boston_housing.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 机器学习工程师纳米学位\n",
    "## 模型评价与验证\n",
    "## 项目 1: 预测波士顿房价\n",
    "\n",
    "\n",
    "欢迎来到机器学习工程师纳米学位的第一个项目！在此文件中，有些示例代码已经提供给你，但你还需要实现更多的功能来让项目成功运行。除非有明确要求，你无须修改任何已给出的代码。以**'练习'**开始的标题表示接下来的内容中有需要你必须实现的功能。每一部分都会有详细的指导，需要实现的部分也会在注释中以**'TODO'**标出。请仔细阅读所有的提示！\n",
    "\n",
    "除了实现代码外，你还**必须**回答一些与项目和实现有关的问题。每一个需要你回答的问题都会以**'问题 X'**为标题。请仔细阅读每个问题，并且在问题后的**'回答'**文字框中写出完整的答案。你的项目将会根据你对问题的回答和撰写代码所实现的功能来进行评分。\n",
    "\n",
    ">**提示：**Code 和 Markdown 区域可通过 **Shift + Enter** 快捷键运行。此外，Markdown可以通过双击进入编辑模式。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 开始\n",
    "在这个项目中，你将利用马萨诸塞州波士顿郊区的房屋信息数据训练和测试一个模型，并对模型的性能和预测能力进行测试。通过该数据训练后的好的模型可以被用来对房屋做特定预测---尤其是对房屋的价值。对于房地产经纪等人的日常工作来说，这样的预测模型被证明非常有价值。\n",
    "\n",
    "此项目的数据集来自[UCI机器学习知识库](https://archive.ics.uci.edu/ml/datasets/Housing)。波士顿房屋这些数据于1978年开始统计，共506个数据点，涵盖了麻省波士顿不同郊区房屋14种特征的信息。本项目对原始数据集做了以下处理：\n",
    "- 有16个`'MEDV'` 值为50.0的数据点被移除。 这很可能是由于这些数据点包含**遗失**或**看不到的值**。\n",
    "- 有1个数据点的 `'RM'` 值为8.78. 这是一个异常值，已经被移除。\n",
    "- 对于本项目，房屋的`'RM'`， `'LSTAT'`，`'PTRATIO'`以及`'MEDV'`特征是必要的，其余不相关特征已经被移除。\n",
    "- `'MEDV'`特征的值已经过必要的数学转换，可以反映35年来市场的通货膨胀效应。\n",
    "\n",
    "运行下面区域的代码以载入波士顿房屋数据集，以及一些此项目所需的Python库。如果成功返回数据集的大小，表示数据集已载入成功。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Boston housing dataset has 489 data points with 4 variables each.\n"
     ]
    }
   ],
   "source": [
    "# Import libraries necessary for this project\n",
    "# 载入此项目所需要的库\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import visuals as vs # Supplementary code\n",
    "from sklearn.model_selection import ShuffleSplit\n",
    "\n",
    "# Pretty display for notebooks\n",
    "# 让结果在notebook中显示\n",
    "%matplotlib inline\n",
    "\n",
    "# Load the Boston housing dataset\n",
    "# 载入波士顿房屋的数据集\n",
    "data = pd.read_csv('housing.csv')\n",
    "prices = data['MEDV']\n",
    "features = data.drop('MEDV', axis = 1)\n",
    "    \n",
    "# Success\n",
    "# 完成\n",
    "print \"Boston housing dataset has {} data points with {} variables each.\".format(*data.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 分析数据\n",
    "在项目的第一个部分，你会对波士顿房地产数据进行初步的观察并给出你的分析。通过对数据的探索来熟悉数据可以让你更好地理解和解释你的结果。\n",
    "\n",
    "由于这个项目的最终目标是建立一个预测房屋价值的模型，我们需要将数据集分为**特征(features)**和**目标变量(target variable)**。**特征** `'RM'`， `'LSTAT'`，和 `'PTRATIO'`，给我们提供了每个数据点的数量相关的信息。**目标变量**：` 'MEDV'`，是我们希望预测的变量。他们分别被存在`features`和`prices`两个变量名中。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 练习：基础统计运算\n",
    "你的第一个编程练习是计算有关波士顿房价的描述统计数据。我们已为你导入了` numpy `，你需要使用这个库来执行必要的计算。这些统计数据对于分析模型的预测结果非常重要的。\n",
    "在下面的代码中，你要做的是：\n",
    "- 计算`prices`中的`'MEDV'`的最小值、最大值、均值、中值和标准差；\n",
    "- 将运算结果储存在相应的变量中。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Statistics for Boston housing dataset:\n",
      "\n",
      "Minimum price: $105,000.00\n",
      "Maximum price: $1,024,800.00\n",
      "Mean price: $454,342.94\n",
      "Median price $438,900.00\n",
      "Standard deviation of prices: $165,171.13\n"
     ]
    }
   ],
   "source": [
    "# TODO: Minimum price of the data\n",
    "#目标：计算价值的最小值\n",
    "minimum_price =np.min(prices)\n",
    "\n",
    "# TODO: Maximum price of the data\n",
    "#目标：计算价值的最大值\n",
    "maximum_price = np.max(prices)\n",
    "\n",
    "# TODO: Mean price of the data\n",
    "#目标：计算价值的平均值\n",
    "mean_price = np.mean(prices)\n",
    "\n",
    "# TODO: Median price of the data\n",
    "#目标：计算价值的中值\n",
    "median_price = np.median(prices)\n",
    "\n",
    "# TODO: Standard deviation of prices of the data\n",
    "#目标：计算价值的标准差\n",
    "std_price = np.std(prices)\n",
    "\n",
    "# Show the calculated statistics\n",
    "#目标：输出计算的结果\n",
    "print \"Statistics for Boston housing dataset:\\n\"\n",
    "print \"Minimum price: ${:,.2f}\".format(minimum_price)\n",
    "print \"Maximum price: ${:,.2f}\".format(maximum_price)\n",
    "print \"Mean price: ${:,.2f}\".format(mean_price)\n",
    "print \"Median price ${:,.2f}\".format(median_price)\n",
    "print \"Standard deviation of prices: ${:,.2f}\".format(std_price)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 问题1 - 特征观察\n",
    "\n",
    "如前文所述，本项目中我们关注的是其中三个值:`'RM'`、`'LSTAT'` 和`'PTRATIO'`，对每一个数据点:\n",
    "- `'RM'` 是该地区中每个房屋的平均房间数量；\n",
    "- `'LSTAT'` 是指该地区有多少百分比的房东属于是低收入阶层（有工作但收入微薄）；\n",
    "- `'PTRATIO'` 是该地区的中学和小学里，学生和老师的数目比（`学生/老师`）。\n",
    "\n",
    "_凭直觉，上述三个特征中对每一个来说，你认为增大该特征的数值，`'MEDV'`的值会是**增大**还是**减小**呢？每一个答案都需要你给出理由。_\n",
    "\n",
    "**提示：**你预期一个`'RM'` 值是6的房屋跟`'RM'` 值是7的房屋相比，价值更高还是更低呢？"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**回答: **增大‘RM’的值，'MEDV'的值会增大，因为按每平米单价来计算，房间越多对应的房子总面积越大，总价格也越大；\n",
    "                  增大‘LSTAT’的值，‘MEDV’的值会减小，因为低收入阶层的可支付能力比较弱，他们可以购买的房子价格也比较便宜，他们的比例上升，说明购买低价                   格房屋的比例上升。\n",
    "                  增大‘PTRATIO’的值，‘MEDV’的值会减小，相比高收入家庭，低收入家庭的子女就读的学校一般师资力量差，老师数量少，所以这个值增大表示低收入                   阶层增加，和增大‘LSTAT’是同样的效果"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 建模\n",
    "在项目的第二部分中，你需要了解必要的工具和技巧来让你的模型进行预测。用这些工具和技巧对每一个模型的表现做精确的衡量可以极大地增强你预测的信心。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 练习：定义衡量标准\n",
    "如果不能对模型的训练和测试的表现进行量化地评估，我们就很难衡量模型的好坏。通常我们会定义一些衡量标准，这些标准可以通过对某些误差或者拟合程度的计算来得到。在这个项目中，你将通过运算[*决定系数*](http://stattrek.com/statistics/dictionary.aspx?definition=coefficient_of_determination) R<sup>2</sup> 来量化模型的表现。模型的决定系数是回归分析中十分常用的统计信息，经常被当作衡量模型预测能力好坏的标准。\n",
    "\n",
    "R<sup>2</sup>的数值范围从0至1，表示**目标变量**的预测值和实际值之间的相关程度平方的百分比。一个模型的R<sup>2</sup> 值为0还不如直接用**平均值**来预测效果好；而一个R<sup>2</sup> 值为1的模型则可以对目标变量进行完美的预测。从0至1之间的数值，则表示该模型中目标变量中有百分之多少能够用**特征**来解释。_模型也可能出现负值的R<sup>2</sup>，这种情况下模型所做预测有时会比直接计算目标变量的平均值差很多。_\n",
    "\n",
    "在下方代码的 `performance_metric` 函数中，你要实现：\n",
    "- 使用 `sklearn.metrics` 中的 `r2_score` 来计算 `y_true` 和 `y_predict`的R<sup>2</sup>值，作为对其表现的评判。\n",
    "- 将他们的表现评分储存到`score`变量中。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# TODO: Import 'r2_score'\n",
    "from sklearn.metrics import r2_score\n",
    "def performance_metric(y_true, y_predict):\n",
    "    \"\"\" Calculates and returns the performance score between \n",
    "        true and predicted values based on the metric chosen. \"\"\"\n",
    "    \n",
    "    # TODO: Calculate the performance score between 'y_true' and 'y_predict'\n",
    "    score = r2_score(y_true,y_predict)\n",
    "    \n",
    "    # Return the score\n",
    "    return score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 问题2 - 拟合程度\n",
    "\n",
    "假设一个数据集有五个数据且一个模型做出下列目标变量的预测：\n",
    "\n",
    "| 真实数值 | 预测数值 |\n",
    "| :-------------: | :--------: |\n",
    "| 3.0 | 2.5 |\n",
    "| -0.5 | 0.0 |\n",
    "| 2.0 | 2.1 |\n",
    "| 7.0 | 7.8 |\n",
    "| 4.2 | 5.3 |\n",
    "*你觉得这个模型已成功地描述了目标变量的变化吗？如果成功，请解释为什么，如果没有，也请给出原因。*  \n",
    "\n",
    "运行下方的代码，使用`performance_metric`函数来计算模型的决定系数。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model has a coefficient of determination, R^2, of 0.923.\n"
     ]
    }
   ],
   "source": [
    "# Calculate the performance of this model\n",
    "score = performance_metric([3, -0.5, 2, 7, 4.2], [2.5, 0.0, 2.1, 7.8, 5.3])\n",
    "print \"Model has a coefficient of determination, R^2, of {:.3f}.\".format(score)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**回答:**我认为已经成功描述了目标变量的变化，因为决定系数是回归平方和与总平方和的比值，这个比值越大，说明残差平方和越小，说明拟合度越高，0.923这个数字已经很接近1了。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 练习: 数据分割与重排\n",
    "接下来，你需要把波士顿房屋数据集分成训练和测试两个子集。通常在这个过程中，数据也会被重新排序，以消除数据集中由于排序而产生的偏差。\n",
    "在下面的代码中，你需要：\n",
    "- 使用 `sklearn.model_selection` 中的 `train_test_split`， 将`features`和`prices`的数据都分成用于训练的数据子集和用于测试的数据子集。\n",
    "  - 分割比例为：80%的数据用于训练，20%用于测试；\n",
    "  - 选定一个数值以设定 `train_test_split` 中的 `random_state` ，这会确保结果的一致性；\n",
    "- 最终分离出的子集为`X_train`,`X_test`,`y_train`,和`y_test`。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training and testing split was successful.\n"
     ]
    }
   ],
   "source": [
    "# TODO: Import 'train_test_split'\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "# TODO: Shuffle and split the data into training and testing subsets\n",
    "X_train, X_test, y_train, y_test = train_test_split(features, prices, train_size=0.8, random_state=0)\n",
    "\n",
    "# Success\n",
    "print \"Training and testing split was successful.\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 问题 3- 训练及测试\n",
    "*将数据集按一定比例分为训练用的数据集和测试用的数据集对学习算法有什么好处？如果用模型已经见过的数据，例如部分训练集数据进行测试，又有什么坏处？*\n",
    "\n",
    "**提示：** 如果没有数据来对模型进行测试，会出现什么问题？"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**答案: **这样会避免出现过度拟合，可以保证后面有真实的数据拿来测试。如果用训练集数据进行测试，那么精确率会达到100%，不利于构建泛化的数据模型。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----\n",
    "\n",
    "## 分析模型的表现\n",
    "在项目的第三部分，我们来看一下几个模型针对不同的数据集在学习和测试上的表现。另外，你需要专注于一个特定的算法，用全部训练集训练时，提高它的`'max_depth'` 参数，观察这一参数的变化如何影响模型的表现。把你模型的表现画出来对于分析过程十分有益。可视化可以让我们看到一些单看结果看不到的行为。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 学习曲线\n",
    "下方区域内的代码会输出四幅图像，它们是一个决策树模型在不同最大深度下的表现。每一条曲线都直观的显示了随着训练数据量的增加，模型学习曲线的训练评分和测试评分的变化。注意，曲线的阴影区域代表的是该曲线的不确定性（用标准差衡量）。这个模型的训练和测试部分都使用决定系数R<sup>2</sup>来评分。\n",
    "\n",
    "运行下方区域中的代码，并利用输出的图形回答下面的问题。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0AAAAIKCAYAAAAK1y/8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzsnXl8FdX5/9/PvVluVkjIRlhEkUVEoAUVERWrVlRQ64J1\nt25Va7WiVuourUtbS9svirhUUbRq1Z+tVMWqWJeKIi4IgqKyQxJ2sid3Ob8/zkwyubkJSUjI9rxf\nr3nNzJkzZ84s99z5zPOc54gxBkVRFEVRFEVRlO6Ar70roCiKoiiKoiiKsrdQAaQoiqIoiqIoSrdB\nBZCiKIqiKIqiKN0GFUCKoiiKoiiKonQbVAApiqIoiqIoitJtUAGkKIqiKIqiKEq3QQWQ0qEQkYtE\nxHimMhFZIyIvi8gUEZE2PPadItKsuPAi8l8R+W8bVSnW8eZEXZ+Gpgl7q06N1PW+qDpVisgyEbmu\nLe9jV8e5rqH2rkdzEJGA8wxMa4djf+R5BiMisklEXhWR0W1wrDtFZL2IhEXko9YuX1EURWkd4tq7\nAorSAGcCG4BEoD9wEvAscLmITDbGVLTBMR8D5jdzn6vaoB6N8Vtgtmf9UuASYDwQ9qQv35uVaoQw\ntm4AvbD1nQFUAw+2V6WUvU4VcBiwrp2O/wlwDSDAvsBtwHsiMtIY811rHEBEjgTuAO4GXgN2tUa5\niqIoSuujAkjpqHwR9WIyV0ReAF4A/gD8srUPaIzZgBVdzdlnrwoNY8z3wPfuuohMdBY/Nsbs1iog\nIonGmKq2ql8sjDE1X8JF5A1gJXAZe1kAtce5t5SOXlcR8QPSlGcOwNgRt9vTIlLseQ4Xishi4Bvg\ncuDXe1Kw514d4CTNMsZs2pMyo8pVFEVRWhl1gVM6DcaYl4B/AZeJSLKbLiLJIvJ7EVktItXO/BYR\nqfN8i0i2iMxyXFSqnPlcEUl0ttdzgRORa0VkhYhUiMgOEVksIj/xbK/nAiciQxyXvZ3Ofh95hIqb\n507HJWeQ445TKiJrReT26Hq3FBGZ6BxjsuM6tw1Y69k+WkT+7anneyJyWIxyjnXOs9SZXhWRA6Lz\nNQXnhXkp1qoXfZym1udGEVnn5FkoIgeLSKGIzPbkucI598Oce7ELeLc55yQik5x7V+zkWeF14RKR\nYSLyiohscdz71onI81FlDBOReSKyy6nv/0TkmKg894lISESGi8jbIlIKPNX8q1vvOp0lIotEpNx5\ndp8TkT5ReS4QkXedcygRkU9F5JyoPK772u0icpuIrMVa8AZ5nrGJIvKwiGwXkc3O85Yeowzv9XPP\ne5CIvCHW3XW1iPxGpK6LpIgcIiIfeq7zDc7+lS25NsaYlUAJsL/nGLki8qiIFIhtH5aLyM+i6hHz\nuRLr7uY+fxu95yoiPUXkIecZrRaRr0Xk6qhyG/yteq7TUBF5y7mfa0TkPGf7xSKy0nlG3xKRfaLK\nbs49vlVErhfbFpU4z+OQ6Osn1h35I+eeFTvLJ3i2xzvPykrnWm4Q20YnROW5V0RWOfd1q4i8LyKH\nNv1OKoqitAy1ACmdjdeAU4ExWBeWOOANYBjWPWwpMBbr4pIJXA8gIhnAh07a74AvgRzgFCAB66JT\nBxE5F/gTMB14H0gCRjhlxERE8oEPsC9XV2PdYH4BvCoik4wxr0ft8jLwBPBnYDJwF7DeSWstZgPz\ngLOBgFPPscA72K/ylwCVTn0XiMghxpilTr7TsFa3l4FzAD/wG+y1H2GMKWhBfQbgsWI1sz5XYy2A\ns4F/AoOd+qU2cKzngGeAB5y6N+mcRGQo8P+Av2PdmkLAIKCfU4YAr2Mthj8HtgF9sa6a7jntA/wP\n2AJcCZQC1wJviMiPjTELvJcAK+4fwT6fXnfGZiMiv8K6Gj7q1L8n9jl+R0RGGWPKnaz7OdfiWyAC\nHI21tiYYY+ZEFftzrNXkV9j7sxlwX7ZnYe/HWcBw4D7sb+rnu6sq9jr/DfgjcBpwD7AG6/KKiPQG\n3gJWA+dhr82NQJ/6xTUNEckG0oCdznoGsNDZfCvWVe8k4G8iEmeMeTSqiOjnaiPwM2x7cxKwHVgX\n1T7dCnyNbXNmikimMWZ6VLn1fqtulbH3aTb2+b8GeEpEDgTGOdcjCfgrMBc40rNvc+7xpcBX2N9e\nMnA/8LKIDDfGRJxrdQP2Xr0A/B6oAEZT+ywA/AM4DnsvF2GfienY38i5Tp7bsW3jb4BlQA/gEBpp\nXxVFUVoNY4xOOnWYCbgIMMD+DWw/3tl+lrN+vrN+ZFS+W7BfqXOc9enYF6cfNHLsO3G8dZz1B4DP\ndlPf/wL/9azfj31Z3t+T5se+OH4WfSzgZ1HlLQX+04zr5ZYTF2PbRGfbszG2/Q9Y4t0PiMcKk+ec\ndR9WjL0WtW8m9sXxvt3U7T7nWsQ5U65T3yBwQgvqEw8UAv8vat9znPOc7Um7wkm7Nypvk84J+6Id\nARIbOLe+Tvk/buT8H8CKgP5R57Qa+DDqOhng50285/cBoUa29wTKsK5Y3vTBzv24ooH9fM59mot1\nqXTTA0791gAJDTxjD0elP4Z1O4suY1qM8z7bkyZYF8lXPGkzsC/ZuZ60VKzIqGzC9foIeNs5t3jn\nOrzmHHuik+du55oNiNp3LrAJ8DX2XDnbrna25XnSznDSfhqV92mgHOjRhN+qe52meNJynLQiIMWT\n/msnPbeBa7G7e/wV4Pekn+ek/9BZ7+Xci783cr2Pi66vk36Jk36As/5WY+XopJNOOrXlpC5wSmfD\ndY1xXdUmYl1FPhSROHcC/oN92Rnr5Psx8Ikx5vNmHOsTYJSIzBTrMpW82z3sl9ePjKf/kjEmjP2a\nPUo8bkEOr0atLyOGe9ge8rJ3xanDYcDzzrp7zQywgNqvxwdiX/Sfjrq2xdhr4/3K3BB+rOAJYsXL\nHcD1xmMJa0Z99sWKqBeijvEStc9Do+fejHP6DCuAXhCR00QkK6qcQqz1534RuUREBsY49pHA+8aY\nmo7/xpigc56HiEggKn90XVvKEdiv989EneMqZ6q5b45b1T9EZBNWHAWxL7313J6worG6gWNGP8dL\ngTQR6dmE+tbsa4xxX8K9v4GxwHvGmCJPvlKaF7DkR9hzq8Z+jPghcIkxxi1jItZyuyHqmr0B9Mbj\nKufQ1Ht1pHPM6Gf2aazF5pBmlFvzmzHGbAZ2AB8YY8o8eb525v3chGbe4zec9splqTN378cRWLH0\nSCP1nIgVk/+K0Sa7ZYD9vZ0qItNFZJyIxDdSpqIoSquiAkjpbLh/7K7rVQ7W9SIYNS1ytvfyzJsV\n4ADbD+NK4FDsi9B2Efl/IjKgkX0yPXXzUogVbxlR6duj1quo6/rSGkTXJ9upy93Uv26XUnvNcpz5\nMzHyHevJ1xhh4GDsi97pWNfDGVK3b09T69PbmW/2HsDYjuINRdyKPvcmnZOxwS1OwN6LvwNFYvvv\nHO5sD2Ffqr/EugN9JyLficglnmM19iz4sS4/LhHnpbY1cM/xA+qf4yCcc3TEyVvAUKwL1XjsvXqG\n2M9gY+6OsZ5jGijHS9gYUxxjX+9+vYm65w5FMdIaYhH23EZjBXBvY8zjnu052I8k0ddrrrM9+llv\nqutnJrA5SlSAfQbc7U0pN2yMKYlKq8aKoOg0qHV1be493t19dK9DY21pDpCCdZP0Xkv3Q4Bbxp3Y\n3/wZWAvwVrF9sKLbSEVRlFZH+wApnY2TsH+snzrr27AuRVMayL/GmW+lmX0GnK/RDwMPO3/KP8b2\nCXoeK4pisR3Ii5Geh7VSRL+w7A2irSPuS86fsH0ZGsq/zZlfD7wXI1+TOqAbYxY7i5+IyIfYr9Qz\nsf24mlMfr+itQWwQix7EJvrcm3xOxpg3gTcdS814nPDGItLfGLPLGPMtcJ7YoBWjsH1jHhORVcaY\nd2j8WQhTV7Q1a/yp3eCe4znYfh/RuILjCOxv4lTPPaKRL/GtWcfmUEDUPXfIbUYZJd5zjME24Dus\nSIjF11HrTb0W24FsEfEZpw+NQ55ne0vKbSrNvce7Y6sz74O9XrHYhu0D+aMGtm+Emg8XdwN3O/28\nTsa2AQnAhS2sn6IoSpNQAaR0GkTkdOyf5F9NbSfu+VjLQqkxJvolxct/gFvFjvuxpLnHNsbsAJ53\nIhQ11rH7XeBXIjLAGLPGqbcf2zn88xhfu/c6xpgdIvIxNqDDjY7Qi8VSbP+HA4wxM1rp2IUicg/w\ne7HjOc1rRn1WY7/6n4nTQd7hDGpdI3dHs8/JGFMJvCUimVjx259a1yCcF9vPnM7h52M7fL+DfRYu\nE5F844RFdlyBpmD7X7QoglkTeA/bT2M/Y8yzjeRzXTqDboKI5AAntlG9WspHwJUikuu6wYlIKtbV\nqrWYj+1/uMoYEy1K9oR3sSH7f4J11XQ5F3uPFsXaqRVp7Xv8Prbel+OJqhjFfGywj0RjzP+aUqix\nwVQeFpFTsL8fRVGUNkUFkNJRGeX0u0jAvnBOwr74vomNGuTyDDb60tsi8idsR/oEYCBWLJ3qiKU/\nY7+IvyUiv8O+wGZhIzJdEcO9BBF5BPslcyHWBWcw9gX3P9F5PfwZ+yL1pojcgf3afpWz70mN7Le3\n+RW2f81rIjIH65KTjbXKBI0xtxljwk7UtRec/k8vYb/u5gGHAyuNMQ+04NgPAFOxkfrmNaM+Qefe\nzRSRh7D9JQYDN2D7HETYDU09JxG5BusqNB/r7pMN3Ix14/laRA7BRrj6BzZQQzzWXa8aGxgDbECM\n87DP5l1OHa/BPs+XNvOaRSMickaM9A3GmI/EhmD+k9iohG9gn+M+2AhgrxtjXsS+zJZhXzynA+nY\nyFxFWDexjsIfseNG/UdEfovtx/JrbFS91nIX/QNWSH8gIn/BBmJIw47tc6gx5vQWlvsvrMh53LkX\n32DbpfOAO4wxbT1YaqveY2PMdhG5HfijY/l8HhvM4QfALmPMbGPMfBH5f9g+QDMA1/K0L7YN/KUx\nZq2IvA58DHyODUAyBms1+nPLT1dRFKVpqABSOipup2E33O5nwE+BF70WAuel+HhgGvar5L7YP/zv\nsZ2rq518O53+G79z8vbCvgQsoNZvPpr/YcXV+VgXq03Yzst3NFRpY8wmERmPDQ/7EJAIfAGc5Olw\n3e44L8mHYl+GHsC+GG3Gvqw85Mn3sogcjX35/xu243YBVhQ+3cJjlztWoL+KyAnGmNebUZ8HHOHy\nS+y9WYINGfw2DfcDij5+U87pc2w0q99jxc827BdvV4htdPZxwzFXYPsDnWickN3OS954bBSvR7Ai\n6XNs5DFvCOyW4KN+x3qwgu4MY8z/iR2vZypwAbWhmt/FsV45z+rp2Jf/l7BCbwa2T92v9rB+rYax\nYcmPBf6C/eCxGTuI7gDsB4zWOMZ25/m7AxtBsjfWXfVrnOAcLSw35LRP9znlZmIDUfyyhR8Pmnv8\nVr/Hxpj7ned/KtYSW40NXHGXJ9sUp/yLsL/pSqwF9w1qXTTfww5pcA1WyK7Dts/3tqReiqIozUEa\n9jZRFEXp+Dgi431s2N1YokDpYjh9WL7Euqx1JMuqoiiK0glQC5CiKJ0GERmMdR9zB5sdjrXkrARe\naceqKW2IiNwLrMBaCbKw/fCGYAfSVBRFUZRmoQJIUZTORAU24trPsAN+bsf2ybrJiSqldE182LDJ\n+dgIel8Ak1rBlVBRFEXphqgLnKIoiqIoiqIo3QYdCFVRFEVRFEVRlG6DCiBFURRFURRFUboNKoAU\nRVEURVEURek2qABSFEVRFEVRFKXboAJIURRFURRFUZRugwogRVEURVEURVG6DSqAFEVRFEVRFEXp\nNqgAUhRFURRFURSl26ACSFEURVEURVGUboMKIEVRFEVRFEVRug0qgJRugYhMEJENbVT2ABExIhLX\nFuUriqJ0FbQtVhSlI6ACSFGaiYisEZFj2+G4U0TkQxEpF5H/7u3jK4qidCTasS3+g4isF5FiEVkr\nIjfv7TooirJnqABSlM7DduAvwH3tXRFFUZRuzOPAgcaYdGAccK6InNbOdVIUpRmoAFLaBOfL3I0i\n8qWIlInI30QkV0ReF5ESEXlLRDI8+V8QkUIR2SUi74nIgU56goh8ISK/dNb9IvI/Ebl9N8dPEpE5\nIrJDRJYDB0dtzxeRl0Rki4isFpFrPNvuFJEXReR5p66fichIZ9tcoD8wT0RKReTXnmLPFZF1IrJV\nRG7Z02sYjTHmLWPMP4BNrV22oihdE22L26Qt/toYU+xJigD7t/ZxFEVpO1QAKW3J6cBxwGBgMvA6\ncDOQjX32rvHkfR0YBOQAnwHPABhjqoHzgOkicgAwDfADd+/m2HcAA53peOBCd4OI+IB5wBKgD3AM\n8CsROd6z/ynAC0Am8HfgnyISb4w5H1gHTDbGpBpj/uDZZzwwxCnvdqe+9RCRaSKys6FpN+elKIrS\nXLQtjsGetMXOvqXABiDFqZuiKJ0EFUBKWzLTGFNkjNkIvA98bIz53BhTCbwM/MDNaIx53BhTYoyp\nAu4ERopID2fbMuB3wD+BG4DzjTHh3Rx7CnC3MWa7MWY98H+ebQcD2caY6caYamPMKuBR4KeePJ8a\nY140xgSBGUAAGLubY95ljKkwxizB/qGPjJXJGHOfMaZnQ9NujqEoitJctC2OwZ60xcaY+4A04IfA\nXGDXbuqkKEoHQgWQ0pYUeZYrYqynQo0rxX0i8r2IFANrnDxZnvxPAvsArxljvm3CsfOB9Z71tZ7l\nfYD8qC99NwO5njw1+xpjItivfPm7OWahZ7kc5/wURVHaGW2L2wBj+Rx7De9qi2MoitI2qABSOgLn\nYN0cjgV6AAOcdPHkmQX8GzheRMY3ocwCoJ9nvb9neT2wOuprX5ox5kRPnpp9HTeNvtT2vTFNOH6D\niMjNjs96zGlPylYURdkDtC1uWVsch3XxUxSlk6ACSOkIpAFVwDYgGbjHu1FEzgdGAxdhfdWfFJHd\nfdH7B/AbEckQkb7ALz3bFgElInKT00HXLyLDRcTbOXe0iJwmdjyJXzn1+8jZVgTs15ITBTDG3OP4\nrMecGtrPqWcA+2frE5GAiMS3tB6KoihRaFu8m7ZYRHwi8nPnfEREDgF+Abzd0nooirL3UQGkdASe\nwrpFbASWU/vnhoj0x4Z+vsAYU2qM+TuwGPjzbsq8yylzNfAfrI82AI7P+iRglLN9K/AY9ouny7+A\ns4AdwPnAaY4POsC9wK2Oy8YNLTnhFnI+1tXiIeAIZ/nRvXh8RVG6NtoWN42fAN8DJcDTwExnUhSl\nkyDG7JEFWVG6HCJyJ7C/Mea89q6LoihKd0XbYkVR2gq1ACmKoiiKoiiK0m1QAaR0WsQO5Ber8+rN\n7V03RVGU7oK2xYqidDbUBU5RFEVRFEVRlG6DWoAURVEURVEURek2xLV3BZpLVlaWGTBgQHtXQ1EU\npdl8+umnW40x2e1dj9ZA22JFUTojXakdVlpOpxNAAwYMYPHixe1dDUVRlGYjImvbuw6thbbFiqJ0\nRrpSO6y0HHWBUxRFURRFURSl26ACSFEURVEURVGUboMKIEVRFEVRFEVRug0qgBRFURRFURRF6Ta0\nmQASkcdFZLOILGtgu4jI/4nIdyLypYj8sEkFL10KPh8MGADPPNOaVVYURelyaFusKIqiKHVpSwvQ\nHGBiI9tPAAY50+XAQ00qtboajIG1a+Hyy/WPV1EUpXHmoG2xoiiKotTQZgLIGPMesL2RLKcATxnL\nR0BPEendrIOUl8Mtt+xBLRVFUbo22hYriqIoSl3asw9QH2C9Z32Dk1YPEblcRBaLSP1BJ9ata5va\nKYqidA+0LVYURVG6FZ0iCIIx5hFjzBhjzJh6G/v3b4caKYqidD8abYuNgSlTYOFCCAbboXaKoiiK\n0jTaUwBtBPp51vs6aU0nEIC7727NOimKonQ3WqctHjMGXnkFxo2Do4+GJ56AwkLrHmdMa9ZXURRF\nUfaI9hRArwAXOBGIxgK7jDEFu90rIQFE7PKoUXDuuW1aSUVRlC7OnrXF++wDjz0GH3wA770Hl10G\ny5fDxRdbITRjBnzyiXWRKy6GcLjNT0hRFEVRGkNMG32ZE5FngQlAFlAE3AHEAxhjZouIAA9goxOV\nAz8zxtT3K49izJgxZvHixXDppfDkk/D55zB8eJucg6IoSmsiIp/GdB9r22O2bVscTXGxFUCvvALP\nPQerV0N+Pvz0p3DSSZCSApmZdkpNhcTE1jpVRVGU3dIe7bDS8WgzAdRW1Pzprl8PAwfCySfDP/5h\nx6NQFEXpwHSlP94GBRBYK09RkQ2R/fHH8Pe/WytQWpoVQmeeCenpNm8gADk5dj05udbCryiK0gZ0\npXZYaTlx7V2BFtOvH1x4ofUz//RTOPjg9q6RoiiKAuD3W6uPa+kZOxZWrYK5c+Fvf4M5c2DSJPjZ\nz2Dffe0HLWPsfpmZ0KuXtRTFdd6/KEVRFKXj0nktQFBrBZo4EV580fqkK4qidFC60pfHRi1AXoyB\nnTutK1woZF3knnrKttkVFTB+PFxyCRx2GEQiNi0YtJagtDTIzraucoFA25+Uoihdnj1phz/99NOc\nuLi4x4DhdJJIyt2UCLAsFApdOnr06M2xMnTuz2v9+sEFF1gr0MKFcNRR7V0jRVEUxYsIZGRYMVNQ\nYEXO1Klw9dW2j9DcudYSdMABNnDCCSdYwWMMVFVZy5Extq9QVhb07Gld5dTtWVGUvUxcXNxjeXl5\nB2RnZ+/w+Xydy4LQjYhEIrJly5ZhhYWFjwEnx8rT+f9B7rjDukn84Q9QVtbetVEURVFiERdnP1qN\nGGGt9ZGItfy8844dzqC6Gm68EY49Fh5/3LbngYAVPBkZEB9vw2p/9ZV1e/7uO9ixQ8ccUhRlbzI8\nOzu7WMVPx8bn85ns7OxdWEtd7Dx7sT5tQ79+cP75MH++DcHayVz6FEVRuhXJydbaM2iQdXerqIDT\nT4d//xseecSG1f79761F//e/t1YjsAIoPb3WmlRSAitXwmefWVFUVKRjDimK0tb4VPx0Dpz71KDO\n6fwCCOC22+yf48yZ9ougoiiK0nERsYEORo60bm3bt1t3t6OOsv2DXnoJJkywQx0ce6y1DK1YUbu/\nz2eDJGRk2CkSsRHnli61QyOsWgVbt0JpqVqIFEXpMhQWFvqHDh06bOjQocOysrJG5uTkjHDXKysr\nmxRC84wzzhiwZMmSRscfuPfee7MfeuihzNao89NPP91z6NChw4YMGTJs4MCBB86YMSOrNcrdUzp3\nEAQvF19s/yxfegkmT7bRhBRFUToQ3TIIQlMoLbWipaLCWnnc9nvjRtuuv/CCte6MG2fb+vHjGw6X\nHQpZMVVdXZsWH2+tRunpkJRk+xN5B9VWFKXbsCft8JIlS9aMHDlya5N3mD07k+nT+1BYmEBeXjW3\n376RK67Y3pJjRzN16tT81NTU8PTp04u86ZFIBGMM/g7wHlxRUSH77LPPQYsWLVoxYMCAYEVFhXz7\n7bcJI0aMqGppmc05vyVLlmSNHDlyQKxtXcMCBHDrrfZP7pFHrCuEoiiK0jlITbUDWu+7r3VtKy62\n6X36wM03w3//C9dfD99+awfBPvlkePnluiLHJS6urnUoI8OKnvJyayVasQKWLIHFi+Hrr63I2rnT\niq9IZK+etqIoXZjZszO57rp9KChIwBgoKEjguuv2YfbsVrGseFm2bFniwIEDDzz55JP3HTRo0IHr\n1q2LP/vss/cZPnz4Afvvv/+BN9xwQ2837+jRo4d8+OGHScFgkLS0tFFXXXVVnyFDhgwbNWrU0I0b\nN8YBXHPNNfnTp0/PcfNfddVVfQ466KADBgwYMPzNN99MASguLvYdf/zxAwcOHHjgxIkT9xs+fPgB\nH374YZK3Xtu3b/cbY8jJyQkBJCUlGVf8rFu3Lu6YY44ZOHjw4GFDhgwZtmDBghSAW2+9NXfQoEEH\nDho06MC77747p6Hz+8c//pE+atSoocOGDTvgpJNO2q+4uLhZmqZzR4Hzst9+cM459mvhBx/AKafo\nCOOKoiidBZ/PDojao4cd4mDLFiuMEhNt2uWXw0UXwauv2iAJ06bBjBk2EuhZZ1nrzrx5Nq2gAHr3\nttHmXI+ApCQ7uUQi1j2uoMAO3Cpi+w+lpFhrUVqaPXYgoB4FiqLU5+KL+7FsWXKD25csSaG6uq6Z\nubLSx7XXDuDxx7Nj7jN8eDmPP76+JdVZvXp14Iknnlh95JFHlgP85S9/2ZCbmxsOBoOMHTt2yKef\nfrpj9OjRld59SktL/RMmTCiZNWvWxksvvbTvgw8+mHXPPfcURpdtjGHp0qUrnnnmmR7Tp0/PP+64\n47697777cnJycoJvvPHG9wsXLkwaP378sOj9+vTpEzryyCOL+/XrN2LcuHHFkyZN2nXppZdu9/v9\nXHbZZfscc8wxxTfffPOWYDBISUmJb8GCBSkvvPBCr88//3x5MBiU0aNHH3DccceVJCcnR7znt3Hj\nxrg//vGPvd9///2VaWlpkZtuuinvnnvuybnvvvvq1b0huo4FCOCWW6xbw5w59g9UURRF6VwkJsL+\n+8OBB1phsnOnnYNt33/yE3jlFXjsMZvv/vtt36FLL7X/AZs2WSGzaZP1DJg3L/ZxfD57LDewQs+e\ndgLYts1am5YtsxHnliyB77+3oqykJLblSVEUxUu0+Nld+h7Sr1+/Klf8ADz++OOZw4YNO+DAAw8c\ntmrVqsCXX36ZFL1PIBCITJkypRhg9OjR5WvWrIk5oOaZZ565E2DcuHHlGzZsSABYuHBh6rnnnrsd\n4LDDDqsYOHBgRax9X3rppTWvvfbaytGjR5fPmDEj75xzztkH4OOPP067/vrrtwLEx8eTmZkZeffd\nd1MnT568IzU11WRkZEROOOGEnQsWLEiNPr8FCxakfvfdd4GDDz546NChQ4e9+OKLvdauXdssq0fX\nsQCBtQKddZYdV+KTTyA3137FUxRFUToX6elw0EHWpXnDBuvalppqt4nAEUfYacUKaxF65ZX6ZVRW\nwp/+ZK1ATUHEiqzoQbXdAVy3batN8/trLUXJydZSpP2KFKX7sDtLTX7+QRQU1BcUvXtXs2jRN61d\nnaSkpBqu7bd3AAAgAElEQVQf3qVLlyY+/PDDuYsXL16RlZUVPuWUU/atqKio1zjFxcXVBALw+/0m\nHA7HbMACgUBkd3ka49BDD6049NBDKy6++OJtw4cPHw6shZpIbU3Ce37GGI466qjif/7zn6ubWxeX\nrmUBEoHf/MZ+1Zs714483smCPCiKoigOfj/k59uxg1JSrACJtr4ccAD88Y8NC4+CAhs8YcoU6xL3\n5z/boAoLF1pPgVBo9/Vw+xW5ViJ3MNaKCivOvvnGWok++QSWL7dpO3bYfkfar0hRuie3374RRzjU\nEAhEuP32jW196J07d/pTUlLCGRkZ4bVr18a/99576a19jLFjx5Y+++yzGQCLFi1KWrVqVT0L0/bt\n232vv/56qru+aNGi5Pz8/Gpn/+I//vGP2QChUIjt27f7JkyYUPLqq69mlJaWyq5du3zz58/v+aMf\n/ag0utyjjz669OOPP05dvnx5Atj+SEuXLu3GFiCwLhFnnmkF0Pnn2z/PrA4RcU9RFEVpCYEADBli\nRcXq1VZYpKdbNzaX3r2t21s0aWlwzDFW7Hz5JbzxRl3R4/dDXh707WvHlevbt+5yr16xxVWsfkXG\nWIFWVFTrtmeMzZOaai1E8fF28vvrTz6fWpAUpavgRntroyhwjXH44YeXDxo0qHLgwIHD8/Pzq0aP\nHl1PROwp06ZN23zmmWfuO3DgwAMHDRpUsd9++1VkZmaGvXmMMXLvvff2vvLKKxMDgUAkJSUl/Nhj\nj60BeOSRR9ZddNFFA+bMmZPt9/uZNWvWmqOPPrr89NNP3/aDH/xgGMDFF1+85ZBDDqlYtmxZHXHT\nr1+/0KxZs9ZOmTJlYDAYFIC77rpr40EHHdTk6HJdJwy2l2XL4NBD4cgj4be/hVGj7Bc8RVGUdkTD\nYLcCoRAUFlorS2KitcyA7etz663W7c0lEIDf/a6uC5x3f3dav95Gg9uwwfbz8ZKUVCuKYk2pqTTK\nvHnWDa+w0LplX3klHH98bKFjjBVJcXF1565wUtGkKHvMXg2D3YUJBoMEg0FJTk42S5cuTZw4ceLg\nNWvWLI2Pj2/vqtXQWBjsrqkKBg+2VqCnnoILL7R/Ov36tXetFEVRlD0lLs4Kj8xMWLPGusWlp9eK\nnFhR4GLt37dv7PIrKmrF0Pr1dUXSokVQVlY3f0ZGfauRu/zpp3DnnbWirLAQ7r3XiraG+iWFw3aq\nqqp1oXOnhkRTfHytWIqPV9GkKEqbs2vXLv9RRx01OBQKiTGGmTNnru1I4md3dE0BlJAA11wDL75o\nw2IPGGDd4JLquScqiqIonZHkZNv/Z/v2Wre4SZOaHvCgIZKSrCv1/vvX32aMjUoXLY42bLB9f956\ny4bWbozKSiuKioqsEEpNbXienNw0oeKKpupqW34kYtcbE01vvgmzZllR1ru37T979tlWIMXF1Z37\nulZ3YUVR9pysrKzwV199taK969FSuqYAAhg61FqB5syxY0dkZFgfckVRFKVrIGL76KSnW6tNYWH9\nfjmtfTx3cNWDDqq/PRyGzZtrxdFvfhO7nNJSG7hhd/j99YWROzUmnGLNExNrxdC8eXD33bWWqU2b\n4MYbrTVt4sTaMZHceVyctSYFArVz10UvWjCpZUlRlE5A1xVAycnw859bK9Djj9svbrt22QH1FEVR\nlK5DfHytpX/1ahssAWpfxiORWvcvkVqrhs9Xm9Ya+P3WmtK7NxxyCMycGTswQ36+HdC1tNS61DU0\ndydvekmJdfHzpjelL68byS4lxYq06Oh3lZU2Ql5OjhWVmZl2Skmx5YdCNk9paa2FKVooucdJTLQC\nyTuPZVlSsaQoSjvRdQUQwLBh1gr0xBNw2WX2q+CIEWrOVxRF6YqkpsLw4bb/jPuS7rqCBYPWRay6\nutZdrKrKvth7Q1W7L+XuC70rlFyx1BzhNHVq7MAMU6faj3TJDQ8i32SMse5/zRFSL78cu6xdu+Cq\nq+qmJSbWiiGvMOrVq/56ZqYVPA2Jpeh6u/2VvELplVds4IoNG2xfqrvusuP7idSW4Z03lu5NUxRF\n8dC1BVBaGvzsZ9YK9PDDtlHdssUGRVAURVG6HiJWZDQHb58Zr2gKh61w8k7V1XZeVla3j41rCfEu\nT5hgBdCDD1r3vLw8uPZaOOGEWlHgfVlv6fm6lp2cnKbt8/HHsS1Tubm2X9C2bbZv1fbt9Ze/+87O\nqxqINpuc3DShlJlZe5/Ky61l69VX4Z57agXj+vVWkK1fb13zGiL62kfjfvR0xas3TaQ2KET0sjvN\nmwf3328tb/n5cPPNVpQ1JI5b27KoKEqr07UFkIi1Ak2ZAn/7G1xxRa0Pd/RI34qiKEr3xPti3ByM\nqRVM0QLKtTKdc479DwoGa6O5eZfdKZrdvTzHetGPfnH3WkG8L/bXXFM3Oh1YMXLjjdaC1pTzLi+v\nK5BiiabCQvjqK7vc0ICz6em1wuirr+rWCez6jBm1/Y68ke6iI941tOytt3vdYi2799Ob/uqrdjgN\nt14bN8L119uPqa4oi2XdAnvNvW5/7uRG7fMGmYgWUF4hpSgOhYWF/gkTJgwB2Lp1a7zP5zOZmZkh\ngC+++GJFIBBo0tg2f/nLX3qddtppu/r37x8COOOMMwbcdtttBSNHjmzyODoNccMNN/T+5z//menz\n+Yzf72f27NlrjjrqqPI9Lbe16doCCOyI3eeea0f+njULfv9724Dtu29710xRFEXpzIjUvsjuKcZY\nIeR9GW9sipU3WlS5kyvI3DxupLy//rV2fKIrroDx423/qWhLVvRLut9fa3VqyhATxlgLz7Ztdtqx\no3bZK5qixY/Lrl1w000tv7bu4LMNCaXG0l5/PbYo+/Of7fkHAvWnpCTrzhcI1AqZSMSGWHfvm3tv\nXLHUkIiCumLJK57mzat9p+nTB26/vdZd0BW8saxb3TEM+jPPwC23wLp1jIAYEUzahtmfzM6c/t70\nPoWlhQl5qXnVtx95+8YrDm75QKh5eXnhr7/+ejnA1KlT81NTU8PTp08vam45c+fOzTrkkEPKXQH0\n4osvrmlpnbzMnz8/9Z133klftmzZ8kAgYDZt2hTnDlTaUoLBIG0RXrvrCyC/30aE++lP4ZFHYO1a\n+yUqJ6d2AD1FURRFaU9cobG3GDasrqiItl65UyhU23fKdQGsrLTzhtzOYommlBRr7Wns4+PRRzfs\nmvf00/Xr4e3X1dy0WNt37aqfVt7Ah+tdu6zFrCn4/VYQJSXFFkzutui5uz0hoXaemGhF0Kef2gBP\n1dX2GBs2wC9/aecnnlhfxMZyE/SOC+XO3Uh+cXG16+52160vWkhFz597Dm67zbou9utnux+cfXbt\nsRuzwrV02bWiei2q7vLLL8O0aVaAAvGwV1yAZn8yO/O6/1y3T2Wo0gdQUFqQcN1/rtsHYE9EUEPM\nnDmz1yOPPJITDAZlzJgxpU8++eS6SCTCmWeeue/y5cuTjDFy4YUXbsnNzQ2uWLEi+ZxzzhkYCAQi\nX3zxxYrDDz988MyZM9cdfPDBFZmZmaPOP//8LW+//XaPpKSkyKuvvvpdnz59QkuXLk0899xz962s\nrPQdf/zxu+bMmZNdUlLyhbcOGzZsiM/MzAy5lqj8/Pwa0++CBQtSpk6d2q+iosKXmJgY+d///rdS\nRMwFF1ywz7Jly5Lj4uLMn/70p/UnnHBC6YwZM7Jee+21HiUlJX6fz8fChQtX3nzzzXnz5s3LqKqq\nkpNPPnnH/fffX7An16vrCyCwkYHOOguef976Ys+YYYXQAQd0v68giqIoihKNa6lozpfWWIIplmgK\nBuuLJi/G2GNfdZV9WY52zbv+evsi3R7/142JsqeesnXd3VRR0fC2srJa61f01FyqqmD6dDu5Fq1Y\nboGxrF3egXSjp4SEWuuTu68b7c9Nc137Fi2CRx+tFWbr1tkgVJ98AuPG2WfDfUYae34ayhOd1lg5\nrlvq22+37Hruhov/dXG/ZZuXNRjJZEnhkpTqSHWdh7YyVOm7dv61Ax7/4vHsWPsMzxle/vgpj69v\nbl0++eSTwL/+9a+en3322Yr4+HjOPvvsfR599NHMwYMHV23fvj1u5cqVywG2bt3qz8rKCs+ePTtn\n5syZ68aNG1cRXVZpaal/woQJJbNmzdp46aWX9n3wwQez7rnnnsKrrrqq/69+9auiiy++eMc999wT\ns/6nnXbarj/84Q+9BwwYMPyII44oPvfcc7dPnDixtLy8XM4///z9nn/++e/Hjx9fvm3bNn9SUlLk\njjvuyEtMTDQrV65cvnjx4sDJJ588aNWqVcsAli9fnrxkyZLl2dnZ4eeff77HunXrEpYsWbLCGMOE\nCRMGvfnmmynHHXdcWax6NIXuIYDi42HgQOuL/dBDthNn377WDJ+Z2d61UxRFUZTOh2sVaCrul/pY\nL6rBIFxwgQ2i4AYcyMuDX/zCipCdO+uWFW3NaEhYefPHCnAQ7RIW7TLWUCS/G2+0odfbCmOsoKmo\nqD+vrLTjGzbEVVfVClDvFMu6tXNnbOuYOzUlxPruqKqyIeFnztzzsmIRH1/7LLqugt7lNhA/TSFa\n/OwufU94/fXX07/88suUgw46aBhAZWWlr2/fvtWnnnrqrlWrVgUuuuiifpMnT971k5/8pHh3ZQUC\ngciUKVOKAUaPHl3+/vvvpwIsWbIk5cILL/wW4JJLLtl+77339oneNzMzM/LVV18tnz9/ftrbb7+d\ndu655w6cPn36+lGjRlXk5+dXjx8/vhygV69eYYCFCxem3njjjYUAY8aMqczJyQl+9dVXiQBHHnlk\ncXZ2dhjgjTfeSH/nnXd6DBs2bBhAeXm5b8WKFYEOK4BEZCLwV8APPGaMuS9qew/gaaC/U5f7jTFP\ntEllcnLg9NPh2WfhgQfstGaNHRdIOxkqitJF6VDtsNK9cV3jGvvPve46uO46IiaCIIhX2DTU/8m7\n3tCy1yLghj739pGKFaAiFLL9oqZNg9mzoajIWn6uvBKOOKK2v1SsCIDRy95z8I6bFH19okN5u5aV\n1NTadLBjTRXE8ADq3dvWr6EyG0qLhTv+U0PiKJaw+sUvGi7vL3+pK1BiCZdoAePt/9TQWFLe6xnL\nRe6442Jfqz1kd5aa/D/lH1RQWlDP3a53au/qRZct+qY162KM4eyzz97617/+tZ658quvvvrqpZde\n6vHggw/mvPjiixnPPvvs2sbKiouLq3k4/X6/CYfDzRJs8fHxTJ48uWTy5Mklw4YNq3jhhRcyR40a\nVc/StDuSk5NrosMYY7jhhhsKrrvuuq3NLach2kwAiYgfeBA4DtgAfCIirxhjlnuy/QJYboyZLCLZ\nwDci8owxprrVK5SUZE3o551nxc+KFdC/v23Q8vNb/XCKoijtTYdrhxUF+zITioQIRoJ2Hg5SEaqg\nIlhBZaiSylAlEWPffXziI84Xh8/nw4+fOF8cfp8fv/jx++y6m+bz+fD5ffjEh+DDJ3F2WQSf+Gqm\nesKq8crCwQdbK5Dbt6Sxvii7297Yfo7wMo54i4Rt9wkTCdvJmJq5b+q1xN92J+KxbphAgOA1v8CY\nEGIMEnFEJODDh0T3lfHOvUQPcOteK9f9rTFhl5dnA2tEk5cHhx0Wu9xY1zxWHleQhUK1Ydi91jqI\nHeb82mvrRz3cC9x+5O0bvX2AAAJxgcjtR96+sbWPdcIJJ5RMmTJl4LRp0zb37t07VFhY6C8pKfGn\npKREkpKSIhdffPGOoUOHVl5++eUDAFJSUiLFxcXN+vo/YsSIsrlz5/a86KKLdj7xxBMx3ac+++yz\nQEJCghk+fHgVwBdffJHcr1+/6h/+8IeVmzZtSvjggw+Sx48fX759+3Zfenp6ZNy4cSVz587tdcIJ\nJ5R+9tlngS1btsQfeOCBVe+++26qt9yJEycW33fffb0vueSS7enp6ZHvv/8+Pjk52fTu3buB8JK7\npy0tQIcA3xljVgGIyHPAKYD3j9cAaWJbolRgO9Dik9ktvXvDqafazpQzZ9qgCOvX1x2PQFEUpevQ\n8dphpctjjCEYCRIMW4FTHa6mMlRpRU6ogqpQFcYYK0IMGEwdIZOemF4jUIwxREyEiIlgsOVWhatq\n1r3bAQTBYOoIHGOMIwPssaBhYeWmRwstVzi5x3TL8h6/scnNa4whQoRIJEKESMz9a3CNNH6xb2vG\nSTPA2UeTmVhFv98/RMKmIqrzc1k/7Uq2n3ocSHXNda1TlsEjBGvFoV/8+PHhE3GEo+DDh09w0n34\nxeekCWJAnLm9Ltg0hLi7bifhV9cjFbUf/E1SEpHfTocDhtqq+BwR6nU9bIqVKta2pjBihO324ESB\nC+6ljztuoIPWjALXEIccckjFtGnTNh199NGDI5EI8fHxZtasWWv9fj+XXXbZAPf3dvfdd28AuOCC\nC7ZeccUVA9wgCE05xoMPPrjuvPPO2++ee+7JP/roo4vT0tLC0XlKSkp81157bf/S0lK/z+cz++23\nX+WTTz65NikpyTz11FOrrrrqqv5VVVW+QCAQ+eCDD1ZOmzZt8/nnn7/P4MGDh8XFxZnHHntsdaxQ\n3medddauFStWBMaMGTMUrIB77rnnVvXu3bvF10zcH3JrIyJnABONMZc66+cDhxpjrvbkSQNeAYYC\nacBZxphXY5R1OXA5QP/+/UevXduo9a5hjIGlS+GJJ6wp9rnnYP/9rRvc/vu3rExFUZQmIiKfGmPG\n7MXjtVo77ORtnbZY6dRETKRG3AQjQapD1XUsOFWRKnBeiBErQOL98fil1mLTZAtMGxEtrNx1gzP3\nrNcIKOdcohGRGquS17oUa7mxtGZZpvbgvGvEmHOusdIbmzuVx9VXrpB10zJffp1+v58dJcwm1r1+\nTl73vKOtc+KKMFcsIbXLUfu4Qi5WOd5tYAVbz6SeLW6HlyxZsmbkyJGt5obVmSguLvalpqZGfD4f\ns2bNyvzXv/6V8cYbb3zf3vVqjCVLlmSNHDlyQKxt7R0E4XjgC+BHwEDgTRF53xhTp5OWMeYR4BGA\nMWPGtFyxidivAJMnw5NPWle4v/3NDmiWmwtpaS0/E0VRlM5Jk9phaMW2WOnQhCPhOhacqnAVFUFr\nvakMVRIMBwH78up+WY7zxRHviychLoFkX4OBsToMtZaP7tUHuEYMtKHOipxzNmvPObtOWs8G8kZb\n01yrlVeg1eSJxM6zu3KA2vM1ddaUZvDee++l3HDDDf0ikQg9evQIP/nkk6vbu057QlsKoI2Ad4S0\nvk6al58B9xn7hH4nIquxXyEXtVmteva008UXw5/+BJ99ZsdDWL3ajoDdktHAFUVROiYdsx1WGiX6\nZc5ddrfFeulrbFtj+SImQmWwso4Fx3XNEgQjBh+1bmFJcUmkJtRxz1faiXnfzGPGRzMoKCmgd1pv\npo6dyuQhk9u7Ws2ql9catjdkyc6KnbvPpMRk0qRJJZMmTVq++5ydg7YUQJ8Ag0RkX+wf7k+Bc6Ly\nrAOOAd4XkVxgCLCqDetkBU6fPjBpEsyZY/sCPfFE7WjUWVltenhFUZS9SMdsh7sR1eFqdlTsYEvZ\nFiJ43I2ivnB73ZGA2C5GTrrXJcsWSG2/lybmqxE4mDrBBFITU/GJfgjs6Mz7Zh63vnMrlSHbsX9T\nySZufedWgHYVQR21XooSTZsJIGNMSESuBt7Ahl993BjzlYhc4WyfDfwWmCMiS7FN9E3GmLb3rczM\ntGMNXHIJ/OEPsHgxjBpVGxa7OQPBKYqidFA6dDvchTHGUBYso6i0iK3l9lKmJKTUuFuJL3YfEaDd\n+8YoHZey6jIKywrZXLqZ373/uxqR4VIZquS2d27j440f1+lLU9MfBl/dNKfPDEJtv5kG0mr65cRK\no7afzazFs2LW654P7iEjKYNAXIAEfwKBuACJ/kQ7j0sk0Z9IYlxim4lvr1WKXIbvQVGRSCQiPp9P\nXYA7OJFIRIAYYQ4tbRYEoa0YM2aMWbx48Z4XtG6dnX7yEztI6lNP2QHB8vJseGxFUZRWZm8HQWhL\nWq0t7kKEIiF2VuxkY+lGKoIVJPgTSIlPUVHTDDqqW1dbYoxhZ+VOisqKKCwtpKi0iMKyQgpLrdgp\nLLNpJdUlTSovJyWnTtACN5gDhjqR57xukG4aUM8yuTeJ98UTiAvUFUoegRTwO+uNLLvCyt3/84LP\nmbNkDtVhJ/jbw2A2mRb9KJcsWfJKXl7esOzs7F0qgjoukUhEtmzZ0qOwsHD5yJEjT46Vp72DILQf\nOTl2YKzLLoN774VFi2ys/02brBtccsfvxKkoiqK0P+XBcraWbaWg1A62mJKQQmZSzGEylEboqO5T\neyLKwpEwW8u31oibwtLCOkKnqKyIotIiqsJVdfbziY+s5CzyUvPYt+e+HNb3MHJScshLzSMvNY/r\n/3M9m8s21zteflo+71z4Tquct0ssd82G0iY9O4nC0vrjAGUnZ/PXE/5KVaiKqlAVleHKmuWqcJWN\nHuhZdkOne5erwlWUVJXULEfn3xuEQqFLCwsLHyssLBwOqK9oxyUCLAuFQpc2lKH7CqBAADIybES4\nxx6zfYHmzrWDfK1fD0OGtHcNFUVRlA5KxEQoripmU/EmiquKifPH0SPQQ/vPNIPKUCVbyrZQVFbE\nlvItTH9vekz3qTv+ewff7/iexLhEEnwJNV/54/3xtdYBfyIJfrstwZ9gl/21FgE3rbnWuMZE2Y8H\n/pjNZZvriJrC0sKatMLSQraWbyVs6g6XEu+LJzc1l7zUPEbkjCBnPytsclNyawROVnIWcb6GX9F+\nPe7XdeoFEIgLMHXs1GadX1PwhpjeHTccdkPMet10+E2M7j261evmEjGRWqHkCKOqkBVHZ7xwRqsd\nZ/To0ZuBmBYFpXPRfQUQ2IFRt2+Hyy+Hu++Gjz6CsWNh2zbYtcv2B1IURVEUh6pQFdvKt7GpdBPh\nSJik+CQyk9Xa4yVa2Gwu28zmss1sKbPLblpxVb1I6zEpC5Yxe/HsVnHHcoVRU8RSYlwib33/VkxR\n9us3f80Nb95Qr/yU+BQrZlJzObzf4eSm5lqxk5JXk54RyNhjt0jXAtXR3AXbq14+8dW4zkWTn5bP\nppJNbXp8pfPRffsAuSxdClVVcNJJtu/P009DMAihEBx0EPi71xgBiqK0HdoHqHNijKGkuoSi0iK2\nlW/D7/OTkpDS6Bf6jk5L3Lq8wsYrZFxh46bFEjbxvniyU7LJSc6x85QcclJyyE62y9kp2fx83s8p\nLKvvPpWfls+CCxbYQVfD1VSHq6kKVdm586W/OlJNdciz7m4LVxEMB2usAnX2ccuKse4eY33x+gav\nx7WHXlvHcpObmqshwjsg0Va8PekDpHQdOm/r3Vr07QsrV8LPfw6//a21Ah12GJSVwdatdoBURVEU\npdsRDAfZUbmDTcWbqAxVkhiXSEbSnn+9b29iuXXdsuAWVu1YxaBegxq02jQkbFwhMzBjIGP7jq0j\nbFxx0xSrxw3jYrtPTR07FRGpsczsTY5+8uiY1oP8tHyuOviqvVqXjoa37w/EHnPKdZ3zia/dfjfR\nVikTMXunw5DSoVELUCQCX3xhl0880Y4R9Pe/2/SSEhseO2HvNriKonRN1ALUOSirLqt56TcYUuJT\n9vqLd2tgjGFH5Q4bVczTT2XOF3OoCFU0uq8rbKKtNNHipmegZ6u+2Ha0KHD1rAdYUfa7o3/X7u5m\noUiI6nA1wXCwjvgAQECM1B0fyjMeVL1xoqLHkoraXlO2Z3wqV9jUWRZH9DjxASJECEfChCKhmihz\nbsj3muO51fLUv07IbakVUdHHa26fu50VOzm036GfGWParkOS0ilQC5A7MOqaNdYKNH06fPghHH64\n3bZhA+y3X3vXUlEURWlDwpEwOyt3sqlkE2XVZSTEJdAjsUeHtfYEw0G2lm+tG1XMiSjmrm8u21wv\nOpZPfHUHXI3ilZ++0ibCpqlMHjK53YWFl47S1yYYDta487kiJd4fT1piGmkJafWsLDXj9HjGl4pe\njs7b0HJD+zUXb+Q4d3LDb9esO9td0RQ24TrLoUiISCRCMBIkFAkRioRq6uY9jlv/mjRPyG9FARVA\nll69YO1aOP10ePRRGxFu3DhIS4OiIusGl5LS3rVUFEVRWpnKUCVby7dSUFJAxERIjk9u1aAGLbFo\nlAfL6wmZ6PDJW8u31gsKkOhPrOlwPypvVN3+Kc68V3Ivjpt7XINuXUOyNAJqNHtTlBljavo6BSPB\nGmtLID5ARiCDtIQ0AvG2s39n64PWnGhyzaEhEdWQ0AI6l+uT0iZ0rl9PWxEXZyPCFRVZK9Cdd8IH\nH8ARR0BSkrUODRsGHfRLoKIoitJ0IiZCSVUJBaUF7KzciV/8pCak4ve1btCbmCGUF9zKppJNDM0a\nWsdy4x30MlZfm/TEdPJS8shJzWFo1tA6wiY3NZfclNwmW22mjp2610IoKw0TMZEay044EkYQjBhS\n4lLISsoiLTGtJjJdaz+bXQkRwS9+/Og1UpqOCiCX7GzYuBFOOw0eecRagcaPtwOibttmw2X36tXe\ntVQURVFaSHW4mu0V29lUvIlgJEggLtCmA5bev/D++iGUw5XM+GhGzbogZCVnkZuaS/+e/Tm4z8Hk\npebVDHrpCpzk+NYbnLujuHV1J8KRcE2UOdcNyyc+UhNSyUzKJCUhpSYUt44lpShtjwogl0DACpyS\nErjiCrj9dnjvPTjqKOsKt2aNHRcoTi+ZoihKZ8EYQ2l1KUVlNoQ1QFpiGqm+tglXXB2u5oN1HzBv\n5TwKS+uHdHZ57vTnaga8jPfHt0ldGqOj9bXpSoQiIapCtr8O2P4ncb440hPTa8SsO95QR+1jpihd\nHX2b99K7t7X2/OQn8PDD1gp05JE2ClxZmXWR69OnvWupKIqi7IZQJMTOip1sLN1IRbCCBH9Cm3Xs\nj5gInxV8xryV85j/7Xx2Vu2kZ6AnyfHJlAfL6+XPT8vnB71/0Or1UPY+rlXHFTsiQoIvgfTEdNIT\n044u3R4AACAASURBVGsG52wPkasoSsOoAPKSmmqnSASuvBJuvRXefRcmTLDWnw0brJUoUH+kYUVR\nFKX9CYaDbCrZRFFZERhITkhuMze3b7d9y7yV8/j3yn+zsWQjgbgAx+57LJOHTObwfocz/7v52tcm\nBm6Ur4iJEDbhOus1YZiJCssMdcI4R28zYuruFxXNDOpHQvPSUH7vNmMM1eFqwpFwTQCKpPgkMpMy\nSU9Mr3Fh62zBCRSlO6K/0mj69oWvv4ZTT7VWoP/7P+sG5/NBfDysWweDB7d3LRVFUZQYFJQUUFBa\nQM9AzzbpS1FYWsi/V/6beSvn8fXWr/GLn3H9xnHt2Gs5dt9jSUmojRjaHfraxBIzbpp3jJmasWbE\nECdxxPvjSfAlkOxPrhngNN4fbzuz+/z4xFdvUE03jLK7HL3NXXeX3b423nmstIa2uWUZY4gQwSc+\negZ6kpqQSmJcIoG4gPbXUZROigqgaNLTITHRLl95Jdx8MyxYAMccY61D27ZBcbHNpyiKonQYwpEw\nRWVF9Ejs0aovpsVVxbzx/RvM+2YeizYuwmAYkTuCW464hRMHnUhWclaD+3amvjbGmDoWGe9yQ2LG\nj5/4OCtmkvxJxPviawSN3+evETTeufZ7URSlvVEBFI13YNRTToHZs+GBB+BHP7JhsFNSYPVqOOgg\nm1dRFEXpEBRXFRMxkVYJGVwVquLdte8y75t5/Hftf6kOVzOgxwCuPuRqJg2exICeA/a8wu1AxESo\nDldTFaqqsZQgVvz4xFdjiUmMT6wRMipmFEXpaqgAikVmph0YVQSuugqmTYO334Zjj7X9f3bsgK1b\nISenvWuqKIqiOBSUFJAUn9Ti/SMmwqKNi5i3ch5vfPcGJdUlZCVn8dPhP2Xy4MkclHNQp3rhjxY7\nBity0hPTyUrKsqGX4xLruJ0piqJ0B1QAxcIdGLWgACZPtlagmTOtFcjns2Gx166FjAzbL0hRFEVp\nVyqCFZRUl5CRlNGs/YwxfL31a15Z+QqvrnyVorIikuOTOW6/4zh5yMmM7Tu2U3Rqj2XZERHSE9PJ\nTs4mOT6ZQFxAQy8riqKgAqhh3IFR/X7bF+imm+Ctt+DHP64dC2jTJthnn/atp6IoisK2im34pemu\nbxuLN9YEM/h2+7fE+eI4ov8R3HT4Tfxo3x/tkSWprYmYCFWhqppBNd0oZa7YcQfVVLGjKIoSGxVA\nDZGYCFlZsGsXTJoEDz1krUDHHmutQOnp1kKUnQ3JrTdCt6IoitI8IiZCYWlhnQhssdhRsYP5389n\n3jfz+LTgUwB+2PuH3HHUHUzcf2KbhcveE2KJHb/PT2pCKjkpOSp2FEVRWoAKoMbIy4MtW2z0t1/8\nAm68Ef7zH5g40fYPSky0YbGHDLHriqIoyl6nuKqYcCTMa9++Vi/k9LH7Hcs7a97hlW9e4YN1HxCM\nBBmYMZDrxl7HSYNPol96v/aufg3hSNi6sYWrAOueF+eLIy0xjdzUXJLjk1XsKIqitAIqgBojJcVa\neior4aSTrBXowQetG5zPZ7dv326tRD17tndtFUVRuiWFpYUsWL2A6e9Nrxl0dFPJJm566yZ84iMY\nCZKTksP5I85n8pDJHJB1QLsLiKaIHbfPjqIoitK6qADaHX36wIoVNjLcL34B118P8+fDiSfa7amp\nNiz2iBG2v5CiKIqy16gMVbKrchcPfPJAjfhxCZswCf4EHj35UQ7JP6RVwmO3BGMMlaHKemInPTGd\nvNQ8kuKTVOwoiqLsRVQA7Y70dBv6uroaTjgBZs2yVqDjj7eCJyEBysth82YbOU5RFEXZa2wv345P\nfBSUFMTcXhmq5LC+h+3lWkEoEqI8WG4HERUhI5ChYkdRFKWDoEH/d4eItQKVl1vBc/XV8N131grk\nkpYG69dDVVX71VNRFKWb4Q1+0Dst9geohtLbgqpQFTsqdrCjYgdVoSryUvMYlj2MMfljGNRrELmp\nuaQnpqv4URRFaWfaVACJyEQR+UZEvhORaQ3kmSAiX4jIVyLyblvWp8VkOONKhMM2AMKgQfDAA3Yd\nrDDy+2HDBggG26+eiqIoUXSZdjgGJVUlVEeqifPFMXXsVIS6/XoCcQGmjp3aZsePmAhl1WU1oscn\nPvbN2JcRuSMYlTeKvul9SUtM0wFGFUVROhht5gInIn7gQeA4YAPwiYi8YoxZ7snTE5gFTDTGrBOR\nnLaqzx4RFwf5+TbsdY8e1gp07bXw2mt2oFSwfYG2bbNTRgbk5FjLkE//+BRFaR+6VDscg6LSIgJx\nAQDG9h2LwZCWkEZpdWlNFLjJQya36jGD4SAVoQpC4RB+n5+MpAyykrNIiU8h3q8DYyuKonQG2rIP\n0CHAd8aYVQAi8hxwCrDck+cc4P8ZY9YBGGM2t2F99gx3YFRjbBS4IUNsX6ATT7TWHxEbCc4YKC21\n0eF8PsjNhV697FhBGrZUUZS9S9dqhz1UharYUbmDngEbgXPhhoUAzDl1DsNzhrfqsSpDlVQEKzDG\nEIgLkJeaR89AT5Ljk9W6oyiK0glpy5a7D7Des77BSfMyGMgQkf+KyKcickEb1mfPcAdGLSuzwubq\nq230t3//u24+ERseOyPDWoU2b4alS2HJEigq0n5CiqLsTbpWO+xhR+UOgJpw1gvXL6RnYk8OyDpg\nj8uOmAil1aU1rm3xvnj2y9iPUb1HMaq3dW1LTUhV8aMoitJJae8ocHHAaOAYIAlYKCIfGWNWejOJ\nyOXA5QD9+/ff65WswR0YFeDYY2HoUBsV7qSTrJtcNH6/jSIHNorc2rWwZo11jcvNte50sfZTFEXZ\nezSpHYaO0xYbYygoKSA1IbVm/cMNH3Jo30NbHOq6OlxNRbCCiIng9/nJDGSSmZxJakIqcT5tpxVF\nUboSbfn5aiPgHWK7r5PmZQPwhjGmzBizFXgPGBldkDHmEWPMGGPMmOzs7Dar8G5xB0atqLBWoF/+\n0gqa/8/encc3dtX3/399tHq3Z8Yzk9mSyUoWyDrZIGUNZIOGEL5Av3wDbWnTQEu/NC0FCqWUX2lp\naWkLFChroQQCzVBCAgkFGgp87exkIZOETFbP7hnPeNd27/n9cXVl2ZZkeZEtW+/n46FIuvdK+ujG\nc6SPzjmfc8stMz82kQiGyK1aBblcUEnu/vvhySdhaAh8v+bhi0jDWbB2GOqnLR7ODJPOpQtzbp4+\n8jT7Rvbxwi0vrPo5nHOMZ8c5nAp6eTznsal9E6etO42zN5zNcauPo6upS8mPiMgKVMsE6B7gRDM7\n1swSwJuA70455mbgIjOLmVkLcD7waA1jmr9Nm4IECOAVr4BTTw16gXK56p+jqSlIhDo7YXAwWGj1\n/vuDKnJjY7WJW0Qa0Ypshw+MHiARmygl3dsXzP950ZYXVXyc53uFoW2D6UGaYk0c33U8Zxx1Bmes\nP4ONHRs1tE1EpAHU7Kct51zOzP4A+AEQBb7knHvEzK7L7/+sc+5RM7sdeAjwgS84535Zq5gWRPHC\nqIlEMBfoHe+AF70oSGY2bIDrr5+oDleJWTBPCIKS2nv3BklQc3PwPF1dwWuIiMzBSmyHM16GQ2OH\nCsUPAHp29bC5YzNbOreUPL54aFt3SzermlbRmmhV746ISIOqaevvnPs+8P0p2z475f7HgI/VMo4F\nZQabNwdD1xKJoCiCGRw5Euzfswc+8IHgdjVJUCgaDXqEIEiunn46uN3ZGcwX6ugIjhERmYWV1g4f\nSR0Bmyh+kPNz3LXrLi474TIgGNqWyqVIeSlw0BRvYnPHZjqSHbTEWwqPExGRxqWfv+Zi1apgDpDn\nwT/+Y1D6ulgqBR//+OwSoGKJxETPz/g4/OpXQZK1dm1Qia6tTSW1RaThOOfYM7SH1nhrYdsvD/yS\n4cwwF265EICh9BBtibZCpbZkLLlU4YqISJ2qeqCzmV1kZr+Vv73WzI6tXVh1LhoN5gINDwfD1krZ\nswe+/e2g8tvUBGk2mpuDhKujAw4fhh074IEHgucP5yKJSENo9HZ4NDtKykuRiE4MDe7p6wGChVAh\nKGG9uWMza1rWKPkREZGSquoBMrO/ALYBzwO+DMSBrwGVZ5yuZN3d0NcXzNXZs2f6fjN43/smjj3n\nHDj77OD6lFNmX/46EpmYL5TLBYuyPvdcUJluw4ZgqFxcq5CLrFRqh+HAyIFJyQ8EBRBOXXsqq5tX\nA8HQOCU+IiJSSbXfwq8CzgLuB3DO7TGz9ppFtRwkErBuHbz97fCRjwTD3kJNTfDhD8Npp8F9901c\nfvCDYH9LC5x5ZpAMnXMOnH56kMhUKxabmC+UTgfzkQBWrw5iamvTfCGRlaeh2+Gsl+Xg+EE6k52F\nbWPZMX6x7xe85Yxg7VbP94hadFqSJCIiUqzaBCjjnHNm5gDMbBbf1lew9euDBVGbm4M5P3v3Tq8C\nd8IJ8MY3Brf37QvKXYcJ0ac+FQyPi0aDXqFzzoFt24Keou7u6mJIJoOLc0FBhsceC3qL1q2DNWuC\nxGq+84XCIXxTr8vdrnRc8bZIJEjmlKyJVKOh2+HB1CA4JhUxuHfPvWT9bGH9n4yXoSPZsVQhiojI\nMlFtAvQtM/tXoMvMfhf4beDztQtrmWhpCUpVv+IV1RU8OOoouPzy4ALBHKIHHphIiG68Eb7ylWDf\n1q0TQ+a2bYNjjqmcyJgF8bS0BIuqHjwYJFyJRDA0LlxotTgRmXq71LZKrxceN5sEq9TxZkGcyeTE\ndVNTkBjFYhNJUiym4g/SyBq2HXbOsXtkNy2Jlknbe/t6SUQTbNu4DYC0l+aotqOWIkQREVlGqkqA\nnHN/b2avBIYIxp9/0Dn3w5pGtlxs3BgUJmhpmfnYqdrb4dd+LbhAUP56x46JhOi//zsopABBb044\nZO7ss4Meo3JzfiKR4LkhmC/k+8G2MHmYej3TtlpzLqiol80GhR18f2Jh2eJEC4IkKOz1SiSCRCmR\nmEiQiq9FVpBGbofHsmOMZ8ZZ3bJ60vaeXT2cveFsmmJNQJAotcTn0BaLiEhDmTEBMrMo8CPn3MuA\nhviwnZX29mAIXLgw6nwkEsHcoDPPhLe9LUgEnn568jyi//qv4NjmZjjjjKB36JxzgtvF84huuaX8\nsLylVC6usKcnOcPkZc8LLmNjQQ+a5wXnaWqPlHqVZAVp9Ha4f6yfeHTyDz6Hxg7x2MHHuP6C6ydt\nVwEEERGZyYwJkHPOMzPfzDqdc4OLEdSyEi6MunPn/BOgqSIROP744PKGNwTb9u8PEqFwLtGnPx0k\nANEonHxykAw5B9/6VlAgAea+OOtchL05vj+RrIS3b7sNPvrR+cUVjVbXu1NNr1J4XKlepXh8cpKk\nXiVZQo3cDuf8HPtH9tPZ1Dlpe++uXoDC+j+e7xGLxFQAQUREZlTtHKAR4GEz+yEwGm50zv1hTaJa\nbrq6gi/Hudzsy1vP1vr1k+cRjYxMnkf0rW9NrkgXSqXg/e8PhtT5/kRSUipRqfZ66jbfn/2aR6lU\nUC78hz8MCjesXx9cwtvr1s2uQl7IbGF6lcIepfA6mQySo+JkaWqPUiwWJK8iC6sh2+HBVJDvRWzy\nv6mevh46kh2ctvY0IJj/owIIIiJSjWq/rX87f5FSwoVRn3suWLR0MbW1wUUXBRcIhuKdfnrJRMSl\n0/hjIxCJQjQCsShE48H9SCTYFsn3sOTvW+F+dNJ+KxxT4n4kMnGJFj3fRz5S+j1ks/DEE/DznweV\n7Eq9x+KkqNTt7u65J5+z7VXy/SBRCnuVpiZJzgXvNxx6V5wsaQiezF1DtsN7hvdMm9fjnKOnr4cL\nNl1ANBL82814GTa0bViKEEVEZJmptgjCV8wsAZyU3/S4cy5bu7CWoTVrgiFdhw9PbAt7IcIv2DXs\nGch4GdJ+hnEvzar1a4nvOzD9mPXdPPgvH8ABBsG1Gc45DMPh8tuL70+/Jv/Y4D3mr/MbIvkv84Zh\nZkSIFO6f+sW1JPb1T4/rqLXs/MYnMDOiY+PE+geI9x8i1n9o4vaBQ8QOHiL61E5iBwcwz5v0HC4S\nwVu9Cn/tGrx1a/HWdeOv7cZbtxZ/3Vr8td249etw7e2YRQqldM2C2BLfu52WT3yWyL79uKPWk3nX\nO+E1ryFmUaIWDY4P/3/CzMMdnQuSo3JD8IqTpbCXauqcpVKFHZQsNaxGbIfHsmOMZkcLi5yGnh18\nlr0je7n2nGsL21QAQUREqlVVAmRmLwW+AjxD8JV3i5m91Tn309qFtswkEkEhgmw2+KKbywW9MalU\nMOclnQ6Gq3ne9DkoxQlSeLvMF13nHBk/S8pLM+6lGMqOMJwbJecHX66NCGPXvZmjP/ppIql04XF+\nU5KB3/9tuhK1HSLi8j1PjinXDg68/a1s/JtPTotr/3XX4Ds/yBmaEqS3HIXbsn7S48PnCB7kEz0y\nmE+SBkj0HyLeP0D84ADx/gESu/poeuAhYoPD0+Lzkgmy3avIrF1Ntns1mbWriB06QusddxLJ5s/h\n3n3EP/iXPD22j0OvuggzI2EJEpEYyWiSZCROMpokHokVEqRYJErMYhNrlJhNVOmbaQhemBylUhN/\nI2HZ8krJUjgUL0yWpv4Nac7SitKI7fDB0YPEItM/pnr6egAK6/+EwmpwIiIilVQ7ZugfgFc55x4H\nMLOTgG8A59QqsGUpHNpUSTjfJEyScrkgOZqaKDmH7/wg2fEzjObGGSLNiEvjRwzLDztLROK0RJuI\nxia+7GauuIT9kTjdn/43Yvv7ya1fy8F3/CbDl728xidgYpFCY0oCZzB2+cXst8i0uMYvezmznra8\nrgXWbcAHUvnLtFjSGWL9h4gezPcg9R8k1j+Qvz5E02NPEfvZISLpzLTHRtMZjv/LT7L1k1/Db2vB\na2sl19qM19ZCrrWZbFsL6daW/P0WvLZmvLZWaG8n2tFJtGMV8Y4ukskWEtF4kCSFl0hs8nyGcLjc\nVKUq5l1+efD3EyZL4Rys4kQJJg/Di8eDS1gyPB6fnixpKN5y0FDtsOd77BvdV3JeT29fL5vaN3FM\n5zGFY+PR+LRKcSIiIqVUmwDFww9dAOfcr8xMnzRzEX7hnPKF1/M90l6adC7NcGqI4dQgI+ODkPMw\n34h4TSS9JjqyDstmg96lTA4sC+RHwYRfgKNRhl/+IoZf+eLJvQAuXAyVieNLXcP0++X2VfvY/P3h\ni85l+KJzg+3hHKFsNpiPZAs7PNAlE2Q3byC7ucK8AOc48fzLsTLFG0ZefAHRkVEiI6PERkdJHDxc\nuB8ZL5V2TeYlE3htLXj5ZGm8tYVcewt+a5AsWXs71t5BtL2TaGcX1tFBtKOT2D33E/unT2DzqZhX\nPGdpdHRygQcovcZSPD55KF4iMXnOktZZWkoN1Q4PpYdwzk0rfuD5HnfuvpNXHf+qwg8uaS9Ne6J9\nKcIUEZFlqNoE6F4z+wLwtfz9NwP31iaklS/n50jn0qS9NEPpIYbSQ4xnxwvzceLROIlogq6OdRND\nqkoJh075Png5yOXLPqfTQYKUTgfzT0Lh/KOp16Fwnku4L5ggk98XmTim+LHhMcX7iy+lHhNehzGm\n0zA6VvqLefhlOxKpTS+FGbkyc6ZyR63jwPv/b/nH5nJERsaChGg0nxSNjBEZGZ1IkoZHJ+1PjIzR\n1H944pii4YAzSqXw/+x9jNx+M15nB35nB35nZ3Dd1YHr6sTl79PRQSQaDeZiYUQsQiQWgVgwTyvs\noTMzLJyn5fuYP46NjsKwh/kO8/yi4wDngjlUBEP8YvEE0fBX98LfjE1eeDcsiFHN7VJ/O5Vul1rc\ntziG8DlXRlW+hmqH9w7vpTnePG37I/2PMJQe4oWbJ4a/ZbwMG9s2LmZ4IiKyjFWbAL0d+H0gLLf6\nM+DTNYlohcl6WdJemvHsOEPpIUYyI6S9dGGuTJjsrGqeQ/W4ckOnlqvi4YFeUTIXDhEcGZk4tni4\nV/GclzkUCjj4jt9k/V//87S5SQff8ZuVHxiL4XcFycec5XKTE6Z8ArXx3R+eOogQAMtkSf7qKaKD\nQ0SHR7EwaZzCRSJ47a3kOtrIdbST65y4TnXm77e3ketsx+toJ9vZhtfVgZ8s+nsK81DAgooZOBxr\nfvBztnz26yQOHCSzrptnf/eNjF36CpqjTbREkzRFksSIELco8UiMKJFJvYDBkxb1HJa6PelNTxne\nN+mNFm2//Xb4zGeCtbLWr4e3vx0uvXRyD1dxMl18u1SxklJJVKn7M7nhhqAE/XPPcTq8YOYHlNUw\n7XDYXq5uWT1tXzj/54LNFxS2OedKJksiIiKlVJsAxYB/ds59HAqrkmu57SmyXpbx3Hjhw3s4M0zW\nC4anGUY8GicZS+qDupwywwMnKZ4/FSZJqdTkHq+pc2HywwKnfbnNC+dGLcWcqSCJ6sTvmrzIY+6o\ndWV7pZ7d/sXgju8HydPgcJAQDQ4RGRwieiR/f2iYSH5f06FBok/tCo6pMHTPTyaC3qWONrzOjvyl\nvdDjFH92Fx3f+xGRbPB3ndx/kK1/93n2D45x+Ne2MYaPBxCxYKSlGdFIlKZoE8lYE82xJhKxRFBA\nIpogHo0HQ5zCXpswgZ3WEznD5dZb4W/+ZmINrH37gvutrRNDBovXqiruPQ0Tr+J9oal/R6HiRK04\noSq+xGLw3e8G61zle2LjzH66W5GGaYcPjR8qlLeeqqevh5O7T2ZNy5pJ21UAQUREqlVtAvRj4GKC\nhfgAmoH/Al5Y9hENZjw7ziMHHsEn+PKUiCZojjXTlmhb4shWmPDLZaXKamFylMtOHhaYTkEmP38q\n/AKb710Yvuhchl964eQv4MUV+yz/n0UqFFBVr1Qkgt/Rjt/RTnZL9cN/LJMhMjQ8kSgNDhEdHA6S\np3ziFN5PPP1cYf/U0uOFMDIZNnzyy2z45Jfn+nZrI5WCd78b/uqvpvf8VLqESXLxdfH+UtuKk+zi\nXqJvfGPyMNT5aYh22Hc++0b2lWw7x7Pj3L/3fq45/ZrCNhVAEBGR2ao2AWpyzhXGHznnRsxMCy7k\n+c7nqcNPEY1E6ahxmWmpQjikqVySFK7R43nB3KlsvmR5Jj93KpcfigcTvQZhr0DxL/+zSYbKFFmY\n9DxFPRrDL7kAshm6P3cDsQMHya3r5uB1b2H4VS8pPxys2lASCbzuNXjda2Y+uCj+yOgYx7/86mA4\n3NTdwL6/fHdwywXziYIuIFfoYTHfFfYXtjmH73v4vo/nPJzvTfTIuODYKBYMpyNG3KJELULUQYRg\nnakIwKc+VTZurrhiYnhlqUvYE1R8HZaxL95W6TlKPU+ZhHEeGqIdHkoP4fleyR6g+/beR9bPcuGW\nCwvb0l66ZKU4ERGRcqpNgEbN7Gzn3P0AZrYNWLCfNZe7AyMHGMmMzG0ejyy+cI2e+Bx/MZ70Bb3o\nUm5f8Zd+mJxMFQ+/Kh6C5fsMX/1qhq+6In/fCxKz8It58dpAU4dpha9Tar7KbOauTDlnflsrufXl\nh+YNX/6KWZ/KanjOI+d7ZF2OnO8VFuwFcAZJS3DaTd8qGZe/YQO59783KAKRvywa5+BlLwvKmC+M\nhmiH943sKzucraevh3gkzraN2wrb0rk0m9o3LVZ4IiKyAlSbAL0L+A8z25O/vwF4Y21CWl7Gs+M8\nO/isfoFsJLZ4Q+HKChMlzytKmjzwi5KpbHb6nKlw3ani3olS81x8f3LSlB/mdfD3rmH9336q+oIR\n03q+3KSrksdNuR0l6AVKEoew6rObeJ6c57H3d97Ipo/9K9GiNZ28ZIKnf/NKBvbeX8hBwRF1kaA4\ngxkxFyUaiRBzUWIWKazVFC1KmCIYkUg03+Nkk7aXTULD29ddN3lu0vys+HY4lUsxmBos+2NS765e\nztpwFi3xiY4vwzT/R0REZqViAmRm5wJ9zrl7zOxk4PeA1wG3A08vQnx1zTnHU4efIhFNlJ2wK1IT\nxXNO5mpqAjX1dnjJZvPDujyGL3kp+B7d//o1Yv2HyK1dw8G3/QbDL9oGQ0Ol4yxW3PNU7na544vv\nF5VojxEjdeUVHEg0Ty5k8Qe/jX/FxXQVl2mPRPAtGLbqA9kIpJ3Dt6A3Kdjugsp3EJS/Mz9/sXxJ\nPMOFBQgjMWKRGLFojKjFiIWL3kbjxCJRYu/8bVo2dtP84b8hsms3Weemr7o7g0ZqhwfGBsr20g2M\nD7CjfwfvuuBdk7Y7nBIgERGZlZl6gP6VYNItwIXAnwHvBM4EPge8vnah1b8Doxr6tlLc8vgtfPzO\nj7N3eC8b2jdw/QXX85rnVbng6HI1l6FwwNfd43y8K8HeYdjQluD6C7bwmuedPnHAEvWODb/jbQy/\n420zHhfJXxaC7/zCJeccGefhuyzOG8fPBdvdpeeQeeXXOXHNiTzUvuHhObxMQ7TDvvPZO7KX1kRr\nyf137roTYNL6Pzk/RyJfTVBERKRaMyVAUefcQP72G4HPOee2A9vN7IHahlbfxrPjPHPkGQ19WwG+\n+/h3+fM7/pxULhimtGd4Dx+44wMAKz8JmqVbHr+FD9zxgYlzNbKHD/zkg2CRhjxX1c4rGkoPFdb+\nmoOGaIeH08Pk/ByxSOmPpZ6+HtoT7Zy27rTCtoyXoT3ZvlghiojICjFjAmRmMedcDngFcO0sHrti\naejb3NWypyXjZRhKDzGYGmQwPThxnb89lB7iSOpI4Zgj6eD2wPjAtOdK5VK898fv5TuPfYd1retY\n17aOda3rWN+6vnC9pmVN2S9ry5lzjtHsKAfHDk669I/189UHvlpIfkKpXIq/+MlfsH90f3Cuii6t\n8VZsqedLLX8N0Q7vH9lPMla6cqNzjp6+Hs7ffP6kf3OZXIbO9s6SjxERESlnpg/PbwD/Y2YHCaoN\n/QzAzE4ABmd6cjO7FPhnIAp8wTn30TLHnQv0Am9yzt1UffhL48DoAYYzw6xunr5Keb2oxyFd03oP\nSvS0OOcYyYwEyUlqiMH04KSkpTihKU5sBtODjGXHKr5+e6KdzqZOOpOddDZ1sqF9Ax3JDr75Kvls\nzwAAIABJREFUyDdLHp/zcwymB3li4AkOjh3Ec5PLGhtGd0v3tC/9hSSpLbjuauqaU/Wxhf5/mMql\npiU1YWJzaOzQpNvjuenFxSIWwXd+iWeG0ewoH+v52LTtLfEW1rVMnJu1rWunna+1LWvLDnsSoAHa\n4XQuzeHUYbqaukru7xvqY/fwbt521uQhjpr/IyIic1ExAXLOfcTMfkxQbei/3MQYjgjBGPSy8quU\n/wvwSmAXcI+Zfdc5t6PEcX9LsKBf3QuHvnUm6/dXx2oSDc/38uWFc5Munu+RczlyXm7Sfs/3yPrZ\nYL+fI+dyE7enPtafvC987Jcf+HLJ3oP3/fh9fOruTxV6ZMp9yYZggdnOZCddTV10JDvY1LGJ05Kn\n0ZHsmJTcdCQ76GrqojMZ3O5IdpTtrfvZcz9jz/Ceads3tm/kpjfcVDhfA+MD7B/dz4HRA4VLeH/v\nyF4e3P9gyd6keCQefPFvmUiKpiZL61rX0ZZoK/SWVPP/ECDrZRkYH6B/rL9kchMmNQfHDjKSGZkW\nG8CqplV0t3TT3dLNWUedxdqWtaxpWUN3S3fh9tqWtXQ1dXHxv19c9lzd8hu3TDo3B0YP0D/az4Gx\n4PZDBx7iwOiBaX8DAK3x1slJUevaSecmTJSa480l30N4zuot6Q/j+vvev2f/yH5Yzwtm+/hGaIcH\nxgcwrGxvYU9fD8Ck9X9CSoBERGS2Zhw+4Zy7s8S2X1Xx3OcBO51zTwGY2Y3AlcCOKce9E9gOnFvF\ncy4p5xxPH366roa++c6nf7SfPcN72DO8h93Du/nMvZ8pmWi8+4fv5n0/fh85P5evc1Ufsn6W09ZN\nJDFdya5pCU14XYsvO9dfcP2kZAOCL1XXX3B94X40EmVt61rWtq6t+FwZLxN86Z+SCISJ0s6BnfT0\n9TCcGZ722OZYcyEpevjAwyX/H/75HX/O9ke3FxKbI6kjJeNoT7QXkppTuk8p3A6TmvD26ubVs5pA\nXulctSXaaEu0cdyq48o+PuzhKz4v/WOTz9cv9v2CA6MHyHjTC6a1J9pL9iY9N/gc33rkW4XHhAlj\nKpfishMvmxYDMOnfwNRtpY6p5rip83x++OQP+buevyPt5cuGR0iUPTkVrOR22DnHvpF9FXsBe/p6\nOKrtKI7tOrawTQUQRERkrmo5fnwT0Fd0fxdwfvEBZrYJuAp4GRU+eM3sWvLj3o8++ugFD7RaB0YP\nMJQZWtShbxkvw76Rfewe3s2eoT2TEp09w3vYN7KPrJ+t6rkcjt8687eIRqJB6d6iS1C6N0bM8veL\njinctlj5xxbfL36sRYnnSwNf/NWL2TNSuvfg45d8fKFPXdXCXoKF6D1IRBNs6tjEpo7KCzOOZkan\nffHfP7K/0FtSaggawHhunFQuxdaurWzbuG1aYhP23NTqV/H5niszoz3ZTnuyneNXH1/2OOccQ+mh\nyb1JU87XfXvu48DogbJ//6lcig/c8YFCz1mDWrB2OH/sgrfFw5lhMl6mbALk+R537bqLlx/38kk9\nRCqAICIic7XUE2j/CXiPc86vNFHaOfc5gnKvbNu2bUm6LioNfZvP0JuRzMi0pGbP8B72Du9l9/Bu\n+kf7J/0KbRhrW9eyqX0Tp68/nUtOuISN7RvZ1L6Jje0b2di+kdd84zVlhyn98Qv/eO4nYZ6uv3Dm\nnpal8prnvWZRh0u1JlppTbSytWtryf0v+8rLyv4/vPH1N9Y4usoW41yZWdDz19TJiWtOLHucc44j\nqSNc+MULy/Zq/ukL/7TwxdnyiweVam+mHTPl2PD+xFXp/YXjgA/+5IMzvNO6UFU7DLVpi/eP7icR\nLd8xtqN/B0fSRyaVvwYVQBARkbmrZQK0G9hSdH9zfluxbcCN+Q/dbuByM8s5575Tw7hmrdLQt0pz\nNV590qsZGB+YlNhMvQymJ89hjkfibGjfwMb2jVx09EVsbAuSmo0dQZJzVNtRFb8sQHVDupbCQva0\nrHT1+v+w3pgZq5pXsaF9Q9mE8W1nz7w2UK189r7PloxrEdV1O5zxMgyMDZQtfgDQu6sXmD7/x+Eq\nzgkTEREpp5YJ0D3AiWZ2LMEH7puA/118gHOuMKDbzP4NuLXekh+oPPTt43d+vORcjff86D3TvsBC\nMNk77K05a8NZbGjbMKn3Zm3r2jlVDCtWz4nGYve0LFf1/P+wHtVrwlgqrkVW1+3w4fHDmJUvfgDB\n/J+T1pxEd0v3pO2GkYyWLpstIiJSSc0SIOdczsz+APgBQfnVLznnHjGz6/L7P1ur115I49lxnj3y\nbNmqb3uH95bc7jmPtz7/rYXenDDJ6Uh2LMq6KEo0lj/9P6xevSaM4euHVeCc76ZXdqihem6HnXPs\nHd5LS7yl7DGpXIr79t7H/37+pJyNnJ8jHo2rAIKIiMxJTecAOee+D3x/yraSH7jOud+sZSxzEQ59\ni0fjZau+VRp6856L3lPrEEUkr14Txtc87zW8ZOtLgh9BPrTx4cV+/Xpth0cyI6S9NC2J8gnQ/Xvv\nJ+NleOGWKfN/VABBRETmYX5jrVa4cOhbpfKs119w/bQha/Uw9Eak1pxz08o+i1TrwOiBGXtwevp6\niEfibNu4bdL2TC5T12uxiYhIfVvqKnB1a6ahb6EXHf0icMHcnrHsWN0MvRFZSM45Ml6GjJfBcx7O\nuWDuBlZYuNbMwBFUSXPBJPVoJErEIkQsQtSC2+G2lcx3/qRL1quuVH2jyHpZDo0fmrF97enr4cyj\nzpz2I5QKIIiIyHwoASqhmqFvoZsfuxkfn2++/psVy/WKLBfOObJ+lnQuTc7PYRjOHO2JdtY3r6ct\n0UZTrIlkNIlZkAB5vhdcOw/P9wrXYdKU9bPkvBwZP0MqkyLncuDyJaPzCRMWvHaYJBlWSJaiFl2U\nxYeLkxbnXOE9hbd95xOEbIVzVVz2OizFHa6DFY/GSUQTtMZbNWSryJHUkUISXc7h8cPs6N/BO89/\nZ8n9tVrrSkREVj4lQCVUu+Cpc46bHr2JM9afoeRHlq1CkhL2Uhi0xdtY17qO9mQ7yWiSplhT2S+r\nEYsQic6uR8c5NylZCpMoz3lkvWwhnqyfJetlSXkpMl4GMysMuwt7nJwFSUjUopgF1xAkMw5XeO7C\nY0rEEpq0gG8k+AEkHo0Tj8QL+8IerULPVlEvV3iR8pxz7BnZU3FoMcCdu+/E4aat/5PzcySjSWIR\nfXyJiMjc6BNkilQuVdXQN4CH9j/EzoGdfPilH16EyKSWfOeT83NkvWzhS234ZXclKe6RCdcNbYm3\n0N3cTXuyPejZiSVr/r7NjJjFZv0ltrh3qbjHKefnCu8t5+eAiWSmkNREooVheOUui1GhsdGNZccY\nz4yzuqXyD0y9fb20xlt5wfoXTNqezqXpSHbUMkQREVnhlAAVCYe+xaKxqobbbH90O82xZq446YpF\niE7mKxzalfWy5PwcDlcYwmRmtCZa6WrqIufnSOfSjGZGJ4aAMbmXIEyQFmto1lxkvSxpL03Wyxbe\nQ3O8mdXNq+lIdhSGsdVr/KVEI1GiRIOCzrIsVVP8AIL5P+dvPn9akpz1skqARERkXpQAFTkweoDB\n9OCMQ98g+BXz1l/dyqUnXEpbom0RopNqhQlO1s/i+d6kX/VbYi10JDtoS7SRjCVJRBMkoomyPRHF\nPUM5P0fOz5HKpUjlUqS9NKlcioyfmTwPpGgeSziMKhyeVSth0pbxgmVmzIxkLMmqplV0NnUWhrEt\np2RHVp6cn6N/rH/GBKZvsI++oT7eesZbp+1TAQQREZkvJUB5qVyKZ448U3Vp1R/s/AGj2VGuPuXq\nGkcmpYTJSJiYFHppLJgc3RJvoTXeSnO8mUQ0UZjHMdskJGKRQpJUjnOuEE+YeKVzQXKUzqVJe2lG\nvBGYWjHamNSLVG18heFeuUyhZyoZTdKZ7KQj2UFzvJlkTHMkpP4MpgYLPxBU0rurF2Da+j8hFUAQ\nEZH50DckZlf1LbT90e1s7dw6bX0KWTjh3I6wglhYJQwIKmslWlndvJqWeEuQ5ESCiluLPY/DzGZc\nlT6c9F+cKGVyGcZz44WepKH00KShduHjYpEYnvMKpaUT0QTtyXY62ztpjjfTFGtSsiPLwp7hPbTE\nyy98Gurp62Fd6zqOW3XcpO0qgCAiIgtBnyLMbugbwDNHnuGePffwxxf+sSZNz1M4LyecvB6Wxg2/\n+LfGW+lIdtAabyURSxQSneU2lKvaSf9h0lfcw5XyUiQiiUKyU838CZF6M5YdYzQ7OmM76zuf3l29\nvGzry6a1ryqAICIiC6HhE6DZDn0D2L5jOxGLcOXzrqxhZPWveK2UsNxwuK34vu/8wqKZwKQkJ2IR\nmmPNrGpaRVuirTDcLB6NN+SvvNFIMBwuSXKpQxFZUAdHD1b1b/rR/kc5kjrChVsunLYv42WUAImI\nyLw13jfMInMZ+pbzc/znY//JS455Cevb1tc4woVTnIxUul+8Jkrxr6/FCz6Gw7QiFplUEa14wn9x\n+eFy5YbjkcrDxkRkZfB8j32j+6pKXnp29QBw4ebpCRCgAggiIjJvDZ0A9Y/1z2roG8DPnv0Z/WP9\nNS1+UK5npVSyEk7+r2aBx6hFiUSCpCURTRQSlTCRCZMZw0omLFrsUUTmotriBxCs/3Pi6hNZ17qu\n5H4VQBARkflq2AQolUvx9OGnZzX0DYLiB2ua1/DSrS+dti/rZfGcVzZ5KR4GFirVsxJWBQt7VsJk\nZWrvSrgCfTUJi4jIUtk7sreqnpt0Ls29e+7ljc9/47R9KoAgIiILpSE/SeYy9A3g4NhB7njmDt5y\nxlumDd3K+TlGMiN0NXVVPQxMPSsistKNZ8cZTg+zumXmnvb7991P2kvzws3Ty1+nc+lZ/2AlIiJS\nSkMmQP1j/QymBqv6QC5282M3k/NzvP6U10/bl8qlOKrtKI7pOmahwhQRWfYOjR+q+oem3r5eohbl\n3E3nTtunAggiIrJQGq67oTD0rWl2vyQ659j+6HbOOuosjl99/LT9WS9LV1PXQoUpIrLseb7H3uG9\ntCXaqjq+p6+HM446o+TxZkZTXPN/RERk/hoqAZrr0DeAB/Y9wJOHn6xY/KCaBf5ERBrFcGYYz3lV\ntbeDqUF+eeCXvGjLi0rud86pAIKIiCyIhkqAwqFv1f4aWWz7o9tpibdw2YmXTduX8TK0xltV0llE\npMje4b00x6orW33nrjtxuJLr/+T8HE2xJhVAEBGRBdEwCdBch74BjGZG+d4T3+PSEy4tmTyNZ8dZ\n07JmIcIUEVkRUrkUg+nBqtft6dnVQ0u8hdPXnT5tXzqXpj3RvtAhiohIg2qIBGg+Q98Abt95O2PZ\nsbLD33zn057Uh7OISGhgbICoVd/e9vb1cv6m80v2pKsAgoiILKSGSIDmM/QNguFvW7u2cs6Gc6bt\ncy5YhFTzf0REAr7z2TOyh9ZEa1XH7xraxbODz5Yc/haqtidJRERkJis+AUrlUjxz5Bk6mub26+FT\nh5/ivr33cfUpV2Nm0/ancilWNa3S+j0iInnD6WE836t6zk7vrl6Akuv/hJKx5ILEJiIisqK/tYdD\n38LFSOdi+6PbiVqUq06+quT+lBckQCIiEtg3sm9WCUtvXy9rW9ZywuoTpu3LelkVQBARkQW1ohOg\n+Q59y3pZvvPYd3jJ1pewtnVt6YMctCXn9vwiIitNOpfm8PjhqocF+86nd1cvF265sGQve8bLqACC\niIgsqJomQGZ2qZk9bmY7zey9Jfa/2cweMrOHzazHzM5YqNee79A3gJ8+91MOjh0sW/wgHOKRjGpo\nhojUp8VuhwfGB2Y1JPjxg48zMD5Qdv0fFUAQEZGFVrMEyMyiwL8AlwGnAr9hZqdOOexp4CXOuRcA\n/x/wuYV47YUY+gawfcd2ulu6eckxLym5fzwXlL8u9auliMhSW+x22DnH3uG9s+oV7+nrAeDCzSqA\nICIii6OWPUDnATudc0855zLAjcCVxQc453qcc4fzd+8ENi/EC8936BtA/2g/P3nmJ7z25NeWXeA0\n62Xpauqa82uIiNTYorbDw5lhMn5mVj889ezq4fhVx7O+bX3J/WamAggiIrKgapkAbQL6iu7vym8r\n523AbaV2mNm1Znavmd3b399f8UUXYugbwHce/w6e83jdKa+reFxrvLoyryIiS2DB2mGYuS3eP7p/\nVkOCM16Ge/fcywu3lK7+lvWyJKNJFUAQEZEFVRdFEMzsZQQfvO8ptd859znn3Dbn3La1a8sUIwiO\n45kjz8x76Jtzju07tnP2hrM5ftXxJY/JeBla4i1le4dERJaTmdphqNwWZ7wMh8eqL34A8Iu9vyCV\nS5Vd/yftpVUAQUREFlwtE6DdwJai+5vz2yYxs9OBLwBXOucOzecF+8f6OTJ+ZF5D3wDu33c/Tx95\numzxA4Dx7DhrmtfM63VERGps0drhw+OHwZjVnMieXT1ELcp5G88ruT/rZVUAQUREFlwtE6B7gBPN\n7FgzSwBvAr5bfICZHQ18G7jGOfer+bzYQg19g6D4QUu8hctOuKzsMb7zaU/ql0kRqWuL0g6HxQ9m\nOyS4t6+X09efXrEtVQEEERFZaDUbWO2cy5nZHwA/AKLAl5xzj5jZdfn9nwU+CKwBPp3/1TDnnNs2\nh9fi2SPPznvoG8BIZoTbdt7G5SdeTmui9Ie5cw4zm9VQDxGRxbZY7fBIZoS0l6YlUX2bOJQe4uED\nD3PdtuvKHqMCCCIiUgs1nVnqnPs+8P0p2z5bdPt3gN+Z7+v0j/UzMD7Ampb5D0m7bedtjGXHKg5/\nS+VSdCY7iUai8349EZFaWox2+MDogVnPh7xr1134zi+7/k/Wy5KMqQCCiIgsvLoogjAf6VyaZ448\nQ2dT54I83/Yd2zlu1XGcddRZZY9J5VKsbl69IK8nIrKcZb0sh8YPzXr4W8+uHlriLZy+/vSS+9Ne\nmo6E5v+IiMjCW9YJ0EJVfQs9OfAkv9j3C64+5eoZJ/LOt9CCiMhKcCR1BIeb9YLQPX09nLvxXBLR\nRMn9KoAgIiK1sqwToENjhxgYH1iwZOSmR28iFonx2pNfW/YYz/eIRWI0xZoW5DVFRJYr5xx7RvbM\nuvdnz/AenjnyTNn1f0JqZ0VEpBaWbQKUzqV56shTCzb0Letlufmxm3np1pfS3dJd9rjx3DirmlbN\n+tdOEZGVZjQ7SiqbKtuLU05vXy8AF24uvf5PSAmQiIjUwrJMgBZ66BvA/zz7PxwaP1Sx+AFAzs+x\nukXzf0RE+kf757QYdM+uHtY0r+GkNSeV3J/1sjTFm1RoRkREamJZJkALPfQN4KYdN7G2ZS0vPubF\nFY9zzqn8tYg0PIejf6x/1u2hc47evl4u3HJh2Z50FUAQEZFaWpYJ0EJWfYOghOtPn/0pV518VcUe\npYyXoSneNOvhHiIiK43vfJxzRGx2HyOPH3qcQ+OHypa/hqAHaCHbeBERkWLLMgHy8BZ0bYjvPPYd\nPOfxulNeV/G4VC7Fmub5rzUkIrISGLOfC1nt/J9kVAugiohIbSzLBGghOefYvmM72zZu49hVx1Y8\n1vd9OpP6VVJEZK56dvVwbNexbGjfUPE4FUAQEZFaafgE6L699/HM4DMzFj9wzuHQ/B8RkbnKeBnu\n2X1PxfLXKoAgIiK11vAJ0PYd22mNt3LpCZdWPC7tpelIduhDWURkjh7c9yDjufGKCVDaS6unXURE\naqqhE6CRzAi37byNK068YsaenVQ2VXF9IBERqaxnVw8Ri3DepvPKHpP1s3QkVQFORERqp6EToO8/\n8X3Gc+NcfWrl4W8QlHxtTcxutXMREZnQ29fLC9a9oGKCY85UAEFERGqqoROg7Tu2c8LqEzhj/RkV\nj/OdTzQSpTnWvEiRiYisLMPpYR7a/xAXbqlc/c3hVABBRERqqmEToJ0DO3lg/wNcfcrVZRfjC41n\nx1nVtGrG40REpLS7dt+F57yK6/+Ea61prqWIiNRSwyZAN+24iVgkxpXPu3LGYzNehtXNqxchKhGR\nlam3r5fmWDNnHnVm2WMyXkYFEEREpOYaMgHKeBlufvxmXr715axpqW5hU5W/FhGZu55dPWzbuI1E\nNFH2GBVAEBGRxdCQCdBPnvkJA+MDVRU/yHpZmmJNJGOalCsiMhf7Rvbx1OGnKpa/BsBpAVQREam9\nhkyAtu/YzrrWdVx09EUzHpvKparuJRIRkel6+3oBZk6AQBXgRESk5houAdo/sp+fPvdTrjr5KmKR\n2IzH5/ychmSIiMxDz64eVjev5qQ1J5U9JuNlaI43qwCCiIjUXMMlQP/52H/iO5+rT5l5+BuAmWn+\nj4jIHDnn6O3r5cLNFxKx8h85GS+jH5tERGRRNFQC5Jxj+6PbOW/jeRzTdcyMx6dzadoT7VX1FImI\nyHRPDDxB/1j/jOv/qACCiIgsloZKgO7Zcw/PDT5XVfEDCOb/qPy1iMjc9fT1AFRc/wdQAQQREVk0\nDZUAbd+xnbZEG5ccf0lVx/vOpy3RVuOoRERWrt6+XrZ2bmVj+8YZj1UBBBERWQwNkwANp4e5/cnb\nueLEK2iON894vO98IhbR/B8RkTnKelnu3nP3jMPfVABBREQWU00TIDO71MweN7OdZvbeEvvNzD6R\n3/+QmZ1dq1i+98T3SOVSvP7U11d1fCqXYlXzKsysViGJiNTcUrbDD+5/kLHs2IzlrzNehs5k50K9\nrIiISEU1S4DMLAr8C3AZcCrwG2Z26pTDLgNOzF+uBT5Tq3i2P7qdk1afxAvWvaCq49O5tOb/iMiy\nttTtcG9fL4Zx/qbzKx6X9bO0J9sX6mVFREQqqmUP0HnATufcU865DHAjcOWUY64EvuoCdwJdZrZh\noQP51aFf8dD+h7j61Ktn1aPTGm9d6FBERBbTkrbDPbt6eP6659PZNEPvjgogiIjIIqplArQJ6Cu6\nvyu/bbbHzNtNO24iHonz68/79aqOz/k5ktEkyZgm5IrIsrZk7fBIZoQH9z044/C3kAogiIjIYlkW\nRRDM7Fozu9fM7u3v75/VYzNehpsfv5mXH/vyqoe0jWfHNfxNRGSK4rb4YP/BisfevftuPOdVVQCh\nJd6iAggiIrJoapkA7Qa2FN3fnN8222Nwzn3OObfNObdt7dq1swriv5/+b46kjlS99g8EPUAzDtkQ\nEal/C9YOw+S2uHttd8UX7unroSnWxNlHVa6pkPEyWgBVREQWVS0ToHuAE83sWDNLAG8CvjvlmO8C\nb8lXIboAGHTO7V3IILY/up31reu5aMtFs3pca0Lzf0Rk2Vuydrh3Vy/bNmybcSixCiCIiMhii9Xq\niZ1zOTP7A+AHQBT4knPuETO7Lr//s8D3gcuBncAY8FsLGcO+kX38/Lmfc+0511Y9vCLjZWhLtBGL\n1OzUiIgsiqVqh/eP7GfnwE6uOvmqKoJUAQQREVlcNf2W75z7PsGHa/G2zxbddsDv1+r1v/3ot/Gd\nz+tPqW7tH4Cx7BhHdx5dq5BERBbVUrTDvbt6AVQAQURE6tKyKIIwF77z+faj3+b8TeezpXPLzA/I\nc87RntBwDBGRuert66WrqYuTu0+ueJwKIIiIyFJYsQnQ3bvvpm+ob1bFD5xzRCxCc7y5hpGJiKxc\nzjl6dvVw4eYLiVjljxgVQBARkaWwYhOg7Y9upz3RziXHX1L1Y8Zz43Q1dc34oS0iIqU9dfgpDowe\nqGr4W9bLKgESEZFFZ8Hw7+XDzIZJ0IePt+BPHiFKjvQ8nrsbqLw4xtKp19jqNS5QbHOl2Mo7xjk3\nu1r+dcrMRkjw3Lza4ghRsqRw+AsYGiz9/+dy6jUuUGxzVa+x1WtcsPSxrZh2WOZuOZY6e9yl3bal\nDqIUM7vXOcU2G/UaFyi2uVJsDeMxtcWzU69xgWKbq3qNrV7jgvqOTRqHxnqJiIiIiEjDUAIkIiIi\nIiINYzkmQJ9b6gAqUGyzV69xgWKbK8XWGOr5XNZrbPUaFyi2uarX2Oo1Lqjv2KRBLLsiCCIiIiIi\nInO1HHuARERERERE5kQJkIiIiIiINIxllQCZ2aVm9riZ7TSz9y5xLM+Y2cNm9oCZ3ZvfttrMfmhm\nT+SvVy1SLF8yswNm9suibWVjMbP35c/h42ZW/UqxCxfbh8xsd/7cPWBmly92bGa2xczuMLMdZvaI\nmf3f/PYlP28VYquH89ZkZneb2YP52P4yv70ezlu52Jb8vK0k9dQO5+NRWzz32Jb834ba4jnHVpdt\nsdphWTacc8viAkSBJ4HjgATwIHDqEsbzDNA9ZdvfAe/N334v8LeLFMuLgbOBX84UC3Bq/twlgWPz\n5zS6yLF9CPiTEscuWmzABuDs/O124Ff511/y81Yhtno4bwa05W/HgbuAC+rkvJWLbcnP20q51Fs7\nnI9JbfHcY1vyfxtqi+ccW122xWqHdVkul+XUA3QesNM595RzLgPcCFy5xDFNdSXwlfztrwCvXYwX\ndc79FBioMpYrgRudc2nn3NPAToJzu5ixlbNosTnn9jrn7s/fHgYeBTZRB+etQmzlLGZszjk3kr8b\nz18c9XHeysVWzqL+W1ghlkM7DGqLq42tnHpo75b8vKktXtC4ylE7LEtiOSVAm4C+ovu7qNwQ1ZoD\nfmRm95nZtflt651ze/O39wHrlya0irHUy3l8p5k9lB+WEXbRL0lsZrYVOIvgl6q6Om9TYoM6OG9m\nFjWzB4ADwA+dc3Vz3srEBnVw3laIejxnaovnp27+bagtnnVMddkWqx2W5WA5JUD15iLn3JnAZcDv\nm9mLi3c65xyVf/VYNPUUS95nCIbQnAnsBf5hqQIxszZgO/Au59xQ8b6lPm8lYquL8+ac8/J/+5uB\n88zs+VP2L9l5KxNbXZw3qRm1xXNXN/821BbPXr22xWqHZTlYTgnQbmBL0f3N+W1Lwjm3O399APhP\ngi7b/Wa2ASB/fWCp4qsQy5KfR+fc/nwD6QOfZ6K7e1FjM7M4wYfaDc65b+c318V5KxUK+Qu3AAAg\nAElEQVRbvZy3kHPuCHAHcCl1ct5KxVZv522Zq7tzprZ47url34ba4vmp17ZY7bDUs+WUAN0DnGhm\nx5pZAngT8N2lCMTMWs2sPbwNvAr4ZT6et+YPeytw81LEl1culu8CbzKzpJkdC5wI3L2YgYWNc95V\nBOduUWMzMwO+CDzqnPt40a4lP2/lYquT87bWzLryt5uBVwKPUR/nrWRs9XDeVpC6aYdBbfF81cO/\nDbXFc46tLttitcOybLg6qMRQ7QW4nKAKy5PA+5cwjuMIqpY8CDwSxgKsAX4MPAH8CFi9SPF8g6BL\nOUswfvZtlWIB3p8/h48Dly1BbP8OPAw8RND4bVjs2ICLCIYGPAQ8kL9cXg/nrUJs9XDeTgd+kY/h\nl8AHZ/rbr4PYlvy8raRLvbTD+VjUFs8vtiX/t6G2eM6x1WVbrHZYl+VyMefqaTiyiIiIiIhI7Syn\nIXAiIiIiIiLzogRIREREREQahhIgERERERFpGEqARERERESkYSgBEhERERGRhqEESCYxM2dm/1B0\n/0/M7EML9Nz/ZmavX4jnmuF1/peZPWpmdxRte4GZPZC/DJjZ0/nbP5rlc/8gXHekwjEfMbOXzTX+\nKc+1y8weNrOHzOx2M1u3APH9tpkdtRDxicjCUzs843OrHRaReVECJFOlgdeZWfdSB1LMzGKzOPxt\nwO865woffs65h51zZzrnziRYg+Dd+fsXz+Z1nHOXOOeGZzjm/c65OyodM0u/5pw7nWD9hPfONz7g\ntwF98IrUL7XDFagdFpH5UgIkU+WAzwF/NHXH1F8OzWwkf/1SM/sfM7vZzJ4ys4+a2ZvN7O78r2bH\nFz3NxWZ2r5n9ysxenX981Mw+Zmb35H9h+72i5/2ZmX0X2FEint/IP/8vzexv89s+SLB43RfN7GPV\nvGEzu9jMfmJmtxIs1IaZ3WJm95nZI2b2O0XH7jKzLjM7If+6X8wfc5uZNeWP+ZqZvbbo+A+Z2S/y\n7+2k/PZ1Zvbj/GP/1cx2h6tnV/BT4IT84/9P0Xv/62rjM7M3AmcC38z/8prIn/sd+fj+tppzJiI1\npXYYtcMiUjtKgKSUfwHebGads3jMGcB1wCnANcBJzrnzgC8A7yw6bitwHnAF8Nn8h9XbgEHn3LnA\nucDvmtmx+ePPBv6vc+6k4hczs43A3wIvJ/ggOdfMXuuc+zBwL/Bm59y7ZxH/NuAdzrlT8vff6pw7\nJx/P9Wa2qsRjngf8k3PuNGAceG2Z597vnDuL4Fxcn9/2YeD2/GNvATZWCs7MDHg18LCZbQb+CngZ\ncBbwovBLzEzxOee+SbCa+Rvzv8KuIljZ/LT8r5t/UykOEVk0aofVDotIjSgBkmmcc0PAV4E/nMXD\n7nHO7XXOpYEngf/Kb3+Y4MM29C3nnO+cewJ4CjgZeBXwFjN7ALgLWAOcmD/+bufc0yVe71zgJ865\nfudcDrgBePEs4p2q1zn3XNH9PzKzB4FeYDNwfInH7HTOPZy/fR+T32exb5c45iLgRgDn3K1ApeES\nPyP4sGwm+LJxPvDfzrmDzrks8HVKv/dq4hsAfODzZnYVMFohDhFZJGqHAbXDIlIjsxnPK43ln4D7\ngS8XbcuRT5rNLAIkivali277Rfd9Jv+duSmv4wAD3umc+0HxDjN7KYv3QVB4HTO7mOCD7ALn3LiZ\n/RxoKvGY4vfsUf7fU7qKYyr5NefckaL4qn3cjPE557Jmtg14JfC/gLcTfBESkaWndljtsIjUgHqA\npCTn3ADwLYJhEaFngHPyt38diM/hqf+XmUXy49GPAx4HfgC83cziAGZ2kpm1zvA8dwMvMbNuM4sC\nvwH8zxziKaUTGMh/6J5G8CvnQvt/wBsAzOxyoGLFoCnuAl5mZmssmCz8Jmb33ofD17OgUlFH/tfP\nPyIYyiEidUDtsNphEakN9QBJJf8A/EHR/c8DN+eHJNzO3H4VfI7gQ7MDuM45lzKzLxAMC7g/P8a6\nn/LjuAFwzu01s/cCdxD8cvk959zNc4inlO8B15rZDoIvBnct0PMW+wvg62b2W8DPgQNUeT6dc7vM\n7M+BnxC891ucc9+bxWt/GfiCmY0TfIG6ycySBD+IXF/xkSKy2NQOqx0WkQVmzk3tCReRWstPOs45\n53JmdhHBJNltSx2XiEijUDss0rjUAySyNLYC38gPG0kDv7e04YiINJytqB0WaUjqARIRERERkYah\nIggiIiIiItIwlACJiIiIiEjDUAIkIiIiIiINQwmQiIiIiIg0DCVAIiIiIiLSMJQAiYiIiIhIw1AC\nJCIiIiIiDUMJkIiIiIiINAwlQCIiIiIi0jCUAImIiIiISMNQAiQNwcxeama7avTcW83MmVmsFs8v\nIrJcqe0VkXqkBEhklszsGTO7eIle+2Izu9/MRs1sl5m9YSniEBFZbEvV9prZG8ysx8zGzOwnJfaf\naWb35fffZ2ZnLnaMIjI7SoBElgkzOxX4OvB+oBM4A7hvSYMSEVn5BoB/Aj46dYeZJYCbga8Bq4Cv\nADfnt4tInVICJDWR/6Xu3Wb2UL634otmtt7MbjOzYTP7kZmtKjr+P8xsn5kNmtlPzey0/PaEmT1g\nZu/M34+a2f8zsw/O8PrNZvZvZnbYzHYA507Zv9HMtptZv5k9bWZ/WLTvQ2Z2k5l9Mx/r/WZ2Rn7f\nvwNHA7eY2YiZ/WnR077ZzJ4zs4Nm9v75nsMSPgD8q3PuNudczjl3yDn3ZA1eR0SWKbW9C9/2Oud+\n5Jz7FrCnxO6XAjHgn5xzaefcJwADXr7QcYjIwlECJLV0NfBK4CTgNcBtwJ8Bawn+9v6w6NjbgBOB\ndcD9wA0AzrkM8H+AD5vZKcB7gSjwkRle+y+A4/OXS4C3hjvMLALcAjwIbAJeAbzLzC4pevyVwH8A\nqwl6Xb5jZnHn3DXAc8BrnHNtzrm/K3rMRcDz8s/3wXy805jZe83sSLlLhfd0Qf7xD5vZXjP7mpmt\nnuE8iEjjUdtbwjza3kpOAx5yzrmibQ/mt4tInVICJLX0SefcfufcbuBnwF3OuV8451LAfwJnhQc6\n577knBt2zqWBDwFnmFlnft8vgb8CvgP8CXCNc86b4bXfAHzEOTfgnOsDPlG071xgrXPuw865jHPu\nKeDzwJuKjrnPOXeTcy4LfBxoIp+AVPCXzrlx59yDBB+AZ5Q6yDn3UedcV7lLheffDFxD8OXmRKAZ\n+OQMMYlI41HbW8I82t5K2oDBKduGgPY5Pp+ILAIlQFJL+4tuj5e43waFoRUfNbMnzWwIeCZ/THfR\n8V8BjgG+75x7oorX3gj0Fd1/tuj2McDGKb/8/RmwvuiYwmOdcz6wK/+clewruj1G/v0toHHgy865\nXznnRoC/Bi5f4NcQkeVPbe/iGQE6pmzrBIYXMQYRmSUlQFIP/jfBsIeLCT44tua3W9ExnwZuBS4x\ns4uqeM69wJai+0cX3e4Dnp7y61+7c644mSg8Nj9sYzMT47+LhzrMmpn9WX4Me8lLhYc+NOW15xWH\niDQ8tb3Vtb2VPAKcbmbF5+z0/HYRqVNKgKQetANp4BDQQtCzUWBm1wDnAL9JMHb9K2Y20y983wLe\nZ2arzGwz8M6ifXcDw2b2nvyE3aiZPd/MiifrnmNmr7NgfYl35eO7M79vP3DcXN4ogHPur/Nj2Ete\nKjz0y8BvmdlxZtZCMCb/1rnGISINT21vFW1vPs4mgmIHETNrMrN4fvdPAA/4QzNL5os6OOC/5xqn\niNSeEiCpB18lGCaxG9jBxIcdZnY0QfnRtzjnRpxzXwfuBf5xhuf8y/xzPg38F/Dv4Y78GPZXA2fm\n9x8EvkDwC2joZuCNwGGCeTevy49JB/gb4AP5IRx/Mpc3PBfOuS8RnKu7CN5bmsmTmUVEZkNtb3Wu\nIRg6+Bng1/K3Pw+FYhGvBd4CHCFIFl+b3y4idcomFy4RETP7EHCCc+7/LHUsIiKNQm2viCwW9QCJ\niIiIiEjDUAIky5YFC/uVmsz6Z0sdm4jISqW2V0SWOw2BExERERGRhqEeIBERERERaRixpQ5gtrq7\nu93WrVuXOgwRkVm77777Djrn1i51HAtBbbGILEcrqR2WuVt2CdDWrVu59957lzoMEZFZM7NnlzqG\nhaK2WESWo5XUDsvcaQiciIiIiIg0DCVAIiIiIiLSMJQAiYiIiIhIw6hZAmRmXzKzA2b2yzL7zcw+\nYWY7zewhMzu7VrGIiDQqtcUiIiKT1bIH6N+ASyvsvww4MX+5FvhMVc/68MMQicDWrXDDDfMMUURk\nxfs3GqUtvuGGIJ56iqseY4L6jUtEZBHUrAqcc+6nZra1wiFXAl91wUqsd5pZl5ltcM7trfjEmUxw\n/eyz8Lu/C74P11yzMEHP1Q03wPvfD889B0cfDR/5CLz5zUsbk4gIi9QW/87vwK5d8OpXL0zQc3Hr\nrfChD0EqNTmuffvgyivBLPiyX3xdi23FbrgBrr0WxsYmYrr22uD2Un5G1GtcUJ+fp/UYk4jMiwWf\neTV68uBD91bn3PNL7LsV+Khz7uf5+z8G3uOcq1hXdZvZ9AOSSWhthbY2aGkJrltbp1/C7W1tlW+H\n9+PxMNiJD7aZPuAgiOFzn1MDKSKTmNl9zrltS/C6W1mMtlgmJ0e5XOljIhFYvRqi0eBiNnE7Gg32\nT71fatts7hdv+8Y3YGRkelydnfDHfwyxWHCJxydul7q/0Jcbb6y/z9N6/oyv18SsXuPKW6p2WOrL\nslgHyMyuJRiawTmlDnjzm4PGfHQUxseDhurIEdizJ7idSgXbx8f5/9m78zC5rvpO+N9Te3VX791q\nqRfJsiV5wXiVbEu2IeBh5GRgnOQJT8C8mZAMMLwTk4DzZgIZh0k8hCVDMpAZgnEYApm8E/JmmYzh\ncWyMMdjVkmzJYGxso82Wu1u9qPfu2u563j9u3epbVbe27qqu6qrvR089VXXvqarTpe5z7++ec34H\nlQR8gQAQDmff2tqAUMi6D4eBJ57IbhgB6/nHPgbs3w8MDFjlfb7sA5nzgOZ8nhtgERE1iJJt8Re+\nsKX1yfLRjxbe99nPWm2/aa7f24+dN+f+3HsAMIzs525lnY+//nX3+pgmcOed6/UwTeu9nc/dthmG\nFVRJub7PWadi7+F8jVvwAwArK8AnP7nh/4KaSCSsUR6/8RvrgaWzx22zvXelyvzoR+u9nc46feAD\nwN/93Xpg6Lw5g0X7PhCwjveBgPs+t/fx+/P32a//1reA//AfrPMaoHF68Rq1d9ERlF0HvLl+FaFG\nUc8eoK8A+L6U8m/Sz08D+JlSwy7yrjru2QNcuGA9thv73JuuA5oGKIrV8K+sAKur1i0etwKkRGI9\nSLKDJjtwUpT1ffZ++zY3V/qL6Oy0rqx1d1s352PnravLuirY3Z3dOHq92Vfeygmk/tf/augrMESt\nqEF7gKrfFtfDZZdZJ1u56lmvQnXavRs4f956bB+Di92XU6aS+xtusIYs5hoeBp5+2jpe2sfNQo+d\n2+znzpt97HXe527L3f4//kfh7/IXf7F4sJn72C24BfKDV+c25739uhdeKFyn0dHiP2+hHsBas88T\n7Jv93LndeT7hfO4sm3uzt/v9+ecl9vaHH7bOrXL19ACf+Yx7UOfcZgd49r3zcSCQvd0ZwBaTE5Qd\nBHBKSl5tbnH17AF6BMB9QohvArgVwErJMee52tqsE3ubffJvD10rR7GgSVWtm6at32tadi/Sz/+8\nNcY8V3c38JGPWD1RS0vrt/l54OxZYHEx/6qSzedbD5J6etaDI/uWG0j19Fi9TLZ//mfriqdzLPwH\nP2gFfu99b36jxl4nolZW/ba4Hv7oj9yHKtWzXoXq9OlPW21vvXz2s+71+tzngMsvr93n5gZzuY+f\neMK6aJdrdBT4n/9z80FhqaDI7f6OO6zRJLmGhoDHH89/H/vznL1v9vmDW7DkPOfI3adp1uvdgsZi\nv9f33uv+GeXcUqn11zp7Hd3Okdy2FbK0BHz4w4X3b0ShQC13yOb589Z3SeRQsxZYCPE3AH4GQL8Q\nYhLAfwLgBwAp5UMAHgXwcwDOAUgA+LWy3jgQsH6Rq9WrsZGgyf5jN02rDvfdt94NDVjByMc/DvyL\nf5HdIAhhNY72Z2raek+UHSA5A6bFRev+9dfX99kNba5w2AqEenqAc+esXiunZBJ44AHr++vsBDo6\nrHv7iksgYM2lCoWse7ex3kS07Wybtniz7M9vpJ7vRqxTPetV6mr9pz/tHph95jPWfT388R+71+mP\n/xi4+urK369UEFfu/r/6K2BiIv/9R0aAL34x//XOIM2tp6ucnjS398jd/853ul8UHhgAHnpoPYiz\nA0JnYOf23BkMuvUaupXPvf30p5X/P1HTq+kQuFo4ePCgPHWqwabeFpvwZ1+9cf6hqqp1lUVR1u9N\nM/vAIGV217QdNDmDpUK373+//LqHw+sBkX2LRPK3dXYCfX1Af7/VkO3YsR5AOW+eEpnVG3xyJFEt\nNdPk24Zsi2n7a8RjRKPWqRGTMxSq10MPWSNQcoMqoPCwxUL7c+e5FdtumsDb3w5Mr3dqcwgcAQyA\nGofb1QxngGTfpMwPlHLH+77jHe5d9v391pW01dXseVD249xt8XjxOgeD+YFSV9d6T5QdMPX1Wbfn\nnrMOHM7eskZosIm2CAMgIqqaRgzMGrFenANELhgAbTdu3b12wgZFsXqXHnnECnTsOUCANbTtE58A\njh5d3+acQOgcmmBnwTEMK2nE2poVFOXeO2+5QVShLENufD7gTW+ygqHctOXl3Oxeq1AoP7NPsZtT\nozXY1JQYABER1YHjGH+9lOqPpQzWu0pUX5zYsd3YvTzBIn+7118P7N1r/bFPTlpjgj/5SeCXfqlw\nN7Izpaqz+7itzerFye1SdqZWBfIDCl23epCcAdNv/ZZ7fXXdCl7icStJhJ1tz87QV8l3Y6cnt+/D\nYSuYyk1lbt8iEWt88N//ff7CjtPTwHves56dJjfzXm4Gvmonk2BQRkREtHnve1/m+PmiEC/VuTbU\nANgDRJtTbOxu7u3qq90nbQ4PA888kz050jlBUlHW05Q7752pypPJ7Me56crjcetmPy/3994Oqkrd\n7EDL7o2y5011da3fd3VZ+4oFU3Yg1ajju2lT2ANERFRfzdQO08axB4g2p5wc/LbPfKZw6tW9e91f\nk9sz5by3M8M4AyZVtbY7U4zb2ffseynXE1GkUlYq80IB0Qc/uB48OW8zM9lBVW7WvUI8nvVeKLtn\nKjeIikSA//2/3RfY/ehHrUDKfk0oZL3Ovg+H8xcKdFvgbzMatWeqUetFREREDYUBEG2djaRetYf8\nVZKm3FYocHJm5lNVYNcu96QRO3cCv1YkI7AzqHLOxXL2PtnP4/HivVLz8+uP7bJu5ueBd72r+M8d\nDK6nNrfTm+c+t9Od24/tWzC4HlDZQZn9OBQCTpwA/tt/Ww/47OGCc3PAu9+dvWidc40pZ+BV6PFm\nNOrq4wBXICciImowHAJHVGi42Ve+YqXtzJ3/5Jw/5XzsXDzOHr5nP87db6+hkHvybwdV//pfA7Oz\n+XXt7QX+8A+twM1OeuF2s/c5yzi3ObML2tsUpborlwtReIG6YgvYOdegyn1sP89dSfxrX3Nffby7\n25r/5vVa9XFLLW/P4XJmUrT32XV1K2u/h/M1ua//9retNcHSAW0zZR9iW0xE2xGHwBHAHiCi0j1T\nXm/1PzM3mMoNqj71KeA3fzO7JygcBh580Bqyl7veQe57FNtmc+t5sYcPFgqe7F4VN/ff775iuds2\nTSu8gF0slv3cuQCe262Y5WWrXkRERERpDICIgKwMMVvC7o0o5AMfsAKeWsxpqSRxRe7twQetzIK5\nhoetQKPUQnW5+wptz/2u7Ho7n9vbDAP4xV907zEbGAC+/OX8DIeGUbg+bsFlqX2Fyv7Jn2z+/4uI\niIiqigEQUaOqVVC2mXk3n/1s4UQWw8PVqR9QfIVwt+2f/jRw3335PWaf/jRw112FX1toJXGgcJDm\n9lq38qZpDa+cmane90JERESbxgCIiMq3kUQWG+FMklCOX/91K4FDo2WB+/zn8wNGIiIiqisGQERU\nma0eLliuRqxXTsCoSakWfwERERHVmqfeFSAiamrvex9w4QJgmngR4ArkREREdcYAiIiIiIiIWgYD\nICIiIiIiahkMgIiIiIiIqGUwACIiIiIiopbBAIiIiIiIiFoGAyAiIiIiImoZDICIiIiIiKhlMAAi\nIiIiIqKWwQCIiIiIiIhaBgMgIiIiIiJqGQyAiIiIiIioZTAAIiIiIiKiluGrdwWIiDZDSglDGtBN\nHYZp3QOAR3jgER54PV4ICHg93sw2IiIial01DYCEEHcD+CIAL4CvSik/m7O/C8BfA9idrsvnpZR/\nWcs6EW0XpjQzNwGROXkXQtS7ajXnFtQY0oCiK1ANFaqhWo9NFZqpQUgBCQkhBKSUJd/fK7zweXyZ\nm9eT/dzeZn/nxW6Nju0wERFRtpoFQEIIL4AvAXgHgEkAJ4UQj0gpX3EU+w0Ar0gp3yWEGABwWgjx\n/0op1VrVi6hWpJQwpQkJCcM0ICGzghj7Zp/QO2/Ok3x7m5Md9Egp4RVe+L3+9ZN4rw9+jz/r5N3u\n+XCeqHuFt65BVKGgJhPMpO81U4Nqqq5Bjf1zeYUXXo8XYV8YEU+k4ro4/z80U4NiKJn/O3u7lBIC\nhb8nCatOmYAp/f9RKJhq97dv7IvbBLbDRERE+WrZA3QLgHNSytcAQAjxTQD3AHAeeCWADmGdjUUA\nLALQc9+IqFbsoKXULS9gcQQq9sm8KU0AKHnSLISjNyenZ8fv8SPoC0JAFAxSck/eU0Yqc8JuShOG\nNODJmd5nf6792ANPVtBkn7T7vf7M9qzgySWYsgMTZ1BjP9YMDSk9taGgJuQLod1T22Chmr035QRT\niq5gf9/+qnxehdgOExER5ahlADQMYMLxfBLArTll/juARwBMAegA8MtSps8iHYQQHwLwIQDYvXt3\nTSpL20+h3hW3wEUzNeuxsR7AaIYGQxpFe0MkJCCRFaTkBgJ+n39Le1WqcfJuBy7Ok3ZDGlkBIQQA\nmR/QSWsjPPDAhJkV1Fhfl6xLUFMv5fx/rCqrZQ3Nq4GqtcMA22IiImoO9U6CcBTACwDeDuAKAE8I\nIZ6RUq46C0kpHwbwMAAcPHiwLmcRVF1uwYp9Qp47TEwzNOgyf7iYPTcml32Cnhu42EPAnD0t22EO\nRy0IIeATm/vzl1K2xHykFlBWOwywLSYiouZQywDoIoBRx/OR9DanXwPwWWldGj0nhHgdwFUAnqth\nvWiL2MGLZmpQdRUJLYGElkBST0I11LyTZ+cV8txhYvYQrHr0uJA7fv/bAtthIiKiHLUMgE4C2C+E\n2AvrgPseAPfmlBkHcBeAZ4QQgwCuBPBaDetEVSSlzAwv0wwNiq4grsWR1JNIasnMRH4Ba4iU32vN\nLQn5QmgPNOdwKKIGw3aYiIgoR80CICmlLoS4D8DjsNKvfk1K+bIQ4sPp/Q8B+M8Avi6EeAnWoKXf\nlVLO16pOVDkppdWDY6iZie3OnpxMpqx0Z4DP40PAG0B7oL1lh5dRfXzr9Lfwpyf+FNNr09jVsQv3\n33Y/3nXlu+pdrbpiO0xERJSvpnOApJSPAng0Z9tDjsdTAP5lLetApZnSzAQ4mqkhqSUzQU7KSEFI\nK8CR0prcbvfkdAW7OAyKGsK3Tn8LDzz1AFJ6CgAwtTaFB556AAAYBLEdJto0e4g2j3lEzaHeSRBo\nizjn4yi6YvXgaMnMfBwn5/olPf6eOtV4azVi70Ej1qlW9dIMDXEtjpgaQ1yNI6al79UY4lp8/bEa\ndy13ZuEMDGlkvWdKT+GBpx7Aj2d/jMHIIHa278TOyE4MRgYx2D6IoC+4qToTUeWklJCQefeV7itU\n3plYp9g2CQnTNGHChGlaSQ9NmFnlM+XsrJiwhnR7hRc+rw8+4ctb88u5FlupG4MpovphANSEVENF\nSk8hpsSwoqwgrsYzJ4ecj5OvEXsPGrFOher1H7/3HzEdm8bBoYOuwUsmcHEEL7n7FEMp6/NDvhAi\ngQja/e2Z+52RnXh1/lXX8ik9hX989R8R1+J5+3pCPZnAaDAymPV4Z8QKliKByhdZJWp0ziybdvp7\nAEWDCecNQF5Q4Qwm7EDCGUDkpdd3rAUGYH0otUTWfdYw65xtUljLFGTeQ4jMGmp2hlA7yHCureYs\nI8T6EG4vvBCe7NfmrsmWFSBBQjEUJPVk1lps9s9Zak04Dzx5iydnFrj2WAtcZxIAebxZCYFyb0RU\nGVGntSk27ODBg/LUqVP1rkbDMEwDKT2FpJbEirKCFWUFmqEBsFJAh3wh+L1+NpBpqqFiem0aE6sT\nmFydxMTqBP76xb/OnNC7yTtYAnkHSOdj58HS7WBs/0sXcC2zmFzMnGg4eYQHveFe16ug9omAxPoJ\nTOa5o7zzNbllMvsLXGndiJAvtB6wBNoR8Vv37YH27O05gU3u9vZAO3we92s2b/vG2zC1NpW3fahj\nCE/96lOIqTHMxmYxE5vBbNy6n4nN4FL8UmbbYnIx7/Vt/jar16h9MKv3yA6QBtsH0RPuKfr39a3T\n38Lnj38es7FZyC9LVc7Ipuh6YlvceNwWbbYXI1YMBSk9BUVXstoWZyCSG1TkBhSF2ja3/fbfRKH9\nlB9MZQWVjrXaMseGQss+5Mg9TmW25wSFzvLllnX2hOWWz93m8XjggScT4HmEJ6uXrNCi4G5r7lWT\nEOJ5KeXBqr4pbTvsAdpGpJSZ3p01dQ3LyWXEtXimVyfgDWSukLcqKSXmEnOZ4GZydRKTK5PW/dok\nZmIzWQd/v8cPzdQKvt+/P/Tv3QMMZ3DgfO5SxhlEZN07ApXcMn/78t+61seUJu7aexeAAsFX7lVP\n5xXP3ADOEXyVKmO/15+f/POC39VX3/XV/ECmSNBSTfffdn9WzxRgBV7333Y/ADXvZtwAACAASURB\nVCASiCDSG8EVvVcUfA9FV3ApfikrQJqNz2I2Zt2OTx7HXHwub6id3+PPCowG2wczvUhnF87iL374\nF+s9XB4Eqv/TU7OzM246b5qhZYYw2wGObuqZv1e7TbF7D+zehc5gJwORBmGvT+eFt6rvW+jCtluw\nVEnZQuULlbXX91MMxXVIIpAdSGUW1Hb5LDt4ygRTjsAqN7jKDZ5ygysigAFQQzNMI5NSejm1jFVl\nNZNa2ufxIegLojvU3XIHs5gaswKa1UlMrExkBTsX1y7m9ebsaN+Bkc4RHBo6hJHOEYx2jlr3XaPY\n0b4Dd/3VXQV7D37r1t/aqh8ryzPjzxSs04Nve7AONbL800//qWC97txzZx1qZLGHBW5mblLQF8Ro\n1yhGu0YLljFMA/OJ+UyQNBubxUx8JhMkvXTpJXw39t2yh/RRa7Ov8ucGNik9lemtUU0VmqnlnyhC\nZIIar8fLzJuUUeicwHVI3jY5fcgdYqhLHdKQBQMrAHkXA+yAE9vmp6ZaYgDUIKSUmSEKq6lVLCvL\nSGpJqyGTQMAXQJu/DV5Pda8UbbVyJtBrhobp2HRecGM/Xk4tZ5WPBCIY7RzF5T2X46173oqRzhGM\ndI1gpHMEwx3DCPlCRetUqvegHhqxTkDj1guwgqBaz4/yeryZ+ULXDV7nWkZKieXUMmbjs7jnm/fU\ntD7UWOyTMHvokmEaWY8zgY2hZIIbyJwTU4GsK9phXxgRT+v26hMB6SGVApvuLVtOLpcuRC2BAVCd\n6KaeSTe9oqxgJbWSuXLh9/oR8AbQE26uDGxuE+g/8eQn8MRrTyASiGBidQIXVy9iOjadN0xtqGMI\no52juOaKazDaZfXgjHRYvTibTcddjd6DamvEOjVyvRqJEAI94R70hHsw1DHk2mNGjc0OWOxgJvex\naqiZBaA107rphtWDY1+0sofzZCbsQ2YN4Qn6gmj3tHYCGiKiemEShC0gpcxc+bODnZSeysyxCHqD\nCPqCTTl8YTm1jLMLZ3Fm4Qw+f/zzSGgJ13IDbQNWUJMemmYHNyOdIxhsH9z2PV/UmnKDfnwFkFOy\nKYZfNHpb7OyNcQYwdnCjGmomSYBzmQDd1NfnJ6SDGetJOvtYehhNZv5GTnYu2pxGTP/fiHVq5Ho1\nsuXkMm4dvfWHUsqb610Xqi/2ANWAPTk1rsaxklrBqrqamczn9/oR9AabrncnqSVxfuk8ziycyQQ8\nZxbP4FL8UsnXCghEfz26BbUk2lr2yUgmC5wp1RIvoU1aVVZxduGsa2+MM0mJnRzAOTk65AtxfZY6\neuT0I/j97/0+Usb6KIHf+97vYXxlHG/Z85aCWefcMtJlkrxUug/Z2e++c/47+Fz0c1l1euB7DyCh\nJXD3vrtdM4PmyktO4xjyuNHXf/vMt/HJ73+y4ZZKINou2ANUJVJKTKxMYCG5ANVQMw1o0BdE0Bts\nmgOqbup4Y+WNrEDn7MJZvLHyRmaiYcAbwL7efTjQewD7+/bjQN8BHOg7gPf+/XsxFSucqpioWa0q\nqxjuGMZQ51DTpF9t1Lb41blXoRgK2vxt9a4K5TBMA3OJOVxcu4ip1SlMrU1hKpa+X5vC+cXzG063\nT5ZIIIIH7nwAe7r3YE/XHvSGe5vm/KMa2ANENvYAVUlMjWEqNoXOYGdTLCwqpcRMbCbTk2MHOueX\nzkM1rIvYHuHB7q7duLL/SrzzwDtxoM8KeHZ37XZNfXz/4cadQE/bk9s6RhIyM0SJWktKT2FVWW26\nHvbtQtEVTMemMbU2ZQU5a1NWoJMOcmZiM5lMprbuUDeGO4axt3svzi2eK/jeX/5XX87+O5fZywhs\nZp/dS5hbXkLiwR8Uzrr5iTs+kfU8dy015zbnsglu5csta2/7s2f/zLVOMTWGjz/58czzSCCCPV17\nsLtrN/Z07cGe7vXH/W39bCepZTEAqpK5+BwC3sCWrHlSjkrGBi+nljMBzumF0zi7eBZnF85iTV3L\nlBlsH8SBvgM4PHoYV/Zdif29+3FF7xUlM6w5NfoEelOamUVka7kI23bhnEPhXJTPfpy1eGLOEKPM\nCA+XyeDOfbll7efOieNZK8bnlLX/b5z3AiKTXcs+WbATi/g9fh7wm9hScsk91S/l2cj8kVVlNdNb\nkxXkrE1hem0ac4m5rPIe4cGO9h0Y6hjCDTtvsHpBO4Yw1DGE4Y5h7OrYldVTV2xB47fvfXt1fvAK\nffWHXy1Yp/ff8P6tr1Da37/y96712hXZhb+85y8xvjKON1bewBvLb+CNlTfwytwr+M7572StY9bm\nb8sER5d1X5YVJA20DbCtpKbWGGfr25xu6phPzqMr2FXvqgBwz7b2wFMPQDVUHOg7gLOLZ7MCHudB\nqzPYiQN9BzI9Ogf6DmB/7350harzs21FquJy2IvKKoYCw7QOCB7hQSQQgYDITI7WTA0mzKyTKueJ\nP7B+gm/3OjgDp3oHUfaEbwmZmQjufJz7MzgDD+fCcn6PH2FvGH6PHz6PDwFvIDP5O3fsvPO7Kfa8\nnLKl3qcQ+//XXiQyrsUR1+JYVVazMwx6/fB7/PB7/S0b6DYLu9e6GXrga63QMWJFWcG1O67NCnKc\ngU5MjWW9T8AbwFBkCEOdQ3jrZW/NBDZDHUPY1bELO9t3wu/1l12vRkyz34h1AgrX67cP/zb29uzF\n3p69ea/RDA1Ta1NZgdH4yjhOz5/Gk68/mdU7F/aFswOj9JC6PV17sKN9R9E2mMkZaDvgHKAqWEgs\n4NziuYYZdlHoKppT0BvEvt592N+7PzN07cq+K0s2bNuR82TYbuCFEOgIdKAr2IX2QDtCvhAC3oDr\nz+5cYM2ZScp50w09k0Gq0A1Y7y2x6+D8DGfwZK9cbQcZ9pCM3M+2X5fbY2K/p8/jg8/rywQuzgDG\nXmvEngjuFd68ieHNRkoJzdQyvw9xNY6YGkNcja9fGRWAT/gywVE1MhByDlDtrSlreGXulYZphxtJ\nSk9hLj6HucQc5uJz+P2nfh8rykrJ13UEOrJ6bIY6rGBnuGMYuyK70NfWV/V2ohFPnhuxTtWul27q\nmF6bxoXlCxhfGceFlQsYX7Z6kSZXJ6GZWqZsyBfK6i1y9iI9e/FZfPKpT+YFZp9626fq+p05vyv5\nZanKGRmsW2WoITAAqoKXZl8CYK0kX0+aoeHZi8/i3z7ybwuW+bO7/wwH+g5gd9fupkwtbZ/gKrqS\n6fmAsA7kncFORAIRhHyhuiSmyARQZn4AZafqzQ2a7KDKA6s3xu9dD2T8XuvkPDdwsZ8zm1VlNGM9\nMEpoCSsw0uLrwyIh4PF4Mt99JcNdGQDV3muLr2FZWUYk0BqLhkopEVNjuJS4ZAU3jgAnsy393Dmc\nuZSH/tVDmaCnI9hRw5+Atgvd1DEdm8b4cnZg9MbKG5hYmcgKjgqJBCL4wE0fQMATQNAXRMAbQMC7\n/jjozb4P+AKu2zdy3tLMyxHQxnEI3CbZi5nW66qjaqg4Pnkcj597HE++9iSWleWsXganoY4hHN13\ntA61rB37hFUztExPSLu/HTvad6wHOw2yxpIdlDTKPDHK5vdagU072rP+nu3hkKqhIqEmrOF0ahxr\nhnVSaQ/Vcw6no62lmzrmEnPoDnXXuyp5Kr1Kb0oTy6llXIpnBzFziTlrm+O58yq7LegNYqB9AANt\nA9jfux9HRo5goH0A/W39GGgfwI62Hfh33/53mI3P5r12qGMIb9v7tqr+/LT9+Tw+jHaOYrRzFLfj\n9qx9hmlgJjaTCYj+4Pt/4PoeMTWGL5z4QlXqYgdPWQGSL4iAxz1w+s7577j+rVBr45nYJi0kF7b8\n5Fo1VIyNj+Gxc4/hydefxJq6hnZ/O96+9+24e9/dWFFW8OAPHmy4McubpRkaFEOxrjal47uwP4y+\ncB86g50I+oKZdTyIqsWeB9Xmb8s6wbYX01QNFUk9mRlKF1Nj63PGBOD3+PMyX1F1raSs4VyN1uPp\nNtfm9773e/jJ3E+wt3tvVmAzn5jHXGIO84l519+XSCCCgbYBDLQP4PrB6zNBzkD7QCawGWgfQEeg\no+T38DtHfqch57XQ9uP1eDHcOYzhzmEcGT2Ch59/2D1pRGQI3/mV70AxlEy7ac/RtOfj2veaYY3i\ncG6zy9rnAbnbne+5qqxmbUvo7guwU2tjALQJpjQxE5vZkiEXKT2F6HgUj517DE9deAoxNYaOQAfu\n2nsX7t53N27ffTsC3kCmfMATaMgxy+XSTT3TgAHWiU3QF0RPqAddoa7MMLZmHMZH24PX40XYE0bY\nH0YX1pOEmNLMHHhTWgoxNYaYFsv6+6TqmonNIOwP17saWebic/jUM5/Ku/KsGiq+/sLXM897Qj2Z\nAGZf776soGagbQA72ndgoG2gqj9fo2fkpHW5cz4zj9PzQu3H9Rranatg0ojD92d62bdaOfOiqfUw\nANqENWUNuqHDG6zNSXhSS+LpN57G4+cfx1MXnkJCS6A72I2jVxzF0X1HcXjkcMGTqkbJtlaOzBAj\nXc0M3Qt6g+gKdqEz2ImwP4ygL8ihY7QteIQHIV8IIV8IncFO7MCOelepqaV0K8isd/KDmBrDyYsn\ncWzyGI5PHMfZxbMFywoIfP/930dfuK9uQyYb8RjhvPDlTOYCwEqBbz9MDzt13gPrafBztzu31Xpu\nZDkBiylNWCsCZNfDbU623QPtFd5MUhav8MLr8Wb2AcBichHLqeXMa0K+UF1+txoxuHYLyoh4RrkJ\ns7FZhPzlr4NTjrgaxw/e+AEeO/cYnn7jaST1JHpCPXjn/nfi6L6juHX41m09x8AeNqQYSmadloA3\ngI5gB7o6uhD2hxHyhRjsEFFZFhOLdRn2qhoqXpx9EccnjuPY5DG8OPsidFNH0BvEwaGDuOeqe/CN\nF76RtzYOACtFdGTnlte5kUgpoRgKFF3JBAcBbwCdwU50BbsQ9AWzFyd13GcllEF+YhkJCdM0YcKE\naZrQoUMa2Vk07YAqs36ZHWAJx1IHLvtyf4Zc5QQsdgZO563QEgrlGowMQjd1JLQElpPLWEwuWmnL\nhXVBMewLb1nvUKMF17lBmTSlWucqUQPgWeYGqYaKpdRSVSbdxtQYnnr9KTx+/nE8/cbTUAwF/W39\n+Pmrfh5H9x3FoaFD2zIgsNNPp/RU5kDh9XjRGezEzshOtPnb6naVioi2v61c+0dKidMLp3Fi8gSO\nTRzDyamTSGgJeIQH1+64Fh+48QM4PHoYN+68MZMRdGf7Ts61ScvM4XQsNt0R6MBA50DWUgRbxS2w\nyh1W5ravVOBSTz6PD53BTnQGO7G7ezdSegpxNY7F5CKWUkswTdMauusPt9yQXDsoW04u49Y/uPUn\n9a4P1d/2O6tuEEvJJatLfYNXVFaVVXzv9e/h8fOPIzoehWqoGGgbwLuveTeO7juKm3fdvO3mt9iT\nEXPX2hnuGM4c4OqdKpyImseaugbN1BDx1GYe5tTaFI5NWEPajk8ex0JyAQCwt3svfuGqX8DhkcO4\nZfiWggtFN+JwoK1gShOKrmQHfv4QesO96Ap2ZYaI1nO+Smah5cbKm1FV9vfc19YHU5pIaAmsplYx\nl5zDUnIJgLWYbdgfrnvwRrTVGABtgJQS02vTaPO3VfS65dQynnztSTx+/nEcmzgGzdSwM7IT77n2\nPbh73924ceeN26YRcsvIZqef7gh2NMyETCJqXnPxuar2IC+nlvHs5LM4NnkMJyZO4MLKBQDAQNsA\njowewZHRIzg8chi7OnaV/Z6NNhyo2pxrr9kXv7weL7qCXdgZ2Zm5+LUdRzE0E4/wIBKIIBKIYKhz\nKLMQ9FJyCYupRRimAQGRmXNL1OzYIm1AXIsjZaTQG+gtWXYxuYgnX3sSj51/DCcmT0A3dQx3DONX\nrvsVHN13FNcNXtfwQY9hGla6SUeSgpAvxIxsRFQ3mqFhIbmArqB770s5UnoKP5z+odXLM3kcL196\nGRISbf423Dp8K+697l4cGTmCfb37eDEnLStRQXrNuXZ/OwYjg4gEIgj7rOFV/L4aW8AbQCAcQE+4\nB3vlXiT1JNaUNcwn5jO9Q36vn8ErNS3+Vm/AfGI+a/xs7kJ3H7zpg/AIDx479xieu/gcDGlgtHMU\n77/h/Th6xVG8ecebG/bgYKfwtSemAuvjipmkgIgaxUpqZX0Se5kM08DLcy9n5vE8P/08VEOFz+PD\nDTtvwH233Icjo0fw5h1v5txElE5UYB8PePFrexNCoM3fhjZ/WyaZQlyNYzm1jIXkQmbeViMMXSSq\nlpqexQoh7gbwRQBeAF+VUn7WpczPAPgCAD+AeSnlW2tZp83STR2zsdnMmG+3he7+8Ad/CAC4rOsy\nfPCmD+LovqO4uv/qhms0nEMXDGkAsNJydgY7saN9RyZJQatNliRqJs3YDgPAdGw6b22c3ItRH7vt\nY7hu8LrMPJ4TF09gVVkFAFzZdyXuvfZeHBk9goNDB7ckkUKjs4c260Z6IVaBuiYqoPrweXzoCnWh\nK9SF3V3ryRQWkguZVNtejxdhX5gXCmjbqlkAJITwAvgSgHcAmARwUgjxiJTyFUeZbgB/DuBuKeW4\nEKLhF8xYTa1msr4A1uRWt9zyA20DeOz/eqyhgh57ccbMvB0BRPyRzNAFztshai7N2g4ntAQSWiJr\n7R+3i1G/88TvZPYPdQzhHZe/A0dGj+C2kdvQ39a/5fVuBIZpwJBG5t6+ug9YiQr6wn3oDHbyaj8B\nsHqHwn5rwef+9n4YpoGElsCKsoKFxAJiagwCAgFfACFfqOGH9BPZatkDdAuAc1LK1wBACPFNAPcA\neMVR5l4A/yilHAcAKeWlGtanKqZj0wj51tf+mV6bdi03n5iv+4HD7sY2pQkhBIK+IHrDvVkHNzZW\nRE2tKdvhxWT+2j+FLkZ1Bbvwd+/+O+zu2l33NrlWTGlCN/Wse8M01n/edKIaKSQCngD8Hj/a/G0I\neANo87dxaDOVzevxoiPYgY5gB0Y6R6DoCuJaOtV2cilzvmH/fhE1qlq2dsMAJhzPJwHcmlPmAAC/\nEOL7ADoAfFFK+Ve5bySE+BCADwHA7t27a1LZcritOL6rYxem1qbyylaSJaja7OwuXo8Xwx3D6ApZ\ni8rx4EbUcqrWDgON0Rab0sRMbAaRQHbq60IXo1aVVezp3rMVVasqKWVeUKObupW6GdYi0kIISCnh\nFd7MFXi/x58ZqpZZlDO9CKdXeJs2CKT6CPqCmYurpjSR1JJYVVYxn5jHcmoZUsq8QNxebNb+Xc6k\nIpfrv9e55Ys9l0Jm77eLpd/H/hwhhLUILhHqnwTBB+BmAHcBCAM4LoQ4IaU84ywkpXwYwMMAcPDg\nwfyll7eI24rj9992P373u7+bmUMD1G+hu7gah2IoaPO3YV/vPnSHujk5lYhKKasdBhqjLV5T1mCY\nRl7b1ogXo3JJKWFIIy+okUifDGbO46wFNwO+AIKeINr97Qj6ggh4A/B7/Zlgxg5u2JNPjcAjPGgP\ntKM90I5dHbtgmEbWQrIAMouiA8jaJh3Ri13Gfq1b+XKe28GO8z4dFNXtPJIaRy0DoIsARh3PR9Lb\nnCYBLEgp4wDiQoinAVwPIO/AW2+mNDEdm86bKPvOA+/Egz94EKppZU7b6oXuTGliTVmDKU30hHqw\nr3cfIoEIr/IREdBk7TAAXEpcch1a85FbPoJPPPmJrG31uhjlpBka4mo8c2U74Akg4AsgEogg6A1m\nhp7l9tTw4hVtd/wdpkZWywDoJID9Qoi9sA6474E11tzp/wD470IIH4AArKEZ/7WGddqwmBqDZmro\n8HRkbT+/dB6r6io+9bZP4d1veveW1UczNMQ0a/Lhzvad2BHZkTU3iYgITdYOq4aKxcQiukPdefsG\n2wcBAL3hXiwll7b8YlQuRVeQ0BIIeAO4rOcy9IR64PP4eHGKiKgB1CwAklLqQoj7ADwOK/3q16SU\nLwshPpze/5CU8lUhxGMAXgRgwkrR+pNa1WkzZuOzCHrzV0eOjkcBAEdGj2xJPVJ6Cgk1gaAviL3d\ne9Eb7uXcHiJy1Wzt8EpqBRDua/9EJ6Lwe/x48t88iTZ/Wx1qZ0loCaS0FNoCbdjfux/d4W4OUSMi\najA1PXOWUj4K4NGcbQ/lPP8vAP5LLeuxWcWuOo5NjGFv914Mdw7X7POllIhrcaiGis5AJ64auAqd\nwU4eVImopGZphwFgKjaFdr/7ej3R8ShuHrq5LsGPs43uCnbh8p7L0RHoYG8PEVGDKjsAEkLcAWC/\nlPIvhRADACJSytdrV7XGUeiqo2qoeO7ic3j3NbUZ+maYBtaUNUhI9Lf1Y2dkJxfrI2phrdwO2z0r\nziyctkvxSzizcAa/ffi3t7ROzjmYfW192BXZxTaaiGgbKCsAEkL8JwAHAVwJ4C9hrRb+1wBur13V\nGkehq47PTz+PlJ7C7aPV/RrsNNYe4cFI5wj62voQ9OUPvyOi1tHq7fB8fL7gpOqx8TEAwJ2779yS\nuhimgVXFWhR7sH0Qg5FBzsEkItpGyu0B+gUANwL4IQBIKaeEEB3FX9Ic4mocSTWJ3rbevH1j42Pw\ne/y4ZfiWqnyWfYUzHAgzjTUR5WrZdtgwDczGZ/PW/rFFJ6LoC/fhyv4ra1oPzdAQU2PwCA9GO0fR\n397PxR6JiLahcgMgVUophbBWmxJCtEwf/0JiAX6v33Xf2MQYbtx546aGPJjSREyNwTANdIe6OXac\niApp2XY4psZgStP1gpApTYyNj+HOPXfWbF6kM6Pb5T2Xozfcy4tTRETbWLkB0P8nhPgKgG4hxAcB\n/DqAv6hdtRqDYRqYic+gM9iZt28hsYBX5l7Bx2772IbeWzd1rKlrELCGUOxo34GwP7zZKhNR82rJ\ndhgAZmOzBYcBvzr3KpZSS7hj9x1V/1xnRrcDfQfQFepi8hkioiZQVgAkpfy8EOIdAFZhjT//pJTy\niZrWrAGsKquQUroe8I5NHAOAiuf/pPQUkloSfo8fl3Vdht5wb8EeJiIiW6u2w6qhYim15Jr8AFhf\niqBaczGdGd3YK09E1JxKBkBCCC+A70op3wag6Q+2TjOxmYITW8cmxtAd6sY1A9eUfJ/MAVVXEQlG\ncGX/lUxjTURla+V2eCm5VDT4iE5EcXX/1ehv69/U59gZ3QzTwED7ALNuEhE1sZIBkJTSEEKYQogu\nKeXKVlSqEaT0FFaVVderjlJKRMejODJ6pOg4cMM0MmPX+9v6Mdg3WHASLxFRIa3aDkspMRObKbi2\nT1yN40fTP8Kv3vCrG/4MZ0a3ne07sSOygxndiIiaXLlzgGIAXhJCPAEgbm+UUv5mTWrVAJaSSxBw\nv+p4dvEs5hJzBYdc2GmshRAY6hjCQNsA01gT0Wa1XDuc0BJIau5ZOAHguYvPQTO1Dc3/cS43sLtr\nN/rb+jkcmYioRZQbAP1j+tYSpJSYXptGJOjeW2OvOZEbACW1JFJ6CkFvEFf0XIGecA8zBRFRtbRU\nOwwA84n5okFJdDyKsC+Mm3fdXPZ7KrqCuBpH0Md2moioVZWbBOEbQogAgAPpTaellFrtqlVfMTUG\nzdQQ8RRec+KKniuwq2NXZptu6lANFVcPXM0Js0RUda3WDttr/7hl4bRFx6O4ZfiWstbiSWgJpPQU\n2v3tuGrgKs7DJCJqYWUFQEKInwHwDQAXAAgAo0KIX5VSPl27qtXPbHy24FVHRVdw8uJJ/PK1v5y3\nvTfcW/RgTUS0Ua3WDhfLwgkAE6sTuLByAfded2/B93AmoOkOd+OKnisQCUR4gYqIqMWVOwTuTwD8\nSynlaQAQQhwA8DcAyh93sE1ohobF5CK6gl2u+5+ffh6KoeCO0ewx56qhFnwNEVEVtEw7DFgXoool\nI7CHIrvN/zGliZgSgyEN9Lf1Y1f/roKJFIiIqPWUGwD57YMuAEgpzwghmnK26EpqBZAoeIUwOh6F\n3+PHoeFDefu4kCkR1VDLtMOKrmAltVJw7R/ACoB2RXbh8u7Ls7YntSQUQ8GuyC7saN/BBDRERJSn\n3ADolBDiqwD+Ov38fQBO1aZK9TUVm0JboPCVwrGJMdy06ybXq4lMnUpENdQy7fBSqnAWTsCac3l8\n8jju3nd33sUqxVCwv3d/0eCJiIhaW7kzQP9vAK8A+M307ZX0tqaS0BJIaImCE2rn4nP46fxP84Zc\nqIaKsD/MTEJEVEst0Q5LKTGzNlN0EdIXZ1/EmrqG23e7L0XA3ngiIiqm3B4gH4AvSin/FMisSt50\n4wrm4/PweQp/JccmjwHIT3+tGir6w5tbhZyIqISWaIfjWhyKoRTtiY+OR+ERHhwZOZK13ZQmPMKD\noLfpvhYiIqqicnuAngTgvKQWBvDd6lenfuyUq+3+wlcdx8bH0BPqwdUDV2dt1w0dHcGOWleRiFpb\n07fDgNXTXmpB0rHxMVy34zp0hbITzyi6gs5gJ7O8ERFRUeUGQCEpZcx+kn7cVCl11tQ1mNIsOIxN\nSomxiTHcPnq7a1pWzv8hohpr+nZYN3Vcil8qmrFtJbWCFy+96Dr8TTEUZuMkIqKSyg2A4kKIm+wn\nQoiDAJK1qVJ9zMaKp1w9vXAa84n5vIOulBIQDICIqOaavh1eTa0CQNEFSo9PHocpTdf015Bgumsi\nIiqp3DlAHwXwd0KIqfTzXQB+uUj5bUXRFSwll9Db1luwjL3mhNv8n45AB4dcEFGtNXU7DAAzsZmS\nCQyi41F0BDpw3eB1efskJBMgEBFRSUV7gIQQh4QQO6WUJwFcBeBvAWgAHgPw+hbUb0ssp5aLXnEE\nrPTX+3v3YzAymLWdQy6IqJZapR1O6SmsqWtFe9PtociHRw7nJazRDA0hX6hoIhsiIiKg9BC4rwBQ\n048PA/g9AF8CsATg4RrWa8tIKTG1NlU05WpKT+Hk1Mm83h/ASp5Q7LVERJvU9O0wACwmFkteiHpt\n+TVMrU0VnP/TGeysVfWIiKiJlLpU5pVSLqYf/zKAh6WU/wDgH4QQL9S24nTTcQAAHv5JREFUalsj\npsagGmrRIObU1Cmohso1J4ioHpq+HZZSYiZWfO0fwBr+BsC1LdZMLS8rHBERkZtSPUBeIYQdJN0F\n4HuOfU0xzqCslKsTY/B7/Dg0dChru2Ea8Hl8BRdOJSKqgqZvh+0LUaWGr42Nj+Gyrssw2jmav1My\nGQ0REZWnVAD0NwB+IIT4P7CyDT0DAEKIfQBWSr25EOJuIcRpIcQ5IcTHi5Q7JITQhRC/VEHdN003\ndcwn54uu/QNYB92bh27O6+nhkAsi2gJN3Q4DwKX4JQR8xS8kqYaK5y4+55r9jdk4iYioEkUDICnl\nHwH4bQBfB3CHlFI6XveRYq9Nr1L+JQA/C+AaAO8VQlxToNznAHyn0spv1kpqBVLKohncLsUv4fTC\nadf5P6quojvUXcsqElGLa/Z2WDd1LCQXSl6Ien76eST1pOvwNzsbZ6k5REREREAZwyeklCdctp0p\n471vAXBOSvkaAAghvgngHgCv5JT7CIB/AHAIW2w6Nl1yzYhjE8cAwP2qI1OuEtEWaOZ2eCW1AkiU\nXEogOh6F3+PHrcO35u1TDAVD4aFaVZGIiJpMLS+XDQOYcDyfTG/LEEIMA/gFAF+uYT1cJbUkYkoM\nQV+waLmxiTH0hntxVf9Vrvs55IKIGlhDt8OAdSGqnAtJY+NjuHHnja6JEkzTZDZOIiIqW73HC3wB\nwO9KKc1ihYQQHxJCnBJCnJqbm6vKBy8kF+D1eIuWMaWJYxPHcGT0SN7QCs3QEPJzzQki2vbKaoeB\n6rfF5V6Imk/M49X5V1174gH2xhMRUWVqefZ+EYAzVc9IepvTQQDfTA996Afwc0IIXUr5T85CUsqH\nkV7v4uDBgxKbZEoTM7EZRAKRouXOLJzBfGIed4zmH3QVQ0FvuHezVSEiqqWqtcNA9dvici5EAVZP\nPOA+FJnZOImIqFK1DIBOAtgvhNgL64D7HgD3OgtIKffaj4UQXwfwbbeDbrWtKWswTKPkgddec+LI\n6JG8fZqpMQMcETW6hm2H7QtR5Qxdi45H0RvuxdUDV+ftUwyF6/8QEVFFahYASSl1IcR9AB4H4AXw\nNSnly0KID6f3P1Srzy5lNjZbcsgFYF11PNB7AIORwbx9AoLzf4iooTVyOxxTY9BNveQw4mJDkQEr\nG2dXBwMgIiIqX00nsEgpHwXwaM421wOulPL9tayLTTVULKWWSqavTmpJnJo6hfe9+X15+6SUkJAM\ngIio4TViOwwAs/FZBL2lL0Sdnj9dcCgyYM3/KZXNk4iIyKneSRC23FJyCUKIkilXT02dgmqoruv/\naKaGiD/CNSeIiDZAMzQsJhbLClyiE9ZQZLf1f2y8GEVERJVoqTN4KSWm10qv/QNYw98C3gAODh3M\n26foHHNORLRRK6kVQJRe+wew5v8c6DuAHe078vaphoqwP1xWIgUiIiJbSwVAcS0OxVDKyhY0Nj6G\ng0MHXVOr6qaOjkBHLapIRNT0pmJTaPeXTn6Q0BJ4fur5gumvVUNFd7D4cGYiIqJcLRUAzSfmy1q3\nZzY2izOLZ1yHv9k45IKIqHIJLYGElijrQtTJiyehmVrB+T+aoaEjyItRRERUmZYJgHRTx6X4pbJS\nrh6bOAbAfc0JU5rwCA/XnCAi2oCFxELZC0g/M/4Mgt6g61BkGxdAJSKiSrVMALSaWoWUsqzEBdGJ\nKPrb+nGg70DePnv+Tzlj14mIaJ0pTczGZ8sa/gZYczEPDR9yXbbAlCaEEGVlkiMiInJqmQBoJjZT\n1pXCUmtOKIaCriATIBARVWpNWYNu6GUlLZham8JrS6/hzt13uu5XDRVdQV6MIiKiyrVEAJTSU1hT\n18qat/PT+Z9iMblYcP6PlFxzgohoI2Zjswj5y5s/OTY+BgAF22Jm4yQioo1qiQBoMbFY9po90XFr\nzYkjo0cKlmECBCKiytiLUId95c3ZeWb8GQy2D2Jf7z7X/aY0yx5KR0RE5NT0AZApTczEZspKfgBY\nY86v7LvSdc0J3dQR9Abh9/qrXU0ioqa2nFouaxFqADBMA8cnj+OO3XcULc+LUUREtBFNHwDF1BhU\nUy0r65C95kShFccVXUFnsLPaVSQiampSSkzHyluEGgBeuvQSVpXVguv/8GIUERFtRtMHQLPx2bKz\nBJ2cKr7mhGqoDICIiCqU0BJIqsmylw+IjkchIHB45LDrfl6MIiKizWjqAEg1VCwmFsu+6jg2Poag\nN4ibh2523S8guOYEEVGFFhILFfXWRMejuHbHtegJ97juV00V3eHualWPiIhaTFMHQCupFUCg7DSp\nYxNjODR0qOC4cgnJMedERBUwTAMz8fLnYa4qq3hx9sWCw98AAJLzf4iIaOOaOgCaik2VnSVoJjaD\nc4vnCs7/UQ0Vbf62stavICIiy5q6BlOaZWfiPDF5AoY0CgZAUkoIIRgAERHRhjVtABRX40hpqbLH\nnJdac8JedI+IiMo3E5spO/U1YA1/a/e34/rB6133a6aGdn972QEVERFRrqY9giwkFsrK/GYbmxjD\nQNsADvQdcN2vGRo6gh3Vqh4RUdNTdAXLyeWy505KKREdj+K2kdsKzhlSdAXdIc7/ISKijWvKAKjS\nMeemNHFs4hiOjB4pOl+ICRCIiMpnr/1TrgvLF3Bx7WLR+T+6qSMSiFSjekRE1KKaMgBaVVYhpSx7\niMSrc69iKbVUcP6PPea83HTaREStTkqJ6bXpioKVsQlrKHKxAEiA83+IiGhzmjIAmonNVHSAtA+6\nR0aOuO5XDAWdgc6KrmQSEbWyuBZHykhVnP56d9du7O7a7brfMA14Pd6y53YSERG5aboAKKWnsKqs\nVjRcLToexVX9V2GgfcB1v6Ir6AoxAQIRUbnm4nMVBSqqoeLZi88WTERjl+kM8mIUERFtTtMFQEvJ\nJQiUf3BMaAn8cPqHRQ+6pjTLnk9ERNTqDNPAXGKu7EWoAeBH0z9CQksUHf7GBAhERFQNTRUASSkx\nEys/+QEAnLx4EpqpFV90D1x0j4ioXJXOwwSsocg+jw+3jdxWsIyEZDIaIiLatKYKgGJqDIquVDbm\nfCKKoDeIm3fd7LrfMA34vX6OOSciKtP02nTFgUp0PIobdt5QNGkCF0AlIqJqaKoA6FL8EgK+ygKV\nsfExHBo+hKDPPcObYijoDHZWo3pERE0vpaewqq5WFKgsJhfx8tzLRXviNUND0BesaH03IiIiN00T\nAGmGhoXkAtr95Q9/m16bxvml87hjtPBBV9VVdAWZAIGIqBxLySV4hbei14yNW5k4i83FVAwF3UHO\n/yEios2raQAkhLhbCHFaCHFOCPFxl/3vE0K8KIR4SQhxTAhx/UY/ayW1AkhUlB3ITn9daP0fq5Jc\nAJWItq+tbIc3Mg8TsNri7lA33jTwpoJldENnbzwREVVFzQIgIYQXwJcA/CyAawC8VwhxTU6x1wG8\nVUr5ZgD/GcDDG/286dg02gLlZxwCrIPuQNsA9vfuL1qOY86JaDva6nbYnodZyTA1KSWi41EcGT0C\nr6d4zxHbYiIiqoZa9gDdAuCclPI1KaUK4JsA7nEWkFIek1IupZ+eADCykQ9KaAnEtXhFiQoM08Cx\n8WO4Y/cdBXuNOOaciLa5LWuHgfTaPxXOwzyzcAZzibmiQ5GllIBgAERERNVRywBoGMCE4/lkelsh\n/xbAP7vtEEJ8SAhxSghxam5uLm//QmKh4iDllblXsKwsc8w5ETWzqrXDQPG2WDd1zCfnK5qHCVjZ\n34DiQ5EVQ0FngAugEhFRdTREEgQhxNtgHXh/122/lPJhKeVBKeXBgYGBrH2GaVhjzis86Nrzf46M\nHilYRjM0jjknopZQqh0GirfFK6kVSCkrDlKiE1Hs792PnZGdBcsouoKuEJPREBFRddQyALoIYNTx\nfCS9LYsQ4joAXwVwj5RyodIPWVPXYEqz5NjxXGPjY7hm4Br0tfUVLcchF0S0jW1JOwwAM7EZtPkr\nm4eZ1JI4NXWqaE88YC2AWmliBSIiokJqGQCdBLBfCLFXCBEA8B4AjzgLCCF2A/hHAL8ipTyzkQ+Z\njc0WXMOnkJgaw49mflT0oGuPOa/0vYmIGsiWtMNJLYk1Za3i9vLk1Emohlp0/R8bL0YREVG11Gx2\nv5RSF0LcB+BxAF4AX5NSviyE+HB6/0MAPgmgD8Cfp4dN6FLKg+V+hqIrWE4toyfcU1HdTl48Cc3U\nio45Vw0VHYEOeERDjBIkIqrYVrTDQHrtnwp74QGrJz7gDeDgUOGP000dfo+/oiQ3RERExdQ0vZmU\n8lEAj+Zse8jx+AMAPrDR919OLW/odWMTYwj5Qrh5180FyyiGgl3hXRutGhFRQ6h1O2xKE9Ox6Q0N\nUYtORHFo6FDRtdYUXeFcTCIiqqpt270hpcT02jQigUjFr42OR3HL8C1FrygaprGh9yYiaiUxNQbN\n1CrOxDkTm8G5xXMlh7+phoquIBMgEBFR9WzbACiuxZEyUvB7/RW97uLqRby+/HrJSbcCgmPOiYhK\nuBS/hKC38rmSmfTXZbTFxXqIiIiIKrVtA6BLsUsbGhN+bOIYABS96mhnleOYcyKiwjRDw2JyseLs\nb4A1FHmgbQAH+g4ULScheTGKiIiqalsGQLqpYy4xt6GDbnQiisH2QVzRc0XBMvaYcy66R0RU2EbX\n/jFMA8fGj+GO3XcUfa1qqGj3t28owQIREVEh2zIAWkmtAEDFGdoM08DxieO4ffftRQ+6isFJt0RE\npWw0+cHLcy9jWVkuOf9H0RV0h7s3Wj0iIiJX2zIAmo5Nb6j35+W5l7GirOCO0RJrTkhs6P2JiFqF\nKU3EtfiGhgpHx6MQEDgyeqRoOUMyGQ0REVXftgyA4lp8QwuU2gfdw6OHi5bjmHMiouIkJDwbPISM\nTYzhmoFr0BvuLVmWbTEREVXbtgyANqqcg65u6gh6gxVnlyMiotJiagwvzLxQcvibKU14hGdDGeaI\niIiKaZkAqNyDbkpPcf4PEVGNnJg8Ad3Uy5r/w2Q0RERUCy0TAD178Vnopl5yzLlmapx0S0RUI9Hx\nKNr8bbhh5w1FyymGwgVQiYioJlomADo2fgxhXxg37bqpeEHJMedERLUSHY/i1uFbSydPYDIaIiKq\nkZYJgKITUdwyfEvRg66UEgADICKiWhhfGcfE6kTJ4W+AlWQh7A9vQa2IiKjVtEQANLk6iQvLF3D7\n7tuLltNMDe3+9orXFyIiotKeGX8GAEoGQJqhIeQLwefxbUW1iIioxbTEmf6xiWMAUHL9H0VX0BXi\nmHMioloYGx/DSOcI9nTtKVpOMdgWExFR7bREABQdj2JnZCcu77m8aDnd1NER7NiiWhERtQ7N0HBi\n8gRuH729ZGY3zdSYjZOIiGqm6QMgwzRwfPJ4WQddgPN/iIhq4cezP0Zci+PO3XeWLiyBsI/zf4iI\nqDaaPgD6yaWfYFVZ5aJ7RER19Mz4M/AKL24bua1oOSklIICgj20xERHVRtMHQNGJKAREyYOuaqhc\ndI+IqEbGxsdw/c7rSw4zVg0VnYFOJqMhIqKaafojzNj4GN60403oDfcWLccECEREtbGYXMRPLv0E\nt48Wz8QJcAFUIiKqvaYOgGJqDC/MvFAy+xtgrTnBRfeIiKrvxOQJSMiy5v8YpoFIMLIFtSIiolbV\n1AHQickTMKRRcv0fwBp3zgQIRETV98z4M+gKduHaHdeWVZ5tMRER1VJTB0BjE2No87fhhp03FC2n\nmzoC3gAC3sAW1YyIqDVIKTE2PobDo4fh9XiLljVMAz6Pj20xERHVVHMHQONjuHX41pIHU0VXuOYE\nEVENnFs8h9n4bFlDkbkAKhERbYWmDYAmVibwxsobZU26VQ2Vk26JiGogOh4FgLKGIqs622IiIqq9\npg2AxibGAJR30AWAsJ+L7hERVVt0IorLey7HUMdQybJMRkNERFuhpgGQEOJuIcRpIcQ5IcTHXfYL\nIcSfpfe/KIS4qVqfPTY+hqGOIezt3ltWeU66JaJmVM92OKWncPLiyZILUTuxLSYiolqrWQAkhPAC\n+BKAnwVwDYD3CiGuySn2swD2p28fAvDlany2buo4Pnkct4/eXnJhU9VQEfaHS07OJSLaburZDgPA\n81PPQzGUsgIg1VDR5m9jW0xERDVXyx6gWwCck1K+JqVUAXwTwD05Ze4B8FfScgJAtxBi12Y/+KXZ\nl7CmrpU35pzzf4ioedWtHQas+T9+jx+Hhg6VLMu2mIiItkotA6BhABOO55PpbZWWgRDiQ0KIU0KI\nU3NzcyU/eGxiDAICh0cOlyyrGzozwBFRs6paOwxkt8Xzc/MlPzw6EcXBoYNlzevRDA2dIbbFRERU\ne9siCYKU8mEp5UEp5cGBgYGS5aPjUbx58M3oDnWX9f4cc05EVJqzLe4f6C9adjY2izMLZ8pORAOw\nLSYioq1RywDoIoBRx/OR9LZKy1RkVVnFi7MvlpX+WkoJCB50iahp1aUdBoBjE8cAAHfuvrNkWVOa\n8AgPgt7gZj+WiIiopFoGQCcB7BdC7BVCBAC8B8AjOWUeAfBv0lmIbgOwIqWc3syHPjv5LAxplD3p\ntiPQUTJRAhHRNlWXdhiweuL72/pxoO9AybKqoaIz2Mm2mIiItoSvVm8spdSFEPcBeByAF8DXpJQv\nCyE+nN7/EIBHAfwcgHMAEgB+bbOfG52Ios3fhusHry9ZVjEU9IX7NvuRREQNqV7tsClNjE2M4S17\n3gKPKH2dLaWnMNg+uNmPJSIiKkvNAiAAkFI+Cuvg6tz2kOOxBPAb1fzMsfEx3DZyG/xef8mypmmi\nPdBezY8nImoo9WiHX5l7BUuppbLn/0jJBVCJiGjrbIskCOUaXxnHxOoE7hgtb9E9Ccn5P0REVTY2\nPgYAZc3FtLEtJiKirdJUAVB0PAoAZV11NEwDPo8PQR8n3RIRVVN0PIqr+69Gf1vxTHGAtXB10Bss\nq9eeiIioGpoqABqbGMNwxzD2dO0pWVYxFK7/Q0RUZTE1hh/N/KisRDSANf+HbTEREW0lYQ3/3j6E\nEGsIYAImjE29kQde6FBhQq9S1QCgH0Dp1QHro1Hr1qj1Ali3jWLdCtsjpSy9mNk2IISIIYDxKrXF\nyqbfJ1u9/58LadR6AazbRjVq3Rq1XkD969Y07TBtXE2TINTIaanIg/WuhBshxCkpWbdKNGq9ANZt\no1i3lvFTtsWVadR6AazbRjVq3Rq1XkBj141aR1MNgSMiIiIiIiqGARAREREREbWM7RgAPVzvChTB\nulWuUesFsG4bxbq1hkb+Lhu1bo1aL4B126hGrVuj1gto7LpRi9h2SRCIiIiIiIg2ajv2ABERERER\nEW0IAyAiIiIiImoZ2yoAEkLcLYQ4LYQ4J4T4eJ3rckEI8ZIQ4gUhxKn0tl4hxBNCiLPp+54tqsvX\nhBCXhBA/cWwrWBchxCfS3+FpIcTROtTtD4QQF9Pf3QtCiJ/b6roJIUaFEE8JIV4RQrwshPit9Pa6\nf29F6tYI31tICPGcEOLH6br9YXp7I3xvhepW9++tmTRSO5yuD9vijdet7n8bbIs3XLeGbIvZDtO2\nIaXcFjcAXgDnAVwOIADgxwCuqWN9LgDoz9n2xwA+nn78cQCf26K6vAXATQB+UqouAK5Jf3dBAHvT\n36l3i+v2BwD+H5eyW1Y3ALsA3JR+3AHgTPrz6/69FalbI3xvAkAk/dgP4FkAtzXI91aobnX/3prl\n1mjtcLpObIs3Xre6/22wLd5w3RqyLWY7zNt2uW2nHqBbAJyTUr4mpVQBfBPAPXWuU657AHwj/fgb\nAH5+Kz5USvk0gMUy63IPgG9KKRUp5esAzsH6breyboVsWd2klNNSyh+mH68BeBXAMBrgeytSt0K2\nsm5SShlLP/WnbxKN8b0VqlshW/q30CS2QzsMsC0ut26FNEJ7V/fvjW1xVetVCNthqovtFAANA5hw\nPJ9E8Yao1iSA7wohnhdCfCi9bVBKOZ1+PANgsD5VK1qXRvkePyKEeDE9LMPuoq9L3YQQlwG4EdaV\nqob63nLqBjTA9yaE8AohXgBwCcATUsqG+d4K1A1ogO+tSTTid8a2eHMa5m+DbXHFdWrItpjtMG0H\n2ykAajR3SClvAPCzAH5DCPEW504ppUTxqx5bppHqkvZlWENobgAwDeBP6lURIUQEwD8A+KiUctW5\nr97fm0vdGuJ7k1Ia6d/9EQC3CCGuzdlft++tQN0a4nujmmFbvHEN87fBtrhyjdoWsx2m7WA7BUAX\n8f+3d/+xV9V1HMefLw3UufwR5tLRBiqYYxkmWFtY0tCKXEMXiaNiSaa2rMFyc2OZc/1izuY/bRUw\nW6WmazZAFjQNTJsD/IF8ETNTm8M5sDHLGjGIV3+cz9cO3y73+/t7D97XYzv7nnPu53zO+3z2ve9z\nP+d87j3w3tryxLKuI2y/Wv7uAX5Ddct2t6QzAMrfPZ2Kr00sHW9H27tLgjwErOB/t7vHNDZJ46hO\nanfbfqCsbkS7tYqtKe3Wy/YbwEbgkzSk3VrF1rR2O8o1rs2Si4euKe+N5OLhaWouTh6OJjuaOkBb\ngSmSJksaDywA1nQiEEknSnpn7zxwGbCjxLOoFFsErO5EfMWRYlkDLJB0nKTJwBRgy1gG1puciyuo\n2m5MY5MkYBXwnO0f1l7qeLsdKbaGtNu7JZ1S5k8ALgX+RDParWVsTWi3t5HG5GFILh6uJrw3kouH\nHFsjc3HycBw13IBfYhjoBMyl+hWWF4FlHYzjLKpfLXkGeLY3FmAC8DDwAvAQ8K4xiudeqlvKB6jG\nzy5uFwuwrLTh88CnOhDbL4AeYDtV8jtjrGMDZlENDdgObCvT3Ca0W5vYmtBu5wNPlxh2ALf097/f\ngNg63m5vp6kpebjEklw8vNg6/t5ILh5ybI3MxcnDmY6WSXaThiNHRERERESMnqNpCFxERERERMSw\npAMUERERERFdIx2giIiIiIjoGukARURERERE10gHKCIiIiIiukY6QHEYSZZ0R235m5JuHaG6fybp\nsyNRVz/7mS/pOUkba+veL2lbmfZKernMPzTIujf0PnekTZnvSpo91Pj71LVLUo+k7ZLWSzp9BOK7\nRtJ7RiK+iBh5ycP91p08HBHDkg5Q9LUfuFLSaZ0OpE7SOwZRfDFwre23Tn62e2xPtz2d6hkEN5Xl\nOYPZj+1P2H6znzLLbG9sV2aQLrZ9PtXzE24ebnzANUBOvBHNlTzcRvJwRAxXOkDR10Hgp8CSvi/0\nvXIo6Z/l7yWSHpG0WtJLkn4gaaGkLeWq2dm1auZIekLSnyVdXrY/VtLtkraWK2zX1ep9VNIaYGeL\neK4u9e+QtLysu4Xq4XWrJN0+kAOWNEfSJkkPUj2oDUlrJT0p6VlJX66V3SXpFEnnlP2uKmV+K+n4\nUuaXkubVyt8q6elybFPL+tMlPVy2/YmkV3ufnt3GH4Bzyvafrx379wYan6SrgOnAfeXK6/jS9jtL\nfMsH0mYRMaqSh0kejojRkw5QtPIjYKGkkwexzQeA64HzgC8AU21fBKwEbqyVmwRcBHwa+HE5WS0G\n/m57JjATuFbS5FL+g8A3bE+t70zSmcBy4ONUJ5KZkubZvg14Alho+6ZBxD8D+Krt88ryItsXlniW\nSjq1xTbnAnfangbsA+Ydoe7dti+gaoulZd1twPqy7VrgzHbBSRJwOdAjaSLwHWA2cAHwkd4PMf3F\nZ/s+qqeZX1Wuwp5K9WTzaeXq5vfbxRERYyZ5OHk4IkZJOkDxf2z/A/g58PVBbLbV9mu29wMvAr8r\n63uoTra97rd9yPYLwEvA+4DLgC9K2gZsBiYAU0r5LbZfbrG/mcAm26/bPgjcDXx0EPH29bjtV2rL\nSyQ9AzwOTATObrHNX2z3lPknOfw46x5oUWYW8CsA2w8C7YZLPEp1sjyB6sPGh4Df2/6b7QPAPbQ+\n9oHEtxc4BKyQdAXwrzZxRMQYSR4GkocjYpQMZjxvdJc7gaeAu2rrDlI6zZKOAcbXXttfmz9UWz7E\n4f9n7rMfAwJutL2h/oKkSxi7E8Fb+5E0h+pE9mHb+yQ9BhzfYpv6Mf+HI7+f9g+gTDsX236jFt9A\nt+s3PtsHJM0ALgXmAzdQfRCKiM5LHk4ejohRkDtA0ZLtvcD9VMMiev0VuLDMfwYYN4Sq50s6poxH\nPwt4HtgA3CBpHICkqZJO7KeeLcDHJJ0m6VjgauCRIcTTysnA3nLSnUZ1lXOk/RH4HICkuUDbXwzq\nYzMwW9IEVV8WXsDgjv3N3v2p+qWik8rVzyVUQzkiogGSh5OHI2J05A5QtHMH8LXa8gpgdRmSsJ6h\nXRV8heqkeRJwve1/S1pJNSzgqTLG+nWOPI4bANuvSboZ2Eh15XKd7dVDiKeVdcBXJO2k+mCweYTq\nrfs2cI+kLwGPAXsYYHva3iXpW8AmqmNfa3vdIPZ9F7BS0j6qD1C/lnQc1QWRpW23jIixljycPBwR\nI0x23zvhETHaypeOD9o+KGkW1ZdkZ3Q6roiIbpE8HNG9cgcoojMmAfeWYSP7ges6G05ERNeZRPJw\nRFfKHaCIiIiIiOga+RGEiIiIiIjoGukARURERERE10gHKCIiIiIiukY6QBERERER0TXSAYqIiIiI\niK7xX2dzkoCxtO3+AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fb88bf5e650>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Produce learning curves for varying training set sizes and maximum depths\n",
    "vs.ModelLearning(features, prices)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 问题 4 - 学习数据\n",
    "*选择上述图像中的其中一个，并给出其最大深度。随着训练数据量的增加，训练曲线的评分有怎样的变化？测试曲线呢？如果有更多的训练数据，是否能有效提升模型的表现呢？*\n",
    "**提示：**学习曲线的评分是否最终会收敛到特定的值？"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**答案: **选max_depth=1。随着训练数据量的增加，训练曲线的评分先迅速降低后逐步趋向稳定，偏差很大，测试曲线先快速上升后也趋于稳定，偏差很大。现在是模型复杂度不够，更多的训练数据不能有效的提升模型的表现。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 复杂度曲线\n",
    "下列代码内的区域会输出一幅图像，它展示了一个已经经过训练和验证的决策树模型在不同最大深度条件下的表现。这个图形将包含两条曲线，一个是训练的变化，一个是测试的变化。跟**学习曲线**相似，阴影区域代表该曲线的不确定性，模型训练和测试部分的评分都用的 `performance_metric` 函数。\n",
    "\n",
    "运行下方区域中的代码，并利用输出的图形并回答下面的两个问题。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAb0AAAFNCAYAAACUvLFdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzsnXecXFX5/9/PlJ3tNZtkUzcJIQGCgt/QxAIifEFBbL8v\naBABMfYCCgiBIJJQFEFUyjcoqBDb10oURToiIARFkRJI22TTy9bZ6XN+f5x7Z+/uzm5mk51t87z3\ndV9z67nn3pm9n/s85znPEWMMiqIoilII+Ea6AoqiKIoyXKjoKYqiKAWDip6iKIpSMKjoKYqiKAWD\nip6iKIpSMKjoKYqiKAWDit44QUT+JCIfz2G/ThGZPRx1UgoPEblQRB4fgnKuEpE7h6BKQ46IHCoi\n/xKRDhH57EjXRxkcKnrDiIhsFJGI88/SKiJPi8inReSAvwdjzGnGmB/nsF+5MWb9gZ7PiyOk7pR2\nrtFdXjSU5+rn/M2ec24XkbtFpCzf5x1NiMhUEbnHuf52EXlVRK4WkZKRrtv+YIy51hjzaQAROUhE\n9rtDsSPEKef30S4i/xSR9xxA9S4D/mKMqTDG3H4A5SgjgIre8HOGMaYCmAncgP0H+uHIVunAcIS0\n3BhTDmzCXqO7bmXv/UUkkIdqnOac/y3A0cCleTgHIuLPR7mDOH+feyciE4BngABwjDGmEjgVmACo\nVW/5q/P7qAF+AvyfiFQNpgDPvZ8JvLw/lcjTb18ZBCp6I4Qxps0Ycz9wFvBxEVkAICIhEblJRDaJ\nyA4RudP7ti4iZ4rIi84b6zoROdVZ/7iIXOjMHyQiT4hIm4jsFpFfeI43InKQM18lIj8RkV0i0iQi\nV7pWp4icJyJPOXVpEZENInLa/lyriCwTkV+IyM9EpAM4R0R8InKFcw27ReTnIlLjOeZ4EXnWsYhf\nFJF35HhftwJ/AY7wlFUsIjeLyGbnnt4uIsWe7Zc7FtIWEfmkc48anW33ichtIvJnEQkDbx+oPBGZ\nKCIPOPXeKyJPes5zhYhsdb6710TkBE/9visi25w63CwiRc62d4v1EFwhItuBu7Jc9leBPcC5xpgm\n5z40GWO+YIx52SnnbSKy2vlNPCcix3jq9ZSIfMO532ER+Z2I1DnfV7uI/F1EZjj7Bpz78wXnN7Fb\nRG6QfrwVYl2BDzv34jUR+ZCzPiQiL4nIZzzlPisiVzjLy0TkR04xTzrrXO/B2537e4jnPA0i0iUi\ndfv4faSAu4FSYJZz7PvEuitbnXuxwFNus4hcIiIvAWHn+3w7cKdTl9kiUu38TnY539XlIiLO8ReK\nyJPO97sXuNJZ94SzrlVE1orIMSLyCc9v6hxPHd4n3f/zm0TkKs+2g5zv41ynrrtE5Gue7QGxruJ1\nzvGrRWTKQN/NuMcYo9MwTcBG4N1Z1m8CPuPM3wLcD9QCFcAq4Hpn29FAG3Ay9oVlKjDf2fY4cKEz\n/zNgibNPMfA2z7kMcJAz/xPg9855GoHXgU84284DEsAnAT/wGWArIIO9RmAZEAfOcOpUAnwF+Jtz\nDcXAD4B7nf2nYx/i/+3sfyqwG6jr55zNwAmeY18Bvu3Z/j3gt9i3/ErgAeBaZ9vpznUdApQ5984A\njc72+4AW4DinLqF9lPct4PtAECgC3uGsPwxoAiY7y7OA2c78dcDTQD0wEfg7cLWz7d1A0tmnCCjJ\ncv2rgasG+E4mOL+bj2CtwY8597fG2f4UsAZrFdYArznLJzr7/xS4y9k34Nyfh519ZwJrgfOc7RcC\njzvz5cAW4FznuP9yzjvP2f5m594eDFzt/B78nt/Mj5z5gwDT65pWAMs9y18BftvP9XvrFAAuBtqx\nv/ujgB3Opx+4AFgHFHl+Wy8A09x779yv8zzl/xT4jVPebOd+fNxz7iT2/8eP/e1fiP3f+piz7gbn\nt/Fd7O/rPc73VeqU8S7n9+Nz7tlu4HTvvQHuxP4fvQWIAXOd7ZcD/wLmOscfgX22DPjdjOdpxCtQ\nSBP9i96zWJESIAzM8Ww7DtjgzP8vcEs/ZT9Ot+j9xHkoTMuyn3H+UfxYITrUs+1TnofDecBaz7ZS\n59jJg71G7APs0V7r3gDe6VmeDkSdf8wlwD299n8EWNTPOZuBTqDDqeNfgCpnm88pd6Zn/7cDb3ju\n1bWebfPpK3p3e7bvq7zrsA/AOb3qOA/7cD0JCPTa1gSc4ll+r3vvsaIXxXkI93P9G9zvvp/t5wNP\n91r3PHCOM/8UcJln263AKs/yB4DVzrwreu/2bP8i8KAz7xWYRcBjvc77Q2CJZ/ky4FVgL85LgOc3\n8yNnPpvoHe9ctzjLLwIf7Of6XeFpxQrG08C7nG134bxgePZfBxzv+W2d22t7RvSwLzdJ4GDP9s8B\nD3vOvT5LfV71LB/p3NM6z7o2YEE/1/N94Fvee4Pn/xL4B/Bhz7W8N0sZ+/xuxuuk7s3RwVTsP309\nVlxecNwercCfnfVghWFdDuVdihXQ50TkZRG5IMs+E7D/sE2edU1OXVy2uzPGmC5ntjyH82djc6/l\nGcAqz3W+5KyfiLUePuJuc7YfC0wZoPzTjW0rPQk4FPs2CzAZ+/b8L09Zf3DOg1Omt26969l73b7K\nc9/aH3FcSpcAGGPWYK2RbwA7HdfhZE8dBvoedhhj4gNc+x6gYYDtvcvPeg7PfCTLcu/v3XtPmsj+\n3cwEju/1PZ7Vq64/AuZgRTbnACtjzN+wYvM2xx05A/jjAIc8ZYypNsZMMMa81RjzqKeOl/WqYwM9\n702234TLROwL5EDfX7bje9/flDFmT6915QAicpzY5otdItKGFc0J3sKMMds9i110f1/9PTNy+W7G\nJSp6I4yIHIX9B3kK+xYaAQ5z/kGrjTFVxjbAg/3nmbOvMo0x240xnzTGTMFab7eL047nYTfWxTLT\ns24G1uWRD3pH3zUDJ3uus9oYU+z8827GWnrebWXGmG/t8yT2YbYS62YE+3CJY9023nvqBjFsw7qu\nXKbvo+4DlmeMaTfGXGSMaQTej32gvtPZdp8x5nisa9MPXO+UuZWBv4fe9643DwMfcNuRstC7/Gzn\nGCze+zTDOUdvNgOP9Poey40xn/fscwfwO+B0ETm2n3P1d/0/Ac7Bugl/aYyJDe4SMnW8plcdS40x\nv8zh/AA7gRQH9v3ti58DvwamO7+zH2BfanOhv2dGLt/NuERFb4QQkUoROR37g77PGPOSMSaNdbfc\nIiITnf2mish/O4f9EDhfRE4SGwgyVUTmZyn7/4mI+yBvwf7Tpb37GNug/0tguYhUiMhMbFvHfXm4\n3GzcCVwn3QESE0Xkfc62e7EP8ZNFxC820ONEtwE+B24B3iMiC5zr/AHwHRGpF8s0ETnF2feXwCdE\nZJ6IlAJX9VcoZO5bv+WJyBkiMscRoDbsAzEtIoc41xDCvthE6P5OfgYsFZEJIlLv1GEw38NN2Df/\nezz3c5qI3Coih2Et0cNE5CwnsOGjWLfYQJbRvrhUbADHDKx78xdZ9rnfOe9HRSToTEeLyDynjucD\nC7Cu9IuAeyV7V5OdgJG+/UvvBT4MfBQrgPvDXcDnROQo57ssd77DnLq8GGMSwK+wv+VyEZnlXMtQ\n/h9VAHuNMVHnxeDsQRz7A2CZ+5sUkSNEpJZ9fDfjGRW94WeV2AjGzdi2q5uxbS4ul2Ebwp8VkXbs\nW/w8AGPMc86+t2AfqE/Q9w0ebKP830WkE/vj/lI/rqMvYNsQ12MtzZ9iI9uGg5uxrttHnPvxNLbe\nGGM2YtuRrgJ2YQN9vkKOv1fHWlxJt4B9Betyeg573/6CbdjHGLMKa208iW1n/JtzzEBWQ7/lYb+r\nR7FtjH8DbjXG/BXrEv0m1sLejg0CWeIccw022OA/wL+xgSyuFZjL9e7Gtv0CPO/cz4ecc603xuwC\n3of9be3BPpRPN8a05HqOLKzCtqP9ExvU86Ms9WrDBiOdg7Wot2OvKyQ2Ovbb2PayLmPMT7DXflOW\ncjqc4/7uuOIWOus3Yt3iMWPM0/tzEcaYZ7FBJndgXxBfd+o7GD6Ltf43Yv8nf8z+i3A2PgNc73yv\nV2Bf1HLlW1hL+hFs8M4KoHig72YI6z0qcRuBFUUBRORwbCBAyLG8FQ9i+5klgFmO6IwoIvITrLB/\nfaTroowN1NJTCh4R+YCIFDlunxuA36vgjX4cd+eZDJ93QhkHqOgpig0x3411K0edZWUUIyLXY13C\n1xljNo10fZSxg7o3FUVRlIJBLT1FURSlYFDRUxRFUQqGMZfxe8KECaaxsXGkq6EoiqKMIl544YXd\nxpj6fe035kSvsbGR1atXj3Q1FEVRlFGEiPROtZcVdW8qiqIoBYOKnqIoilIwqOgpiqIoBYOKnqIo\nilIwqOgpiqIoBYOKnqIoilIwqOgpiqIoBUPeRE9E7haRnSLyn362i4h8V0TWisi/ReQt+aqLoiiK\nokB+Lb0fAacOsP007MCbc4HF2EEcFUVRFCVv5E30jDFPAnsH2OVM4CfG8ixQLSIN+aqPoiiKooxk\nm95UYLNnudlZ1wcRWSwiq0Vk9a5du4alcoqiKMr4Y0wEshhjVhhjFhpjFtbX7zOfqKIoijLaWbkS\nGhvB57OfK1cOy2lHMuH0FmC6Z3mas05RFEUZz6xcCYsXQ1eXXW5qsssAixbl9dQjaendD5zrRHEe\nC7QZY7aNYH0URVGUoSKdhmQS4nFoaYENG+Bf/4InnoCLLuoWPJeuLliyJO/VypulJyI/A04AJohI\nM3A1EAQwxtwJPAC8B1gLdAHn56suiqIoBcnKlVZINm2CGTNg+fKellQ63T0Z0/MzlYLOTmhthfZ2\naGvrOd/eDh0ddnLnOzu7p3DYClk4bMvKhU2b8nMfPIgxJu8nGUoWLlxodDw9RVEKEq9I9TelUlZs\n7rsPrr0WYrHu44NBePe7Ydo0K1JeYco25aIPxcVQVgbl5T2nsjI7lZb2na6/HvZmCe6fORM2btyv\nWyMiLxhjFu5rvzE3iKyiKMqoYV+WFOQuVMlkzykSscLQ0mIn18rq7MxuZXnno9Hs9U0k4E9/svOl\npT3FqqICGhr6Cpb7WVJi50tK7HJ5uf0MBrvLd0WyqAj8/p6fwaCd/H6oroYvftFeo0tpqb1/eUZF\nT1EUZSCMsaLkipM7/eIX8OUvdz+4m5rgwguhuRlOPdUKV1dXtyBlEyjvcq7C5VJaClVVUFlpP2fN\n6p6vrIRbb81+nAg89ZQVHxF7fb0/wQpUIND9WVTU/en328nny/65Ly680Irnvl4Y8oC6NxVFKQz6\nEy9vwIX7mUjYybW6UikrRK2t3dN111mB6k0gALW1dttghauy0k7V1d3z3vWuVVZaagXGvRaRvmWf\neSZs3953/dSp8M9/HrhwjTLUvakoyvggmwvx7LP7CpfrInQFK5t4uXR2WpehG5zhBma4LkTXnehO\nbW25tW+BPc/b395TrHoLW1WVdRV6hcv76bW4vLhuwqKi7sm1xFwRc+dvuqlntwCwYnnjjVDA/Z1V\n9BRFGR1427Xc+Z/9DC65pK8Lce1aOO00uy4SsaLkipg7eSMO3cltH/MKoJdgEGpq7FRbC1OmdC/3\nnj71Kdixo28ZDQ1wxRUDW2HGWFF2Bay0tKeIuQLWexoMrqtwBFyIoxl1byqKkl9c96F3isWs6y8e\nt/OxmBUC1wXZ0gK7d9v+XC0tfcsMBKy10tLSvwvR57NuwmyCVVubfVtZWbdIeS1JryUGdp8//9lG\nIXrPX1wMy5bBhz/cLWL9CVg2MVT2G3VvKoqSX3qLWSplRSwa7Ra1WKzbXReLwZ49Vsz27u2edu+2\n63ftstPevfaYgUgm4bjj+gpXbW33fGWlFT5vfb1teO6yG8CRTPZ0Y7pWV0mJnQ+FeorY4YfbEPul\nS2HzZrWkxghq6SlKodO7zWzZMjjrrJ6ClkhYN6LXMnNdhOGwFa7+xGz3bitmbW19z+33Q12dtdp6\nTxMnwtVX2+N7M2UKPPpodhFzBdO1pNxnnBuJ6Fpgroi5bWDedjG1xMYcaukpitI/rpCtXNmzv1RT\nE3zyk7BuHRx7bLeY7dljXYl79nRPrmXm7WvlUlTULVyzZ8Mxx/QVtPp6a5F526qM6RawZBK+9CVr\nPfV2IS5ebNvsXGusvLxbxNw2sd4i5hsT+fWVPKOWnqKMV1Kp7ijGWMxG8blTLAY7d8IFF2RvM+uP\n8vK+wpVtqqzsaSl5hcz9zNY/zOfraYWFQrBqlRW+LVtsJpFrr4VzzhmTYfVK/sjV0lPRU5SxTDrd\nHZofj3enlIpEutvQNm2y09attu1p0yb7GY8PXPaSJX0FrqSke3vvQI/e4fbez2Cwp5AVF1tLLJtr\nUVH2A3VvKsp4wQ1vd4UtEunOmRiNWjff5s3WNekVtk2bevbRKiqybXazZsEJJ9jPW26xwtibhgb4\n0Ie6xcwNRHGjLMEKVShkhdAVMm+fMa+QafuYMkpQ0VOU0YAxPTtV9xa2cLhbzLZssfOu0Hmzgvj9\n1gXY2AhHH20/GxttlGFDg92eTvfstH3DDX3bzC66yEZGhkJ26i1kgYAKmTImUdFTlOHAGyE5bRpc\ndRWccUZ3Nvto1Apdc7Pdp7m5W9yamnpaYyJWwBob4b3v7Ra2xkabYioY7JtayxibRssYK1ilpbbd\n7bOftcd84xv2nBp2r4xztE1PUfKB2wE7HrdDvFx6aU9rKhiEd73LBoY0N1th650nsb6+20rzCtuM\nGdb68lqHboCIa32JdA/jUl7eN2WVoowztE1PUYaDVKpb3Lq6bE5HN0P+xo2wfr3Nddg7a0giAQ8+\naF2IjY22e4BX3GbMsGLlhvB7ha2rq7ubgGux9Q7Z9w73oihKBhU9RcmFdLpb3CKR7gE4w2EbPLJ+\nvZ02brR5ITdt6j+/o4sI/P3vdt4rbO58a6sVvZKS7jHOiot7Cpu2qynKoFDRUxQvxnSLWzRqxa2z\n01pXO3b0FLd16+y8G/ovAtOnw9y5cMop9nPuXJuYeNu2vueaNKlb2IqLu92RbtorV9xU2BRlyFDR\nUwoTtxuAK26dnXYKh23QyLp1sGGDndavt8ve8P+GBitob30rHHywnZ8zxwqWGx3pBpF86lN9IyRL\nSmy6rze/2QqbZgtRlGFBRU8Z/7jiFov1FLfW1m5rzRW4det65oisq7OC9qEPWXE76CC7XFHRs/9c\nItGdZNnvt+5IN2v/ggU2FddVV+kQL4oywmj0pjI+WLnSjmG2ebMNwb/kEjj55O7AEtdaa2rqnvcm\nMq6stGJ20EHdltvcuTZrvxsl6QqcN2VWWVn3aNZuthENIlGUYUejN5Xxj5v5/yc/sSLnug+bm+Hi\ni627saOjZ3taaald/453dFtuBx9s02y5ZboCJ2LzUrrh/3V1VuTcDtva3qYoYw4VPWVs4AaYuKNk\nt7baNrbXX7cWXu8uAamUtehOO63baps711qBPl9PcQNbHti2tupqa7m5kZJFRSpuijJOUNFTRiep\nlBW4ri4rSG1tNihk3Tr45z/hH/+A1aut+3KgMm64odstmU7blF1utGRFhRU3b7SkBpQoyrhGRU8Z\nHXj7v7W0dEdKbtgA//oXvPACPP98t0U2c6ZNwXXMMbbz944dfcucNMmWW1ZmBc47AraKm6IUJCp6\nyvCTTlt3ZDRqRay11VpixtiO3i++aEXuuee6g02mTrVpu445xmYvmTzZ7u+OMpCtS8A3vwlHHDEy\n16goyqhERU/JP8lk96gBLS1WpNwIyN27rbvy+edtdhLXYps40faBO+YYO02fbte7/ercAJOaGvj8\n563l9/Wva5cARVEGREVPGVrcgJNotDvgxLXAfD67/M9/WoH7+9/tSAJgIyNdgTvmGJt/UsRahZFI\nz8wlDQ22i0Fpafego+edZydFUZQBUNFTDoxUyoqaG3DS3m7XGWND+js6bNCJK3IbN9rjqqvhqKPg\nggusyB10UHeEZDxuy0mnrVDW1FhRLC217XGKoij7iYqeMjjccdk6O62L0Rs9GQpZwVq9ulvk1q61\n28rLrch95CNW5ObN6w4mSaWs69Md96201I45V1lp2+Y06ERRlCFCRU/JnbvvhiuvtOO+TZoEX/6y\nDS7xitxrr9l9S0vhLW+B97/fityhh/Ycx83tc2eMdVHW1tqprEwzmiiKkjc0DZmyb5JJuO02uOwy\nK1YubjAKWCvvyCO7oysPP7yneLn97hIJu1xeDhMmdHcl0M7fiqIcAJqGTDlwjLEuzA0bbJcAr+C5\n28vL4Y477GgB3vY2tztBJGIFze+37XJuEmYdvVtRlBFAnzxKdiIRm5y5pcXmsty+Pft+4TAcfbSd\nd7smuIOnVlbaSEs3pZdac4qijDAqekpPUimboLm52Vp2P/yhHcHA57PRlL2ZPNkKI9hsJ/X1NjKz\ntHRErDljDKLiqihKP6joKd20tdkkzbEY/O1vNr3X7t024nL+fLjuup5ZT4qLu0czcK25YSSZThJL\nxoilYnTGOulMdBKOh0mbNEX+IoK+IEF/kIAvQJG/yK7zB/GJD7/48fv8mU+f+PCJRokqynhHRU+x\nItfUZEcM370brr8ennnGDn56xx02KAWstff979usKdOm2awnH/tY3quXNumMuEUSETriHXTGO0mk\nnKAYwQqcL0hlqDJzTMqk7DHJiF1Op0gba62KCBh7rBvM5RNfRhgDvgBBXzAjll6B9AqlX/xqWSrK\nGEJFr5BJp62Abdpk2+Huu8+6M4uL4eqr4ayzbABKImH75n3wg3DRRdZ1mQeMMcRTceKpONFklI5Y\nBx2JDqIJa12KCIJQ5C+iJFBCeVF5v2X5xY8f/6DPnzJWGCOJCGETzoinK4yCgAAGDAYEAhIg6Lei\n29uq7E8ofeJTsVSUEUBFr1BxRxOPRm3GlOXLbUqw97/fDsg6YYKNwGxrswEo7ijiQ/Sg9romw/Ew\nHfEOwvEwxhgrJpARj5qSmiE5574QEQLi/EsMQi9dKzKZThJLxeiId2TWect2LUuwgunDlxHKgC+A\n3+fvIZ4BX8CKpMf96gqmO41H4XR/A2mT7jEvSMY9rSj7i4peoRGPw+bNsHOnTfX1rW/BI4/YNGD3\n3tsdiRmJ2KmhAaZM2e8O465rMp6K05Xo2qdrciw+xH3iw+cf/IPYa1km0gmiqSjpuH3Quxamj+5y\nDT2DdIwx+MWfsShdt6w7BX1BAv5Av4Lpimku9TSYTL16z/cWJ3fedSen0qnMdbrzBkMynSSdTpMi\nRTqd7mNVQ/fLgvsiBBD0BykJlFAcKKY0WEooEMq8KKgoKvsir6InIqcCt2Lfm39gjLmh1/Yq4D5g\nhlOXm4wx9+SzTgWLMbBrl819mUrBr34Ft99uLbevftUmaw4GrZuzvd0Gphx+uO1Tl1PxhkQ6QSwZ\ny7gmOxOdRBKRjEswV9dkodDDstxP0iadmbztl+5kjEGQrILpEvAFCPitSKbTadKkrRg5QuVtB3XL\nc9tCe5Ttcfliut3RIpIRIkEyFqq7LUAAX6DnuoFwreqWaAu7unZ1R+waMGIo8tnfWEmwhNJgKUX+\noowgBn3BMflipQwdeRM9EfEDtwEnA83A8yJyvzHmFc9unwNeMcacISL1wBoRWWmMieerXgVJOGw7\nmHd2wquvwrJl1rV58slwxRXWkjOmO8nz7Nm260E/D4d9uSZdN1TQHxw212ShMhRRpz2CfAT8+AkE\nrIWYiwgNN36fbR8NkT35eDKdJJFO0BXpYkfYGarKI8jF/mJKgiUZa9Ftf3WtxdF2vcrQkk9L72hg\nrTFmPYCI/Bw4E/CKngEqxP7KyoG9QDKPdSosEgk7KOu2bVb4br0VVq2ykZf/+79wwgl2v2jUbp80\nyW4rKspeXCrBto5tbA9vt5aCE8RR5C8as65JZf/ds6MV1707kChGk9GebnYHESEUCFEaKM0IoyuI\nrvtYGdvk8xucCmz2LDcDx/Ta5/vA/cBWoAI4yxiTpQe0MiiMgb17rXWXTMIf/wjf+Y7tmvDZz8Kn\nPmUjNFMpa90VF9vuCRUVWYtLmzS7u3bT1NqEMYaq4ioVOGXM4opiNoyxbY1diS7aY+0k0/Yd3HWf\nigjFgW5LsSRYkjXwSBm9jPRry38DLwLvAuYAD4nIX40x7d6dRGQxsBhgxowZw17JMUVXl+1z19Zm\n2++WLYOXX7ajkC9dCrNm2f06OqzoNTZaV2aW4XuMMbTH2tnQsoFoMkplcaW+6SrjGpFu13w23Lbr\nzngnbdE2kulk1j6fmaQIvqJMe6nbtuhG6nqDi9xuLfoymX/y+QTbAkz3LE9z1nk5H7jB2F/KWhHZ\nAMwHnvPuZIxZAawAO8pC3mo8lkkmbX5MN33YnXfCL35hux7ccgucdppto4vHreDV1cHMmf0OyhqO\nh9nUtom2aBvloXJqi2qH+YIUZfQhIpmuNAPhtpFmonK9wUUYxDji5gT9gA3C8eHrE3jj/fRmE+ot\nmvmgd1Ru74jdbNG87rUn08kenyljo3ST6WSPYKlQIMRh9YcNm+DnU/SeB+aKyCys2J0NfLTXPpuA\nk4C/isgkYB6wPo91Gp+4IyEkEvD447YbQlsbfPzj8IUv2EjMdNqObF5UZMe2q6rKWlQ8FWdL+xZ2\ndO4gFAhRW6pipyiDxe8bfHIE6NmNJZqMEk6Hewgm9OzG4UbUApnMQUW+oh7p94L+IIJkyshVjNyu\nJTb+RzJ9THuf243g7R216+1H6gZFueu8wVJt0bZMANxwkDfRM8YkReTzwIPYLgt3G2NeFpFPO9vv\nBK4FfiQiL2Fv2WXGmN35qtO4Ixq1rsy9e62Vd911dkDXI4+Er3/d5ssEG7WZSMD06TZYxd/3nzGV\nTrGjcwfNHc348FFTUjOmXC2r1qzi5mdvZlvHNhoqGrj42Is5Y94ZI10tRRkU+9uNxdu3s7d1mUpb\n8eotPN7uJO78aI/cHQry2kBjjHkAeKDXujs981uBU/JZh3FJKmXThzU3W3flPffAj39sA1GWL7fp\nwny+bldmba11ZWZJCG2MYW9kL02tTSTSCSpDlWOuIX7VmlVc+diVRJM2XdnWjq1c+diVACp8SkEg\nIvuVeq+AZW0gAAAgAElEQVQQ0aiEsUZ7e/dICM8+awd33b4d/ud/7IgHNTU2erO11Q7tM2+eXZfl\nja0j1kFTWxOdsU7KQ+WU+8dWh/GuRBev73mda5+8NiN4LtFklMsfuZzfvfY7SoOlmY7K+5pKAiWU\nFpVSFiyz2T78oSF521VLVFFGByp6Y4VYzKYP27XLujNvvBH++lfrwvzOd6xLE2z0ZjQKU6faFGJZ\nxrSLJqNsbt/M7vBuSoOlo77dzhjDzvBOXtv9Gq/tfo1Xd7/Ka7tfY2Prxh7pqXqTSCfoiHewI7yD\nrkRXZoqlYv0e0xtBsotjsCQjjF5RLQuWWeEMllJaZLe9sPUFVrywInNetUQVZeQQbzqiscDChQvN\n6tWrR7oaw0c6bYWuqcm2y/3iF7ZjeTAIX/oSLFpkhc0dCaGy0nZDyDISQjKdZHvndra0byHgC1AR\nyt4vbyRJpBJsaN2QETZ32hvZm9lnasVUDqk/hEMmHML8CfO55olr2Bne2aesKRVTeOzjj/VZn0qn\niCQjhONhIslID0Hsb4ok7H7hRM9jIokI4USYrkQX8dTgEgn5xc8h9YdQFaqiMlRJZaiS6uJqKkOV\n3euKK6kOOeuKqygLlg15O4taocpI0hJp4aipRx1wZiERecEYs3Bf+6mlN5rp7LSuzEgEXnrJ9rlr\naoL3vhcuu8wGpeQwEkLapNnTtYem1iZSJkVVcdWoSMrbEevoYbm9tvs13tj7RkY8ivxFzK2dy4mN\nJ2YEbt6EeZkx81wiiUiPNj2A4kAxFx97cdbz+n1+yovKhzz/ZzKd7CmOzvw5vz0n6/4pk6K2pJb2\naDtbO7bSFmvr0SE6a93FT0WowgphcWVGIKuKq3qIZbZ1pcHSPoKp7aFKoaGiNxpJJGyQyvbt1nq7\n5Rb485+tBXfPPbajOVgx7OqybsypU7OOhNAWbWNj20aiiSgVoYoR6VxujGFLxxYrcLte5bU99nNL\nR3e3zZriGg6pP4Rz3nRORuBmVc/qt5OwF/fhPNLWims997agp1RMYWvH1j77T6mYwl1n3NVjnTEm\nkw2kLdZGW7QtM98ea6ct2tY973xubtucmU8PkNDIHckiI4TFlTy/5fms7aHXP3U9s2pmUVFkr6ei\nqCKn70JRRjvq3hxNrFxpE0Bv3gwTJ8JRR8Gjj9pozc98Bj7xCdvPLpm0YlhSYpNDl/e1WLoSXWxu\n20xLtIWyYBmhQPZO6IMhFzdYLBlj7d61fdyTHfEOwLaRNVY3Mn/CfCtu9fOZXzefiWUTx2V4NPS1\npsBaostOXDakwmyMIZwIZ4SxLdZGe7Sd9lg7rbFW2qPdQtkea6c12srLu17OufySQElGACtDlZQX\nlff49ApkZaiyz77ZLM1cUPfr+Ga43ZsqeqOFlSth8WJruXk55BD43vdsHzvomT5swoQ+6cPiqTjb\nOraxrXMbRf6iIXPhZX1w+4s5983nUl1czau7X2XN7jWsa1lHytjBU0uDpcyrm8f8CfMzIje3bi6l\nwfyMvD6aGa0P7hN/fGJWK7SupI5r33WtHb0+1kFH3H62x9t7LLvz7bF2EulEljN04xd/tzA6glgR\nqqCyqJLyUDmVRb2EMlTOi9te5I7Vd/QIPsrHC4Mycqjo7YNxK3qNjba9rjdTpsBjj9nozc5OawFO\nn95nJIRUOmWTQrc1IciQj3rQ38PRZXL5ZObXzbeWmyNwM6pmjIq2Q6V/htIKjSVjtMfauwXREcMe\nIums64x39tm3M96Z87mCviDHTjuWquIqaoprqApVUV1cTVWx/XSnqlAVFaEK/R2OYjSQpVDZtCn7\n+m3bbJqx4mI47DAbnenBGENLpIWmtiYSqQQVoYq8dC7f1rGt323PfOIZaktGd7cHJTtD2R4aCoSo\nD9RTX1a/X3VJpVOEE+EeQvmx334s676JdILWaCtNbU20Rltpj7Vn3Q9sCiw3MrY6VN1DHDMi2Wt9\ndXF1TpGyo9WCHwt47930qulcd9J1LDp8Ud7Pq6I3Wpg2zbbl9WbSJJgxw372cmV2xjvZ1LqJ9ng7\n5UXllBXlNsr5YDDG8POXf97v9ikVU1TwxjhnzDtjVDyo/T5/JtDGZaAgoF/9z68yy6l0KhP40xpt\npS1mP93JXW6LtrGraxdv7H2D1mgr4US43/oEfIG+FqRHHJvamli1ZlXGrbu1YytLHl1CR7yD9817\nH8WB4hEflWS0ivLv1/yepY8tzXgYNrVtYvGqxQB5Fz51b44GjIGLLrKDvHopLoY77oDzzuuxOpaM\n0dzezM7wzkyn6HywM7yTJY8u4cmmJ5lbO5dNbZu0bUUZVvIdBJRIJTJi2RJtseIYGVg0W6OtRJKR\nnMoP+AKE/CFKgiWE/CGKA8VZp1Cg1zZ/McVB+xkKhDKjvLvzffYPFPfJHpTLvUubNLFkjGgySjwV\nJ5qMEkvZ5Vgylpn3bosl7RRNRTPzvffLeoyn3P7af2dWzWTjlzfu13epbXpjiR074L//G15/3Y5+\nsGOH7YJwww2287lDMp20SaHbm/H7bFBAviIeH1z7IEsfX0okEeHS4y/lo4d/lD++/sdR+daojG9G\no7USS8Z4851v7jcj0GXHX5Z5+LtTLBkjkoxkBCOaiNrPLPvsL65ghgIhdnftzgSVefGJj9Jg6YDi\nM5jzhfwhQoFQRtSL/EUZgXa3uXVy1931j7uylicI6av3bxxxbdMbK8TjcPvt8K9/2c7np5xiozLd\nwV6xLsY9XXtoast/5/KOWAfLnlzG79b8jgUTF/DNk7/JnJo5wOhxgymFxWj83YUCIRoqGvp1vV5w\n5AX7XbYxpofF1GfyCKZrpWWbfv3qr7OWnzZpPnjIB/sIUQ/BcpcDRVn3Kw4UE/QF9/ul+49v/DHr\nvZtRlf9BwlX0Rpr//Md2STjySDjzTNvhfNq0zOb2WDsbWzcSSUQoLyrPawfh57Y8x2UPX8aOzh18\n7qjP8ZmFn9EOyYrSDxcfe/GgMgHliohkRKaK7ONe5sIzzc/0K8pL3r7kQKp4wGS7d6XBUpaftDzv\n59Y43pGkvd0OBdTWBldfbbskzJoFwSCRRITX97zOyztt5+Gakpq8CVAsGePGv93Iub89l6AvyE8/\n9FO+eMwXVfAUZQDOmHcGy05cxpSKKQjClIopo6qN++JjL6Y40HM4saEQ5aGg972bUTWDFWesGJbo\nTW3TGynSaZs8etEiOOcc+MpX7OCuhx3GjvBONrRsoCgwdJ3L++O13a9xyV8u4fW9r3P2grO57PjL\nxmzn8bRJk0gleoww7Y7G7HXDuINjerdl209RxjqjsT20N9pPr1DYts224dXVwRe/aDOxHH44XckI\nG1o2UFVcldfBXFPpFHe/eDe3Pnsr1cXVrDh9Be9sfGfezjfUuFFn8VSctElnhKyiqLufosGQTqcx\nzl86nSaNFUSTtiNNG0xm1Gl3HgE3PkFEMMZYUfSsd+cNxgplr3lc7XTm3TLcffzixyc+/D776RMf\nfvGr6CpDymhsDx1pVPRGgmgUvv99eOUV+OY37brJk0mXlbJ2538oCZbkVfA2t2/maw9/jdVbV3PK\n7FO45sRrRnVfu1Q6RSwVy4y+YIzJJHeeVD4pM9hrkb9oSETDGJOJyss273pH9jWfrYy0SZNKp4in\n4sRTcRLpBPFk93zapLOKqMFkBNIVTK9oKoqSGyp6I8GLL9r+d0cfDe95j7Xypk1ja8dWooko1SXV\neTmtMYbfvPoblv11GYJw47tv5Mx5Z44q6yKZTlpBSDoChyHoD1JRVEFDeYMVuIAVuHzhdXcyzLfG\nFcWUSfX4TJt0Rijjqbi9T+k4kViEFCnESMaidK/BGNPHmnRFM58vVYoymlHRG25aWuC66yAchqVL\nbQLp2bPpNDE2t22mpqQmL6fdG9nLVY9dxcPrH+aoKUdx47tvZGrl1LycK1cSqUTmIe4+pEP+kE1C\nXFGZ6dBbSAE1PvHh8/sIkvs1G2OyimTKpHrcY/eFIpKKkEgleliRXoL+IEX+ogMKSVeU0YqK3nCS\nTML998OqVXDBBbZrgs9Hqq6WdbtepqyoLC+uqsc2PMaSR5fQHmvn0rdeynlHnDfsb/ped54Y27ZV\nHCymuriaiqIKK3CB0IinbRqLiAgBCQzq3rntmL3FMp6K0xnvJBwP9xyfTyAgAYL+IEFfUC1FZcyi\nT5jhpLkZrr/e5tH87GczwSvNHVuIpWJUFw+tWzMcD3PDUzfwy1d+yby6edxz5j3MmzBvSM/RG2OM\nbadKxTPWhDGG0mApdSV1VIQqMh1c9cE5crjBNH780OtrcBNGe7/LeCpuR4SPhwknwj1GdxckI4aF\nZJUrYxMVveGiq8t2Ql+zxo6EbgxMmkR7MM3Wlq1DHkjyj23/4NKHLqW5vZkL33IhXzrmS0PeDuZm\njnBdZ4JgxFAWKGNCyQQrcE4WBxW4sYeIUOQvyvxuvL/RTNurkzmkM95JOBGmM9rZ3R5qIOAPqKtU\nGVWo6A0HxsDq1bBiBbz1rXDyyRAOk5wymbV7XqMiNHQ5NOOpOLc9dxsr/rGChvIG7v3AvRw19agh\nKRvsw64z3mlD8EUoLypnUskkyovKM3n2NJpw/BPwWXdq7z6dbl9JN8lwV6IrM1xQJpoVo65SZcRQ\n0RsO9u61bs1YDK66KpN5ZVNkOymTGjILbO3etVzy0CW8susVPnjIB1ny9iVD2rk9mowSSUSYXTM7\n46bUt3fFi098mRRahLrXG2N6WIeuGIbjYRLpRCbq1Cc+gv4gAV9ArUMlL6jo5ZtEAn79a/jzn+HT\nn7YjoQMtZX527N0xJG7NtElz77/v5aanb6IsWMb3T/s+J885+YDL9dIZ7wQDCyYuyMu4fcr4RsRp\n9/MHKaOsR5Syt9+i11XaGmvFG1ga8AX6iKCblGCgeXfZrYdm4hkc3v6p/fVBTZt0v31avX1YEfp0\nrxluz5CKXr7ZsAFuvNEOFfSpT0FXF/H5c1nXumFI3JrbOrZx+SOX80zzM5zYeCLXnnjtfo9c3R8t\n0RbKg+XMrZub1/5xSmHi9/kp8ZVQEiyhiiomMQno2WbsjsfmRpOmTbrHvPcBm04760ln5lOk7AM5\nTZ8yej+Uvdl4+mTa8ezTO3lA5pgsZPpPZun4mSmnHzIZgbJtG+C8Pc7pXtc+Pnt0X3HWeTMI+cSH\niODDh8/n9P30+RCku0+ou42efUTdl4zen8OdYEFFL590dtrglfXr4bbbIJnE1NezMbUnEySwvxhj\n+MPrf+CaJ64hZVJce+K1/L9D/9+QvrGmTZqWSAuTyicxs2qmtr0ow4p3tIGKUEXez5drph133j2m\nh2j2V3Y/4+7t73G5HO89NpvY7OvTFarxhopevkin4Zln4O674YQT4J3vhHCYvXWl7OnYSF1p3X4X\n3Rpt5ZonruGBNx7gyMlHcuO7b2Rm9cyhqzs2IKYj1kFjdSOTyyePyx+/onjxukaHOxOPMnyo6OWL\nXbusWzOVgiuvhM5OYtMbWBfeTFXx/o+R9dSmp7j8kcvZG9nLRcdexIVvuXDIO3RHEhFiqRiHTDgk\nbynRFEVRRgIVvXwQi8H//R888gh84QtQX49Jp9kYCBNIDS5zhkskEeGmp2/ivpfuY07NHO58750c\nNvGwIa96R6wDv/hZMHHBmB1iSFEUpT9U9PLBG2/At74FM2bAhRdCOMyu2ZNoie2ktnTw0Zov7XiJ\nSx66hA2tGzj3zefyleO+0mdwyAPFGENrtJWqUBVzaudoZg1FUcYlKnpDTVubHTZo0ya46y5IJIjU\nVrIxsXvQbs1kOsmdq+/kjtV3UFdSxz1n3sNbp791yKucSqdojbbSUN7AjOoZ2rlcUZRxi4reUJJK\nwVNPwY9/DKecAscfT7q9jfXl/pwyT3hHOa4vrScUCLG5fTOnzz2dpe9cekBtgf3hJhieUzuHiWUT\nh7x8RVGU0YSK3lCyY4cdFFYErrgC2trYMbGUThOhpmjgIYNWrVnFlY9dSTQZBWBn104AFi1YxNIT\nlualuuG4TRx8aP2hVIYq83IORVGU0YT6sYaKSAR+/nN48kk7gkJtLeEgNAU6c7LQbn725ozgeXms\n6bF81Ja2aBs+n4/DJx2ugqcoSsGglt5QYIwdPeHmm2HOHDjvPFLtrayb5KMkxzHytnVsG9T6/a+q\noSXaQm1JLbNrZuv4dYqiFBT6xBsKWlps5pUtW2x7XjTKtjJDNBSgOliSUxEN5Q1s7dzad31Fw5BV\nM5lO0hZtY1rlNKZVTtMO54qiFBzq3jxQkkn461/hvvvg9NPhqKPoiHfQXGEGFXhyzLRj+qwrDhRz\n8bEXD0k1Y8kY7dF2Dq47mOlV01XwFEUpSNTSO1Cam23mlWAQLruMZOte1lWmKSubkLOwNLc38+C6\nBzm49mA6451s69xGQ0UDFx97MWfMO+OAq+iOf7dg0oIhHWpIURRlrJFX0RORU4FbAT/wA2PMDVn2\nOQH4DhAEdhtj3pnPOg0p4bANXnnmGbj8cqiqorljM/GqWqoDoX0fj21ju+qxqwC48/Q7mVo5dUir\n2BptpTRYytzauXaMM0VRlAImb6InIn7gNuBkoBl4XkTuN8a84tmnGrgdONUYs0lExk5HMWPg5Zfh\n1lth3jw45xza92xlW30RtSUDd0/w8qtXfsXTm5/m6ndePaSClzZpWiOtTCibwKzqWTpCgqIoCvm1\n9I4G1hpj1gOIyM+BM4FXPPt8FPiNMWYTgDFmZx7rM7Ts2mUzr2zfDt/+NolwB2uLwlTUzsnZrbm9\nczs3/O0Gjp56NGcvOHvIqpZIJWiPtTOzaiYNFQ3afqcoiuKQz0CWqcBmz3Kzs87LwUCNiDwuIi+I\nyLl5rM/QEY/D44/Dz34GH/gAHHkkmzq3kJpYn/MYecYYlj62lGQ6yfJ3LR+y1F/RZJRwPMz8CfOZ\nUjlFBU9RFMXDSAeyBID/Ak4CSoBnRORZY8zr3p1EZDGwGGDGjBnDXsk+bNpkE0qXlsIll7B392Z2\nVhdRVzkh5yJ+v+b3PNH0BJe/7XJmVA3NNXXEOvCJjwWTdIQERVGUbOTT0tsCTPcsT3PWeWkGHjTG\nhI0xu4EngTf3LsgYs8IYs9AYs7C+vj5vFc6J9nb46U9h9Wr48peJl4ZYl9xFZUPug7juCu/iur9e\nx5GTj+Rjb/rYAVfJGENrxAasHDbxMBU8RVGUfsin6D0PzBWRWSJSBJwN3N9rn98DbxORgIiUAscA\nr+axTgdGOg0vvWTb8g47DHPWWWzc9QYydRrBYO7Rmtc8cQ2RZITlJy0/4ACTVDrF3she6svqmTdh\nXs7uVUVRlEIkb+5NY0xSRD4PPIjtsnC3MeZlEfm0s/1OY8yrIvJn4N9AGtut4T/5qtMBs2OHFbzd\nu+H229nTupU9JVBXl3vWlD+t/RMPrX+Ir771q8ypmXNA1Ymn4nTGOpldM5uJZRO1/U5RFGUf5LVN\nzxjzAPBAr3V39lr+FvCtfNZjSIhG4bHH7Ijo//M/xObPZf2WF6ha8Jaci9gb2cu1T17LgokLOP+I\n8w+oOl2JLuKpOIfUH5KXIYcURVHGI5qGLFeammzwSkUF5stfZv3O1wg0TCVQnHv72bInl9ER6+D6\nk64/oETP7dF2RITDJx6ugqcoijIIVPRyoaXF5tZ88UX46lfZGYjSSpTyyblHXT68/mH++MYf+cxR\nn+HguoP3qxrGGPZ27aUiVMGhEw6lJMdk1oqiKIpFRW9fJJPwr3/B7bfDkUcSOfO9bNy1lurGQ8Cf\nWxBKa7SVrz/+deZPmM/ityzer2q4AStTKqcwt24uQX9wv8pRFEUpZEa6n97oZ9s2G7zS2kr6hz9g\n3e7XKaqZgL+qOucibnjqBvZG9rLijBX7JVbxVJyOWAcH1R5EfdkId9lQFEUZw6ilNxBdXfDII/Cb\n38CiRWyfXkNnrJ2yGblHXT7R9AS/fe23fPK/Psmh9YcOugrheJhoIsqCiQtU8BRFUQ4QtfT6wxhY\nvx6+/W2oqyP8uU+yaffrVE+bC6Hc+uR1xDpY+thSDqo9iM8d9blBV6E10kpxsJhD6w/VERIURVGG\nABW9/tizB1auhP/8h/SNN7A2sZOSUDm++twHgvjW099iZ3gn3z31u4PqNJ42aVqjrdSV1DGrZtYB\nRXoqiqIo3ah7MxuJhI3UvOMOOPpotrz7GKKdrZQ0HpRz8Mozm5/hFy//gvOOOI83T+6TWW1AWiIt\nTK+czkG1B6ngKYqiDCE5i56IvE1Eznfm60VkVv6qNcI0N9vglXCY8BVfpXnvRqrrpkJFRU6Hh+Nh\nljy6hMaqRr50zJcGderOeCfVxdVMqdAREhRFUYaanMwIEbkaWAjMA+7BjnJ+H3B8/qo2QnR2wsMP\nw+9/T/qC83m93k95pAiZmvsAr7c8ewtbO7Zy3wfvozhQnPNxqXSKRCrBofWHquApiqLkgVwtvQ8A\n7wPCAMaYrUBuZs9YIp2GN96Am2+GSZPYfP6HSHS2EWqYnnPwyuqtq7n33/dyzpvOYeGUhYM6fXus\nnRlVMwYllIqiKEru5Cp6cWOMAQyAiJTlr0ojyK5dNnjltdcIX3IRW1OtVBVXw4TcxsmLJCJc8cgV\nTKucxsXHXTyoU0eTUYqDxUwqn7Q/NVcURVFyIFfR+6WI/C9QLSKfBB4G7spftUaAWAz+8Q9YsYL0\nW4/jteMOoiIGMmNGzsEr3/37d2lqa2LZu5YNakw7Ywxd8S7m1MwZshHUFUVRlL7k1KZnjLlJRE4G\n2rHtekuNMQ/ltWbDTVOTDV6JRtn8lcWYSISimgk5B6+8uP1FfvSvH3HWYWdx3LTjBnXqtmgbDRUN\nlBeV70/NFUVRlBzZp+iJiB942BhzIjC+hM6lrc0GrzzwAJFPns/WiSHq4gHIMXglnopzxSNXMLFs\nIpcef+mgTh1PxQn4AkytzD1QRlEURdk/9ulLM8akgLSIjM8xbFIpWLMGbrkFM2UKr3zkJCqjwOTJ\nOQev3PbcbaxrWce1J147aGutI9bB7NrZ2h9PURRlGMj1SdsJvCQiD+FEcAIYY76Yl1oNJ9u322GD\n1q5l601Xgz9AMBDKOXjl5Z0vc9c/7uKD8z/IO2a+Y1Cn7oh1UFdaR3Vx7smrFUVRlP0nV9H7jTON\nLyIRG7zywx8Sf/tb2XzMfGqjwJxpOQWvxFNxLn/kcmpLavna2742qFMn00lSJsXMqpn7WXlFURRl\nsOQayPJjESkC3NFP1xhjEvmr1jBgjA1e+d73MKkUr37hI1QmA1BVDpWVORWx4oUVrNmzhtvfe/ug\nRzDviHUwq2aWJpJWFEUZRnLNyHIC8GNgIyDAdBH5uDHmyfxVLc+0tMBf/gIPPcTuxYtITWugJJKG\nKVNyOnzN7jXcufpOTp97OifNOmlQp+5KdFFeVE59qQ4VpCiKMpzk6t78NnCKMWYNgIgcDPwM+K98\nVSyvGAOvvw633EJy+lTWn30qNZE0TJoExfvOhpJMJ7n8kcupCFWw5B1LBnXqtEkTTUQ5ePLBmmpM\nURRlmMm1J3TQFTwAY8zr2PybY4+VK6GxEY47DjZuZMfb/4vKUAUEAlCfm+V19z/v5uVdL7P0nUup\nLakd1OnbYm1Mr5o+qM7riqIoytCQq6W3WkR+gE0yDbAIWJ2fKuWRlSth8WI7IrpDw28exDdzBh0X\nnptT8Mq6lnV877nvccrsUzh1zqmDOn0sGSPkCzG5fPKgq64oiqIcOLlaep8BXgG+6EyvOOvGFkuW\n9BA8AF80xoR7fplT8EoqnWLJI0soDZSy9J1LB+WeNMbQGe9kdu1s/L7c0popiqIoQ0uull4AuNUY\nczNksrSMvbDDTZuyrg7s2AU5CNi9/76Xf27/J988+ZvUlw0uCKU91s6k8klUhnKLDFUURVGGnlwt\nvUeAEs9yCTbp9Nhixoysq5NTG/Z5aFNrE7c8ewsnNp7I+w5+36BOm0glEITpldMHdZyiKIoytOQq\nesXGmE53wZkfe5EYy5dDac9qp0uK2X3lwMMApU2aKx+9koAvwDUnXDPoqMv2WDuza2YT9I/N2B9F\nUZTxQq6iFxaRt7gLIrIQiOSnSnlk0SJYsQIzYwZGIDFlMjtuWUbHh88Y8LCf/ednPLf1Ob72tq8N\nery7zngntSW11JTUHEjNFUVRlCEg1za9LwP/JyJbneUG4Kz8VCnPLFqE+ehHeP6Vh6iZ1LjPtrzm\n9mZuevomjp9+PB8+5MODOlUqnSKRSjCzfqb2yVMURRkFDGjpichRIjLZGPM8MB/4BZAA/gxsGIb6\n5Y+q6n0KnjGGpY8tBeDaE6/dL7dmY3UjxYF9d3hXFEVR8s++3Jv/C8Sd+eOAK4DbgBZgRR7rNSr4\n1au/4m+b/8Ylb71k0OPdRRIRSoIlg47yVBRFUfLHvtybfmPMXmf+LGCFMebXwK9F5MX8Vm1k2dG5\ngxueuoGjpxzN2QvOHtSxxhi6El28adKb8EmuzaaKoihKvtnXE9kvIq4wngQ86tk2bkc9dd2ayXSS\nZe9aNmjhaou1Ma1yGmVFZXmqoaIoirI/7Eu4fgY8ISK7sdGafwUQkYOAtjzXbcS4//X7ebzpcS5/\n2+XMrB7ceHfxVJyABGio2HffP0VRFGV4GVD0jDHLReQRbLTmX4wxxtnkA76Q78qNBLvCu1j+5HKO\nnHwkH3vTxwZ9fEesg0PrDyXgG7eGsKIoyphln09mY8yzWda9np/qjCzGGK554hoiyQjLT1o+6ByZ\nHbEO6svqBz2grKIoijI8aJSFhz+v+zMPrX+ILxz9BebUzBnUscl0krRJM6Mqe6ozRVEUZeRR0XPY\nG9nLN574BgsmLuCCIy8Y9PFt0TZmVc+iyF+Uh9opiqIoQ4GKnsPyJ5fTEevgunddN+j2uHA8THVx\nNXWldXmqnaIoijIUqOgBD69/mD+88Qc+vfDTzJswb1DHpk2aeCpOY3WjphpTFEUZ5RS86LVF2/j6\n4yR5h3MAABmGSURBVF9n/oT5fOq/PrVfx0+vnE5JsGTfOyuKoigjSl5FT0ROFZE1IrJWRL42wH5H\niUhSRAaX0XkIuOGpG9gb2ct177pu0EP/RJNRioPFTK6YnKfaKYqiKENJ3kTPGV39NuA04FDgIyJy\naD/73Qj8JV916Y8nmp7gN6/9hk++5ZMcNvGwQR1rjCEcDzO7eramGlMURRkj5PNpfTSw1hiz3hgT\nB34OnJllvy8AvwZ25rEufQjHwyx9bClzaubwuaM/N+jj26JtNJQ3UBGqyEPtFEVRlHyQT9GbCmz2\nLDc76zKIyFTgA8AdeaxHD1a+tJJZt87ipHtPYnvndt4z9z2D7mYQT8Xx+/xMq5qWp1oqiqIo+WCk\n/XLfAS4zxqQH2klEFovIahFZvWvXrv0+2cqXVrJ41WI2tW3KrLvrH3exas2qQZXTEetgds1sTTWm\nKIoyxsin6G0BpnuWpznrvCwEfi4iG4EPA7eLyPt7F2SMWWGMWWiMWVhfv//j0y15ZAldia4e66LJ\nKDc/e3POZXTEOqgrraOmpGa/66EoiqKMDPk0VZ4H5orILKzYnQ181LuDMWaWOy8iPwL+YIz5Xb4q\n5LXwvGzr2JbT8al0ipRJMbNqcCMvKIqiKKODvFl6xpgk8HngQeBV4JfGmJdF5NMi8ul8nXcg+suL\nmeswQG2xNhqrGwkFQkNZLUVRFGWYyGujlDHmAeCBXuvu7Gff8/JZF4DlJy1n8arFPVycxYFiLj72\n4n0e25XooqKogvrS/XevKoqiKCPLSAeyDCuLDl/EijNWMKNqBoIwpWIKy05cxhnzzhjwuLRJE01G\nmVUzS1ONKYqijGEKLvxw0eGL+MiCj/D8ludzDkZpi7UxrXIapcHSPNdOURRFyScFZentD7FkjJAv\nREN5bu1+iqIoyuhFRW8AjDF0xjuZXTt70KOoK4qiKKMPFb0B6Ih1MKl8EpWhypGuiqIoijIEqOj1\nQyKVAGB65fR97KkoiqKMFVT0+qEj3sGsmlmDHm5IURRFGb2o6GWhM95JTXENtSW1I10VRVEUZQhR\n0etFKp0imUoys3qm9slTFEUZZ6jo9aI91s70qukUB4pHuiqKoijKEKOi5yGSiFASLGFS+aSRroqi\nKIqSB1T0HIwxRBIRZtfMxid6WxRFUcYj+nR3aIu10VDRQHlR+UhXRVEURckTKnpAPBUnIAGmVk4d\n6aooiqIoeURFD5t5ZU7tHAK+gsu/rSiKUlAUvOi1x9qpL6unqrhqpKuiKIqi5JmCFr1kOknapPsd\nUV1RFEUZXxS06LVH25ldPZsif9FIV0VRFEUZBgpW9IwxVIYqqSutG+mqKIqiKMNEwYpeWVEZs2pm\naaoxRVGUAqIgwxV94uPguoMJBUIjXRVFURRlGClYS08FT1EUpfAoWNFTFEVRCg8VPUVRFKVgUNFT\nFEVRCgYVPUVRFKVgUNFTFEVRCgYVPUVRFKVgUNFTFEVRCgYVPUVRFKVgUNFTFEVRCgYVPUVRFKVg\nUNFTFEVRCgYVPUVRFKVgUNFTFEVRCgYVPUVRFKVgUNFTFEVRCgYVPUVRFKVgUNFTFEVRCgYVPUVR\nFKVgyKvoicipIrJGRNaKyNeybF8kIv8WkZdE5GkReXM+66MoiqIUNnkTPRHxA7cBpwGHAh8RkUN7\n7bYBeKcx5nDgWmBFvuqjKIqiKPm09I4G1hpj1htj4sDPgTO9OxhjnjbGtDiLzwLT8lgfRVEUpcDJ\np+hNBTZ7lpuddf3xCeBPeayPoiiKUuAERroCACJyIlb03tbP9sXAYoAZM2YMY80URVGU8UQ+Lb0t\nwHTP8jRnXQ9E5E3AD4AzjTF7shVkjFlhjFlojFlYX1+fl8oqiqIo4598it7zwFwRmSUiRcDZwP3e\nHURkBvAb4GPGmNfzWBdFURRFyZ970xiTFJHPAw8CfuBuY8zLIvJpZ/udwFKgDrhdRACSxpiF+aqT\noiiKUtiIMWak6zAoFi5caFavXj3S1VAURVFGESLyQi5Gk2ZkURRFUQoGFT1FURSlYFDRUxRFUQoG\nFT1FURSlYFDRUxRFUQoGFT1FURSlYFDRUxRFUQoGFT1FURSlYFDRUxRFUQoGFT1FURSlYFDRUxRF\nUQoGFT1FURSlYFDRUxRFUQoGFT1FURSlYFDRUxRFUQoGFT1FURSlYFDRUxRFUQoGFT1FURSlYFDR\nUxRFUQqGwEhXQFEUJV8kEgmam5uJRqMjXRVliCguLmbatGkEg8H9Ol5FT1GUcUtzczMVFRU0NjYi\nIiNdHeUAMcawZ88empubmTVr1n6Voe5NRVHGLdFolLq6OhW8cYKIUFdXd0CWu4qeoijjGhW88cWB\nfp8qeoqiKHliz549HHHEERxxxBFMnjyZqVOnZpbj8XhOZZx//vmsWbNmwH1uu+02Vq5cORRV5vf/\nv717j66qvhI4/t0QNAkBQlSkyJiwOgh5J5c08miEGHnojCIYhACDASWYIhQ6rCm1LrXMQrOsi4La\nxUN5iI0wCqXQVlCLIGWoQKAhIIEJ1WCR8F4GITQ1uOePe3NNSELCTS43l7s/a2XlnHPP+Z19Tlay\n8zuP316/nqSkJBITE4mJieGNN95okXZbC7unZ4wx1fLz4ec/hy++gDvvhLlzYdw4j5u75ZZbKCws\nBOD5558nLCyMWbNm1VpHVVFV2rSpvw+yfPnyRvczdepUj2OsqbKyktzcXAoKCujWrRuVlZUcPXq0\nWW02dnzXW+uIwhhjfC0/H3Jy4OhRUHV+z8lxLm9hR44cISYmhnHjxhEbG0tZWRk5OTmkpKQQGxvL\nnDlz3Ov+8Ic/pLCwkKqqKsLDw5k9ezaJiYn069ePU6dOAfDMM88wf/589/qzZ88mNTWVXr16sWPH\nDgAuXrzII488QkxMDJmZmaSkpLgTcrXy8nJUlYiICABuvvlm7rrrLgBOnDjB8OHDSUhIIDExkZ07\ndwLw0ksvERcXR1xcHK+++mqDx7dx40b69euHw+Fg9OjRXLx4scXPa1NYT88YExhmzIAr/sjX8skn\nUFlZe1lFBTz+OLz+ev3bJCWBK9lcq0OHDrFy5UpSUlIAyMvLIyIigqqqKtLT08nMzCQmJqbWNuXl\n5QwcOJC8vDx+8pOfsGzZMmbPnl2nbVVl165dbNiwgTlz5rBp0yZeffVVunbtytq1a9m3bx8Oh6PO\ndl26dGHo0KFERkaSkZHBgw8+yOjRo2nTpg1Tp05l8ODBPPXUU1RVVVFRUcHOnTvJz89n9+7dVFVV\nkZqayqBBgwgJCal1fKdOnSIvL4/NmzcTGhrK3LlzWbBgAU8//bRH5645rKdnjDFQN+E1tryZvv/9\n77sTHsCqVatwOBw4HA6Ki4s5ePBgnW1CQkK4//77AejTpw+lpaX1tj1y5Mg662zfvp0xY8YAkJiY\nSGxsbL3brlixgg8//JCUlBTy8vLIyckBYOvWrUyZMgWAoKAgOnbsyPbt23nkkUcICQmhQ4cOPPzw\nw/z5z3+uc3w7duzg4MGD9O/fn6SkJPLz8xuM3dusp2eMCQyN9ciiopyXNK8UGQlbt7Z4OO3bt3dP\nl5SUsGDBAnbt2kV4eDjjx4+v97H8m266yT3dtm1bqqqq6m375ptvbnSdq0lISCAhIYGxY8cSHR3t\nfpjlWp6crHl8qsqwYcN46623rjmWlmY9PWOMAedDK6GhtZeFhjqXe9n58+fp0KEDHTt2pKysjPff\nf7/F9zFgwADeeecdAPbv319vT/L8+fNs27bNPV9YWEhkZCQA6enpLFq0CIDLly9z/vx50tLSWLdu\nHZcuXeLChQusX7+etLS0Ou3279+fjz/+mM8++wxw3l8sKSlp8WNsCuvpGWMMfPeUZgs+vdlUDoeD\nmJgYevfuTWRkJAMGDGjxfUybNo0JEyYQExPj/urUqVOtdVSVF198kcmTJxMSEkJYWBjLli0D4LXX\nXmPy5MksXryYoKAgFi9eTGpqKllZWfzgBz8AIDc3l/j4eI4cOVKr3dtvv52lS5cyevRo96saL7zw\nAj179mzx42yMqOp132lzpKSkaEFBga/DMMb4geLiYqKjo30dRqtQVVVFVVUVwcHBlJSUMGTIEEpK\nSggK8r++T30/VxHZo6opDWzi5n9Ha4wx5ppduHCBjIwMqqqqUFV3jy3QBN4RG2NMAAoPD2fPnj2+\nDsPn7EEWY4wxAcOSnjHGmIBhSc8YY0zAsKRnjDEmYFjSM8YYL0lPT6/zovn8+fPJzc296nZhYWEA\nHD9+nMzMzHrXGTRoEI29vjV//nwqKirc8w888ABfffVVU0K/qsOHDzNo0CCSkpKIjo52D1XmDyzp\nGWOMS/7+fKLmR9HmF22Imh9F/v7mVVjIyspi9erVtZatXr2arKysJm3frVs31qxZ4/H+r0x67733\nHuHh4R63V2369OnMnDmTwsJCiouLmTZtWrPbvHz5crPbaApLesYYgzPh5fw+h6PlR1GUo+VHyfl9\nTrMSX2ZmJn/84x/do5CUlpZy/Phx0tLS3O/NORwO4uPjWb9+fZ3tS0tLiYuLA+DSpUuMGTOG6Oho\nRowYwaVLl9zr5ebmussSPffccwC88sorHD9+nPT0dNLT0wGIiorizJkzAMybN89dEqi6LFFpaSnR\n0dFMnjyZ2NhYhgwZUms/1crKyujevbt7Pj4+HnAmrlmzZhEXF0dCQoK71NDmzZtJTk4mPj6eSZMm\nUekaxDsqKoqf/vSnOBwO3n33Xf72t78xbNgw+vTpQ1paGocOHfL43DfEq+/picgwYAHQFnhDVfOu\n+Fxcnz8AVADZqrrXmzEZYwLTjE0zKDzRcGmhT459QuXl2hUVKr6p4PH1j/P6nvpLCyV1TWL+sIYH\nso6IiCA1NZWNGzcyfPhwVq9ezaOPPoqIEBwczLp16+jYsSNnzpyhb9++PPTQQw0O6rxw4UJCQ0Mp\nLi6mqKioVmmguXPnEhERweXLl8nIyKCoqIjp06czb948tmzZwq233lqrrT179rB8+XJ27tyJqnL3\n3XczcOBAOnfuTElJCatWreL111/n0UcfZe3atYwfP77W9jNnzuTee++lf//+DBkyhIkTJxIeHs6S\nJUsoLS2lsLCQoKAgzp07xz/+8Q+ys7PZvHkzd911FxMmTGDhwoXMmDEDcBba3bvX+Wc/IyODRYsW\n0bNnT3bu3MmPfvQjPvroowbPrye81tMTkbbAr4H7gRggS0RirljtfqCn6ysHWOiteIwx5mquTHiN\nLW+qmpc4a17aVFWefvppEhISuO+++/jyyy85efJkg+1s27bNnXyqqyBUe+edd3A4HCQnJ/Ppp5/W\nO5h0Tdu3b2fEiBG0b9+esLAwRo4c6S4J1KNHD5KSkoCGyxdNnDiR4uJiRo0axdatW+nbty+VlZX8\n6U9/YsqUKe6RXiIiIjh8+DA9evRwF6N97LHHag1qPXr0aMA5YsyOHTsYNWoUSUlJTJkyhbKysqse\nhye82dNLBY6o6mcAIrIaGA7U/GkMB1aqcwDQT0QkXES+p6otf6TGmIB2tR4ZQNT8KI6W1y0tFNkp\nkq3ZWz3e7/Dhw5k5cyZ79+6loqKCPn36AJCfn8/p06fZs2cP7dq1Iyoqqt5yQo35/PPPefnll9m9\nezedO3cmOzvbo3aqVZclAmdpovoub4LzfuOkSZOYNGkScXFxHDhwwKP9VZcg+vbbbwkPD69Tzb2l\nefOe3h3A32vMH3Mtu9Z1EJEcESkQkYLTp0+3eKDGGDM3Yy6h7WqXFgptF8rcjOaVFgoLCyM9PZ1J\nkybVeoClvLycLl260K5dO7Zs2cLR+mr51XDPPffw9ttvA3DgwAGKiooAZzmg9u3b06lTJ06ePMnG\njRvd23To0IGvv/66TltpaWn87ne/o6KigosXL7Ju3bp6SwI1ZNOmTXzzzTcAnDhxgrNnz3LHHXcw\nePBgFi9e7K7hd+7cOXr16kVpaam78sJbb73FwIED67TZsWNHevTowbvvvgs4e8L79u1rckxN5RcP\nsqjqElVNUdWU2267zdfhGGNuQOPix7HkwSVEdopEECI7RbLkwSWMi29+aaGsrCz27dtXK+mNGzeO\ngoIC4uPjWblyJb17975qG7m5uVy4cIHo6GieffZZd48xMTGR5ORkevfuzdixY2uVJcrJyWHYsGHu\nB1mqORwOsrOzSU1N5e677+aJJ54gOTm5ycfzwQcfEBcXR2JiIkOHDuWXv/wlXbt25YknnuDOO+8k\nISGBxMRE3n77bYKDg1m+fDmjRo0iPj6eNm3a8OSTT9bbbn5+PkuXLnVXdq/v4Z7m8lppIRHpBzyv\nqkNd8z8DUNUXa6yzGNiqqqtc84eBQVe7vGmlhYwxTWWlhW5MzSkt5M2e3m6gp4j0EJGbgDHAhivW\n2QBMEKe+QLndzzPGGOMtXnuQRVWrROQp4H2crywsU9VPReRJ1+eLgPdwvq5wBOcrCxO9FY8xxhjj\n1ff0VPU9nImt5rJFNaYVmOrNGIwxxphqfvEgizHGeMpbzy0Y32juz9OSnjHmhhUcHMzZs2ct8d0g\nVJWzZ88SHBzscRtevbxpjDG+1L17d44dO4a933vjCA4OrjXu57WypGeMuWG1a9eOHj16+DoM04rY\n5U1jjDEBw5KeMcaYgGFJzxhjTMDw2jBk3iIip4Grj8zq/24Fzvg6CD9l585zdu48Z+fOcy117iJV\ntdHBmf0u6QUCESloyhhypi47d56zc+c5O3eeu97nzi5vGmOMCRiW9IwxxgQMS3qt0xJfB+DH7Nx5\nzs6d5+zcee66nju7p2eMMSZgWE/PGGNMwLCk14qIyL+IyBYROSgin4rIj30dk78RkbYi8lcR+YOv\nY/EnIhIuImtE5JCIFItIP1/H5C9EZKbr9/WAiKwSEc9HQ77BicgyETklIgdqLIsQkQ9FpMT1vbM3\nY7Ck17pUAf+pqjFAX2CqiMT4OCZ/82Og2NdB+KEFwCZV7Q0kYuewSUTkDmA6kKKqcTgLZo/xbVSt\n2gpg2BXLZgObVbUnsNk17zWW9FoRVS1T1b2u6a9x/uG5w7dR+Q8R6Q78G/CGr2PxJyLSCbgHWAqg\nqv9U1a98G5VfCQJCRCQICAWO+zieVktVtwHnrlg8HHjTNf0m8LA3Y7Ck10qJSBSQDOz0bSR+ZT7w\nX8C3vg7Ez/QATgPLXZeG3xCR9r4Oyh+o6pfAy8AXQBlQrqof+DYqv3O7qpa5pk8At3tzZ5b0WiER\nCQPWAjNU9byv4/EHIvLvwClV3ePrWPxQEOAAFqpqMnARL19iulG47j8Nx/mPQzegvYiM921U/kud\nrxN49ZUCS3qtjIi0w5nw8lX1t76Ox48MAB4SkVJgNXCviPzGtyH5jWPAMVWtvqqwBmcSNI27D/hc\nVU+r6jfAb4H+Po7J35wUke8BuL6f8ubOLOm1IiIiOO+rFKvqPF/H409U9Weq2l1Vo3A+SPCRqtp/\n3E2gqieAv4tIL9eiDOCgD0PyJ18AfUUk1PX7m4E9BHStNgCPuaYfA9Z7c2eW9FqXAcB/4OylFLq+\nHvB1UCYgTAPyRaQISAJe8HE8fsHVO14D7AX24/ybaqOzNEBEVgF/AXqJyDEReRzIAwaLSAnOnnOe\nV2OwEVmMMcYECuvpGWOMCRiW9IwxxgQMS3rGGGMChiU9Y4wxAcOSnjHGmIBhSc+YBoiI1nzBXUSC\nROS0pxUcROQhEfHZSCcislVEDotIkauawmsiEt6M9rJFpFuN+VIRubVlojXGOyzpGdOwi0CciIS4\n5gcDX3ramKpuUFWvvoPUBONUNQFIACpp3ovA2TiH3jLGb1jSM+bq3sNZuQEgC1hV/YGIpIrIX1yD\nNO+oHtHEVV9tmWs63lVnLdTVM3rNtXyFiCwUkU9E5DMRGeSqNVYsIitq7ONCjenM6s+aun1DVPWf\nOAfnvlNEEl1tjheRXa5BERaLSNvqGETkV66acZtF5DYRyQRScL7QXljjH4NpIrJXRPaLSG8Pzrcx\nXmVJz5irWw2McRUGTaB21YtDQJprkOZn+W4UkwXAv4rICGA5MEVVK+ppuzPQD5iJcyimXwGxQLyI\nJDUhtmZtr6qXgX1AbxGJBkYDA1Q1CbgMjHOt2h4oUNVY4GPgOVVdAxTg7Dkmqeol17pnVNUBLARm\nNeEYjLmugnwdgDGtmaoWuco8ZeHs9dXUCXhTRHriHBm+nWubb0UkGygCFqvq/zbQ/O9VVUVkP3BS\nVfcDiMinQBRQ2Eh4zd0eQFzfM4A+wG7nEJKE8N3Av98C/+Oa/g3OQZUbUv3ZHmBkE/ZvzHVlSc+Y\nxm3AWTNtEHBLjeX/DWxR1RGuxLi1xmc9gQtc/Z5Xpev7tzWmq+erfzdrjhMY7MH2DXJdvozHOUBy\nF+BNVf1ZY9tx9dIv1XFcbkoMxlxvdnnTmMYtA35R3ZOqoRPfPdiSXb3QVYn8FZzVyG9x3f/y1EkR\niRaRNsCIZrRTi6uE1YvA31W1CNgMZIpIF9fnESIS6Vq9DVB9DGOB7a7pr4EOLRWTMdeDJT1jGqGq\nx1T1lXo+egl4UUT+Su1eza+AX6vq/wGPA3nVycQDs4E/ADtwVuZurupKCgdw3qsbDqCqB4FngA9c\nn38IfM+1zUUgVUQOAPcCc1zLVwCLrniQxZhWzaosGGOuSkQuqGqYr+MwpiVYT88YY0zAsJ6eMcaY\ngGE9PWOMMQHDkp4xxpiAYUnPGGNMwLCkZ4wxJmBY0jPGGBMwLOkZY4wJGP8Pt+ph6pRJK9YAAAAA\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fb84fced650>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "vs.ModelComplexity(X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 问题 5- 偏差与方差之间的权衡取舍\n",
    "*当模型以最大深度 1训练时，模型的预测是出现很大的偏差还是出现了很大的方差？当模型以最大深度10训练时，情形又如何呢？图形中的哪些特征能够支持你的结论？*\n",
    "  \n",
    "**提示：** 你如何得知模型是否出现了偏差很大或者方差很大的问题？"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**答案: **当最大深度为1时，模型的预测出现了很大的偏差，当深度为10，出现了很大的方差。深度为1时，训练分数和验证分数都比较低，两者差距比较小，说明偏差大方差小。当深度为10，分数比较高，偏差小，但是训练分数明显高于验证分数，方差大。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 问题 6- 最优模型的猜测\n",
    "*你认为最大深度是多少的模型能够最好地对未见过的数据进行预测？你得出这个答案的依据是什么？*"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**答案: **最大深度是3的模型能够最好地对未见过的数据进行预测。3以后随着数据量的增加会出现数据过拟合现象，训练分数高于测试分数。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "-----\n",
    "\n",
    "## 评价模型表现\n",
    "在这个项目的最后，你将自己建立模型，并使用最优化的`fit_model`函数，基于客户房子的特征来预测该房屋的价值。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 问题 7- 网格搜索（Grid Search）\n",
    "*什么是网格搜索法？如何用它来优化学习算法？*\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**回答: **网格搜索算法是一种通过遍历给定的参数组合来优化模型表现的方法。通过找到最优的参数组合来优化算法。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 问题 8- 交叉验证\n",
    "- 什么是K折交叉验证法（k-fold cross-validation）？\n",
    "- 网格搜索是如何结合交叉验证来完成对最佳参数组合的选择的？\n",
    "- 优化模型时，使用这种方法对网格搜索有什么好处？网格搜索时如果不使用交叉验证会有什么问题？为什么？\n",
    "- GridSearchCV中的[`'cv_results'`](http://scikit-learn.org/stable/modules/generated/sklearn.model_selection.GridSearchCV.html)属性能告诉我们什么？\n",
    "\n",
    "**提示：** 在下面 fit_model函数最后加入 `print pd.DataFrame(grid.cv_results_)` 可以帮你查看更多。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**答案: **1.把初始数据采样按一定比例分成训练集和测试集（例如8:2），再把训练数据分割成K个子样本，每次交叉验证中将某个子样本作为验证模型的数据，其他K-1个样本合集作为训练集；交叉验证重复K次，使得每个子样本验证一次，然后平均K次的结果或者使用其它结合方式，最终得到一个单一估测。<br/>\n",
    "2.先用网格搜索选取几组参数组合，然后在每个参数组合下使用交叉验证算出验证分数，最后比较哪个参数组合下的验证分数高，哪个参数组合就是最佳参数组合。<br/>\n",
    "3.可以更准确的反应网格搜索选取的每个参数组合的验证分数。不使用交叉验证可能会找错最优参数组合。如果不使用交叉验证，比如只做一次分割，这样分割后数据的分布是否和原始数据集的分布相同等因素比较敏感，不同的划分会得到不同的最优模型，而且分成训练集，验证集和测试集后用于训练的数据更少了。交叉验证通过对k个不同分组训练的结果进行平均来减少方差，因此模型的性能对数据的划分就不那么敏感。<br/>\n",
    "4.可以告诉我们一个字典中每列的关键词和值。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 练习：训练模型\n",
    "在最后一个练习中，你将需要将所学到的内容整合，使用**决策树演算法**训练一个模型。为了保证你得出的是一个最优模型，你需要使用网格搜索法训练模型，以找到最佳的 `'max_depth'` 参数。你可以把`'max_depth'` 参数理解为决策树算法在做出预测前，允许其对数据提出问题的数量。决策树是**监督学习算法**中的一种。\n",
    "\n",
    "此外，你会发现你的实现使用的是 `ShuffleSplit()` 。它也是交叉验证的一种方式（见变量 `'cv_sets'`）。虽然这不是**问题8**中描述的 K-Fold 交叉验证，这个教程验证方法也很有用！这里 `ShuffleSplit()` 会创造10个(`'n_splits'`)混洗过的集合，每个集合中20%(`'test_size'`)的数据会被用作**验证集**。当你在实现的时候，想一想这跟 K-Fold 交叉验证有哪些相同点，哪些不同点？\n",
    "\n",
    "在下方 `fit_model` 函数中，你需要做的是：\n",
    "- 使用  `sklearn.tree` 中的 [`DecisionTreeRegressor`](http://scikit-learn.org/stable/modules/generated/sklearn.tree.DecisionTreeRegressor.html) 创建一个决策树的回归函数；\n",
    "  - 将这个回归函数储存到 `'regressor'` 变量中；\n",
    "- 为 `'max_depth'` 创造一个字典，它的值是从1至10的数组，并储存到 `'params'` 变量中；\n",
    "- 使用 `sklearn.metrics` 中的 [`make_scorer`](http://scikit-learn.org/stable/modules/generated/sklearn.metrics.make_scorer.html)  创建一个评分函数；\n",
    "  - 将 `performance_metric` 作为参数传至这个函数中；\n",
    "  - 将评分函数储存到 `'scoring_fnc'` 变量中；\n",
    "- 使用 `sklearn.model_selection` 中的 [`GridSearchCV`](http://scikit-learn.org/stable/modules/generated/sklearn.model_selection.GridSearchCV.html) 创建一个网格搜索对象；\n",
    "  - 将变量`'regressor'`, `'params'`, `'scoring_fnc'`, 和 `'cv_sets'` 作为参数传至这个对象中；\n",
    "  - 将 `GridSearchCV` 存到 `'grid'` 变量中。\n",
    "  \n",
    "如果有同学对python函数如何传递多个参数不熟悉，可以参考这个MIT课程的[视频](http://cn-static.udacity.com/mlnd/videos/MIT600XXT114-V004200_DTH.mp4)。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# TODO: Import 'make_scorer', 'DecisionTreeRegressor', and 'GridSearchCV'\n",
    "from sklearn.metrics import make_scorer\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "def fit_model(X, y):\n",
    "    \"\"\" Performs grid search over the 'max_depth' parameter for a \n",
    "        decision tree regressor trained on the input data [X, y]. \"\"\"\n",
    "    \n",
    "    # Create cross-validation sets from the training data\n",
    "    cv_sets = ShuffleSplit( n_splits = 10, test_size = 0.20, random_state = 0)\n",
    "\n",
    "    # TODO: Create a decision tree regressor object\n",
    "    regressor = DecisionTreeRegressor()\n",
    "\n",
    "    # TODO: Create a dictionary for the parameter 'max_depth' with a range from 1 to 10\n",
    "    params = {'max_depth': range(1, 11)}\n",
    "\n",
    "    # TODO: Transform 'performance_metric' into a scoring function using 'make_scorer' \n",
    "    scoring_fnc = make_scorer(performance_metric)\n",
    "\n",
    "    # TODO: Create the grid search object\n",
    "    grid = GridSearchCV(estimator = regressor, param_grid = params, scoring = scoring_fnc, cv = cv_sets)\n",
    "\n",
    "    # Fit the grid search object to the data to compute the optimal model\n",
    "    grid = grid.fit(X, y)\n",
    "\n",
    "    # Return the optimal model after fitting the data\n",
    "    return grid.best_estimator_\n",
    "    #print pd.DataFrame(grid.cv_results_)\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 做出预测\n",
    "当我们用数据训练出一个模型，它现在就可用于对新的数据进行预测。在决策树回归函数中，模型已经学会对新输入的数据*提问*，并返回对**目标变量**的预测值。你可以用这个预测来获取数据未知目标变量的信息，这些数据必须是不包含在训练数据之内的。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 问题 9- 最优模型\n",
    "*最优模型的最大深度（maximum depth）是多少？此答案与你在**问题 6**所做的猜测是否相同？*\n",
    "\n",
    "运行下方区域内的代码，将决策树回归函数代入训练数据的集合，以得到最优化的模型。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parameter 'max_depth' is 4 for the optimal model.\n"
     ]
    }
   ],
   "source": [
    "# Fit the training data to the model using grid search\n",
    "reg = fit_model(X_train, y_train)\n",
    "\n",
    "# Produce the value for 'max_depth'\n",
    "print \"Parameter 'max_depth' is {} for the optimal model.\".format(reg.get_params()['max_depth'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Answer: **最优模型的最大深度是4，和我猜测的3比较接近。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 问题 10 - 预测销售价格\n",
    "想像你是一个在波士顿地区的房屋经纪人，并期待使用此模型以帮助你的客户评估他们想出售的房屋。你已经从你的三个客户收集到以下的资讯:\n",
    "\n",
    "| 特征 | 客戶 1 | 客戶 2 | 客戶 3 |\n",
    "| :---: | :---: | :---: | :---: |\n",
    "| 房屋内房间总数 | 5 间房间 | 4 间房间 | 8 间房间 |\n",
    "| 社区贫困指数（％被认为是贫困阶层） | 17% | 32% | 3% |\n",
    "| 邻近学校的学生-老师比例 | 15：1 | 22：1 | 12：1 |\n",
    "\n",
    "*你会建议每位客户的房屋销售的价格为多少？从房屋特征的数值判断，这样的价格合理吗？为什么？* \n",
    "\n",
    "**提示：**用你在**分析数据**部分计算出来的统计信息来帮助你证明你的答案。\n",
    "\n",
    "运行下列的代码区域，使用你优化的模型来为每位客户的房屋价值做出预测。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predicted selling price for Client 1's home: $391,183.33\n",
      "Predicted selling price for Client 2's home: $189,123.53\n",
      "Predicted selling price for Client 3's home: $942,666.67\n"
     ]
    }
   ],
   "source": [
    "# Produce a matrix for client data\n",
    "client_data = [[5, 17, 15], # Client 1\n",
    "               [4, 32, 22], # Client 2\n",
    "               [8, 3, 12]]  # Client 3\n",
    "\n",
    "# Show predictions\n",
    "for i, price in enumerate(reg.predict(client_data)):\n",
    "    print \"Predicted selling price for Client {}'s home: ${:,.2f}\".format(i+1, price)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**答案: **我会按预测到的价格给客户建议。这个价格合理，客户1的特征应该是属于中等收入人群，他的房屋价格 $391,183.33和分析数据中Mean price和Median price都比较接近；客户2的特征应该是属于偏低等收入人群，他的房屋价格$189,123.53和分析数据中Minimum price比较接近；客户3的特征是属于偏高等收入人群，他的房屋价格$942,666.67和分析数据中的Maximum price比较接近。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 敏感度\n",
    "\n",
    "一个最优的模型不一定是一个健壮模型。有的时候模型会过于复杂或者过于简单，以致于难以泛化新增添的数据；有的时候模型采用的学习算法并不适用于特定的数据结构；有的时候样本本身可能有太多噪点或样本过少，使得模型无法准确地预测目标变量。这些情况下我们会说模型是欠拟合的。执行下方区域中的代码，采用不同的训练和测试集执行 `fit_model` 函数10次。注意观察对一个特定的客户来说，预测是如何随训练数据的变化而变化的。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Trial 1: $391,183.33\n",
      "Trial 2: $424,935.00\n",
      "Trial 3: $415,800.00\n",
      "Trial 4: $420,622.22\n",
      "Trial 5: $413,334.78\n",
      "Trial 6: $411,931.58\n",
      "Trial 7: $399,663.16\n",
      "Trial 8: $407,232.00\n",
      "Trial 9: $351,577.61\n",
      "Trial 10: $413,700.00\n",
      "\n",
      "Range in prices: $73,357.39\n"
     ]
    }
   ],
   "source": [
    "# 如果你在 fit_model 函数里使用了 print pd.DataFrame(grid.cv_results_) 请注释掉这行之后再运行本代码\n",
    "vs.PredictTrials(features, prices, fit_model, client_data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 问题 11 - 实用性探讨\n",
    "*简单地讨论一下你建构的模型能否在现实世界中使用？* \n",
    "\n",
    "**提示：** 回答几个问题，并给出相应结论的理由：\n",
    "- *1978年所采集的数据，在今天是否仍然适用？*\n",
    "- *数据中呈现的特征是否足够描述一个房屋？*\n",
    "- *模型是否足够健壮来保证预测的一致性？你需要参考上面敏感度部分的结果来解释你的结论。*\n",
    "- *在波士顿这样的大都市采集的数据，能否应用在其它乡镇地区？*"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**答案: **1.1978年采集的数据对今天并不适用，因为房屋价格已经发生很大的改变。<br/>\n",
    "2.不够，我觉得还需要房屋面积，装修成本，周围商圈大小等特征。<br/>\n",
    "3.不够健壮。因为采用不同的训练和测试集执行 fit_model 函数10次，Range in prices高达 $73,357.39。<br/>\n",
    "4.不能，大都市和乡镇地区有很大区别，大都市的房屋平均价格是要高于乡镇地区，乡镇地区地广人稀，相同的房间数可能会比波士顿价格便宜很多。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 可选问题 - 预测北京房价\n",
    "\n",
    "（本题结果不影响项目是否通过）通过上面的实践，相信你对机器学习的一些常用概念有了很好的领悟和掌握。但利用70年代的波士顿房价数据进行建模的确对我们来说意义不是太大。现在你可以把你上面所学应用到北京房价数据集中`bj_housing.csv`。\n",
    "\n",
    "免责声明：考虑到北京房价受到宏观经济、政策调整等众多因素的直接影响，预测结果仅供参考。\n",
    "\n",
    "这个数据集的特征有：\n",
    "- Area：房屋面积，平方米\n",
    "- Room：房间数，间\n",
    "- Living: 厅数，间\n",
    "- School: 是否为学区房，0或1\n",
    "- Year: 房屋建造时间，年\n",
    "- Floor: 房屋所处楼层，层\n",
    "\n",
    "目标变量：\n",
    "- Value: 房屋人民币售价，万\n",
    "\n",
    "你可以参考上面学到的内容，拿这个数据集来练习数据分割与重排、定义衡量标准、训练模型、评价模型表现、使用网格搜索配合交叉验证对参数进行调优并选出最佳参数，比较两者的差别，最终得出最佳模型对验证集的预测分数。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "### 你的代码\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你成功的用新的数据集构建了模型了吗？他能对测试数据进行验证吗？它的表现是否符合你的预期？交叉验证是否有助于提升你模型的表现？"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**答案：**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "如果你是从零开始构建机器学习的代码会让你一时觉得无从下手。这时不要着急，你要做的只是查看之前写的代码，把每一行都看明白，然后逐步构建你的模型。当中遇到什么问题也可以在我们论坛寻找答案。也许你会发现你所构建的模型的表现并没有达到你的预期，这说明机器学习并非是一项简单的任务，构建一个表现良好的模型需要长时间的研究和测试。这也是我们接下来的课程中会逐渐学到的。"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}