys5f51iAldhYF6FPDvOmdR7NWTkrx85R/GQ3WENcUT57VjW6i8EBgDU/6l5FRXG9U0CmyxOyjwRrCL00ModKvhDnmqaBfCbHQ6qfUuTDSoYPgF4H1PONbQl7xUGcl+EJ6KyvM7nGvxcqRVp6q3kjIylOfG8BB4aWKZbQEK9kbbWBurp89sUY6AuXsiC9gVbJ1CvGzoOj0+Evo90UZ8kRYnTxJt6+E31SuZdgTJm5kOBSp53i4lqBV4Mb+ffT64hyK1JNVvdzes4Mhb5i9sUZOmZVUFpLEzbMXSjHKNTr+uISFX+ufwvgeg1Sg58NRVxkOQELJy1m0adLQBOA/ZVD6QpbOP3QFVZJX+On4bIz6HyTQkzMvcPJ1GmpB4hk8s30t6VD2fIbkFj9mqUauyUP3h6PM++7QWdvO12v0fDjiVnADvL0WpS9kZ8W0N0s1Oj4fZ9Ff9E27e5cK9N0VIbNiom/ib7iKf+80hISSl7L03h0mvdqHE1DofX8E/ymT8qfSKDOEXaRwC/54Bm209ERf6pZFoOAunk7VlHINzVPOM1WF4/MqGI68M+m20+H/GPd/B+jU4zwTWUXczo671v9n2Y1ckz1GY7GfXi3Kc5EVzC/20+ytxCMtqs0Rfh1dy0cTb7LHP5cBLcyqfDuvhRbxieGteM/iotcch3XH2/j2D/p5duMyfnLbFrrK4hiaihTi/LwkY4VaLBt/0eSm3Uf4Dy/vYkn7mWwg6RHIeg1ZrSKrVDAlzmqvm7PeYSGKEvWRLPa7AhjfK0N9POdW6Ls7hPp0DmW7a9yld7SdWg1ZoSD8uO30WYhOe0alOoACOqfUEtYZ7Xw5/SohpzBuaO3iSa4qnuSH4Ss5rlXQocZ5OLiKP029imcan2TcyfOJ7Nto0mGN0YlfuhH3MU+LjeD2/CG+F76ard5GcoqHV3xNXGG0EZBnzkwSeM3byHO+JWSEF1XarDc6+Hh2B9V2Ck2ORfPBRsEUKv1qiIA00M7TZyoBA5WfBTfwZGAZJipzrSG+kN7KMrN3iuGsLiT4aPsr2Ah8jokqbW7o3881A4fwOiaqdNgydJgNw8fxOuYZISKfY7JhuNlN13RMEp4Qc/KDLE+1ETWz+G2Dheku/vLoQ6jSxutYBLIF/qz5MRQpx71cn295GoHEZ59jVndcDlH/bWGkKqj/X24Rl7G1lB1S6L07TOdn4m56FxA6VKDkpez0bHlAzUqqHkiRXuVjZL0Pxy/o+mSM7FIvNf82QnRnHmGOKokL3PKsVRqpdT6GrguSWutj8Z/3UrL1TKKpAMqeyzByhZ+eD0ZxvAq9H4pilajM/cdhvD3muDtcWxfwzgAAIABJREFUqpBZ5uPUn5e61elUgVJwqLkvRWTvLLweo30z+K4Q8juCud8dclXbRtu3Awr97wvR8bkSl5QIBI4XKX0+gzpDSezxb0fjKVK49zs+kIRL0nPUUWKmnDjpXLOOt8ei4b8Pcfh71VgxlWKNRvPflpO40k/tz0bwt5qI0V24FG4GQWGeh+RGP8M3BrBCKkv+Yy9aesITEkvnmdOb4E0peeA967ll6yHiqRxCShxFcLK+nJ+/9wp3Tvwt4bIa9/JkmjUnOiicFndY2NF3OS87jsXtvWjW5WFONBMniEkNUyVI5/UOnXduouo4zO0dYu3xtjO+K0vNXIbxuLeK9blWAo5Brx4FIOQUuTF9GICtwYWszrcxzxji7UADacXHiDfAinwnR33VBJ0ikWKeFYVOujwlZBQfXvvsuzhFQjSb54Mv7+SObft4dsMyXly7mJaaclJBH1mvl6JHw1RVHEUgpER1JKrt4LEs/EWTQNEgmsuzoLOfjUdPcc3+ZioSZxdU6byqHOevolTm0+P9YX01BhmJ/q0EYp+BzEravyWovN3Ge40fTAftxynUX0zUkpYLdcy/jsOo2pQMK1h/GYMRB/1bScTBWbilhSBq5/hieitxZ2qOv4rNEquP9+UO8sPwlaSEj2NaJc16BcvM3jObAm4uHJvyeUpfI9FkkXuyO9ntmUNWeDmml2OdJZI/pAR4wzefhOJHAOuMDv40/dq4N2Zy+/oomS5kFae99nRwMypcMmFGeHgksIqn/EsxUZlvDfKHme0sH/2dp8e7/fbU6/gck4mSiuB1LLxMv7sTgM8x8DluSVWvbVJRTBIxc/ht9/906VaKG79XKdHt4pTPmj2zB8fTbxHZVSC5JUD/HSEGbw4SbHbdyrZPkGvyUKyZmNP0AZvqX44QPnh2N7XAdU3P+x9DtPxlGemVXqRHkLgqQOKqAGpG4u23EIbE8QmMMg07dJpgjXr2J6SY0PCtIYplKonrgjhehf73hhi4NUTocBFPn4VUBPl6jfyCUbU8IVAzNhVPZKj9aWJWZDd9yCa6q8DIeh+D7w4yfG3Adbl3W0iPILfAQ6Fuom+0hE31Aymiu8/OE3F0yNfrWDEV2y+wwypWSMEOKYxs9I8T/qyAwsB7QuTrddS0jZZx0DIOSl6iFCWhs/S/AGJv5Gn8fwc59eUSilUadlCh7/0R+u6OoCccPEMWOG6WhFGuTiECerus8YyCMYRyRbbsPsEbaxvpqIzzib/7JDdsP0o0k6erIs7LVyyifDjNvK4hWueUzdyxlwGX1bjfsuMwt+w4fDkvcU78fz/69WVpVwL3iFtZSx9flrsuur1A0eSLj77CFx995YLOrzFHOOCvJWwXx9PZtEnSojVmksO+GgrCzWWfZwzyZnABt6T283RkNVuyxxlRA2fVC5dAcr2f1PIJwmH4aJH49hyBosldr+/lvW/up6ckSktNGd2lMYYjQbI+V8tfcSQ+08KqU7DWKax4tIt5h4eoHUwST081jrYQOKpAs6ZKo/5i0RrCPy/yJ/e+cvY4n6bydx9/N3/1T8+ysHX6BaRywMD74YtfXG402og705diHTOqlXaalOZjUAlyQitj6Wm72cnHnwsCCDkF6uwkx5RK0sKHJVQ3+2nScRLoUOMc0ypACPxOkQ/k9lJyljDLbK49FRKvdOOmOeHhCf8KnvIvoyB0Gq0hPpXZzhqz85wG7mIw+fwSM0PJyMza/hcCqQlq7nONdc+HIhTmekiv8pFedeax/hMGtfcmqXw4PaNxFEBse56FX+uj84/iJK70u/XFhcAOC3Jhz/TnGRL/KQN96NwbFTXtsPBvBuj4rMXgLSEKdZqra7/yTKIwliTQUqTy8TQ1/5qcldQuuDvpyodSRN/O0/2xKPn5HjIrfGRWnHmsr9Wg5r4Rqu8bOSc73SxROfn1chJbAnCOBYwTVOi7O0Lf3ad9YUv0YZvNm1rPStgTQOWvU+hDNt0fizKy3ocdVlyCY6l6VtldJe8QOlqc4pIfa2/z3hb+6MHXefz6lZycU859t21EdRzKExmu3nWCO17ax9PXLP/3adwvNywEw1qYkFMg4Fy6ilC/b1hQ7MMSCioOUTuPLm1uTU2IIDQYA9hCcWVqjUGido6QU6DGTHJj+hB15jB5xYPXsdiYPUnIOW13I6BYoZFc58eIawxvDjDn5wnib+fG3WK67VA/kKB+YPriQRLouTPKgS/VsOF7bcRbp1e1a6srpa22hCv2thIonF/qmmo7fPa+rVT3j5zXeReChdYA6jnc2BFZoN5KcEIroyA0+tQIJioeLsyTJADfqBveEQrTJcU5CHrUCAnFJfHUWwkWm/0XdL3poEkHHRsHweP+ZTwaWEFa8THfHOTzmdfPcMX/vsL2K+BA3T8nCO8vMHx9kOQVfgp1GnZIRUs7+NpMYttzlLyUJbK3MGPsdjKCxwwavzlAarWPkfV+0qt95Bs8WFEF2+sy9LWkja/TIthsEDpQIHTUmJEgJ3B31nP/cZiS17IkrgwwstFPfp6OFVIQNugjNoFmg+iOPLE384QOF88rJW9Mga/mZ0lCh4sMXx8geUWAfL2GHVZR0w6+DpPY23lKXsoS3ZWfOa6tCHenfA7Dfk6oYjw8ci4IG0pezRI8XmRknY/UOj/pVT4KNRp2WEGqAqXooA/a+DssgkeLhA8UCB4x8PSdufrxGxbve3Evq4900FUZI+fzoDqSWCpHY8cA0UyecLbAFftbaewYuLDfdhH4vTXuEhjRAvy04jpuTe5hZW56RvKksMzoZzEuBjO2NxyLb04+frKUw9jxMwmzjJ137vbk+LR8qSZCHZvlhc4pbS4uTsStVSQLi1NdpQuL7u51adHVqI84bryt3pyG+CWh/MUMJduyGGUabzw3/xLd+ZRL4CiCE/PKObiwltWHO/AY1mhhlInjHGWS4pUzIZjjjMbo1u8/NSUNcaxdIeVE/Eu6Ltqxltxzp/otxuRez4ZKOz2u8zAdBFBrj6DgeiNGFB95oU8hiY3dH7hjcVgJclSv4KRWxqASIq14yQudolApCJ0+NTzpvDPvzkSlR40gR5nxi6z+cy5Azhce6ZZ3ft63iAcDa8grHuabg3wx/RqLrDOrCv6+QnoEqK6rO/ZmnvDeAi/98RYU1eGDD++kpyLKv3zkSpb3dPP+vbvpL4nwq/+wnptfPMLSwz0kowEefP869q2qQ7Edrnyzhbse3YPHcL0etqrwwsYlvLl5Pj31UVQp2fBoK3c/tgePaYGEB+5ax3A4xKqdHTx9/QryV+pc9cYJ7n5kN17THUMS6JhTwoN3r6V5QQXBnMHNzx/m5hePENuWJ7K7wM8/s4nB40GWHO/hyVtXUNR1Vh3r5CMPv02kd3Yplaf3jdQEigXR7XnCBwrUhpM4XldAR9huZTot7ZyzYmPL/DL+9TNXcsPzR7hmazOLv9I3o4E+tKKGBz68no/+23aajvdPDX86zCrNrmV+GT/44rV8/N/eYulLPdhBV55WqsIVYnNcYSClKFFzrst/TMZlTAfFGbMho+HGxs5B5ncOuuoXo/K6YnR+WXqyh8UtvVPSPMfaGpuDxj6P2abTU0IdMXn+EgjpzErv4vfWuAP06TGO+Wu4eeTAWY9xEDxDA80ijlfabBO1XCm7WE0/D4mFZPDwp3I3y3FFYY5SwhNiASeJkUHHj8UmeviQPEKIsy9BE/j4nlhLBp0/lnuZzwgO0EOIp5jP26KaHDp1pLlVtrCJbryzFDmZDWbj2r2YttWiRC1KUC4Ph8FWFb721Ts5tKiGTMDLq1c0ESgYfPa+rVz/1nEA+ktC/Jc/v4OTc8spTWT53H2vseJIFwJ49F2r+dUdG+grDfPdbzzA6iPuYkcK+ONvfoR1B9s5sKiWgdIQjW0DfOUnL1CWyFLUVX55x0Ze27iARCzAUCxEeSLDN/774yxrPltRHIl/GjLb6QjJ4nioo4jmiqmcNt9ZKLRqpTwUWM0uTx0FoY+/5GPEurGX3eTcmQm2EGQU7/jnsgtkv58NGja7vHN4wbeYvHDTf/zSJCiN8zcSQNIbJKd7kULgswzihcyMtSPOBwVVJ+EPYSkKipSUZ1N4ZpOiKhgnhgoJWk4SaStwaGk1vlMW2aCXzmAcZb7E0hRSYR+9lVG8BYu8X+ebf/0eBPCeZw5Q8Hl47I5VdFdH+dL/fBmPZSOkZOe6ucw7NcTVW5vpro7x2B2rCJ4y+NCDrhRpPuPh1S0LSYgA1794jERJgEfuXEPO7+FT925DdSR9lRH+5hvvo75jmDue3M9gWYh/u+dK+isifORXb6PlHTKml1eWLKJPRLj5F0eRAh6+ay3GhzX+4/deRLfOk0Q5aZcjADUnUXPTzwmvXt/EiaYKPv2jN874zvCq9FZHyIS8CItZSbv6gybBE0WCJ4oETp7/mAMoejS6a6IUPPpozH52573SuIxXG5fisSx21zXgs0zuOLSDOw7twmuZ9IVj/GzdNeyY04jHtrj25GE+vOd1CrqH/3Tbx/jDt19iyymXW9MeL+PL7/sk33riZzQM97OntoFfrrmKltIKalIJ7jy4g2tPHkK3bY6V1/Cda2/nhuaDPL1kDTndyyd3vMxto/VczoUZjbsEWr3lDOgR1mVa0XAoCJ2j/hoyqo+NmRN4pI0E9gbm4pUWS/NdSCChBun1xEipASyh4HEsSq00c4uDU5i5KcXP4UAtjYU+JIJObwk5xYvXMakxElSaI+PHF4VGtyfOoBbm9chi8oqHg4E5ZEcnNa80acr3UjKJEOYg2CmquEm2caNs4wnRSC9BrpJdvCzq+YVYwn+VW1GAdiIoSG6VLYQwaBNRnqYBTTh8Uh6cto96CfDPYiUjePiMPMB8RtzJCy8/FSvIoPNeeZIgBgdFOfeKZThScA0daNPs/oy4SnKtn+j+At4By2W4LvCSbfQQ25nH2+++CNlGD4VqnfjbOVTDLQlaLNfIzfdgRhT3pekxCbYYrmGe/FwV6L8hRLDFwN9lkp3voVCjI1WBnrAIHyuipy5s1ydxS1NmFnoxYypK0SHYarjfnGXyVm2Hb/6Px3j8plUcWVDNn/zsFcKZAtqkyWf3inq++oPf0NA5yP23r+dX793A0uYeFMvhruf2cvtLB/iDH37+DIZqKuznwKJaPnvfa4RyBv/tj9/FY+9axafv38auFXN5acsi/v6/PgQS/v5z7+KmrUdYeuLsLmZ3fpvZCE3e2Tvj/p8JmCi86lvAj0ObSSl+NGlTbaeoslOUOjliTp6ALOKVNjaCp/3L6NaiZ72eRGBNWgCcTZToQuHG2ZdjoRJ38iQVP4c81dwXXMfn02/MmHs/GZ2RMv5h43vpD0bRbJulgx18ds9viBYvXthl4hql3Lf8Wo6V1tITivP9Z3/I0sHOC2prTucwL96wGEtTGCgPU9M9wkjET9GrkYgH0C2bWDLHa1c10V0d49tfe5i6riQAkXSeez+2mRMLKlh6tAfVkfyXv3t6vG0JnGoo5cDyWj744M4JwmHe4GO/2M6i5r7x3d22zY3c8ptD1HUneeTONfjzBl/+7guUDmeRQOlQlkfuXM3mt1poOumGZAJ5gw/fv4M1+zrcdlSFl69d6P6OnssTwpLAm1fOP+/Fw7mw/EA3yw/8dqp5SiF4a24TH9q7jXsefYU9dfN5aslaVna3s2CwlwdXbqKoavy3J3+OoWl8+7r34TcNPrZ7K2u6W3lq6Vq2nDqGBN6Yt5g5iUHmJgboipby8MpNbGk9yl+88jgHquv59fKNVKRHWNnThhSC42XVbGg/wTeeux9HKPjN2YWgZ7VzfzvUxKMlG/hRyw+J2AW6PXF+UnkDbd5y/qH1X2gsDpBRfPyk8kYW5btZmu9iQAvzUOkmdgcbMBUNVToYQqPEyvCe5F5uSe4dH8Rdnjjfqb6NG1IHGdJCtHvLMIROQdGZVxjgo4NbWZLvQgFSqp/Xw4s56q+hzVtOTvHyengxu4MNAMStLBH7TUryU3ctEWlwE6cIY/CcaGCuTHEHJ8hLjZdFPUVU/NhcTzs3yrZxo2tJwVFRwl4qkEzJykBF0kWIe8Uykvj4nNzHIiZizvuooI0IX5K7WckAAlgt+/meWMs2UcN62UuEMx9Ufo7O4f+nisXf7KPymTRWQKH9E3E6PxRn2dd6qH0wiaMJuu+KklwbYPWhThTTJrXcR9unSkiu8SMsCapAH7GpejzFnF8mpsTWpCY4+O0aah9IEj5epOvuKGZMxYyq+LtNFv9tH9H9FyYIYsZVTny5nMFrgiBAzTn4O01XxOIsC3QB6KaNZjsojoPHsMfdj2NYebiTq3e4ZYXXHOrg/veunyIZLOT0rnSPYbF5dwsrj7kTw5pDHXRUlwCuAIVmOXiLFlIIPIaFpZ3L4e4aUWOGXTS4KXNj7ejYZwi2HNUr+UVwHSnhwydNriq0cGPhOA3WIBE5lfmbFD62eefTzdmNu4Icj8sDZMX0BK0LhSVU/I7B9cUTrDC6eTKwjGN6JVu9jZQ4Oe7J7sAnZ8fMer5hFSfjVXx96/2EjDwe2yZgXhoBmjE0JPv487ceZVvdYr69+Q8uqq26zgS5gIf+ijDdNVGWHerm7Y3z6KmK0VFXQulQlmgqz76VdeT9Hp64beX4ucPxIDm/Tn95mKVHe5BAV02Mg8trGSwLUfBqdFXHCOYMpJjQ3/fnTea1DY23s3ZPO8+9axl9lRHqupPsX1FLQ+sgpcMTc93a3W3c+7FNdNbFxo17OFWgqXmCRFrdM4Lh0cj7pxdfuRhkgx72rp1D6/wy9q2dQzhV4Mefv2r8+3t++iZeY9J7LQQ7N8zl+OJKpCKo6Uqy6Y0WAvmJcXyysYyXbl48/vmOX++joi895f04uKKGE03lXPFmK/tX19FTEyWUKbL4UA+Lj/Sh2WdfZAyUhdh6XRORkTzXvXT8rMfOTQxy4/H91KUSWD3tvNy4jIzXh60oPLdoNfOH+3h45Sa3Hzxedtc18JHdW7n5+H6+evvH6YnECBaL7Kmdx03NB1Adh45YKUcravFYFi2llWQ9XtrjZXRFS1jZ42ZPlWdTXN16hHmJwWnv62yY0bgLoMpMIpD0aTEidi99ehTdsdAcm2ZfDY3FAfr0KHnFQ1PBjfUGHINyM8WHBrcxxxhCkw79eoR/rrieB0s3sTp7imozOX6djOrjlcgyrk4d4X3Dz+GVJgf9c3iwdDOPx9czv9CPX5pE7RzvSu7jutRhHo+v461wE+8feotleXdFrkqHuHWmOzJCEQ8OKpI4eWIU0JCEMHEQmKPGXSB5kxoOUM6g8JNHp4UoZeRPM+6SHBr3i8UM4efzci+NJKdcc7eoZIAAPxYr8Y4SqWwUeghSTp7iWQyElnbw9VsUqnWk7hJ8CtU6+rBFaqmXmtEazoUandCJIoohMcpUTv5ZGflaDwu+O0DoeBHbr9DzviinPlOCPmJR80hqKrtahcHrgpilKnP/ZRh/p4njVbDCCr4LKEk5hvZ74vTcHqHm0RGqH3OZskNbgrR9usTNW71AVA1OpMpptoOlzE5pXbMdSpITY0I3bcxR/fq1B9t59YomvvI3HyCcK1Dbm2T14c5zSqECJJTAlPEwHQbV4PhuPSiNKWleDoK3PHMZUFyFsyZzgI9nd1DuTJ/6KBEY0yjYTYYqHaJyIg+6U4uf8/gLwR/k9vPe/EFCskhAGvw4fCW9SpjnfEsodXLcmds/K436zkgZ1ZkEa3pbLlusXpWSoFkkUsxdVOlMgKreFIFskRONFQyWhtn01kna55ZwZHEVvVURqvpG0CyHbNCLr2jin2ScavNJFpwcoL7DNdTNCyr47pduoLIvxYpD3VT1JDnRWEHRqzG5Wo0iJR5jYsz4Cia2IjA1d+zm/Z4p1xlbICMlxqT0Y822CeQnNhFitAT25cjBtlWFdMiL4VEpejXCk/t9mkfw5pb5jMT8VHeNkA15ee49S2luKudz//T6+DH+vEF8OEvzokq2b27gyq0nqeibmjZ7oqmCX310PXvXzqHg14kPZdm5cS4v37CQP/zxNtbtbD/90gBkQl7u/+h6TjaVc89P30I5R3nwWD6LZzTV0o2Xu3FwRwgyXh+L+rtpGnS5TSt72inLuvPtgoFeGob7eWnBCpb2dWKoGit62hBAQdPxWSbL+jqI5V2v1ea2Zpb3Ttyv3zQIFc9/4TurnXuVkcTvFGn1VbCg6Br3MiuDIxSO+6u5ZWQfvZ4YBUWnqeD+uKBT5M7ETrfIyugU11Dso8VXwWPx9bR6K6YYdwdBfXGQOxK7qDVcUle5meK4v5pdwfmYiobfNvFIm0orhQNE7DyqdCiz0tQZ566EpSHH70NFjk9AY+5VCeTQ+Eexlr1UsFl2s5EeIhg8yELyTF3l2gh2U4WBQpwC/tPyc123vI8oRdbRd0b+blwWCJ4lhq+lbfwdBrl5Oo5XwQ4qmFGV6L4CmQVeHBWssEK+RqfqmRSKKRncEGR4Q4Cmbw9Q9WRqfIfs7zLJLPRw8k/LqXo6PXX3rrp1piueS1PxQgYxShyZNL+cN8yQQseHYgTaDBp+MDQeQvD1mKSW++i7LXLWc4UExXGwVRVbVcaJJGNT0EwG91w427mhbIHhWIhP3/8Gy4934THtceLTudCmlWAXlWkr6YE7Plq1EhyEO0btzJQ4fV7oDKhhbBQ0HJaYvecw7IwqyoWm+XYCHmzmWEk0aWMJlYN6FXmhTyt2c6GosVNEZBEFyUajjf5cmF8E15ERXh7zL6fCTrOl2DIt4Seje/np6pt5s24R/7u9M42O6zzv+++9y+wzGOw7QILgAq4mKVEiKdK2ZImWFNvxlshxZMWO0qTN6XFPs/UkOaf9lJ70tM1J49RLtiqxLadSlNiy49iSKVEkJWvhIi4gSBAk9nUAzD5z17cf7hAbARCSSTdR5vcJmLlz79w7977P+z7L/xmOVuMoKh/5+d8H4JfePsrHrryO5jrYisqrzZt5dusBhmK1NGZn+UTPa7x/4CJ6KWZuKyonW7bwbNeBuYnCpy69yqHBi+ju7ctluYFuO6wbnKavo5ZURYCmsRSdVye5vKkeS9fYdc4rAWyYSHN9fQ2feu4UfmPxM6+WVoTf+cgu8kEfX/zSUcI5EyElr+7vLBn3eSxNJRULEk97E7bJuhh+0yaS8wx1w0SKqbrIXHKWBFJxr1Iill4sdqP8ZHObNRNNF3nghR4sXeXlBzbTeXWKJ/7itbn39SXeuN7NdfzBb/0DLUOzSCH4s187xCv3b+LR716gZcizD42jaX72797mjf3rOLerecVjp+JBwjmDX//jl4mmiyRqI/zBf36YN+5dR1f3GKG89xx494ak6Nd46gv3cmlbI7/81RPsOjO86nUSJYO+FEVK2pMJKgtZPtB3wROyEcpcwpwmXT568S2+vucQtqLQNpugJudNTmJGgaiRZ+v4EJunvIWxKwSau/g6vZv7eW3G3Zol7BhcC9RRyPgY9VXSYCapsdNcCrZgozCux1GlS5MxM/9lpGRGi5BRg5hCxRUKhtCRQlBUFhtLgaTVTNBozs6dSNgxqLRzZNQg7h2b389zhjrOUM+n5BU+xeU5A/dt0clyulDtpPiEvMLfiU38L7GH35FvUEWx9C5UYFBJkSPyOo3c7E1Y6Yy0jEtwyCK9LYDjF2Q3+hGWpPLNPMOPxXGDCnZUxapSCQ6a4EimD4YJjNtEe4ooC8YU/7hN9EKR9LYAqe0Bqt5afCbBEYv4mfkuat7v9s6u20KyXX7coEL0iuGJcpRe19MuFecKqxt3oGEyzdR9YX5wuIv6RIYtfeM0Tq2ue1/0aQw1VTEdD2HoGj0b6lEdl7bRGWLZ1VvHZsMBpqojpGIBrq6rQ7Nd6hNpaqczq2aknvK18on82+hyeVGUK1ot42psTvCmzZld9Hs7CE/vvTQwB+TqE4q3fG2klWXqlRcggEYnTYuTpF+rJqkE+WFgC48WLqLfgY4dKpJHCxeZVCJ8N7SdKTXK34Z2U+3m2GJN3HT9ArbFxy//mCN9Z/jyXR8m5Q/xn159DiTU5FOoroMrBCdbuviL3R/i4NAlPt39Khfq2viz3Q+hSJcPDHh5L6+0beP/7LqfQ4Pd/Fz3Sc7Xt/PVPUdQpMvhwYu3NTEPvGu7qXeS1+7pwG9Y1E+m2dQ7wQsf2kosXaC5FF9//ytXOHFgA98/sp3DJ3rRLAfDr2P6VFpGZlEcieK4aLaL4fc6IY42V3Kpq5GG8cXx72zUz/H7Orn3jetYusr3j2yjeSRJ25A3xj74wiWe+tx+fnzPejb1TmD6NJ79+F7WDUzT2ffTL72CkvfgRpy91CnTZ62c+7H/xDU29E7N3fu7Tw1y8tAGL/RQMu4Cz/umOqtbgYpkgXte66d2ysuSq5zJs6lngunqMMWAPmfcNcuhEPTx1JMHOPe+Fn7ly8fZfWroXVsY3XV47MwJ/mH7PmxFpbKQYzYYoWtymLuH+rzz7L/C1+59kJc7t/O5t44RNr2w28apMTZNjvLNPYe4e+gqUghMVePQtR6a0j9Z6+Y1GfcKp0CNnWXQV0NGCTDui3MwfZmQa/J6ZCOjvkqm9BhN5uzcijiphjgZ3cwbkQ1k1CC69Fze01p0gbjmgi8iXWJ2cZFLTymttn9aEn5FNBwU2knNfbtBogwTJbokNi6QNJFlP2OEpM2fiD38jdjGE/ICcbwBf4+c4Iyoo4dq6snPnZtdSq7SVpiyKIYkMGIxdX8ENyBIvS+IP2ETP5Vn5FMVZDb7kZpAOBLfjAsCio06as5dVmwhMGkjXCg267DEuKsFiZ66fQO/UV3Sqp5abKyELfGtoZ3k9ssjfOKfztK9sZF80E/HoBdnuvvtfvwLXJQtY7M8dPwSqitJR/yc2dbKcEOcDx+7yGh9nGQsROi4QUWmwEPHu2kfno/TpEGwAAAZsElEQVRdbr8ySuOkl/TYu66OjsEpXt/dgeK6FPw+fLbDb3/lB4QKK2fk9mk1vBjYzM8ULtwkK5sUAV4MbGZGCYGUtDrJm+rNA9Im4hoIKbFRGFa9QFFgiYfHBbr1Bp4L7VqTtG+Lk2S3OcywGsdC5TvB7VS6efaZAwSXiYdLvCz8nOIj4hrvaBIg8FTufiF/ilk1xMuBTvr0Wp4O7eXXM8epdxfHRTXp0pr2fs+KYg5bqGyaHlm0TVoP8KP1O9k71sfnzr1EyDbZOdnPZDjOidat7BvtxRUKR9ftZN9oL4+ff5mgbbJjaoDxcCUnWrvYN9pL2Lr9Pbs39k7w15+9l4de7CZUsFh/PUGyIkgwb9Iw4Rnmrp4xPvV3pzl5sJPz25vxmzamrlKbyPLLf3USzbF44KXLXO2s40//7QeJpQoUgjp7Tw8w0rS4QZVqu1zqauTczhZSFUHyQR+Pf/3HxDLeAuLQyV6uddTyzCf3Ek/lKfp1bE3lsf/7JrWJOyPwc7tpHVysixEoWkjBorDCWgkULaqn5s9bkRK/YTNbGV5UQusqgu8/uo2erY3c/2LPmgx7azLBfdd7CJqeHYgaBQ5fu0R9JokiJff3XqCikOds8zqmw1Ea0kmaUvPGOWBbPH7qGDOhCFvH548XL+b5/Bsv8+r6zVytbkB3HTZOjRI1vHG6Kp/lyOWzhO+UW14A7cYUr8S6SGtBckqAWjtNqzGNKVQuB5uY0SKsMyZRpIuFwiuxLr5ec4h7slf56OwpKpw8PtfmlVgX3666a5ljyBVdnD8t1pEmgsnzYgOmVMmi86ponouXr8ROpvi8PM+XxW5qyfMJruDHYTcTdNHEM2IzkzJEI1nS+Bikgj1MsI+xZeOTAghM2DgBBSeokOnyE7lsEO0x0NMOyb0h9FkH37SDniwN2Ku0mJWlG3tZBa1SXedto7SrZRdOazhMuGhx/2uXuf+1y4tef+DVxf9vGEywoWT4q1J5fu57KysFPvb8W4v+v+dsPwCZsJ/nHt7Nh05c4t4z11FcyXBDnN//zY8xWxFaFKdciFaqDvnb8G6m1Aj7jes0lWrah9Q4LwU2ctK/HktoBKTFA8XLVC0pS9Nx6LSmOOFfT17xcdbXzD8Gu/hg8SpxWcBBMKZW8Lq/naOBTUyoUSrcAill9S5TIWnxYOEyF/UGrmh1jKsx/jJyL6fNVrZbozQ7KfzSxkQlqYQYUSu4plVT7eb4eOH8qmp2KxGWJo/n3iStBDjta+WUr5WnIvv4YubYmhPsblDUdK5UN6G6LhNhz9g5QqGvqoHafBpL0TBVjd6qRvrjtYxGvMRIR1G4WtlIfS65pn7t74bOq5P85h+9QNugN2hXpIv8xz9+kWBhPsYuJHzkH8+xvXuUyboojqoQLFg0jSXxG942Oy4M88UvHWWyNoqQkqbRJNGMwVBL5aLcgIp0gc9+83UG2quxVYWm8RQd1+aTqnymw2effp2rG2pJxkNolkPr8Cytw/MG86EXurnnjeuLzmPT1Ul+5S+OUz+xukfsp0GgcPtCRoq7OEdhJfIhHz7L4b5Xejm7p5VTd7ex7/WBVT+zMTE+F08HqCzk+NjF+XHF79gcGLjCgYEry35elS4PdL9NPxVcFjF8GAyJKEJCsGDT3D3EOgZICT8tMkMfURKijoPZER46c5puUclGJDWs3civeXq0vjjJCxU7GPZVI5DE7TwVTp64XaAn2ExaDXJX1utUNaOFeTvURsg1+EziJI3W7FwYd3VpkDuDKHkAbqAsiJ0oc/F3z83+pDzHt8QWvip2UUWRD8kBdosJXmQdpRQKwHNJ3tinAPYxTkL28JzYTKtMc4BRYsLkSXmOH7KOV0QrKXyEsGgnXcqGXtna+SdstJxLoVknt85Hw3fTnrt+wCK1M0j4uoF/3EKfdRASQkMW+XU+7IobZ+MhgWK9hlQEweE7r+LnL3V+Muq0xd9D88rj/jmh2d61u7qulp09Ixg+jRP7OvGbNtHsygIfnVaCe83rfCu8h+8Ft3HM34mvtOI20MgqfmwUhJQ8kr/I4WLfTS5qAdxnXuOE3cF5vYkZJcQ3wnfzndAOQq6FJVQMoZERfkyh8nChm5g0eDq055Yr+HZnhn+XOcEfxT7IoFrJpBLhR4FNnPSvL8nHSq9sTiiYaBhC4y5zcEW9+lshgAYnzWdyp5hRQvSrVZz0r6fOyfB47s1lyz1XQgqBrah0zI6za6J/7vU9433U5lOELIOipuMoChtnJtkxOT8o7xnroy6fImi/i/tcSk8i1ZYIZ/mSzXDB4vCJ+e5fAjjw42uLtrnhlt50dZJNV5dXB1QkbL4yweYri2WQa6eXrLYlNI+laF6hXE0A4bzJrvMjK57WjYz5hZ+pmc5RMz0/2bwhtLWghP3mY9nSWwR4+Xq3TCZVXK+7oxS32vanlAywAJ/l8KEfXOJ9Z4b5n7/9If72F+6mOpGjoy+x6jmda2vh2Xvu5vf+/nl8tve8Lz23hf/bikIyHCJSNAhYXuvgSRFiRETZISfpo5K7GWOCEAOigkNyiAnCTIgwUWkiEcwS4JqIc17U0iizd8a4bzAmcIXC2fA64naOuJPzEgnMSa4EG0FCu5FAIHGEgqnoRByDYEnKVAIpNcjb4TbWpq+zOgIIShNTaOQV34o3kILkUa7xiLw+Z4z/SL40Z1gf4RpH6OfKpnqElBzsHWGfGMdRBFIR+C0bISUHqiZ4c/06uq6MEckZfI0foEiJrQhPEU2DI2KAhxjAb9mYPo3vHtnJJ58/zS/SzWOiB8On4TdsVCmRivCEFGwX3zLNbQJjFr4Zm8ThCFIXRHq8GE3F+SIDX6hCOJJwb6kWXULtS1lGP15BaleQirMFFEOCgHybj/TOIIExi9jF2++qXEq0x0BLO2S2Big06wRHvJm5GdeY3fP/r7fxcgQMm195+jjf/Ng+vvg/fh4lK9nUO8Hv/un3ieZWfojWO9M8XLhEuz3LX4f3MalGyAjPoCtIfNKhWuZ4uHDJi8uv4JGKuwV+I32UL0UPc1mvoyB0JkUUFFBx8UmHCrfAo8VufjZ/jrd8beg4WLd4bBU8edz/Ovs8T4f38Ia/nazwYwiN/JxIzvwxKt08DU4G3ztcZS895lZrnMdyp/nzyH4SSph/Cm6l1slypNizZne/7jjU5tKsS07yiZ7Xlk0s8jk21fkMHcmJFbd5p1S8WeTgzr55gZbbKBGQ1oLE7NVzP26FCxioqEh03FJdj8BF4CtJAlsoXskl4OB17LuxvSz9L5D4loQDHQQv00oLGTYwrylyQ+I4dqbI/r3XsITKBarpc6JEFiQROwjUJcs2xZXUTmYYbq0kURvBb9i4iqAiuXw/hpWQgOVTcRSB4deQQmAENIoBHdVx0WznHScLClcSKNpE00V+7UvH+MPf+zDPPLaXX/7aSWqmVm7UtX1wmOfuvssT8gGGqyq53NRITSbDroEhpmJRLrY0Eyka7O4f4FJLEy9t7aJjcooHz13Atb3fQMfFRqFDJjklGuiQSaooEMYiiokfhzAmcYpIBHl0Irzzro1rNu4NVhKftHk71M7BzGWiTgEFSWdhgh9HNtFiThN2vdVOxCnSYkxzMd7C8VgXG4rj5BU/r0U3kVf83K7ZWmdhHLNa42R0C35p43et0oQjQczxHqb52ej8MRfH9cHwq0xVR2gd9dxZfRtqmaiN4ioKXVfGqEtkuLqhFrdUejXYXIUUEE8VSFRHWDc4zdFDm4mnCsTTeTb1TdC3rpZwzjPItqZwaXMj6UiQSK7I9kujdG9uZCYepm46w7bLN7c49c06+CcdEh+IoGWckggMxC4WsWIK+Xadmpezc4lwsfNF6l7IMPLJCoTpEr1s4IQUxh+JYdSqbPzvU8v2pV4LTkBQbNBw/QpmqcGCWa2R7gqgGi5qXhIslc5peZe2byS59qtVXPv3NdR/P41wILk3SGZbAGGtFj4At0J4bShnXW+VEBLIiEBJumCCjAqkBsKQiHzp/aD3vnDAiQvweZ+XChAUYEikX6AkJfjArVRQZlyEBe2JaX7n2R+QV4KE/r6INrz8qC6APeYwjU6aveYgYWlyjzlApz3FaV8r17Uq0iJIUJo0OWl2W8O02bMIJDl0fHiufA0XpzQJKKJR6Rb43dQPOeVr44yvmZzwIRAo0mGzPcX7rBEqnDxTIkSdm+GDxV6SIrionh1gWESpkgU0XHy4FNAIS5N/k32Nhwo9nNcbmdKi9Ki1VJTK2CpkkYhTZI81QrvtuZrzaASxsVAooBPAZpM1RV5cBaDezbDS86sAB4zrpJUA3XoDLoJZJURKCa5ZLS9sFTk0dJFj7dvZODPGxplRXCGYDUapy6VoyM4QNQvcN9TNydYuNsyOs2FmDFdRmAlEaMgmqc95z7GtqBiqRk4PeOVKvgA53Y/PsdDcxQZO8O4Mek7144jSpM6xkEKguw6GouF3bbJagON123lk9E0coWALlZwWIOQYBJ2VPQydfVNz8soucJEaUviopshmZjhDHeCNZduY5hoVzOIniM0OEgwQY5QIcQy2k2CYKP3EiGHOaW4sxECjl0qmCbKLKQxUeqhCxWULMyiO5Ax1TBGihgLXqaCKIgFspgnQRmZR7okAHv3OeZ5+fB9/+HtHqJzJo7iS3/hvL+Az136hLV3l2x/fRbIyyFBbFbmQj+9+bCen7m6nbiLD/hN91E29u/wCAdRNZHjiL1/jq79+mG9/fBef/Zs3FpUXLkSBOY+OFJAMh0iFgjx1+D7+6it/zvnWFrqbm9ndX/ImSQhYFg3JJKrrUkQnhkmlLFJAZxeT7JSTpf16H7lHztuCdull1DfLDMh3Pmlds3FXpUubkaA72EKjmUSX3mxwgzFOUdGJOQWCpeYtIdfkgfRFklqY56r24ZcWPtehw5jg8anj/HHjI+/way7PxuIYH519i1eiXVwMteBzbVrMaR6femXOuK8F3bRxNJXRxjgdAwmutddQ9OsEixaDzVXUTmdpGk9xZkcrmXCA4abKObGJ6201tIzOcvTQFp78mxNUJXMICU3jSZ77mT08fPQiow1xfviBbdQmMgw1V9I8lqR7cxNSiLkJxVKEC6F+k4kPR6l5KYuW9x6caI+BYnutCYMj84ODnnTo+N/TDH02zvAvVOL6vFshMGbR8afT1P9g5XaqtyK7yc/V/1CLVWrJCDBzb4h8m45alIQGLHb81rxyVMu3ZrGiCpNHokzfF0bLOMQuFml5OsnA56uWPYYEnHqF4gN+tCEH3ykLFMh/NIAy7eI/Y6EkXDJPhtAv2+jdNuqMS/6RAErGRZ1w8b1tYe7WcSsUlJT02mfGBPYGDW3ARj9nY23VEbbEDQqCRw2Kh/0gJVb7rVbDkl/KvXHT6zVunoeKl5f5REn7W1QwrkTocGcZFRE2ujNcUyppd1NcVaqolTnqZZY6O8W9jsEmJ4GOy5tqM3c5IyRFgD6lim61lkPOAA/kLpFQwihLBHHeVJtpd5MIoF5mSYgwaeFnoztNH3FiVo7t5igxNUOHM0ONzNMsM5zU2mixZ0mIEJNKmGkRYr89xJBSwbCIscOd4MPFSzxcvLTq9bmBjstHChf5SOHimrZfis+xeejaWdL+EM917ff0u6WLIl0+3f0q9blZfI7Nkb4zZH1Bnu06sGibn794gvpcEkPVOLp+B683bWY8EifrC/LUzvupyac5ONzDB/rP43N/8uX5c+0Hac9O0lCYRXdtMnqQDZkxLsTb2Z4cYCpQQcoXAmDaH+VU1UaiVp4Zf4xHR95Ak8uvxh780fz1niXAGeo4zFBJiwNOUc89jNFAjgRBrhNjC7OcpIl68vQTY5wwteSRwChhrlHBXpbviCiBKCYThBgiwghRAtjk8NFHHAONHD6imLgIglgcp5mtTJPEzzpujt8fOnaVytk8I81xbE0hlinOlQPWTWT49NNvseHq4qz+1oEZnvjL12jv9xJghZSE8iaqK6mezvO+M/MKg6G8OVfiuv3cCKG8Qc0CQ69bDgeO95EP+wmXygfrx9P84lOv0zbgTWYVCVsvjPHkV04wWR+dW5Xfiqw/wHi8gtp0Br/lTQa2joziKAo9zY3cc7WP6myWhmSKjskpdNclikmnnMVApY7cqmGQhbzbQPY7Skn81YkXyagBGsz5+vSNhXH+cOAbxJwCcXu+Z3RHcZwnJ48yo0UwhYrftamxM0ScIv9l6BkaFtS4t5sJ/mDwW9Rai28QgeTR2dPck7lKxCkuec+bRHxy+nXuS/dQUHxeDMoxqLcWi8ncCtWVRLPFuQxNzXFZNzSNKwR2STCiKpkjWOpSprgSS1dwVAXT561kozmDjdcn53ILKjIFVNebAOWCPlpHZth3ph/hSioyRR481s1gcxUvvL+Lzv7ly1aan0lS9WrO6zVcQks57PnCkOeWv7ZAmEJC6LrJhj9J0PxsCjuszGWoB8ZslCUrZmFJ7vrcIFrauaUjJTRg0vGlxIoCNEu7Smkpl44vT9P4fBo7rKBYksCYhWJIql/LEbq+wopFE1hdGuqIi7Ak5k4dJPjfsFAyLihgr9eIfKOAknYp7vdRfL8PZVqiTjrovTZuVMHq0hAFiZrwDH7xfj96r42x34e5R0eZdXEaVdRZFzcmCJy0MO++M8mcAWxGRJSYMJhQoqxzk4wrEVrcNI4QFKWGicqIiNHlThHCQiIoCg0bletKJW1uipg0mRFBrig1uAjyQicq56+jRFAnc1xTqkiLAG1ukhklyIiIMaWEaXeTjIsICpIYBkkRpFlmMNCwUJkRQYLSwhYKWeEjIUJMKSH022AAl+MLb/9o2cQ3AdTk0zx+7uWSUQ6gSEnYLNKQS87VDtflUzx+7iUmlmzTmPW8JbrrsH1ykIbszWNBXS51k1rgu+VqtIGHh98k5Bj0RptI+GO05aYYD1axJT1Mc36agYi3ys6rAQqqj90zfXw/1kxaD1Fl3nrVWUQliEUzuTnXbACH9aSIY3KNCoLY1JMjjI2Byl1MMEyUY7SymVm2M00FJsdppouZm5oe6bjUksfFU1/MobOeFGn8FFFJo9NArqQHotNEjiw6g0RLrvybBxHNcXnfmeFFBvkGVTN5Hvxhz02vN0xkePT5eZlv3XYX/b8SnVen6FwyUdBsl11vL85HqJ7J8/D3Fk88fZazosjNDQq6zpsb1nO5uZHv7N3Dgcu99NXVYakqriKwFYWh6ir6a2voq69HCkHIMElEIzxz7z6eOHaCoGVRs0xR9Z1izcZdwE1CMV7c22J74eYfTwGq7SzV9s0375biYn3gkGvOKcwt3X+DlaLBWiWhxDXpMN59TacEBlqrOXZgI46i0Dw6i2Y76JaDU+oUtPSYG69P8pUnDlM1m6Nh0puQhPLzA2027OfooS0MNcU5emgLd53p5/TONl54fxfVM1lqXurm9T3rmaiLoTkrD56BCZvAklaDAoifXf4GEYCectFTt46tCwmVp9Z2o+kpl8rTa78pBaDlvNDAUlaTtFXHHaJfy5P7dBCt3/Zm0Q5edykHr1NX2kWddOcOFDhqEjxqIEyJuUPHrRQE/8nAvFsHAcqMi5KR3j5qwXfWIvCSgTAkbo2C3aiCKcG5M4k9aXw4QiEjfDS6aV7UN5DFx3YxyaiI0a/GqXVyRDE4qbZx0PES266LSmqVHPUyxxm1kUkRZqc7TpUskMZPbEl9fUhanFabCGCz2UnQo9QwS5BOOc2grCAmDfJCR5GSC2o9nc4MaeFnSFQQUz1BpYhrEpMGC+V1b0d+zHK0p1Z+ZgUQsYp0zo6vuk3UKhJdYRtVurSlE7Sl35lk5zslZJtUm17Jny5dLFWnoPrIaX4knsdHIrAUDSkgZuWJ2gUCjjeRWgueURU8xVaayPEQ/QQWpD82k+EylTzDJpTS9qeop58Y4JVa9lDFJapwVvg9bRRO0oyD4NNMUUWR79GBhssR+tmAyzfYgh+HrUyjINnDFMdp4gFWN4zvBQKWxb29fey5PoDmOPgch88dfxVXCD5/7Dh+22bXwCCbRsfRHQfdcYjn83zm5I+xVXVudf/TRMjVxB6EuDMj3j8zbnWS2bCfoaZK+tbXsv+ta3NNGqQivKxQFmdJrrQ/S1O9DG28mI2lqfis26+m9S8RCdidKsUPeoNi+NtFRFaS/XwIJARfKKL1O2SfCBH9c89DJHXI/GoYLInvrIU67pL/uSAi5U0AlLzEd9qi8GgA31kLDIm1Q8epEqgJl+B3DQo/48etVHCjCpGv51Gnbu8KXjKfcASeEuNKyvUrJYWu9PrSLN2l9+BK99Vq+5NAGj8X1HqC0jP2HXL2Dpn4f/l8c/37+cz1YwggqwV4ru0AQdtElzYHJ7u5VNHKq7VdHJzqpiWXYNofY1tqkJO1W9k73UuldetchBt3y8L7CG4ebxa+L/EMtl5KnnPxpK/VFaZrcsF+lAX/yyX36417J42Py1RhoHIfI/8qx7AbV2Xp77D0uVz62u3/IssHE8rGfQ0MNcXp7ainZibLtp7RRf3Cy9w+JIDq/XGjJv9Gi8lla/QXfOZGMpRU8Gr3VzuOBtgLHso1fOZfEw6ChAiRQ6dFpm8S6SmzMssZxPci/SVxr71MELyFDkiZO0zZuJcpU6ZMmTLvMVYw7mVvW5kyZcqUKfMeY/WVe5kyZcqUKVPmXxzllXuZMmXKlCnzHqNs3MuUKVOmTJn3GGXjXqZMmTJlyrzHKBv3MmXKlClT5j1G2biXKVOmTJky7zHKxr1MmTJlypR5j/H/AMibesAlWt/BAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JzCCSaTCGeok"
      },
      "source": [
        "# 3. Vectorization of text\n",
        "\n",
        "Pre-process text needs to be transformed into a vetor matrix of numbers before a machine learning model can undertsand it and learn from it.This can be done by a number of tecniques:\n",
        "\n",
        "## 1. Bag of Words\n",
        "\n",
        "The bag of words is a representation of text that describes the occurrence of words within a document. It involves two things:\n",
        "- A vocabulary of known words.\n",
        "- A measure of the presence of known words.\n",
        "\n",
        "**Why is it is called a “bag” of words?** Its called bag of words because any information about the order or structure of words in the document is discarded and the model is only concerned with whether the known words occur in the document, not where they occur in the document.\n",
        "\n",
        "### Bag of Words - Countvectorizer Features\n",
        "\n",
        "`Countvectorizer` converts a collection of text documents to a matrix of token counts. It is important to note that `CountVectorizer` comes with a lot of options to automatically do preprocessing, tokenization, and stop word removal. However, all the pre-processing of the text has already been performed by creating a function.Only vanilla version of `Countvectorizer` will be used.\n",
        "\n",
        "## 2. TFIDF Features\n",
        "\n",
        "A problem with the bag of words approach is that highly frequent words start to dominate in the document (e.g. larger score) but may not contain as much “informational content” this will lead to more weight to longer documents than shorter documents.\n",
        "\n",
        "To avoid that, one approach is to rescale the frequency of words by how often they appear in all documents so that the scores for frequent words like “the” that are also frequent across all documents are penalized. This approach to scoring is called \"Term Frequency-Inverse Document Frequency\", or TF-IDF for short, where:\n",
        "\n",
        "**Term Frequency**: is a scoring of the frequency of the word in the current document.\n",
        "\n",
        "`TF = (Number of times term t appears in a document)/(Number of terms in the document)`\n",
        "\n",
        "**Inverse Document Frequency**: is a scoring of how rare the word is across documents.\n",
        "\n",
        "`IDF = 1+log(N/n), where, N is the number of documents and n is the number of documents a term t has appeared in.`"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kNsaGojSGeol",
        "outputId": "57d0a9c2-e63d-49a4-ad3a-cdb71d96f552",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "# Vectorize the text using CountVectorizer\n",
        "count_vectorizer = CountVectorizer()\n",
        "train_cv = count_vectorizer.fit_transform(train_df2['text'])\n",
        "test_cv = count_vectorizer.transform(test_df2[\"text\"])\n",
        "\n",
        "## Keeping only non-zero elements to preserve space \n",
        "print(train_cv[0].todense())"
      ],
      "execution_count": 186,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[0 0 0 ... 0 0 0]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sM9cNPKqGeon"
      },
      "source": [
        "# Vectorize the text using TFIDF\n",
        "tfidf = TfidfVectorizer(min_df=2, max_df=0.5, ngram_range=(1, 2))\n",
        "train_tf = tfidf.fit_transform(train_df2['text'])\n",
        "test_tf = tfidf.transform(test_df2[\"text\"])"
      ],
      "execution_count": 187,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "i4rc88vwGeoo"
      },
      "source": [
        "# 4. Build a Text Classification Machine Learning model\n",
        "\n",
        "Lets create a simple classification model using commonly used calssification algorithms and check how the models performs."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CsPPAhiqGeop"
      },
      "source": [
        "#Split the CountVector vectorized data into train and test datasets for model training and testing\n",
        "X_train_cv, X_test_cv, y_train_cv, y_test_cv =train_test_split(train_cv,train_df.target,test_size=0.2,random_state=2020)"
      ],
      "execution_count": 188,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "x2YaGUWcGeos"
      },
      "source": [
        "#Define a function to fit and predict on training and test data sets\n",
        "def fit_and_predict(model,X_train,y_train,X_test,y_test):\n",
        "    \n",
        "    '''Input- model=model to be trained\n",
        "              X_train, y_train= traing data set\n",
        "              X_test,  y_test = testing data set\n",
        "       Output- Print accuracy of model for training and test data sets   \n",
        "    '''\n",
        "    \n",
        "    # Fitting a simple Logistic Regression on Counts\n",
        "    clf = model\n",
        "    clf.fit(X_train, y_train)\n",
        "    predictions=clf.predict(X_test)\n",
        "    confusion_matrix(y_test,predictions)\n",
        "    print(classification_report(y_test,predictions))\n",
        "    print('-'*50)\n",
        "    print(\"{}\" .format(model))\n",
        "    print('-'*50)\n",
        "    print('Accuracy of classifier on training set:{}%'.format(round(clf.score(X_train, y_train)*100)))\n",
        "    print('-'*50)\n",
        "    print('Accuracy of classifier on test set:{}%' .format(round(accuracy_score(y_test,predictions)*100)))\n",
        "    print('-'*50)"
      ],
      "execution_count": 189,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "l69IV13wGeou"
      },
      "source": [
        "# Create a list of the regression models to be used\n",
        "models=[LogisticRegression(C=1.0),SVC(),MultinomialNB(),DecisionTreeClassifier(),\n",
        "        KNeighborsClassifier(n_neighbors=5),RandomForestClassifier()]"
      ],
      "execution_count": 190,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cmZ9WegeGeow",
        "outputId": "d984f0a9-3dbf-4a5e-8e8c-a79d03fd9cd2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "# Loop through the list of models and use 'fit_and_predict()' function to trian and make predictions\n",
        "for model in models:\n",
        "    fit_and_predict(model,X_train_cv, y_train_cv,X_test_cv,y_test_cv)"
      ],
      "execution_count": 191,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.79      0.89      0.84       849\n",
            "           1       0.84      0.71      0.77       674\n",
            "\n",
            "    accuracy                           0.81      1523\n",
            "   macro avg       0.81      0.80      0.80      1523\n",
            "weighted avg       0.81      0.81      0.81      1523\n",
            "\n",
            "--------------------------------------------------\n",
            "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
            "                   intercept_scaling=1, l1_ratio=None, max_iter=100,\n",
            "                   multi_class='auto', n_jobs=None, penalty='l2',\n",
            "                   random_state=None, solver='lbfgs', tol=0.0001, verbose=0,\n",
            "                   warm_start=False)\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on training set:96.0%\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on test set:81.0%\n",
            "--------------------------------------------------\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.77      0.93      0.84       849\n",
            "           1       0.88      0.65      0.75       674\n",
            "\n",
            "    accuracy                           0.80      1523\n",
            "   macro avg       0.82      0.79      0.79      1523\n",
            "weighted avg       0.82      0.80      0.80      1523\n",
            "\n",
            "--------------------------------------------------\n",
            "SVC(C=1.0, break_ties=False, cache_size=200, class_weight=None, coef0=0.0,\n",
            "    decision_function_shape='ovr', degree=3, gamma='scale', kernel='rbf',\n",
            "    max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
            "    tol=0.001, verbose=False)\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on training set:95.0%\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on test set:80.0%\n",
            "--------------------------------------------------\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.81      0.85      0.83       849\n",
            "           1       0.80      0.74      0.77       674\n",
            "\n",
            "    accuracy                           0.80      1523\n",
            "   macro avg       0.80      0.80      0.80      1523\n",
            "weighted avg       0.80      0.80      0.80      1523\n",
            "\n",
            "--------------------------------------------------\n",
            "MultinomialNB(alpha=1.0, class_prior=None, fit_prior=True)\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on training set:92.0%\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on test set:80.0%\n",
            "--------------------------------------------------\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.76      0.82      0.79       849\n",
            "           1       0.75      0.68      0.71       674\n",
            "\n",
            "    accuracy                           0.76      1523\n",
            "   macro avg       0.76      0.75      0.75      1523\n",
            "weighted avg       0.76      0.76      0.76      1523\n",
            "\n",
            "--------------------------------------------------\n",
            "DecisionTreeClassifier(ccp_alpha=0.0, class_weight=None, criterion='gini',\n",
            "                       max_depth=None, max_features=None, max_leaf_nodes=None,\n",
            "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
            "                       min_samples_leaf=1, min_samples_split=2,\n",
            "                       min_weight_fraction_leaf=0.0, presort='deprecated',\n",
            "                       random_state=None, splitter='best')\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on training set:99.0%\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on test set:76.0%\n",
            "--------------------------------------------------\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.65      0.97      0.78       849\n",
            "           1       0.90      0.34      0.49       674\n",
            "\n",
            "    accuracy                           0.69      1523\n",
            "   macro avg       0.78      0.65      0.63      1523\n",
            "weighted avg       0.76      0.69      0.65      1523\n",
            "\n",
            "--------------------------------------------------\n",
            "KNeighborsClassifier(algorithm='auto', leaf_size=30, metric='minkowski',\n",
            "                     metric_params=None, n_jobs=None, n_neighbors=5, p=2,\n",
            "                     weights='uniform')\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on training set:76.0%\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on test set:69.0%\n",
            "--------------------------------------------------\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.77      0.90      0.83       849\n",
            "           1       0.83      0.66      0.74       674\n",
            "\n",
            "    accuracy                           0.79      1523\n",
            "   macro avg       0.80      0.78      0.78      1523\n",
            "weighted avg       0.80      0.79      0.79      1523\n",
            "\n",
            "--------------------------------------------------\n",
            "RandomForestClassifier(bootstrap=True, ccp_alpha=0.0, class_weight=None,\n",
            "                       criterion='gini', max_depth=None, max_features='auto',\n",
            "                       max_leaf_nodes=None, max_samples=None,\n",
            "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
            "                       min_samples_leaf=1, min_samples_split=2,\n",
            "                       min_weight_fraction_leaf=0.0, n_estimators=100,\n",
            "                       n_jobs=None, oob_score=False, random_state=None,\n",
            "                       verbose=0, warm_start=False)\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on training set:99.0%\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on test set:79.0%\n",
            "--------------------------------------------------\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ym3ooZd7Geox"
      },
      "source": [
        "# Split the TFDIF vectorized data into train and test datasets for model training and testing\n",
        "X_train_tf, X_test_tf, y_train_tf, y_test_tf =train_test_split(train_tf,train_df.target,test_size=0.2,random_state=2020)"
      ],
      "execution_count": 192,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jX6yOJ47Geoz",
        "outputId": "04df46ed-f80a-4d8e-9e1f-d234a3d71a9d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "# Loop through the list of models and use 'fit_and_predict()' function to train and make predictions on the TFDIF vectororized data\n",
        "for model in models:\n",
        "    fit_and_predict(model,X_train_tf, y_train_tf,X_test_tf,y_test_tf)"
      ],
      "execution_count": 193,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.78      0.91      0.84       849\n",
            "           1       0.86      0.67      0.75       674\n",
            "\n",
            "    accuracy                           0.80      1523\n",
            "   macro avg       0.82      0.79      0.79      1523\n",
            "weighted avg       0.81      0.80      0.80      1523\n",
            "\n",
            "--------------------------------------------------\n",
            "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
            "                   intercept_scaling=1, l1_ratio=None, max_iter=100,\n",
            "                   multi_class='auto', n_jobs=None, penalty='l2',\n",
            "                   random_state=None, solver='lbfgs', tol=0.0001, verbose=0,\n",
            "                   warm_start=False)\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on training set:88.0%\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on test set:80.0%\n",
            "--------------------------------------------------\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.77      0.92      0.84       849\n",
            "           1       0.86      0.65      0.74       674\n",
            "\n",
            "    accuracy                           0.80      1523\n",
            "   macro avg       0.81      0.78      0.79      1523\n",
            "weighted avg       0.81      0.80      0.79      1523\n",
            "\n",
            "--------------------------------------------------\n",
            "SVC(C=1.0, break_ties=False, cache_size=200, class_weight=None, coef0=0.0,\n",
            "    decision_function_shape='ovr', degree=3, gamma='scale', kernel='rbf',\n",
            "    max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
            "    tol=0.001, verbose=False)\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on training set:96.0%\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on test set:80.0%\n",
            "--------------------------------------------------\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.77      0.92      0.83       849\n",
            "           1       0.86      0.65      0.74       674\n",
            "\n",
            "    accuracy                           0.80      1523\n",
            "   macro avg       0.81      0.78      0.79      1523\n",
            "weighted avg       0.81      0.80      0.79      1523\n",
            "\n",
            "--------------------------------------------------\n",
            "MultinomialNB(alpha=1.0, class_prior=None, fit_prior=True)\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on training set:87.0%\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on test set:80.0%\n",
            "--------------------------------------------------\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.77      0.80      0.79       849\n",
            "           1       0.74      0.69      0.71       674\n",
            "\n",
            "    accuracy                           0.76      1523\n",
            "   macro avg       0.75      0.75      0.75      1523\n",
            "weighted avg       0.75      0.76      0.75      1523\n",
            "\n",
            "--------------------------------------------------\n",
            "DecisionTreeClassifier(ccp_alpha=0.0, class_weight=None, criterion='gini',\n",
            "                       max_depth=None, max_features=None, max_leaf_nodes=None,\n",
            "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
            "                       min_samples_leaf=1, min_samples_split=2,\n",
            "                       min_weight_fraction_leaf=0.0, presort='deprecated',\n",
            "                       random_state=None, splitter='best')\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on training set:98.0%\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on test set:76.0%\n",
            "--------------------------------------------------\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.62      0.98      0.76       849\n",
            "           1       0.93      0.24      0.38       674\n",
            "\n",
            "    accuracy                           0.66      1523\n",
            "   macro avg       0.77      0.61      0.57      1523\n",
            "weighted avg       0.76      0.66      0.59      1523\n",
            "\n",
            "--------------------------------------------------\n",
            "KNeighborsClassifier(algorithm='auto', leaf_size=30, metric='minkowski',\n",
            "                     metric_params=None, n_jobs=None, n_neighbors=5, p=2,\n",
            "                     weights='uniform')\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on training set:72.0%\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on test set:66.0%\n",
            "--------------------------------------------------\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.77      0.90      0.83       849\n",
            "           1       0.84      0.67      0.74       674\n",
            "\n",
            "    accuracy                           0.80      1523\n",
            "   macro avg       0.81      0.78      0.79      1523\n",
            "weighted avg       0.80      0.80      0.79      1523\n",
            "\n",
            "--------------------------------------------------\n",
            "RandomForestClassifier(bootstrap=True, ccp_alpha=0.0, class_weight=None,\n",
            "                       criterion='gini', max_depth=None, max_features='auto',\n",
            "                       max_leaf_nodes=None, max_samples=None,\n",
            "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
            "                       min_samples_leaf=1, min_samples_split=2,\n",
            "                       min_weight_fraction_leaf=0.0, n_estimators=100,\n",
            "                       n_jobs=None, oob_score=False, random_state=None,\n",
            "                       verbose=0, warm_start=False)\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on training set:98.0%\n",
            "--------------------------------------------------\n",
            "Accuracy of classifier on test set:80.0%\n",
            "--------------------------------------------------\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-W8_tHlPGeo3"
      },
      "source": [
        "# 5. Results\n",
        "\n",
        "Based on the above results, lets print the various classifier performance results:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IcxDK0boGeo3",
        "outputId": "12d4220b-4dc4-476e-ee6a-a03021a37deb",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 235
        }
      },
      "source": [
        "# Printing model performance results.\n",
        "results_dict={'Classifier':['Logistic regression', 'SVC', 'MultinomialNB', 'DecisionTreeClassifier',\n",
        "                            'KNeighborsClassifier','RandomForestClassifier'],\n",
        "              'F1-Score':[0.81, 0.40, .80, .75,0.65,0.76],'Accuracy':['81%', '56%', '80%','75%','69%','77%']} \n",
        "results=pd.DataFrame(results_dict)\n",
        "results"
      ],
      "execution_count": 194,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Classifier</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>Accuracy</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Logistic regression</td>\n",
              "      <td>0.81</td>\n",
              "      <td>81%</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>SVC</td>\n",
              "      <td>0.40</td>\n",
              "      <td>56%</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>MultinomialNB</td>\n",
              "      <td>0.80</td>\n",
              "      <td>80%</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>DecisionTreeClassifier</td>\n",
              "      <td>0.75</td>\n",
              "      <td>75%</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>KNeighborsClassifier</td>\n",
              "      <td>0.65</td>\n",
              "      <td>69%</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>RandomForestClassifier</td>\n",
              "      <td>0.76</td>\n",
              "      <td>77%</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "               Classifier  F1-Score Accuracy\n",
              "0     Logistic regression      0.81      81%\n",
              "1                     SVC      0.40      56%\n",
              "2           MultinomialNB      0.80      80%\n",
              "3  DecisionTreeClassifier      0.75      75%\n",
              "4    KNeighborsClassifier      0.65      69%\n",
              "5  RandomForestClassifier      0.76      77%"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 194
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XPsth1b3Geo4"
      },
      "source": [
        "It seems from all the classifiers tested, `LogisticRegression()` classifier performed best with `CountVectorizer()`."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pC9ODnAMGeo5",
        "outputId": "a338df77-efb5-4f3a-ae5f-c95086da68cd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "# Fitting 'LogisticRegression()' with CountVectorizer() fit dataset\n",
        "clf_logreg = LogisticRegression(C=1.0)\n",
        "clf_logreg.fit(X_train_cv, y_train_cv)\n",
        "pred=clf_logreg.predict(X_test_cv)\n",
        "confusion_matrix(y_test_cv,pred)\n",
        "print(classification_report(y_test_cv,pred))\n",
        "print('Accuracy of classifier on training set:{}%'.format(round(clf_logreg.score(X_train_cv, y_train_cv)*100)))\n",
        "print('Accuracy of classifier on test set:{}%' .format(round(accuracy_score(y_test_cv,pred)*100)))"
      ],
      "execution_count": 195,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.79      0.89      0.84       849\n",
            "           1       0.84      0.71      0.77       674\n",
            "\n",
            "    accuracy                           0.81      1523\n",
            "   macro avg       0.81      0.80      0.80      1523\n",
            "weighted avg       0.81      0.81      0.81      1523\n",
            "\n",
            "Accuracy of classifier on training set:96.0%\n",
            "Accuracy of classifier on test set:81.0%\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wJ0tAwB5Geo6",
        "outputId": "21dc23eb-ff0a-48f1-f7eb-c6f6af2e97a7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 102
        }
      },
      "source": [
        "clf_logreg"
      ],
      "execution_count": 196,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
              "                   intercept_scaling=1, l1_ratio=None, max_iter=100,\n",
              "                   multi_class='auto', n_jobs=None, penalty='l2',\n",
              "                   random_state=None, solver='lbfgs', tol=0.0001, verbose=0,\n",
              "                   warm_start=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 196
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1mLh74e1Geo7"
      },
      "source": [
        "### Model Performance Metrics\n",
        "\n",
        "After implementing a model and getting some output in forms of a probability or a class, the next step is to find out how effective is the model based on some metric using test data sets.Different performance metrics are used to evaluate different Machine Learning Algorithms.\n",
        "\n",
        "This problem is a classification problem hence I have claculated confusion matrix, F1-score and accuracy as a performacne metric to evaluate model performance.\n",
        "\n",
        "**1.Confusion Matrix:**\n",
        "\n",
        "The confusion matrix, is a table with two dimensions (“Actual” and “Predicted”), and sets of “classes” in both dimensions. Our Actual classifications are columns and Predicted ones are Rows.\n",
        "![image.png](attachment:image.png)\n",
        "\n",
        "The Confusion matrix in itself is not a performance measure as such, but almost all of the performance metrics are based on Confusion Matrix and the numbers inside it.\n",
        "\n",
        "**3.F1-Score:**\n",
        "\n",
        "F1-score combines precision and recall relative to a specific positive class. The F1-score can be interpreted as a weighted average of the precision and recall, where an F1-score reaches its best value at 1 and worst at 0.\n",
        "\n",
        "**2.Accuracy**:\n",
        "\n",
        "Accuracy in classification problems is the number of correct predictions made by the model over all predictions made.\n",
        "\n",
        "Accuracy is good measure when target class variable in the training dataset is nearly balanced.The distribution of the target class in provided training set is 57% for `0` class to 43% for `1` class which is about equally distributed.\n",
        "Due to this reason i chose 'accuracy' to be model performance evalution metric for this problem."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qnlfieitGeo8"
      },
      "source": [
        "### Model Improvements\n",
        "\n",
        "The result from `LogisticRegression()` classifier seems good enough than rest of the classifiers to be used as a final model.Based on the above results, it seems the classifier is not over fitting but lets try to improve the model by tuning some hyper parameters of the classifier."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1oPGu_ZYGeo8",
        "outputId": "67f44382-14bd-4826-9688-9b2d250df99e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "# Create the list of various hyper parameters to try\n",
        "solvers = ['newton-cg', 'lbfgs', 'liblinear']\n",
        "penalty = ['l2']\n",
        "c_values = [100, 10, 1.0, 0.1, 0.01]\n",
        "logreg= LogisticRegression()\n",
        "\n",
        "# Define and fit grid search\n",
        "grid = dict(solver=solvers,penalty=penalty,C=c_values)\n",
        "cv = RepeatedStratifiedKFold(n_splits=10, n_repeats=3, random_state=1)\n",
        "grid_search = GridSearchCV(estimator=logreg, param_grid=grid, n_jobs=-1, cv=cv, scoring='accuracy',error_score=0)\n",
        "grid_result = grid_search.fit(X_train_cv, y_train_cv)\n",
        "\n",
        "# Summarize and print results\n",
        "print(\"Best: %f using %s\" % (grid_result.best_score_, grid_result.best_params_))"
      ],
      "execution_count": 197,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Best: 0.795676 using {'C': 0.1, 'penalty': 'l2', 'solver': 'liblinear'}\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "t-Hc0hc5Geo9"
      },
      "source": [
        "From above results, it seems orignal classifier with default parameters performed best so that will be used as the final model for test data predictions."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2yfHINBkGeo-"
      },
      "source": [
        "# Lets apply pre-processing function to clean and pre-process text data before vectorizing\n",
        "test_df_final=test_df.copy()\n",
        "test_df_final['text'] = test_df_final['text'].apply(lambda x : pre_process_text_combined(x))"
      ],
      "execution_count": 198,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PFUU6WPWGepB"
      },
      "source": [
        "# Lets fit the test data with Countvectorizer() method to vectroize the data\n",
        "test_vector = count_vectorizer.transform(test_df_final[\"text\"])"
      ],
      "execution_count": 199,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "c_cQD490GepD"
      },
      "source": [
        "Lets create a function to generate predictions for the provided test data for submission."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LezXHWexGepD"
      },
      "source": [
        "# Define a function to generate predictions and store in a.csv file for submission on Kaggle\n",
        "def submission(sub_file,model,test_vector):\n",
        "    \n",
        "    '''Input- sub_file=Location of the file submission file\n",
        "              model=final fit model to be used for predictions\n",
        "              test_vector=pre-processed and vectorized test dataset\n",
        "       Output- submission file in .csv format with predictions       \n",
        "    \n",
        "    '''\n",
        "    sub_df = pd.read_csv(sub_file)\n",
        "    sub_df[\"target\"] = model.predict(test_vector)\n",
        "    sub_df.to_csv(\"submission.csv\", index=False)"
      ],
      "execution_count": 200,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "j6XpGQdoGepF"
      },
      "source": [
        "# Use Submission() function to generate submission file for posting on Kaggle\n",
        "sub_file = \"https://raw.githubusercontent.com/pavstar619/Data-Science-Capstone/master/sample_submission.csv\"\n",
        "test_vector=test_vector\n",
        "submission(sub_file,clf_logreg,test_vector)"
      ],
      "execution_count": 201,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SCwp6Zc05xO3",
        "outputId": "e5bef2d9-3cd3-4ed2-a0df-d23c87b55053",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17
        }
      },
      "source": [
        "''' from google.colab import files\n",
        "files.download(\"submission.csv\") '''"
      ],
      "execution_count": 202,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "\n",
              "    async function download(id, filename, size) {\n",
              "      if (!google.colab.kernel.accessAllowed) {\n",
              "        return;\n",
              "      }\n",
              "      const div = document.createElement('div');\n",
              "      const label = document.createElement('label');\n",
              "      label.textContent = `Downloading \"${filename}\": `;\n",
              "      div.appendChild(label);\n",
              "      const progress = document.createElement('progress');\n",
              "      progress.max = size;\n",
              "      div.appendChild(progress);\n",
              "      document.body.appendChild(div);\n",
              "\n",
              "      const buffers = [];\n",
              "      let downloaded = 0;\n",
              "\n",
              "      const channel = await google.colab.kernel.comms.open(id);\n",
              "      // Send a message to notify the kernel that we're ready.\n",
              "      channel.send({})\n",
              "\n",
              "      for await (const message of channel.messages) {\n",
              "        // Send a message to notify the kernel that we're ready.\n",
              "        channel.send({})\n",
              "        if (message.buffers) {\n",
              "          for (const buffer of message.buffers) {\n",
              "            buffers.push(buffer);\n",
              "            downloaded += buffer.byteLength;\n",
              "            progress.value = downloaded;\n",
              "          }\n",
              "        }\n",
              "      }\n",
              "      const blob = new Blob(buffers, {type: 'application/binary'});\n",
              "      const a = document.createElement('a');\n",
              "      a.href = window.URL.createObjectURL(blob);\n",
              "      a.download = filename;\n",
              "      div.appendChild(a);\n",
              "      a.click();\n",
              "      div.remove();\n",
              "    }\n",
              "  "
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "download(\"download_ebfdda20-c976-44e6-b282-126ed56c214f\", \"submission.csv\", 22746)"
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Yl8nSf3qGepH"
      },
      "source": [
        "# 6. Challenges\n",
        "\n",
        "Working on this data science capstone project went pretty smoothly without many issues. Even though every step in developing this project was smooth, following complications were faced:\n",
        "\n",
        "- There are only about 7000 data records for training which seems low and hence accuracy of model suffers.\n",
        "- I took the NLP challenge for my capstone project and this area is new and challenging for me since i had to start from scratch. \n",
        "- Learning new area and working on a capstone project was little difficult due to the fact that I don't have much time left  on my nano degree.  \n",
        "- New topic and less time also restricted me to test few more models from deep neural network area.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qkGjZAduGepH"
      },
      "source": [
        "# 7. Conclusion\n",
        "\n",
        "In this capstone project, I took a Kaggle challenge to classify tweets into disaster tweets in real or not?. First, I have analyzed and explored all the provided tweets data to visualize the statistical and other properties of the presented data.Next, I performed some exploratory analysis of the data to check type of the data, whether there are unwanted features and if features have missing data. Based on the analysis, I decided to drop ‘location’ column since it has most the data missing and really have no effect on classification of tweets. The ‘text’ columns is all text data along with alpha numeric, special characters and embedded URLs.The ‘text’ column data needs to be cleaned and pre processed and vectorized before it can be used with a machine learning algorithm for the classification of the tweets.\n",
        "\n",
        "After pre processing the train and test data,the data was vectorized using `CountVectorizer` and `TFIDF` features and then it was split into training and test data, and then various classifiers were fit on the data and predictions were made.Out of all classifiers tested, LogisticRegression() performed the best with the test accuracy of 81%. An effort was made to tune some hyper parameters of the final classifier to see if accuracy can be improved.It turned out that classifier with default parameters preformed little better than tuned model .\n",
        "\n",
        "Finally, `LogisticRegression()` classifier with default parameters was selected as final model for this NLP classifications challenge and predictions were made on the test data. Kaggle submission³ of the predictions resulted in an accuracy of 79%."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "g0eK9oD3GepH"
      },
      "source": [
        "# 8. Improvements\n",
        "\n",
        "I believe there are definitely opportunities to try out various machine learning models to see if there are better performing models could be found to solve this problem with higher than 79% accuracy.\n",
        "\n",
        "Trying neural network and deep neural network models such as `RNN(Recurrent Neural Network)` might result in higher accuracy and better performance but may be very costly in terms of compute resources used.\n",
        "\n",
        "Another way to improve the accuracy of the models would be to have more training data than given."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_PZlXhPAGepI"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WRyPmMjpGepJ"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}