E07_Recomendação_usando_Word2Vec.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "E07 Recomendação usando Word2Vec.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xJReOFtyGFSs"
      },
      "source": [
        "# E07 Sistema de Recomendação usando Word2Vec\n",
        "\n",
        "Neste exercício você irá desenvolver um sistema de recomendação usando _embeddings_ criados com Word2Vec. A partir de uma base de dados de _e-commerce_ onde estão registradas as compras de um site de vendas pela internet, vamos avaliar quais mercadorias foram compradas juntamente a quais outras mercadorias. O pressuposto aqui é que quando o cliente realiza uma compra online, ele possui um interesse temático em mente e acaba comprando itens que se correlacionam de alguma forma. Ainda que em alguns casos possam ser incluídos itens menos relacionados, em sua vasta maioria assumimos que as compras seguem algum tipo de correlação entre si. Por exemplo, um cliente pode estar fazendo compras para organizar uma festa infantil ou uma viagem, e de acordo com esse interesse, acaba comprando itens relacionados a essas atividades.\n",
        "\n",
        "Sendo assim, neste exercício, nosso interesse estará em tentar criar um embedding usando o algoritmo Word2Vec, usando como contexto de vizinhança a sequência de produtos comprados por um mesmo cliente em uma específica nota fiscal. Cada ordem de compra (carrinho de compras) vai conter uma lista de produtos, e os códigos de identificação destes produtos serão nossas \"palavras\", sendo que a sequência de produtos comprados formam as \"frases\" para nosso treinamento.\n",
        "\n",
        "Como neste exercício usaremos a biblioteca _gensim_, você perceberá que a parte mais trabalhosa do exercício estará na preparação dos dados, já que a ferramenta que cria o _embedding_ já estará pronta para uso. Então nesse exercício, você ficara encarregado da preparação dos dados.\n",
        "\n",
        "Para resolver o exercício, siga os passos, lendo as instruções e comentários, um a um, e preenchendo código onde for requisitado."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KEHafkstxrSx"
      },
      "source": [
        "## Passo 1: Baixar Dados\n",
        "\n",
        "Aqui usaremos uma planilha do Microsoft Excel com dados de compras de usuários em um sistema de varejo. O nome do arquivo é `retail.xlsx`."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "y9J6QmO3xJ3t",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "d8747fee-f7f0-466f-ad6d-21405d943b82"
      },
      "source": [
        "# O comando abaixo baixa o arquivo correspondente a essa planilha\n",
        "# do Excel para o computador remoto do Google Colab.\n",
        "\n",
        "!gdown https://drive.google.com/uc?id=1NK-2z0l-qTplDJJ2SHpTVBGRP3zWAK-n"
      ],
      "execution_count": 53,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading...\n",
            "From: https://drive.google.com/uc?id=1NK-2z0l-qTplDJJ2SHpTVBGRP3zWAK-n\n",
            "To: /content/retail.xlsx\n",
            "\r0.00B [00:00, ?B/s]\r5.77MB [00:00, 57.4MB/s]\r23.7MB [00:00, 111MB/s] \n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XFZELJRCyLut"
      },
      "source": [
        "Usaremos a biblioteca `pandas` para fazer a leitura dos dados"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "A_HXamB_yLIi"
      },
      "source": [
        "import pandas as pd"
      ],
      "execution_count": 54,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qQqCAb-3yZRN"
      },
      "source": [
        "Aqui fazemos a leitura dos dados para um objeto `DataFrame` do `pandas`."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "33NLQ2NgyciL"
      },
      "source": [
        "# Esta linha de código pode demorar cerca de 1 min para rodar\n",
        "df = pd.read_excel('retail.xlsx')"
      ],
      "execution_count": 55,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0FLMkSqUyrWa"
      },
      "source": [
        "Agora podemos observar os dados para entender como estão organizados."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TE4oGyTqyqdT",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 419
        },
        "outputId": "bbc2031a-4bce-4738-c4f4-2cd61fdb3c23"
      },
      "source": [
        "df"
      ],
      "execution_count": 56,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>InvoiceNo</th>\n",
              "      <th>StockCode</th>\n",
              "      <th>Description</th>\n",
              "      <th>Quantity</th>\n",
              "      <th>InvoiceDate</th>\n",
              "      <th>UnitPrice</th>\n",
              "      <th>CustomerID</th>\n",
              "      <th>Country</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>536365</td>\n",
              "      <td>85123A</td>\n",
              "      <td>WHITE HANGING HEART T-LIGHT HOLDER</td>\n",
              "      <td>6</td>\n",
              "      <td>2010-12-01 08:26:00</td>\n",
              "      <td>2.55</td>\n",
              "      <td>17850.0</td>\n",
              "      <td>United Kingdom</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>536365</td>\n",
              "      <td>71053</td>\n",
              "      <td>WHITE METAL LANTERN</td>\n",
              "      <td>6</td>\n",
              "      <td>2010-12-01 08:26:00</td>\n",
              "      <td>3.39</td>\n",
              "      <td>17850.0</td>\n",
              "      <td>United Kingdom</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>536365</td>\n",
              "      <td>84406B</td>\n",
              "      <td>CREAM CUPID HEARTS COAT HANGER</td>\n",
              "      <td>8</td>\n",
              "      <td>2010-12-01 08:26:00</td>\n",
              "      <td>2.75</td>\n",
              "      <td>17850.0</td>\n",
              "      <td>United Kingdom</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>536365</td>\n",
              "      <td>84029G</td>\n",
              "      <td>KNITTED UNION FLAG HOT WATER BOTTLE</td>\n",
              "      <td>6</td>\n",
              "      <td>2010-12-01 08:26:00</td>\n",
              "      <td>3.39</td>\n",
              "      <td>17850.0</td>\n",
              "      <td>United Kingdom</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>536365</td>\n",
              "      <td>84029E</td>\n",
              "      <td>RED WOOLLY HOTTIE WHITE HEART.</td>\n",
              "      <td>6</td>\n",
              "      <td>2010-12-01 08:26:00</td>\n",
              "      <td>3.39</td>\n",
              "      <td>17850.0</td>\n",
              "      <td>United Kingdom</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>541904</th>\n",
              "      <td>581587</td>\n",
              "      <td>22613</td>\n",
              "      <td>PACK OF 20 SPACEBOY NAPKINS</td>\n",
              "      <td>12</td>\n",
              "      <td>2011-12-09 12:50:00</td>\n",
              "      <td>0.85</td>\n",
              "      <td>12680.0</td>\n",
              "      <td>France</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>541905</th>\n",
              "      <td>581587</td>\n",
              "      <td>22899</td>\n",
              "      <td>CHILDREN'S APRON DOLLY GIRL</td>\n",
              "      <td>6</td>\n",
              "      <td>2011-12-09 12:50:00</td>\n",
              "      <td>2.10</td>\n",
              "      <td>12680.0</td>\n",
              "      <td>France</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>541906</th>\n",
              "      <td>581587</td>\n",
              "      <td>23254</td>\n",
              "      <td>CHILDRENS CUTLERY DOLLY GIRL</td>\n",
              "      <td>4</td>\n",
              "      <td>2011-12-09 12:50:00</td>\n",
              "      <td>4.15</td>\n",
              "      <td>12680.0</td>\n",
              "      <td>France</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>541907</th>\n",
              "      <td>581587</td>\n",
              "      <td>23255</td>\n",
              "      <td>CHILDRENS CUTLERY CIRCUS PARADE</td>\n",
              "      <td>4</td>\n",
              "      <td>2011-12-09 12:50:00</td>\n",
              "      <td>4.15</td>\n",
              "      <td>12680.0</td>\n",
              "      <td>France</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>541908</th>\n",
              "      <td>581587</td>\n",
              "      <td>22138</td>\n",
              "      <td>BAKING SET 9 PIECE RETROSPOT</td>\n",
              "      <td>3</td>\n",
              "      <td>2011-12-09 12:50:00</td>\n",
              "      <td>4.95</td>\n",
              "      <td>12680.0</td>\n",
              "      <td>France</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>541909 rows × 8 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "       InvoiceNo StockCode  ... CustomerID         Country\n",
              "0         536365    85123A  ...    17850.0  United Kingdom\n",
              "1         536365     71053  ...    17850.0  United Kingdom\n",
              "2         536365    84406B  ...    17850.0  United Kingdom\n",
              "3         536365    84029G  ...    17850.0  United Kingdom\n",
              "4         536365    84029E  ...    17850.0  United Kingdom\n",
              "...          ...       ...  ...        ...             ...\n",
              "541904    581587     22613  ...    12680.0          France\n",
              "541905    581587     22899  ...    12680.0          France\n",
              "541906    581587     23254  ...    12680.0          France\n",
              "541907    581587     23255  ...    12680.0          France\n",
              "541908    581587     22138  ...    12680.0          France\n",
              "\n",
              "[541909 rows x 8 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 56
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HChM_88szGTN"
      },
      "source": [
        "A tabela acima mostra a estrutura da planilha. Podemos observar os campos:\n",
        "\n",
        "- **InvoiceNo**: Este é um identificador único para cada compra.\n",
        "- **StockCode**: Identificador único para cada produto.\n",
        "- **Description**: Descrição do produto.\n",
        "- **Quantidade**: Quantidade daquele produto, naquela compra.\n",
        "- **InvoiceDate**: Dia e hora da compra.\n",
        "- **CustomerID**: Identificador único do cliente."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oN-jtPHe0VKz",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c4c6c046-ae02-499c-c0f3-1992c0dbb0f5"
      },
      "source": [
        "df.shape"
      ],
      "execution_count": 57,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(541909, 8)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 57
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TfS_FBWK0Y_Y"
      },
      "source": [
        "Acima vemos que essa base de dados possui 541.909 registros. Cada registro representa uma linha, ou seja, um produto comprado (cuja quantidade pode ser maior que 1). Diferentes linhas podem representar diferentes itens de uma mesma compra."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zwgPe3f90tSw"
      },
      "source": [
        "## Passo 2: Remoção de Dados Nulos\n",
        "\n",
        "Neste passo vamos remover da base de dados os dados relativos a compras de produtos onde algum dos dados de interesse estejam faltando."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AeSIJZ9x1Cuv"
      },
      "source": [
        "Começamos observando a quantidade de dados nulos"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jJmozJlf1Aze",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "1c87b0b1-da2d-4142-e95b-f8d699d928d7"
      },
      "source": [
        "df.isnull().sum()"
      ],
      "execution_count": 58,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "InvoiceNo           0\n",
              "StockCode           0\n",
              "Description      1454\n",
              "Quantity            0\n",
              "InvoiceDate         0\n",
              "UnitPrice           0\n",
              "CustomerID     135080\n",
              "Country             0\n",
              "dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 58
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KWgIGnJL1M9b"
      },
      "source": [
        "Acima podemos observar há 1.454 registros sem a descrição do produto, e 135.080 registros sem a identificação do cliente comprador."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "H7RR0IK82HuT"
      },
      "source": [
        "No código abaixo você deve remover todos os registros onde aparece pelo menos um dos campos nulos. Para entender como fazer isso, veja esse trecho da videoaula sobre pandas (especificamente a partir de 19m41s no vídeo, até mais ou menos 21m).\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-77dPKr1Wcz3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 321
        },
        "outputId": "05986dbe-a8e1-411c-b799-48afb4ec9ea0"
      },
      "source": [
        "from IPython.display import YouTubeVideo\n",
        "YouTubeVideo('usmbTUx9zdY', start=1181)"
      ],
      "execution_count": 59,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "\n",
              "        <iframe\n",
              "            width=\"400\"\n",
              "            height=\"300\"\n",
              "            src=\"https://www.youtube.com/embed/usmbTUx9zdY?start=1181\"\n",
              "            frameborder=\"0\"\n",
              "            allowfullscreen\n",
              "        ></iframe>\n",
              "        "
            ],
            "text/plain": [
              "<IPython.lib.display.YouTubeVideo at 0x7f9738b4ae10>"
            ],
            "image/jpeg": "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\n"
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 59
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1ZrVo84f2OmJ"
      },
      "source": [
        "# Insira aqui seu código para remover os dados nulos.\n",
        "# Ao final do código, a variável df deve ser um DataFrame\n",
        "# do pandas contendo apenas os registros onde nenhuma\n",
        "# coluna é nula. O resultado deve ser sobrescrito na\n",
        "# na própria variável df. Há duas formas de fazer isso\n",
        "# no pandas\n",
        "#\n",
        "#  Opção 1 Usando inplace: df.função(argumentos, inplace=True)\n",
        "#  Opção 2 Usando atribuição: df = df.função(argumentos)\n",
        "\n",
        "df.dropna(inplace=True)"
      ],
      "execution_count": 60,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pJL2FDvJVOF1",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "cea8512f-2a5f-49ee-cd46-57a058f6131f"
      },
      "source": [
        "# Verifica novamente. Dessa vez os valores\n",
        "# da soma devem estar zerados para todas\n",
        "# colunas.\n",
        "\n",
        "df.isnull().sum()"
      ],
      "execution_count": 61,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "InvoiceNo      0\n",
              "StockCode      0\n",
              "Description    0\n",
              "Quantity       0\n",
              "InvoiceDate    0\n",
              "UnitPrice      0\n",
              "CustomerID     0\n",
              "Country        0\n",
              "dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 61
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nlj5sCY7SNs8"
      },
      "source": [
        "## Passo 3: Dicionário de Produtos"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xd7p7gROY2Dq"
      },
      "source": [
        "Agora vamos preparar um dicionário de códigos e respectivas descrições dos produtos. Esse dicionário servirá para que nós, humanos, possamos inspecionar o resultado de nossas operações no _embedding_, já que o _embedding_ vai ser treinado com base nos códigos dos produtos. Como os códigos não permitem inspeção intuitiva da natureza de cada produto, vamos precisar desse dicionário para entender qual é a descrição do produto correspondente a cada código."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CVIvmK9v3O96"
      },
      "source": [
        "### Passo 3.1: Conversão dos Códigos para Strings\n",
        "\n",
        "Começamos convertendo os códigos de produto (coluna **StockCode**) para string, para usar como \"palavras\" no treinamento de um modelo word2vec no `gensim` mais tarde.\n",
        "\n",
        "Para fazer isso você pode usar o método `.astype(tipo)` do pandas. Com esse comando, você consegue converter uma coluna específica do DataFrame para o tipo de dado que quiser. Por exemplo, se tivéssemos um DataFrame `d` com uma coluna chamada `idade` no formato `int` e se quiséssemos converter para o formato `float`, bastaria executar o comando:\n",
        "\n",
        "`# Converte idade para float`\n",
        "\n",
        "`d['idade'] = d['idade'].astype(float)`\n",
        "\n",
        "No nosso caso, queremos transformar o campo `StockCode` do DataFrame `df` para o formato string (`str`). Após a alteração, os elementos da coluna `StockCode` de `df` deverão ser strings.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Wvr77qNr2yKh"
      },
      "source": [
        "# Escreva abaixo seu código para converter os códigos\n",
        "# de produtos (campo 'StockCode') para o formato string\n",
        "# no DataFrame df.\n",
        "\n",
        "df['StockCode'] = df['StockCode'].astype(str)"
      ],
      "execution_count": 62,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0HLn63a1TYfX"
      },
      "source": [
        "### Passo 3.2\n",
        "\n",
        "Agora criaremos um novo _DataFrame_, de nome `products`. Selecionaremos apenas as colunas `StockCode` e `Description` do _DataFrame_ original, `df`.\n",
        "\n",
        "Para entender como selecionar colunas específicas, veja o vídeo abaixo a partir de 21m07s até 21m35s."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "suIsqNd2Z-qL",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 321
        },
        "outputId": "2db7bd0d-36c3-4b74-a738-73fbcb4a6dcd"
      },
      "source": [
        "YouTubeVideo('usmbTUx9zdY', start=1267)"
      ],
      "execution_count": 63,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "\n",
              "        <iframe\n",
              "            width=\"400\"\n",
              "            height=\"300\"\n",
              "            src=\"https://www.youtube.com/embed/usmbTUx9zdY?start=1267\"\n",
              "            frameborder=\"0\"\n",
              "            allowfullscreen\n",
              "        ></iframe>\n",
              "        "
            ],
            "text/plain": [
              "<IPython.lib.display.YouTubeVideo at 0x7f97363a0b00>"
            ],
            "image/jpeg": "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\n"
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 63
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DdB8mxOuy4bl"
      },
      "source": [
        "Escreva seu código abaixo, criando o `DataFrame` de nome `products` que seleciona apenas as colunas `StockCode` e `Description`."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mZhbpw2PSUUL"
      },
      "source": [
        "# Escreva aqui abaixo seu código para separar apenas\n",
        "# as colunas de código de produto, nome 'StockCode'\n",
        "# e descrição, nome 'Description'. O novo DataFrame\n",
        "# deve ser gravado na variável de nome products.\n",
        "\n",
        "products = df[['StockCode','Description']].copy()\n",
        "#len(products)"
      ],
      "execution_count": 64,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VzMELP7uzeUM"
      },
      "source": [
        "### Passo 3.3 Remoção de Itens Duplicados\n",
        "\n",
        "Obviamente muitos produtos foram comprados mais de uma vez. Em nosso dicionário de códigos e descrições, como a finalidade é simplesmente identificar a descrição que corresponde a um determinado código, então não faz sentido ter itens duplicados.\n",
        "\n",
        "No código abaixo faça uma remoção de todas entradas onde o código do produto (coluna `StockCode`) seja duplicado.\n",
        "\n",
        "Para entender como fazer isso, veja o vídeo abaixo a partir de 21m41s até 22m10s."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Y20WJ3up0nLO",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 321
        },
        "outputId": "d7f9ca28-4426-4eec-8971-6a3938ae0459"
      },
      "source": [
        "YouTubeVideo('usmbTUx9zdY', start=1301)"
      ],
      "execution_count": 65,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "\n",
              "        <iframe\n",
              "            width=\"400\"\n",
              "            height=\"300\"\n",
              "            src=\"https://www.youtube.com/embed/usmbTUx9zdY?start=1301\"\n",
              "            frameborder=\"0\"\n",
              "            allowfullscreen\n",
              "        ></iframe>\n",
              "        "
            ],
            "text/plain": [
              "<IPython.lib.display.YouTubeVideo at 0x7f973639b400>"
            ],
            "image/jpeg": "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\n"
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 65
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BtdlOtbOzIJ8"
      },
      "source": [
        "# Escreva abaixo o código para remover os itens duplicados\n",
        "# do DataFrame de produtos. Ao fazer isso garanta que o nome\n",
        "# do DataFrame seja o mesmo, ou seja, o resultado deve ser\n",
        "# gravado na própria variável products.\n",
        "\n",
        "products.drop_duplicates(inplace= True, subset='StockCode', keep='last')\n",
        "#len(products)"
      ],
      "execution_count": 66,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "d_Zh1f1P1MfN"
      },
      "source": [
        "### Passo 3.4 Criação do Dicionário\n",
        "\n",
        "Agora você vai criar um dicionário nativo do Python, onde as chaves serão os códigos dos produtos (`StockCode`) e os valores serão as respectivas strings de descrição (`Description`).\n",
        "\n",
        "Para entender como converter este DataFrame em um dicionário, assista o vídeo abaixo, a partir de 22min35s até 23min03s."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QkNkTizRvFu_",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 321
        },
        "outputId": "ea61ab43-d767-4799-d886-9a4ac0b48bfd"
      },
      "source": [
        "YouTubeVideo('usmbTUx9zdY', start=1355)"
      ],
      "execution_count": 67,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "\n",
              "        <iframe\n",
              "            width=\"400\"\n",
              "            height=\"300\"\n",
              "            src=\"https://www.youtube.com/embed/usmbTUx9zdY?start=1355\"\n",
              "            frameborder=\"0\"\n",
              "            allowfullscreen\n",
              "        ></iframe>\n",
              "        "
            ],
            "text/plain": [
              "<IPython.lib.display.YouTubeVideo at 0x7f973639ba58>"
            ],
            "image/jpeg": "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\n"
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 67
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "An0BgKa91RTl"
      },
      "source": [
        "# Escreva aqui seu código para criar um dicionário do\n",
        "# Python a partir do DataFrame products. O nome do\n",
        "# dicionário deverá ser prodcuts_dict\n",
        "\n",
        "products_dict = products.groupby('StockCode')['Description'].apply(list).to_dict()\n",
        "#d"
      ],
      "execution_count": 68,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0GtvB9h7Sm-1",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3b4a0ac2-ddd1-4332-b37b-268c1737f11d"
      },
      "source": [
        "# Teste do dicionário: na linha de código abaixo\n",
        "# consultamos um código de produto no seu dicionário.\n",
        "# Se você cumpriu todos passos corretamente até\n",
        "# aqui, então o resultado deve ser uma lista contendo\n",
        "# uma única string, que é a descrição do produto\n",
        "# correspondente àquele código.\n",
        "\n",
        "products_dict['84029E']"
      ],
      "execution_count": 69,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['RED WOOLLY HOTTIE WHITE HEART.']"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 69
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XzH4dqdM2bPg"
      },
      "source": [
        "## Passo 4: Preparação dos Dados\n",
        "\n",
        "Neste passo vamos preprarar os dados transformando o histórico de compras de cada consumidor numa espécie de \"frase\", onde cada \"palavra\" é um produto comprado."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qev247Zd3fq2"
      },
      "source": [
        "### Passo 4.1: Contagem dos Clientes Únicos\n",
        "\n",
        "Nesta base de dados temos o registro de cada compra de cada cliente ao longo de um ano. Isso significa que não apenas podemos ter o mesmo produto sendo comprado por diferentes clientes, mas também teremos o mesmo cliente, fazendo diferentes compras em diferentes momentos.\n",
        "\n",
        "Para avaliar nosso sistema de forma mais justa e próxima de um uso real, vamos tentar separar dados de validação e dados de treinamento de forma que não usemos compras de um mesmo cliente nos dados de validação e nos dados de treinamento (mesmo que em momentos diferentes). Em outras palavras, vamos separar _todas as compras_ de certos clientes para treinamento e _todas as compras_ de certos outros clientes para validação, ou seja, a separação será pelo cliente que realizou a compra. Assim todas compras de um determinado cliente estarão exclusivamente nos dados de validação ou exclusivamente nos dados de treinamento, nunca em ambos. Isso servirá para que possamos avaliar se o sistema de recomendação funciona, mesmo quando treinado para clientes diferentes daqueles para os quais estamos recomendando no momento: quer dizer que podemos treinar o sistema com as compras dos clientes escolhidos para o grupo de treinamento e avaliar as recomendações num grupo de clientes que o sistema nunca viu antes.\n",
        "\n",
        "Para fazer isso, seria interessante começar confirmando exatamente quantos clientes únicos temos nessa base de dados.\n",
        "\n",
        "A partir do DataFrame `df` selecione a coluna `CustomerID` e usando os métodos `unique()` e a conversão `tolist()`, gere uma lista com todos os ids de clientes únicos. Essa lista deverá se chamar `customers`.\n",
        "\n",
        "O vídeo abaixo, a partir de 23m59s até 25m16s, mostra como isso pode ser feito."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uVQZtVkT0Mqh",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 321
        },
        "outputId": "68db3f21-8aed-40ce-bdb3-08eea7007cae"
      },
      "source": [
        "YouTubeVideo('usmbTUx9zdY', start=1439)"
      ],
      "execution_count": 70,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "\n",
              "        <iframe\n",
              "            width=\"400\"\n",
              "            height=\"300\"\n",
              "            src=\"https://www.youtube.com/embed/usmbTUx9zdY?start=1439\"\n",
              "            frameborder=\"0\"\n",
              "            allowfullscreen\n",
              "        ></iframe>\n",
              "        "
            ],
            "text/plain": [
              "<IPython.lib.display.YouTubeVideo at 0x7f97363a0d30>"
            ],
            "image/jpeg": "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\n"
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 70
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SLGQA2VZ3lpc"
      },
      "source": [
        "# Escreva abaixo o código que cria uma lista \n",
        "# coletando os ids da coluna CustomerID de df,\n",
        "# selecionando apenas ids únicos (não recolhe\n",
        "# ids repetidos). Esta lista deve ser uma lista\n",
        "# nativa do Python, de nome customers.\n",
        "\n",
        "customers = df[\"CustomerID\"].unique().tolist()"
      ],
      "execution_count": 71,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Em6vNYCjzU8K",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c287106b-fe94-4863-bc6a-e39ae5fee01c"
      },
      "source": [
        "len(customers)"
      ],
      "execution_count": 72,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "4372"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 72
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2ZTGaq3n35dn"
      },
      "source": [
        "Se tudo deu certo até aqui, então verificamos que há exatamente 4.372 clientes na nossa base de dados. Para cada um desses clientes vamos verificar o histórico de compras, criando 4.372 sequências de compras."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cIMP9AOiHWOr"
      },
      "source": [
        "### Passo 4.2: Embaralhamento dos Dados\n",
        "\n",
        "Agora escreva código para embaralhar a ordem dos ids dos clientes na lista `customers`, usando o método `shuffle`. O resultado deve estar na própria variável `customers` (não mude o nome da variável)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MvRoZx2wHOXv"
      },
      "source": [
        "# No espaço abaixo, importe o módulo random\n",
        "# e chame a função shuffle embaralhando os\n",
        "# itens na lista customers\n",
        "\n",
        "import random\n",
        "random.shuffle(customers)"
      ],
      "execution_count": 73,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cQYdfD3RJI75"
      },
      "source": [
        "### Passo 4.3: Separação dos Dados de Treinamento e Validação\n",
        "\n",
        "Agora vamos separar dados de treinamento e de validação. Vamos separar 10% dos dados para validação e 90% dos dados para treinamento. Mas esses percentuais serão separados em termos de clientes. Isso significa que faremos o seguinte:\n",
        "\n",
        "1. Definiremos uma lista de clientes de treinamento, contendo 90% dos itens da variável `customers`. Vamos chamar essa variável de `customers_train`.\n",
        "2. Definiremos uma lista de clientes de validação, contendo 10% dos itens da variável `customers`. Vamos chamar essa variável de `customers_val`.\n",
        "3. Em nosso dataset `df`, selecionaremos todas as linhas onde o campo `CustomerID` contenha um código de cliente presente na lista `customers_train`. Essa seleção será um novo dataset de nome `df_train`.\n",
        "4. Faremos o mesmo com os  dados de validação,  selecionando as linhas de `df` onde `CustomerID` está contido em `customers_val`. Essa seleção chamaremos de `df_val`.\n",
        "\n",
        "Dica: para os passos 3 e 4 acima, use o método `.isin(nome_da_lista)`. Veja o vídeo abaixo a partir de 25m19s até o final do vídeo.\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ckUuDo_JYF5l",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 321
        },
        "outputId": "975ac33b-1a9e-4d6b-f2f7-d0958a27c113"
      },
      "source": [
        "YouTubeVideo('usmbTUx9zdY', start=1519)"
      ],
      "execution_count": 74,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "\n",
              "        <iframe\n",
              "            width=\"400\"\n",
              "            height=\"300\"\n",
              "            src=\"https://www.youtube.com/embed/usmbTUx9zdY?start=1519\"\n",
              "            frameborder=\"0\"\n",
              "            allowfullscreen\n",
              "        ></iframe>\n",
              "        "
            ],
            "text/plain": [
              "<IPython.lib.display.YouTubeVideo at 0x7f970dda30b8>"
            ],
            "image/jpeg": "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\n"
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 74
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Kq6yEiaQJIjr"
      },
      "source": [
        "# Calcule a quantidade de clientes que usaremos\n",
        "# para treinamento. Essa quantidade deve ser 90%\n",
        "# do total de clientes na variável customers.\n",
        "# Transforme o resultado em um número inteiro.\n",
        "# O resultado deve ser gravado na variável de\n",
        "# nome train_size.\n",
        "\n",
        "train_size = int(0.9 * len(customers))\n",
        "\n",
        "# Use o slicing do Python para separar\n",
        "# os consumidores em duas listas: uma para\n",
        "# treinamento e a outra para validação. Para\n",
        "# definir o ponto de corte do slicing, use\n",
        "# a variável train_size definida acima. A lista\n",
        "# de consumidores para treinamento deve se\n",
        "# chamar customers_train e incluir exatamente\n",
        "# o trecho da lista customers começando\n",
        "# do início, com quantidade igual a train_size.\n",
        "# Os demais devem itens ser colocados na lista\n",
        "# de nome customers_val (lista ids de validação)\n",
        "\n",
        "customers_train = customers[:train_size]\n",
        "customers_val = customers[train_size:len(customers)]\n",
        "\n",
        "# Agora, baseado nos ids separados nas listas\n",
        "# customers_train e customers_val acima, crie\n",
        "# dois datasets a partir do dataset original df,\n",
        "# selecionando em df_train as linhas que contem\n",
        "# ids da lista customers_train no campo CustomerID,\n",
        "# e em df_val as linhas que contem  ids da lista\n",
        "# customers_val.\n",
        "\n",
        "df_train = df[df['CustomerID'].isin(customers_train)]\n",
        "df_val = df[df['CustomerID'].isin(customers_val)]"
      ],
      "execution_count": 75,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IPSvCp8LKB5n"
      },
      "source": [
        "### Passo 4.4: Função que compila sequência de compras de cada cliente.\n",
        "\n",
        "Com essas listas de clientes criamos abaixo as sequências de compras de acordo com os históricos de cada cliente."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Cp8qT1FJKPY2"
      },
      "source": [
        "def compile_orders(customers, df):\n",
        "  ''' Essa função coleta todas compras do histórico\n",
        "      de cada cliente. O parâmetro customers é a lista\n",
        "      de ids de clientes e o parâmetro df é o objeto\n",
        "      DataFrame do pandas com os dados de cada compra.\n",
        "      O valor retornado é uma lista de listas, onde cada\n",
        "      lista interna contém a sequência de códigos de produto\n",
        "      de cada compra, na ordem que se apresentava no\n",
        "      histórico.\n",
        "  '''\n",
        "\n",
        "  # Essa será a lista de compras -- lista de listas de produtos\n",
        "  orders = []\n",
        "  \n",
        "  # Coletaremos as compras de cada cliente\n",
        "  for customer in customers:\n",
        "\n",
        "    # Aqui convertemos os códigos de produtos das compras\n",
        "    # de cada cliente para uma lista única\n",
        "    order = df[df['CustomerID'] == customer]['StockCode'].tolist()\n",
        "\n",
        "    # Essa lista de itens será a lista de produtos que será adicionada\n",
        "    # à lista de compras principal.\n",
        "    orders.append(order)\n",
        "    \n",
        "  return orders"
      ],
      "execution_count": 76,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CaIkbOMiM2ne"
      },
      "source": [
        "# Aqui separamos as listas de listas de compras. Este código\n",
        "# demora cerca de 1 minuto para rodar\n",
        "orders_train = compile_orders(customers_train, df_train)\n",
        "orders_val = compile_orders(customers_val, df_val)"
      ],
      "execution_count": 77,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VCqVu7rsQDeQ"
      },
      "source": [
        "## Passo 5: Treinar o Embedding Word2Vec\n",
        "\n",
        "Esta é a última etapa, onde utilizamos a lista de compras (lista de listas com códigos de produtos) para treinar nosso embedding. Essa lista de listas funcionará de forma análoga a uma lista de frases, onde cada frase é uma lista de palavras: lista de listas de palavras. A diferença é que aqui as strings que seriam palavras foram substituídas por códigos de produtos."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "u49iedUAQThF"
      },
      "source": [
        "# Aqui importamos o módulo GenSim\n",
        "import gensim\n",
        "\n",
        "# Importamos o NumPy\n",
        "import numpy as np\n",
        "\n",
        "# Código para ignorar warnings\n",
        "# (só para 'despoluir' as saídas\n",
        "#  por motivos didáticos, ignorar\n",
        "#  warnings não é prática recomendada\n",
        "#  exceto quando o código já foi\n",
        "#  testado e os warnings examinados\n",
        "#  e determinados como seguros)\n",
        "import warnings;\n",
        "warnings.filterwarnings('ignore')"
      ],
      "execution_count": 78,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CWoNEmF9Q99o"
      },
      "source": [
        "# Criamos o objeto da classe Word2Vec. Já no\n",
        "# construtor passamos os parâmetros de treinamento,\n",
        "# pois o modelo será treinado ao criar o objeto.\n",
        "\n",
        "model = gensim.models.Word2Vec(orders_train, \\\n",
        "                               size=50, \\\n",
        "                               window=5, \\\n",
        "                               workers=10, \\\n",
        "                               iter=10)"
      ],
      "execution_count": 79,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aRUDgNq6Rd-_",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "964c6a55-f19a-4c08-932e-71dc9f06dbcc"
      },
      "source": [
        "# Mostra os dados básicos do objeto\n",
        "\n",
        "print(model)"
      ],
      "execution_count": 80,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Word2Vec(vocab=3168, size=50, alpha=0.025)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rHWIan6WBhU4"
      },
      "source": [
        "## Passo 5: Resultados\n",
        "\n",
        "Finalmente podemos agora examinar como ficou o resultado.\n",
        "\n",
        "### Passo 5.1: Produtos semelhantes\n",
        "\n",
        "Para facilitar nossos testes, definimos abaixo uma função que retorna os 10 produtos mais semelhantes a um produto específico, passado como argumento."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "D8o6X7LbSxM3"
      },
      "source": [
        "def most_similar(product_id):\n",
        "  ''' Essa função mostra os 10 produtos mais semelhantes\n",
        "      ao produto cujo código foi passado como argumento.\n",
        "      Na verdade o método most_similar() do objeto Word2Vec\n",
        "      já faz isso. Aqui apenas acrescentamos as descrições,\n",
        "      para facilitar examinar o resultado, já que olhando\n",
        "      apenas os códigos ficaria difícil de dizer se o\n",
        "      conjunto de produtos semelhantes faz sentido.\n",
        "  '''\n",
        "\n",
        "  # Primeiro mostramos a descrição do produto consultado\n",
        "  print('Most similar to:', products_dict[product_id])\n",
        "\n",
        "  # Aqui buscamos a lista de códigos mais semelhantes\n",
        "  # ao apresentado no embedding.\n",
        "  s = model.most_similar(positive=product_id)\n",
        "\n",
        "  # Para cada código, mostramos a descrição\n",
        "  for product, prob in s:\n",
        "    print(products_dict[product], prob)"
      ],
      "execution_count": 81,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VPUMtcNgJY8Q",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "cee506f0-db8c-4589-d22a-b6e30d7efcb2"
      },
      "source": [
        "model.wv.index2entity[1000:1010]"
      ],
      "execution_count": 82,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['23407',\n",
              " '10135',\n",
              " '21832',\n",
              " '22819',\n",
              " '37370',\n",
              " '23213',\n",
              " '22588',\n",
              " '22639',\n",
              " '21983',\n",
              " '85040A']"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 82
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wxfEa_-bKAQd",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "9223186f-4143-4cd8-bdcf-d95a29e6680d"
      },
      "source": [
        "products_dict.keys()"
      ],
      "execution_count": 83,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "dict_keys(['10002', '10080', '10120', '10123C', '10124A', '10124G', '10125', '10133', '10135', '11001', '15030', '15034', '15036', '15039', '15044A', '15044B', '15044C', '15044D', '15056BL', '15056N', '15056P', '15058A', '15058B', '15058C', '15060B', '16008', '16010', '16011', '16012', '16014', '16015', '16016', '16020C', '16033', '16043', '16045', '16046', '16048', '16049', '16052', '16054', '16151A', '16156L', '16156S', '16161G', '16161M', '16161P', '16161U', '16162L', '16162M', '16168M', '16169E', '16169K', '16169M', '16169N', '16169P', '16202A', '16202B', '16202E', '16206B', '16207A', '16207B', '16216', '16218', '16219', '16225', '16235', '16236', '16237', '16238', '16244B', '16248B', '16254', '16258A', '16259', '17001', '17003', '17007B', '17011F', '17012A', '17012B', '17012C', '17012D', '17012E', '17012F', '17013D', '17014A', '17021', '17028J', '17038', '17084A', '17084J', '17084N', '17084P', '17084R', '17090A', '17090D', '17091A', '17091J', '17096', '17107D', '17109D', '17129F', '17136A', '17164B', '17165D', '17174', '17191A', '18007', '18094C', '18097A', '18097C', '18098C', '20615', '20616', '20617', '20618', '20619', '20622', '20652', '20653', '20654', '20655', '20657', '20658', '20659', '20661', '20662', '20663', '20664', '20665', '20666', '20667', '20668', '20669', '20670', '20671', '20674', '20675', '20676', '20677', '20678', '20679', '20681', '20682', '20684', '20685', '20686', '20694', '20695', '20696', '20697', '20698', '20699', '20700', '20701', '20702', '20703', '20704', '20705', '20707', '20711', '20712', '20713', '20716', '20717', '20718', '20719', '20723', '20724', '20725', '20726', '20727', '20728', '20731', '20733', '20735', '20748', '20749', '20750', '20751', '20752', '20754', '20755', '20756', '20757', '20758', '20759', '20760', '20761', '20762', '20763', '20764', '20765', '20766', '20767', '20768', '20769', '20770', '20771', '20772', '20773', '20774', '20775', '20777', '20778', '20780', '20781', '20782', '20785', '20793', '20794', '20795', '20796', '20798', '20801', '20802', '20803', '20816', '20818', '20819', '20820', '20821', '20823', '20826', '20827', '20828', '20829', '20830', '20831', '20832', '20835', '20836', '20837', '20838', '20839', '20840', '20845', '20846', '20847', '20848', '20851', '20854', '20856', '20857', '20861', '20866', '20867', '20868', '20869', '20878', '20886', '20892', '20893', '20894', '20897', '20898', '20901', '20902', '20903', '20906', '20910', '20914', '20931', '20932', '20933', '20934', '20935', '20936', '20941', '20956', '20957', '20961', '20963', '20964', '20966', '20967', '20969', '20970', '20971', '20972', '20973', '20974', '20975', '20977', '20978', '20979', '20980', '20981', '20982', '20983', '20984', '20985', '20986', '20992', '20996', '20997', '20998', '21000', '21001', '21002', '21003', '21011', '21012', '21014', '21015', '21018', '21025', '21026', '21027', '21028', '21030', '21031', '21032', '21033', '21034', '21035', '21038', '21039', '21040', '21041', '21042', '21043', '21051', '21054', '21055', '21056', '21058', '21059', '21060', '21061', '21062', '21063', '21064', '21065', '21066', '21067', '21068', '21069', '21070', '21071', '21078', '21080', '21084', '21086', '21087', '21088', '21089', '21090', '21094', '21095', '21096', '21098', '21100', '21106', '21107', '21108', '21109', '21110', '21111', '21112', '21114', '21115', '21116', '21117', '21121', '21122', '21123', '21124', '21125', '21126', '21128', '21129', '21131', '21132', '21135', '21136', '21137', '21143', '21144', '21145', '21147', '21154', '21155', '21156', '21157', '21158', '21159', '21161', '21162', '21163', '21164', '21165', '21166', '21167', '21169', '21171', '21172', '21174', '21175', '21179', '21181', '21187', '21188', '21189', '21190', '21191', '21192', '21194', '21195', '21196', '21197', '21198', '21199', '21200', '21201', '21202', '21204', '21205', '21206', '21207', '21208', '21209', '21210', '21211', '21212', '21213', '21215', '21216', '21217', '21218', '21219', '21220', '21221', '21222', '21224', '21226', '21228', '21231', '21232', '21238', '21239', '21240', '21242', '21243', '21244', '21245', '21246', '21248', '21249', '21250', '21251', '21253', '21257', '21258', '21259', '21260', '21261', '21262', '21263', '21264', '21265', '21268', '21269', '21270', '21272', '21275', '21277', '21278', '21279', '21280', '21281', '21284', '21285', '21286', '21287', '21288', '21289', '21291', '21292', '21293', '21294', '21306', '21307', '21310', '21311', '21313', '21314', '21316', '21317', '21318', '21319', '21320', '21324', '21326', '21327', '21328', '21329', '21331', '21332', '21333', '21336', '21340', '21344', '21348', '21349', '21351', '21352', '21354', '21355', '21356', '21357', '21358', '21359', '21360', '21361', '21363', '21364', '21365', '21366', '21367', '21368', '21369', '21370', '21371', '21372', '21373', '21374', '21375', '21376', '21377', '21378', '21379', '21380', '21381', '21382', '21383', '21385', '21386', '21389', '21390', '21391', '21392', '21393', '21394', '21395', '21397', '21398', '21399', '21400', '21401', '21402', '21403', '21407', '21408', '21411', '21412', '21413', '21414', '21415', '21416', '21417', '21418', '21420', '21421', '21422', '21424', '21425', '21426', '21427', '21428', '21429', '21430', '21439', '21441', '21442', '21445', '21446', '21447', '21448', '21452', '21454', '21455', '21456', '21457', '21458', '21459', '21460', '21461', '21462', '21463', '21464', '21465', '21466', '21467', '21468', '21469', '21470', '21471', '21472', '21473', '21476', '21479', '21481', '21484', '21485', '21486', '21488', '21494', '21495', '21497', '21498', '21499', '21500', '21503', '21504', '21506', '21507', '21508', '21509', '21518', '21519', '21520', '21523', '21524', '21527', '21528', '21530', '21531', '21533', '21534', '21535', '21537', '21538', '21539', '21544', '21547', '21555', '21556', '21557', '21558', '21559', '21561', '21562', '21563', '21564', '21576', '21577', '21578', '21579', '21580', '21581', '21584', '21586', '21587', '21588', '21591', '21592', '21609', '21613', '21614', '21615', '21616', '21617', '21618', '21619', '21620', '21621', '21622', '21623', '21624', '21625', '21626', '21627', '21628', '21629', '21630', '21631', '21632', '21633', '21634', '21635', '21636', '21637', '21638', '21640', '21641', '21642', '21643', '21644', '21645', '21647', '21648', '21650', '21651', '21654', '21655', '21656', '21657', '21658', '21661', '21662', '21664', '21666', '21667', '21668', '21669', '21670', '21671', '21672', '21673', '21675', '21676', '21677', '21678', '21679', '21680', '21681', '21682', '21683', '21684', '21686', '21688', '21689', '21690', '21692', '21693', '21694', '21695', '21696', '21698', '21700', '21703', '21704', '21705', '21706', '21707', '21708', '21709', '21710', '21711', '21713', '21714', '21715', '21716', '21717', '21718', '21719', '21720', '21721', '21722', '21723', '21724', '21725', '21726', '21730', '21731', '21733', '21735', '21736', '21738', '21739', '21741', '21742', '21743', '21744', '21745', '21746', '21747', '21749', '21750', '21754', '21755', '21756', '21757', '21758', '21761', '21763', '21769', '21770', '21773', '21774', '21775', '21777', '21781', '21784', '21785', '21786', '21787', '21788', '21789', '21790', '21791', '21792', '21793', '21794', '21801', '21802', '21803', '21804', '21807', '21808', '21809', '21810', '21811', '21812', '21813', '21814', '21815', '21816', '21817', '21818', '21819', '21820', '21821', '21822', '21823', '21824', '21826', '21827', '21828', '21829', '21830', '21832', '21833', '21836', '21839', '21841', '21843', '21844', '21845', '21846', '21847', '21849', '21850', '21851', '21864', '21865', '21866', '21867', '21868', '21870', '21871', '21872', '21873', '21874', '21875', '21876', '21877', '21878', '21879', '21880', '21881', '21882', '21883', '21884', '21888', '21889', '21890', '21891', '21892', '21894', '21895', '21896', '21897', '21898', '21899', '21900', '21901', '21902', '21903', '21904', '21905', '21906', '21907', '21908', '21910', '21911', '21912', '21913', '21914', '21915', '21916', '21917', '21918', '21922', '21925', '21926', '21927', '21928', '21929', '21930', '21931', '21932', '21933', '21934', '21935', '21936', '21937', '21942', '21943', '21944', '21945', '21946', '21947', '21948', '21949', '21955', '21967', '21972', '21973', '21974', '21975', '21976', '21977', '21980', '21981', '21982', '21983', '21984', '21985', '21986', '21987', '21988', '21989', '21990', '21991', '21992', '21993', '22021', '22023', '22024', '22025', '22026', '22027', '22028', '22029', '22030', '22031', '22032', '22033', '22034', '22035', '22037', '22038', '22039', '22040', '22041', '22042', '22043', '22044', '22045', '22046', '22047', '22048', '22049', '22050', '22051', '22052', '22053', '22055', '22057', '22059', '22060', '22061', '22062', '22063', '22064', '22065', '22066', '22067', '22068', '22069', '22070', '22071', '22072', '22073', '22074', '22075', '22076', '22077', '22078', '22079', '22080', '22081', '22082', '22083', '22084', '22085', '22086', '22087', '22088', '22089', '22090', '22091', '22092', '22093', '22094', '22095', '22096', '22097', '22098', '22099', '22100', '22101', '22102', '22103', '22104', '22105', '22106', '22107', '22108', '22109', '22110', '22111', '22112', '22113', '22114', '22115', '22116', '22117', '22118', '22119', '22120', '22121', '22123', '22124', '22125', '22127', '22128', '22129', '22130', '22131', '22132', '22133', '22134', '22135', '22136', '22137', '22138', '22139', '22141', '22142', '22144', '22145', '22146', '22147', '22148', '22149', '22150', '22151', '22152', '22153', '22154', '22155', '22156', '22157', '22158', '22161', '22162', '22163', '22164', '22165', '22166', '22167', '22168', '22169', '22170', '22171', '22173', '22174', '22175', '22176', '22178', '22179', '22180', '22181', '22182', '22183', '22184', '22185', '22186', '22187', '22188', '22189', '22190', '22191', '22192', '22193', '22194', '22195', '22196', '22197', '22198', '22199', '22200', '22201', '22202', '22203', '22204', '22207', '22208', '22209', '22210', '22211', '22212', '22214', '22215', '22216', '22217', '22218', '22219', '22220', '22221', '22222', '22223', '22224', '22227', '22228', '22229', '22230', '22231', '22232', '22233', '22236', '22241', '22242', '22243', '22244', '22245', '22246', '22247', '22248', '22249', '22250', '22251', '22252', '22254', '22255', '22256', '22257', '22258', '22259', '22260', '22261', '22262', '22263', '22264', '22265', '22266', '22267', '22268', '22269', '22270', '22271', '22272', '22273', '22274', '22275', '22276', '22277', '22278', '22279', '22280', '22281', '22282', '22283', '22284', '22285', '22286', '22287', '22288', '22289', '22291', '22292', '22293', '22294', '22295', '22296', '22297', '22299', '22300', '22301', '22302', '22303', '22304', '22305', '22306', '22307', '22308', '22309', '22310', '22311', '22312', '22313', '22314', '22315', '22316', '22317', '22318', '22319', '22320', '22321', '22322', '22323', '22324', '22325', '22326', '22327', '22328', '22329', '22331', '22332', '22333', '22334', '22335', '22336', '22337', '22338', '22339', '22340', '22341', '22342', '22343', '22344', '22345', '22346', '22348', '22349', '22350', '22351', '22352', '22353', '22354', '22355', '22356', '22357', '22358', '22359', '22360', '22361', '22362', '22363', '22364', '22365', '22366', '22367', '22371', '22372', '22374', '22375', '22376', '22377', '22378', '22379', '22380', '22381', '22382', '22383', '22384', '22385', '22386', '22389', '22390', '22391', '22393', '22394', '22395', '22396', '22398', '22399', '22400', '22402', '22403', '22405', '22406', '22407', '22408', '22409', '22410', '22411', '22412', '22413', '22414', '22415', '22416', '22417', '22418', '22419', '22420', '22421', '22422', '22423', '22424', '22425', '22426', '22427', '22428', '22429', '22430', '22431', '22432', '22433', '22434', '22435', '22436', '22437', '22438', '22439', '22440', '22441', '22442', '22443', '22444', '22445', '22446', '22447', '22448', '22449', '22450', '22451', '22452', '22453', '22454', '22456', '22457', '22458', '22459', '22460', '22461', '22464', '22465', '22466', '22467', '22468', '22469', '22470', '22471', '22472', '22473', '22474', '22475', '22476', '22477', '22478', '22479', '22480', '22481', '22482', '22483', '22485', '22486', '22487', '22488', '22489', '22491', '22492', '22493', '22494', '22495', '22496', '22497', '22498', '22499', '22500', '22501', '22502', '22503', '22504', '22505', '22507', '22508', '22509', '22510', '22511', '22512', '22513', '22514', '22515', '22516', '22517', '22518', '22519', '22520', '22521', '22522', '22523', '22524', '22525', '22526', '22529', '22530', '22531', '22532', '22533', '22534', '22535', '22536', '22537', '22538', '22539', '22540', '22541', '22543', '22544', '22545', '22546', '22547', '22548', '22549', '22550', '22551', '22553', '22554', '22555', '22556', '22557', '22558', '22559', '22560', '22561', '22562', '22563', '22564', '22565', '22566', '22567', '22568', '22569', '22570', '22571', '22572', '22573', '22574', '22575', '22576', '22577', '22578', '22579', '22580', '22581', '22582', '22583', '22584', '22585', '22586', '22587', '22588', '22589', '22591', '22592', '22593', '22594', '22595', '22596', '22597', '22598', '22599', '22600', '22601', '22602', '22603', '22604', '22605', '22606', '22607', '22608', '22609', '22610', '22611', '22613', '22614', '22615', '22616', '22617', '22618', '22619', '22620', '22621', '22622', '22623', '22624', '22625', '22626', '22627', '22628', '22629', '22630', '22631', '22632', '22633', '22634', '22635', '22636', '22637', '22638', '22639', '22640', '22641', '22642', '22643', '22644', '22645', '22646', '22647', '22649', '22650', '22651', '22652', '22653', '22654', '22655', '22656', '22659', '22660', '22661', '22662', '22663', '22664', '22665', '22666', '22667', '22668', '22669', '22670', '22671', '22672', '22673', '22674', '22675', '22676', '22677', '22678', '22679', '22680', '22681', '22682', '22683', '22684', '22685', '22686', '22687', '22688', '22689', '22690', '22691', '22692', '22693', '22694', '22695', '22696', '22697', '22698', '22699', '22700', '22701', '22702', '22703', '22704', '22705', '22706', '22707', '22708', '22709', '22710', '22711', '22712', '22713', '22714', '22715', '22716', '22717', '22718', '22719', '22720', '22721', '22722', '22723', '22725', '22726', '22727', '22728', '22729', '22730', '22731', '22732', '22733', '22734', '22735', '22736', '22737', '22738', '22739', '22740', '22741', '22742', '22743', '22744', '22745', '22746', '22747', '22748', '22749', '22750', '22751', '22752', '22753', '22754', '22755', '22756', '22757', '22758', '22759', '22760', '22761', '22762', '22763', '22764', '22765', '22766', '22767', '22768', '22769', '22770', '22771', '22772', '22773', '22774', '22775', '22776', '22777', '22778', '22779', '22780', '22781', '22782', '22783', '22784', '22785', '22786', '22788', '22789', '22791', '22792', '22794', '22795', '22796', '22797', '22798', '22799', '22800', '22801', '22802', '22803', '22804', '22805', '22806', '22807', '22808', '22809', '22810', '22811', '22812', '22813', '22814', '22815', '22816', '22817', '22818', '22819', '22820', '22821', '22822', '22823', '22824', '22825', '22826', '22827', '22828', '22829', '22830', '22831', '22832', '22833', '22834', '22835', '22837', '22838', '22839', '22840', '22841', '22842', '22843', '22844', '22845', '22846', '22847', '22848', '22849', '22851', '22852', '22853', '22854', '22855', '22856', '22857', '22858', '22859', '22860', '22861', '22862', '22863', '22865', '22866', '22867', '22868', '22869', '22870', '22871', '22872', '22873', '22874', '22875', '22876', '22877', '22878', '22879', '22880', '22881', '22882', '22883', '22884', '22885', '22886', '22887', '22888', '22889', '22890', '22891', '22892', '22893', '22894', '22895', '22896', '22897', '22898', '22899', '22900', '22902', '22903', '22904', '22905', '22906', '22907', '22908', '22909', '22910', '22911', '22912', '22913', '22914', '22915', '22916', '22917', '22918', '22919', '22920', '22921', '22922', '22923', '22924', '22925', '22926', '22927', '22928', '22929', '22930', '22931', '22932', '22933', '22934', '22935', '22936', '22937', '22938', '22939', '22940', '22941', '22942', '22943', '22944', '22945', '22946', '22947', '22948', '22949', '22950', '22951', '22952', '22953', '22954', '22955', '22956', '22957', '22959', '22960', '22961', '22962', '22963', '22964', '22965', '22966', '22967', '22968', '22969', '22970', '22971', '22972', '22973', '22974', '22975', '22976', '22977', '22978', '22979', '22980', '22981', '22982', '22983', '22984', '22985', '22986', '22987', '22988', '22989', '22990', '22991', '22992', '22993', '22994', '22995', '22996', '22997', '22998', '22999', '23000', '23002', '23003', '23004', '23005', '23006', '23007', '23008', '23009', '23010', '23012', '23013', '23014', '23015', '23016', '23017', '23018', '23019', '23020', '23021', '23022', '23023', '23024', '23025', '23026', '23027', '23028', '23029', '23031', '23032', '23033', '23034', '23035', '23036', '23037', '23038', '23039', '23040', '23041', '23042', '23043', '23044', '23045', '23046', '23047', '23048', '23049', '23050', '23051', '23052', '23053', '23054', '23055', '23056', '23057', '23058', '23060', '23061', '23064', '23065', '23066', '23067', '23068', '23069', '23070', '23071', '23072', '23073', '23074', '23075', '23076', '23077', '23078', '23079', '23080', '23081', '23082', '23083', '23084', '23085', '23086', '23087', '23088', '23089', '23090', '23091', '23092', '23093', '23094', '23096', '23099', '23100', '23101', '23102', '23103', '23104', '23106', '23107', '23108', '23109', '23110', '23111', '23112', '23113', '23114', '23115', '23116', '23117', '23118', '23119', '23120', '23121', '23122', '23123', '23124', '23125', '23126', '23127', '23128', '23129', '23130', '23131', '23132', '23133', '23134', '23135', '23136', '23137', '23138', '23139', '23140', '23141', '23142', '23143', '23144', '23145', '23146', '23147', '23148', '23149', '23150', '23151', '23152', '23153', '23154', '23155', '23156', '23157', '23158', '23159', '23160', '23161', '23162', '23163', '23164', '23165', '23166', '23167', '23168', '23169', '23170', '23171', '23172', '23173', '23174', '23175', '23176', '23177', '23178', '23179', '23180', '23181', '23182', '23183', '23184', '23185', '23186', '23187', '23188', '23189', '23190', '23191', '23192', '23193', '23194', '23196', '23197', '23198', '23199', '23200', '23201', '23202', '23203', '23204', '23205', '23206', '23207', '23208', '23209', '23210', '23211', '23212', '23213', '23214', '23215', '23216', '23217', '23218', '23219', '23220', '23221', '23222', '23223', '23224', '23225', '23226', '23227', '23228', '23229', '23230', '23231', '23232', '23233', '23234', '23235', '23236', '23237', '23238', '23239', '23240', '23241', '23242', '23243', '23244', '23245', '23247', '23249', '23250', '23251', '23252', '23253', '23254', '23255', '23256', '23263', '23264', '23265', '23266', '23267', '23268', '23269', '23270', '23271', '23272', '23273', '23274', '23275', '23280', '23281', '23282', '23283', '23284', '23285', '23286', '23287', '23288', '23289', '23290', '23291', '23292', '23293', '23294', '23295', '23296', '23297', '23298', '23299', '23300', '23301', '23302', '23303', '23304', '23305', '23306', '23307', '23308', '23309', '23310', '23311', '23312', '23313', '23314', '23315', '23316', '23317', '23318', '23319', '23320', '23321', '23322', '23323', '23324', '23325', '23326', '23327', '23328', '23329', '23330', '23331', '23332', '23333', '23334', '23335', '23336', '23337', '23338', '23339', '23340', '23341', '23342', '23343', '23344', '23345', '23346', '23347', '23348', '23349', '23350', '23351', '23352', '23353', '23354', '23355', '23356', '23357', '23358', '23359', '23360', '23365', '23366', '23367', '23368', '23369', '23370', '23371', '23372', '23373', '23374', '23375', '23376', '23377', '23378', '23379', '23380', '23381', '23382', '23388', '23389', '23390', '23391', '23392', '23393', '23394', '23395', '23396', '23397', '23398', '23399', '23400', '23401', '23402', '23403', '23404', '23405', '23406', '23407', '23408', '23409', '23410', '23411', '23412', '23413', '23414', '23415', '23416', '23417', '23418', '23419', '23420', '23421', '23422', '23423', '23424', '23425', '23426', '23427', '23428', '23429', '23430', '23431', '23432', '23433', '23434', '23435', '23436', '23437', '23438', '23439', '23440', '23441', '23442', '23444', '23445', '23446', '23447', '23448', '23449', '23451', '23452', '23453', '23454', '23455', '23456', '23457', '23458', '23459', '23460', '23461', '23462', '23463', '23464', '23465', '23466', '23467', '23468', '23469', '23470', '23471', '23472', '23473', '23474', '23475', '23476', '23477', '23478', '23479', '23480', '23481', '23482', '23483', '23484', '23485', '23486', '23487', '23489', '23490', '23491', '23492', '23493', '23494', '23495', '23496', '23497', '23498', '23499', '23500', '23501', '23502', '23503', '23504', '23505', '23506', '23507', '23508', '23509', '23510', '23511', '23512', '23513', '23514', '23515', '23516', '23517', '23518', '23519', '23520', '23521', '23522', '23523', '23524', '23525', '23526', '23527', '23528', '23529', '23530', '23531', '23532', '23533', '23534', '23535', '23536', '23537', '23538', '23539', '23540', '23541', '23542', '23543', '23544', '23545', '23546', '23547', '23548', '23549', '23550', '23551', '23552', '23553', '23554', '23555', '23556', '23557', '23558', '23559', '23560', '23561', '23562', '23564', '23565', '23566', '23567', '23568', '23569', '23570', '23571', '23574', '23575', '23576', '23578', '23579', '23580', '23581', '23582', '23583', '23597', '23598', '23630', '23660', '23661', '23664', '23681', '23691', '23692', '23694', '23695', '23697', '23702', '23843', '35001G', '35001W', '35004B', '35004C', '35004G', '35004P', '35095A', '35095B', '35241', '35265', '35271S', '35400', '35443', '35471D', '35591T', '35597A', '35597B', '35597D', '35598A', '35598B', '35598C', '35598D', '35599B', '35599D', '35607A', '35607B', '35610A', '35610B', '35610C', '35637A', '35637C', '35638A', '35638B', '35645', '35646', '35647', '35648', '35649', '35650', '35651', '35653', '35809A', '35809B', '35810A', '35810B', '35815P', '35816P', '35817P', '35818B', '35818P', '35819B', '35819P', '35832', '35833P', '35909A', '35909B', '35910A', '35910B', '35911A', '35911B', '35912B', '35913B', '35914', '35915B', '35915C', '35916A', '35916B', '35916C', '35920', '35921', '35922', '35923', '35924', '35933', '35953', '35954', '35957', '35958', '35961', '35964', '35965', '35966', '35967', '35968', '35970', '35971', '35972', '37327', '37330', '37333', '37340', '37342', '37343', '37351', '37370', '37379A', '37413', '37423', '37444A', '37444B', '37444C', '37446', '37447', '37448', '37449', '37450', '37461', '37462E', '37464', '37467', '37468', '37471', '37475', '37476', '37479B', '37479P', '37482P', '37487B', '37488A', '37489A', '37489B', '37489C', '37489D', '37491B', '37491C', '37491D', '37495', '37500', '37501', '37503', '37509', '40001', '40003', '40005B', '40016', '40046A', '44089A', '44089C', '44091A', '44092B', '44092C', '44228', '44234', '44235', '44236', '44242A', '44242B', '44265', '45013', '46000M', '46000P', '46000R', '46000S', '46000U', '46115B', '46118', '46126A', '46138B', '46775D', '46776A', '46776B', '46776C', '46776D', '46776E', '46776F', '47013A', '47013C', '47016', '47021G', '47310M', '47341A', '47341B', '47343A', '47344B', '47348A', '47351B', '47367B', '47369A', '47369B', '47420', '47421', '47422', '47469', '47471', '47480', '47481', '47503A', '47503H', '47503J', '47504H', '47504K', '47518F', '47556B', '47559B', '47563A', '47566', '47566B', '47567B', '47570B', '47574A', '47574B', '47578A', '47579', '47580', '47585A', '47586A', '47589', '47590A', '47590B', '47591B', '47591D', '47593A', '47593B', '47594A', '47594B', '47599A', '47599B', '48111', '48116', '48129', '48138', '48173C', '48184', '48185', '48187', '48188', '48194', '51008', '51014A', '51014C', '51014L', '51020A', '51020B', '62018', '62043B', '62074B', '62086A', '62094B', '62096A', '62096B', '62097B', '70006', '70007', '71038', '71050', '71053', '71101E', '71143', '71215', '71270', '71279', '71403', '71406C', '71459', '71477', '71495A', '71495B', '71496A', '71496B', '71510', '72024U', '72051S', '72122', '72127', '72128', '72130', '72131', '72132', '72133', '72134', '72140E', '72225C', '72232', '72349B', '72351A', '72351B', '72369A', '72586', '72598', '72709', '72741', '72760B', '72780', '72783', '72798C', '72799C', '72799E', '72799F', '72800B', '72800C', '72800D', '72800E', '72801C', '72801D', '72801G', '72802A', '72802B', '72802C', '72803A', '72807A', '72807B', '72807C', '72811', '72812', '72815', '72816', '72817', '72818', '72819', '72821', '75011', '75013B', '75049L', '75131', '75172', '75178', '77079', '77101A', '78033', '78034B', '78124', '79000', '79026B', '79030D', '79030G', '79051A', '79062D', '79063C', '79063D', '79066K', '79067', '79071B', '79144B', '79144C', '79149B', '79151B', '79157B', '79157V', '79160', '79161A', '79163', '79164', '79190A', '79190B', '79190D', '79191B', '79191C', '79191D', '79192A', '79302M', '79320', '79321', '79323B', '79323P', '79323W', '79329', '79331', '79336', '79337', '79403', '79406', '81950B', '81950V', '81952B', '81952V', '81953B', '81953P', '82001S', '82011A', '82011B', '82011C', '82095', '82482', '82483', '82484', '82486', '82494L', '82551', '82552', '82567', '82578', '82580', '82581', '82582', '82583', '82599', '82600', '82605', '82613A', '82613B', '82613C', '82613D', '82615', '82616B', '82616C', '84006', '84012', '84016', '84029E', '84029G', '84030E', '84031A', '84031B', '84032A', '84032B', '84033', '84050', '84051', '84077', '84078A', '84086B', '84086C', '84192', '84199', '84201B', '84201C', '84206A', '84206B', '84206C', '84212', '84218', '84226', '84227', '84228', '84231', '84247E', '84247G', '84247K', '84247L', '84247N', '84249A', '84251B', '84251C', '84251G', '84279B', '84279P', '84306', '84313B', '84313C', '84341B', '84347', '84352', '84356', '84358', '84360', '84375', '84378', '84380', '84387A', '84388', '84402B', '84406B', '84415A', '84415B', '84422', '84429A', '84452', '84457', '84459A', '84459B', '84461', '84462', '84465', '84466', '84497', '84499', '84507B', '84507C', '84508A', '84508B', '84508C', '84509A', '84509B', '84509C', '84509E', '84509G', '84510A', '84510C', '84510E', '84519A', '84519B', '84520B', '84522', '84527', '84531A', '84531B', '84534B', '84535A', '84535B', '84536A', '84536B', '84539', '84548', '84549', '84550', '84551', '84558A', '84559A', '84559B', '84562A', '84563A', '84563B', '84568', '84569A', '84569B', '84569C', '84569D', '84575A', '84576', '84580', '84581', '84584', '84592', '84593', '84595E', '84596B', '84596E', '84596F', '84596G', '84596J', '84596L', '84597B', '84597C', '84598', '84600', '84609', '84613A', '84613C', '84614A', '84616', '84617', '84620', '84625A', '84625C', '84629', '84631', '84632', '84637', '84638', '84658', '84659A', '84660A', '84660B', '84660C', '84661A', '84661B', '84661C', '84663A', '84665', '84666', '84673A', '84673B', '84674', '84675', '84678', '84679', '84683', '84685', '84686', '84687', '84688', '84689', '84691', '84692', '84705C', '84706D', '84706F', '84707A', '84707B', '84708B', '84709B', '84711A', '84711B', '84712B', '84715', '84723', '84725', '84731', '84732B', '84732D', '84741C', '84743C', '84744', '84745A', '84745B', '84746', '84748', '84750A', '84750B', '84751B', '84754', '84755', '84760S', '84763', '84766', '84773', '84789', '84792', '84795B', '84795C', '84796A', '84796B', '84798A', '84798B', '84799', '84800L', '84800M', '84800S', '84801A', '84801B', '84802A', '84802B', '84803B', '84804A', '84804B', '84805A', '84805B', '84806A', '84806B', '84809A', '84809B', '84813', '84816', '84817', '84818', '84819', '84820', '84821', '84823', '84824', '84826', '84827', '84828', '84832', '84836', '84839', '84840', '84842', '84843', '84846A', '84847', '84849A', '84849B', '84849D', '84854', '84856S', '84857B', '84857C', '84858C', '84859A', '84859B', '84859C', '84865', '84870B', '84870C', '84872A', '84873A', '84874B', '84875B', '84875D', '84876B', '84876D', '84877A', '84877B', '84877D', '84879', '84880', '84881', '84882', '84884A', '84898F', '84905', '84906', '84907', '84910A', '84912A', '84912B', '84913A', '84913B', '84915', '84916', '84917', '84919', '84920', '84921', '84922', '84924A', '84924F', '84925D', '84925F', '84926A', '84926D', '84926E', '84926F', '84927A', '84927D', '84927E', '84927F', '84929', '84931A', '84931B', '84944', '84945', '84946', '84947', '84949', '84950', '84951A', '84951B', '84952A', '84952B', '84952C', '84963A', '84963B', '84968A', '84968B', '84968C', '84968D', '84968E', '84968F', '84969', '84970L', '84970S', '84971S', '84974', '84975', '84976', '84978', '84984A', '84984B', '84984D', '84985A', '84987', '84988', '84989A', '84990', '84991', '84992', '84993A', '84993B', '84997A', '84997B', '84997C', '84997D', '85006', '85008', '85014A', '85014B', '85015', '85016', '85017A', '85017B', '85017C', '85018D', '85019A', '85019B', '85019C', '85020', '85023B', '85023C', '85024B', '85024C', '85025B', '85025C', '85026B', '85027L', '85028L', '85028S', '85030', '85031A', '85031B', '85031C', '85032A', '85032B', '85032C', '85032D', '85034A', '85034B', '85034C', '85035A', '85035B', '85035C', '85036A', '85036B', '85036C', '85038', '85039A', '85039B', '85039C', '85040A', '85040B', '85042', '85045', '85047', '85048', '85049A', '85049B', '85049C', '85049D', '85049E', '85049F', '85049G', '85049H', '85053', '85054', '85055', '85059', '85061W', '85062', '85063', '85064', '85065', '85066', '85067', '85068', '85071A', '85071B', '85071C', '85071D', '85078', '85084', '85086A', '85087', '85088', '85089', '85092', '85093', '85094', '85095', '85096', '85098B', '85099B', '85099C', '85099F', '85103', '85104', '85106', '85107', '85109', '85111', '85114A', '85114B', '85114C', '85116', '85118', '85123A', '85124B', '85124C', '85125', '85126', '85127', '85129B', '85129C', '85129D', '85130A', '85130B', '85130C', '85130D', '85131A', '85131B', '85131C', '85131D', '85132A', '85132B', '85132C', '85135A', '85135B', '85135C', '85136A', '85136B', '85136C', '85141', '85144', '85145', '85146', '85150', '85152', '85159A', '85159B', '85160B', '85161', '85163A', '85163B', '85167B', '85168B', '85169A', '85169B', '85169C', '85169D', '85170A', '85170B', '85170C', '85170D', '85172', '85173', '85174', '85175', '85176', '85177', '85178', '85179A', '85179C', '85180A', '85180B', '85183A', '85183B', '85184C', '85185B', '85185D', '85186A', '85186C', '85187', '85188A', '85188B', '85189', '85194L', '85194S', '85195', '85197', '85198', '85199L', '85199S', '85200', '85202', '85203', '85204', '85205B', '85206A', '85206B', '85208', '85211', '85212', '85213', '85214', '85215', '85216', '85224', '85225', '85227', '85230A', '85230B', '85230E', '85230F', '85230G', '85231B', '85231G', '85232A', '85232B', '85232D', '90000A', '90000B', '90000D', '90001A', '90001B', '90001C', '90001D', '90002D', '90003B', '90003C', '90003D', '90003E', '90010A', '90010B', '90010E', '90011A', '90011B', '90012A', '90012B', '90013A', '90013C', '90014A', '90014B', '90014C', '90016A', '90016B', '90018A', '90018B', '90018C', '90019A', '90019B', '90019C', '90021', '90022', '90023', '90024B', '90024C', '90024D', '90024F', '90025A', '90025F', '90026A', '90026B', '90026C', '90026D', '90027A', '90027D', '90028', '90029', '90030A', '90030B', '90030C', '90031', '90032', '90033', '90035A', '90035C', '90036A', '90036B', '90036C', '90036D', '90036E', '90036F', '90037B', '90038A', '90038B', '90038C', '90039A', '90039B', '90039C', '90039D', '90040A', '90040B', '90040C', '90041', '90042A', '90043', '90046', '90048', '90049', '90050', '90051', '90054', '90055', '90057', '90058A', '90058B', '90059A', '90059B', '90059C', '90059D', '90059E', '90059F', '90060B', '90060D', '90062', '90063A', '90063B', '90064A', '90064B', '90065A', '90065B', '90067A', '90069', '90070', '90071', '90072', '90073', '90074', '90076', '90078', '90079', '90081A', '90081C', '90082A', '90082B', '90082D', '90083', '90084', '90085', '90086', '90087', '90089', '90093', '90094', '90096', '90098', '90099', '90100', '90101', '90102', '90103', '90104', '90108', '90114', '90116', '90118', '90119', '90120B', '90120C', '90120D', '90122A', '90122B', '90122C', '90123A', '90123B', '90123C', '90123D', '90124A', '90124B', '90124C', '90125A', '90125B', '90125C', '90125D', '90125E', '90128B', '90129A', '90129B', '90129C', '90129D', '90129E', '90129F', '90130A', '90130B', '90131', '90133', '90135', '90136', '90138', '90141A', '90141B', '90141C', '90141D', '90141E', '90142D', '90143', '90144', '90145', '90146', '90147', '90148', '90149', '90151', '90152A', '90152B', '90152C', '90154', '90155', '90156', '90157', '90158', '90159', '90160A', '90160B', '90160C', '90160D', '90161A', '90161B', '90161C', '90161D', '90162A', '90162B', '90162C', '90162D', '90163A', '90163B', '90164A', '90164B', '90165B', '90166', '90167', '90168', '90169', '90170', '90173', '90174', '90175A', '90175C', '90175D', '90176A', '90176B', '90176C', '90176D', '90176E', '90177A', '90177B', '90177C', '90177D', '90177E', '90178A', '90178B', '90179A', '90179C', '90180A', '90180B', '90181A', '90182C', '90183A', '90183C', '90184A', '90184B', '90184C', '90185A', '90185B', '90185C', '90185D', '90186A', '90186B', '90187B', '90188', '90189A', '90190A', '90190B', '90190C', '90191', '90192', '90194', '90195A', '90195B', '90196A', '90196B', '90197B', '90198A', '90198B', '90199A', '90199B', '90199C', '90199D', '90200A', '90200B', '90200C', '90200D', '90200E', '90201A', '90201B', '90201C', '90201D', '90202A', '90202B', '90202C', '90202D', '90204', '90205A', '90205C', '90206A', '90206C', '90208', '90209A', '90209B', '90209C', '90210A', '90210B', '90210C', '90210D', '90211A', '90211B', '90212B', '90212C', '90214A', '90214B', '90214C', '90214D', '90214E', '90214F', '90214G', '90214H', '90214I', '90214J', '90214K', '90214L', '90214M', '90214N', '90214O', '90214P', '90214R', '90214S', '90214T', '90214U', '90214V', '90214W', '90214Y', '90214Z', 'BANK CHARGES', 'C2', 'CRUK', 'D', 'DOT', 'M', 'PADS', 'POST'])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 83
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BJyuZm-6UKeP",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0dc012a5-16a6-495f-d55b-4840ae7f0fbb"
      },
      "source": [
        "# Aqui vamos verificar o produto mais semelhante\n",
        "# ao produto de código 23188.\n",
        "\n",
        "most_similar('23188')"
      ],
      "execution_count": 84,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Most similar to: ['VINTAGE  2 METRE FOLDING RULER']\n",
            "['BUNDLE OF 3 SCHOOL EXERCISE BOOKS  '] 0.7492429614067078\n",
            "['CLASSIC BICYCLE CLIPS '] 0.7391677498817444\n",
            "['SET 8 CANDLES VINTAGE DOILY'] 0.7351871132850647\n",
            "['BICYCLE PUNCTURE REPAIR KIT '] 0.7292545437812805\n",
            "['BUNDLE OF 3 ALPHABET EXERCISE BOOKS'] 0.7267224788665771\n",
            "['IVORY RETRO KITCHEN WALL CLOCK'] 0.7207701206207275\n",
            "['CLASSIC CHROME BICYCLE BELL '] 0.7113609313964844\n",
            "['POPPY FIELDS CHOPPING BOARD'] 0.7054341435432434\n",
            "['TRADITIONAL PICK UP STICKS GAME '] 0.6969001889228821\n",
            "['BLUE RETRO KITCHEN WALL CLOCK'] 0.6852618455886841\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "G8LcgPfnCw4m"
      },
      "source": [
        "## Passo 5.2: Sugestão de produtos\n",
        "\n",
        "Agora sim, podemos fazer nosso sistema de sugestões de compras. Na variável orders_val temos os conjuntos de itens em cada compra, de cada cliente. Por exemplo, abaixo, podemos listar as compras da ordem número 32, no código abaixo:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "41eY2zY1UjIT",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c098c525-ea8c-4ec3-c9a8-352b5ef481ab"
      },
      "source": [
        "# Lista os códigos dos produtos comprados\n",
        "# na ordem de compras número 32.\n",
        "\n",
        "orders_val[32]"
      ],
      "execution_count": 85,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['23399',\n",
              " '23400',\n",
              " '23403',\n",
              " '23427',\n",
              " '23432',\n",
              " '23322',\n",
              " '23321',\n",
              " '20846',\n",
              " '21313',\n",
              " '84946',\n",
              " '23088',\n",
              " '22114',\n",
              " '22835',\n",
              " '22469',\n",
              " '84406B',\n",
              " '23072']"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 85
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jaQ58JsNDY5i"
      },
      "source": [
        "Para sugerir compras, basta recuperarmos os vetores de embedding de cada um desses itens de uma ordem de compra, calcular um vetor médio (soma todos vetores e divide pela quantidade) e buscar os produtos mais semelhantes, como fizemos antes!"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9Me8QvWIVXi-"
      },
      "source": [
        "def get_suggestions(orders):\n",
        "  ''' Nesta função, passa-se como argumento uma lista\n",
        "      de códigos de produtos, que representa as compras\n",
        "      de um cliente, em um determinado momento. Com base\n",
        "      nos embeddings desses itens, calculamos um vetor\n",
        "      médio, que captura a \"semântica\" das compras e\n",
        "      baseado nesse vetor, buscamos os produtos mais\n",
        "      similares para sugerir. Apenas sugerimos os\n",
        "      produtos que já não estiverem na lista.\n",
        "  '''\n",
        "  # Esse será o array que armazenará o embedding médio\n",
        "  v = np.zeros((50))\n",
        "\n",
        "  # Essa variável vai receber a quantidade de produtos,\n",
        "  # para calcular a média mais tarde.\n",
        "  count = 0\n",
        "\n",
        "  # Aqui mostramos as descrições dos produtos já na lista\n",
        "  # para facilitar a conferência.\n",
        "  print('Orders so far:')\n",
        "  for product in orders:\n",
        "    print(products_dict[product])\n",
        "    # Para cada produto no embedding (aqui testamos,\n",
        "    # para evitar incluir um produto que não esteja\n",
        "    # no embedding treinado, pois isso geraria um erro)\n",
        "    if product in model.wv:\n",
        "      # somamos o vetor de embedding daquele produto\n",
        "      # no array v\n",
        "      v += model.wv[product]\n",
        "      # e contamos +1 em count\n",
        "      count += 1\n",
        "  # Dividmos o resultado por count para calcular\n",
        "  # o vetor médio, no espaço de embeddings\n",
        "  v /= count\n",
        "  # E por fim buscamos os 10 produtos mais semelhantes\n",
        "  # como fizemos originalmente.\n",
        "  s = model.similar_by_vector(v)\n",
        "  print('Suggestions:')\n",
        "  for product, prob in s:\n",
        "    # Mostramos apenas os produtos que já não estiverem\n",
        "    # na lista original.\n",
        "    if product not in orders:\n",
        "      # Mostramos as descrições para facilitar a leitura\n",
        "      print(products_dict[product], prob)"
      ],
      "execution_count": 86,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-i31ZVGuFCH7"
      },
      "source": [
        "Abaixo testamos o código de sugestões definido acima."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MEku0QXUXhxp",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3c541627-e0b7-40ca-ad2f-6d078bed291e"
      },
      "source": [
        "get_suggestions(orders_val[32])"
      ],
      "execution_count": 87,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Orders so far:\n",
            "['HOME SWEET HOME HANGING HEART']\n",
            "['SHELF WITH 4 HOOKS HOME SWEET HOME']\n",
            "['LETTER HOLDER HOME SWEET HOME']\n",
            "['STOOL HOME SWEET HOME ']\n",
            "['PRETTY HANGING QUILTED HEARTS']\n",
            "['LARGE WHITE HEART OF WICKER']\n",
            "['SMALL WHITE HEART OF WICKER']\n",
            "['ZINC HEART LATTICE T-LIGHT HOLDER']\n",
            "['GLASS HEART T-LIGHT HOLDER ']\n",
            "['ANTIQUE SILVER T-LIGHT GLASS']\n",
            "['ZINC HEART FLOWER T-LIGHT HOLDER']\n",
            "['HOT WATER BOTTLE TEA AND SYMPATHY']\n",
            "['HOT WATER BOTTLE I AM SO POORLY']\n",
            "['HEART OF WICKER SMALL']\n",
            "['CREAM CUPID HEARTS COAT HANGER']\n",
            "['MARIE ANTOINETTE TRINKET BOX SILVER']\n",
            "Suggestions:\n",
            "['HOME SWEET HOME BLACKBOARD'] 0.8711927533149719\n",
            "['HOME SWEET HOME 2 DRAWER CABINET'] 0.835765540599823\n",
            "['HOME SWEET HOME KEY HOLDER'] 0.8346580266952515\n",
            "['SET OF 2 TRAYS HOME SWEET HOME'] 0.828395426273346\n",
            "['HOME SWEET HOME 3 PEG HANGER '] 0.8246036767959595\n",
            "['RUSTIC MIRROR WITH LACE HEART'] 0.8196208477020264\n",
            "['HEART MIRROR ANTIQUE WHITE'] 0.8156135678291321\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-DzmHZYJVu5N",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0eb4c816-ac34-4a80-d705-6c9230c10531"
      },
      "source": [
        "get_suggestions(orders_val[67])"
      ],
      "execution_count": 88,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Orders so far:\n",
            "['SET OF 3 CAKE TINS PANTRY DESIGN ']\n",
            "['TOTE BAG I LOVE LONDON']\n",
            "['LETS GO SHOPPING COTTON TOTE BAG']\n",
            "['ROMANTIC PINKS RIBBONS ']\n",
            "['RIBBON REEL STRIPES DESIGN ']\n",
            "['RETROSPOT WOODEN HEART DECORATION']\n",
            "['ZINC METAL HEART DECORATION']\n",
            "['IVORY HANGING DECORATION  BIRD']\n",
            "['HEART IVORY TRELLIS SMALL']\n",
            "['IVORY HANGING DECORATION  HEART']\n",
            "['HEART IVORY TRELLIS LARGE']\n",
            "['HANGING HEART ZINC T-LIGHT HOLDER']\n",
            "['GYMKHANA TREASURE BOOK BOX']\n",
            "['MOUSE TOY WITH PINK T-SHIRT']\n",
            "['FELTCRAFT CUSHION BUTTERFLY']\n",
            "['FELTCRAFT CUSHION RABBIT']\n",
            "['FELTCRAFT CUSHION OWL']\n",
            "['SET OF 3 NOTEBOOKS IN PARCEL']\n",
            "['CLOTHES PEGS RETROSPOT PACK 24 ']\n",
            "['36 PENCILS TUBE RED RETROSPOT']\n",
            "['36 PENCILS TUBE RED RETROSPOT']\n",
            "['JAZZ HEARTS PURSE NOTEBOOK']\n",
            "['CREAM HANGING HEART T-LIGHT HOLDER']\n",
            "['HOME BUILDING BLOCK WORD']\n",
            "['BATH BUILDING BLOCK WORD']\n",
            "['ASSORTED COLOUR MINI CASES']\n",
            "['APRON MODERN VINTAGE COTTON']\n",
            "['SET/4 MODERN VINTAGE COTTON NAPKINS']\n",
            "['GIN AND TONIC DIET METAL SIGN']\n",
            "['HOT BATHS METAL SIGN']\n",
            "['AREA PATROLLED METAL SIGN']\n",
            "['HAND OVER THE CHOCOLATE   SIGN ']\n",
            "['METAL SIGN HIS DINNER IS SERVED']\n",
            "['NO JUNK MAIL METAL SIGN']\n",
            "['PLEASE ONE PERSON METAL SIGN']\n",
            "['COOK WITH WINE METAL SIGN ']\n",
            "['PAPER CHAIN KIT EMPIRE']\n",
            "['COFFEE MUG CAT + BIRD DESIGN']\n",
            "['CHILDRENS APRON APPLES DESIGN']\n",
            "['ADULT APRON APPLE DELIGHT']\n",
            "['OVEN MITT APPLES DESIGN']\n",
            "['BLACK TEA TOWEL CLASSIC DESIGN']\n",
            "['CHILDRENS APRON SPACEBOY DESIGN']\n",
            "['RED TEA TOWEL CLASSIC DESIGN']\n",
            "['VICTORIAN GLASS HANGING T-LIGHT']\n",
            "['SMALL DECO JEWELLERY STAND']\n",
            "['SMALL WHITE HEART OF WICKER']\n",
            "['LARGE WHITE HEART OF WICKER']\n",
            "['GIN AND TONIC DIET METAL SIGN']\n",
            "['AREA PATROLLED METAL SIGN']\n",
            "['LAUNDRY 15C METAL SIGN']\n",
            "['HAND OVER THE CHOCOLATE   SIGN ']\n",
            "['BEWARE OF THE CAT METAL SIGN ']\n",
            "['N0 SINGING METAL SIGN']\n",
            "[\"I'M ON HOLIDAY METAL SIGN\"]\n",
            "['MOUSE TOY WITH PINK T-SHIRT']\n",
            "['DOG TOY WITH PINK CROCHET SKIRT']\n",
            "['BATH BUILDING BLOCK WORD']\n",
            "['GARDENERS KNEELING PAD KEEP CALM ']\n",
            "['GARDENERS KNEELING PAD CUP OF TEA ']\n",
            "['LUNCH BAG  BLACK SKULL.']\n",
            "['LUNCH BAG APPLE DESIGN']\n",
            "['LUNCH BAG PINK POLKADOT']\n",
            "['LUNCH BAG VINTAGE DOILY ']\n",
            "['LUNCH BAG RED RETROSPOT']\n",
            "['LUNCH BAG SPACEBOY DESIGN ']\n",
            "['HANGING ENGRAVED METAL HEART']\n",
            "['VINTAGE EMBOSSED HEART']\n",
            "['ALUMINIUM STAMPED HEART']\n",
            "['TRAVEL CARD WALLET PANTRY']\n",
            "['TRAVEL CARD WALLET VINTAGE LEAF']\n",
            "['BLACK TEA TOWEL CLASSIC DESIGN']\n",
            "['PARTY METAL SIGN ']\n",
            "['CREAM HANGING HEART T-LIGHT HOLDER']\n",
            "['LARGE WHITE HEART OF WICKER']\n",
            "['LUNCH BAG ALPHABET DESIGN']\n",
            "['LUNCH BAG VINTAGE DOILY ']\n",
            "['LUNCH BAG VINTAGE LEAF DESIGN']\n",
            "['LUNCH BAG CARS BLUE']\n",
            "['TRAVEL CARD WALLET PANTRY']\n",
            "['TRAVEL CARD WALLET UNION JACK']\n",
            "['TRAVEL CARD WALLET KEEP CALM']\n",
            "['TRAVEL CARD WALLET VINTAGE LEAF']\n",
            "['TRAVEL CARD WALLET SUKI']\n",
            "['LUNCH BAG RED RETROSPOT']\n",
            "['LUNCH BAG PINK POLKADOT']\n",
            "['LUNCH BAG APPLE DESIGN']\n",
            "['RIBBON REEL POLKADOTS ']\n",
            "['ZINC METAL HEART DECORATION']\n",
            "['ADULT APRON APPLE DELIGHT']\n",
            "['VICTORIAN SEWING BOX SMALL ']\n",
            "['GIN AND TONIC DIET METAL SIGN']\n",
            "['METAL SIGN TAKE IT OR LEAVE IT ']\n",
            "['POTTERING IN THE SHED METAL SIGN']\n",
            "[\"I'M ON HOLIDAY METAL SIGN\"]\n",
            "['BEWARE OF THE CAT METAL SIGN ']\n",
            "['WOOD 2 DRAWER CABINET WHITE FINISH']\n",
            "['RED RETROSPOT PICNIC BAG']\n",
            "['SMALL PARISIENNE HEART PHOTO FRAME ']\n",
            "['LARGE WHITE HEART OF WICKER']\n",
            "['VICTORIAN GLASS HANGING T-LIGHT']\n",
            "['VINTAGE LEAF MAGNETIC NOTEPAD']\n",
            "['SKETCHBOOK MAGNETIC SHOPPING LIST']\n",
            "['PANTRY MAGNETIC  SHOPPING LIST']\n",
            "['SET 36 COLOUR PENCILS SPACEBOY ']\n",
            "['SET 36 COLOUR PENCILS DOLLY GIRL']\n",
            "['SET 36 COLOUR PENCILS LOVE LONDON']\n",
            "['LARGE WHITE HEART OF WICKER']\n",
            "['IVORY DINER WALL CLOCK']\n",
            "['POPCORN HOLDER']\n",
            "['PLAYING CARDS I LOVE LONDON ']\n",
            "['PLAYING CARDS JUBILEE UNION JACK']\n",
            "['PLAYING CARDS KEEP CALM & CARRY ON']\n",
            "['SET OF 3 CAKE TINS PANTRY DESIGN ']\n",
            "['WOODEN STAR CHRISTMAS SCANDINAVIAN']\n",
            "['WOODEN TREE CHRISTMAS SCANDINAVIAN']\n",
            "['JUMBO BAG VINTAGE CHRISTMAS ']\n",
            "[\"JUMBO BAG 50'S CHRISTMAS \"]\n",
            "['TRIPLE PHOTO FRAME CORNICE ']\n",
            "['PLEASE ONE PERSON METAL SIGN']\n",
            "['METAL SIGN DROP YOUR PANTS']\n",
            "['HAND OVER THE CHOCOLATE   SIGN ']\n",
            "['SET OF 2 CERAMIC CHRISTMAS REINDEER']\n",
            "['VINTAGE CHRISTMAS STOCKING ']\n",
            "['SILVER MUG BONE CHINA TREE OF LIFE']\n",
            "['IVORY WICKER HEART LARGE']\n",
            "['TRAVEL CARD WALLET KEEP CALM']\n",
            "['TRAVEL CARD WALLET I LOVE LONDON']\n",
            "['TRAVEL CARD WALLET UNION JACK']\n",
            "['TRAVEL CARD WALLET VINTAGE LEAF']\n",
            "['SET OF 3 WOODEN TREE DECORATIONS']\n",
            "['SET OF 3 WOODEN SLEIGH DECORATIONS']\n",
            "['SET OF 3 WOODEN STOCKING DECORATION']\n",
            "['DOUGHNUT LIP GLOSS ']\n",
            "['LOVE HOT WATER BOTTLE']\n",
            "['ALARM CLOCK BAKELIKE IVORY']\n",
            "['ALARM CLOCK BAKELIKE RED ']\n",
            "['VICTORIAN GLASS HANGING T-LIGHT']\n",
            "['HAND WARMER SCOTTY DOG DESIGN']\n",
            "['HAND WARMER RED LOVE HEART']\n",
            "['SCOTTIE DOG HOT WATER BOTTLE']\n",
            "['RETROSPOT HEART HOT WATER BOTTLE']\n",
            "['METAL MERRY CHRISTMAS WREATH']\n",
            "['VICTORIAN SEWING BOX SMALL ']\n",
            "['NURSERY A,B,C PAINTED LETTERS']\n",
            "Suggestions:\n",
            "['LADIES & GENTLEMEN METAL SIGN'] 0.6195432543754578\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "D6rk6No3XkpS",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "123dbd12-c9d7-40a7-c417-85037b91eace"
      },
      "source": [
        "get_suggestions(orders_val[88])"
      ],
      "execution_count": 89,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Orders so far:\n",
            "['DOLLY GIRL LUNCH BOX']\n",
            "['SPACEBOY LUNCH BOX ']\n",
            "['JUMBO BAG PINK VINTAGE PAISLEY']\n",
            "['PACK OF 12 TRADITIONAL CRAYONS']\n",
            "['PARTY BUNTING']\n",
            "['PINK GINGHAM CAT WITH SCARF']\n",
            "['GREEN REGENCY TEACUP AND SAUCER']\n",
            "['PINK REGENCY TEACUP AND SAUCER']\n",
            "['REGENCY TEA PLATE GREEN ']\n",
            "['REGENCY CAKESTAND 3 TIER']\n",
            "['MEDIUM CERAMIC TOP STORAGE JAR']\n",
            "['FRENCH FLORAL CUSHION COVER ']\n",
            "['CLASSIC FRENCH STYLE BASKET GREEN ']\n",
            "['VINTAGE BEAD PINK PURSE ']\n",
            "['SCOTTIE DOGS BABY BIB']\n",
            "['PINK MONTE CARLO HANDBAG']\n",
            "['WHITE TEA,COFFEE,SUGAR JARS']\n",
            "['PANTRY WASHING UP BRUSH']\n",
            "['HOMEMADE JAM SCENTED CANDLES']\n",
            "['BAKING SET 9 PIECE RETROSPOT ']\n",
            "['RETROSPOT TEA SET CERAMIC 11 PC ']\n",
            "['WOOD STAMP SET HAPPY BIRTHDAY']\n",
            "[\"WRAP 50'S  CHRISTMAS\"]\n",
            "['VINTAGE CHRISTMAS BUNTING']\n",
            "['VINTAGE CHRISTMAS STOCKING ']\n",
            "['GREEN REGENCY TEACUP AND SAUCER']\n",
            "Suggestions:\n",
            "['16 PIECE CUTLERY SET PANTRY DESIGN'] 0.7345637083053589\n",
            "['REGENCY TEA PLATE ROSES '] 0.7210922241210938\n",
            "['REGENCY TEA PLATE PINK'] 0.7179951071739197\n",
            "['REGENCY TEAPOT ROSES '] 0.7101251482963562\n",
            "['REGENCY SUGAR BOWL GREEN'] 0.6866954565048218\n",
            "['REGENCY MILK JUG PINK '] 0.6832743287086487\n",
            "['REGENCY CAKE FORK'] 0.6806017756462097\n",
            "['ANTIQUE IVORY WIRE BOWL SMALL'] 0.67032790184021\n",
            "['REGENCY CAKE SLICE'] 0.6616465449333191\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}