Movie_Rating_Prediction/Movie_Rating_Prediction.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "0CaIbEZrgx0c"
   },
   "source": [
    "# Predição de Classificação de Filme / *Movie Rating Prediction*\n",
    "\n",
    "# Desafio de Ciência de Dados / *Data Science Challenge*\n",
    "\n",
    "# Entrega / *Delivery*: Notebook\n",
    "\n",
    "# Desenvolvedor / *Developer*: Solano Jacon"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 1267,
     "status": "ok",
     "timestamp": 1661397622826,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "K6gAaw3Ngx0d",
    "outputId": "12deca9f-4c04-4ee7-c0b2-59c8ae13635a"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[nltk_data] Downloading package stopwords to\n",
      "[nltk_data]     /home/solanojacon/nltk_data...\n",
      "[nltk_data]   Package stopwords is already up-to-date!\n",
      "[nltk_data] Downloading package punkt to\n",
      "[nltk_data]     /home/solanojacon/nltk_data...\n",
      "[nltk_data]   Package punkt is already up-to-date!\n",
      "[nltk_data] Downloading package wordnet to\n",
      "[nltk_data]     /home/solanojacon/nltk_data...\n",
      "[nltk_data]   Package wordnet is already up-to-date!\n",
      "[nltk_data] Downloading package omw-1.4 to\n",
      "[nltk_data]     /home/solanojacon/nltk_data...\n",
      "[nltk_data]   Package omw-1.4 is already up-to-date!\n"
     ]
    }
   ],
   "source": [
    "# The installation of the openpyxl library would be necessary for the command pd.read_excel(file, engine='openpyxl') to work\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import string\n",
    "import nltk\n",
    "import warnings\n",
    "from IPython.display import Image, display\n",
    "from sklearn.preprocessing import OneHotEncoder\n",
    "from nltk.corpus import stopwords \n",
    "from nltk import word_tokenize\n",
    "from nltk.stem import WordNetLemmatizer\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.naive_bayes import MultinomialNB\n",
    "from sklearn.model_selection import cross_validate\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score\n",
    "from sklearn.metrics import classification_report\n",
    "from sklearn.metrics import precision_recall_curve\n",
    "\n",
    "nltk.download('stopwords')\n",
    "nltk.download('punkt')\n",
    "nltk.download('wordnet')\n",
    "nltk.download('omw-1.4')\n",
    "\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "gPknSn6zgx0k"
   },
   "source": [
    "## Leitura dos dados / Data reading"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 337
    },
    "executionInfo": {
     "elapsed": 2222,
     "status": "ok",
     "timestamp": 1661397625046,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "jz5Y1J1Wgx0k",
    "outputId": "06ca5d83-308b-4c6e-d0b7-630e2f1fcd92"
   },
   "outputs": [
    {
     "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>show_id</th>\n",
       "      <th>type</th>\n",
       "      <th>title</th>\n",
       "      <th>director</th>\n",
       "      <th>cast</th>\n",
       "      <th>country</th>\n",
       "      <th>date_added</th>\n",
       "      <th>release_year</th>\n",
       "      <th>rating</th>\n",
       "      <th>duration</th>\n",
       "      <th>listed_in</th>\n",
       "      <th>description</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>81145628</td>\n",
       "      <td>Movie</td>\n",
       "      <td>Norm of the North: King Sized Adventure</td>\n",
       "      <td>Richard Finn, Tim Maltby</td>\n",
       "      <td>Alan Marriott, Andrew Toth, Brian Dobson, Cole...</td>\n",
       "      <td>United States, India, South Korea, China</td>\n",
       "      <td>2019-09-09</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>41.0</td>\n",
       "      <td>90 min</td>\n",
       "      <td>Children &amp; Family Movies, Comedies</td>\n",
       "      <td>Before planning an awesome wedding for his gra...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>80117401</td>\n",
       "      <td>Movie</td>\n",
       "      <td>Jandino: Whatever it Takes</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Jandino Asporaat</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>2016-09-09</td>\n",
       "      <td>2016.0</td>\n",
       "      <td>52.0</td>\n",
       "      <td>94 min</td>\n",
       "      <td>Stand-Up Comedy</td>\n",
       "      <td>Jandino Asporaat riffs on the challenges of ra...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>70234439</td>\n",
       "      <td>TV Show</td>\n",
       "      <td>Transformers Prime</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Peter Cullen, Sumalee Montano, Frank Welker, J...</td>\n",
       "      <td>United States</td>\n",
       "      <td>2018-09-08</td>\n",
       "      <td>2013.0</td>\n",
       "      <td>82.0</td>\n",
       "      <td>1 Season</td>\n",
       "      <td>Kids' TV</td>\n",
       "      <td>With the help of three human allies, the Autob...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>80058654</td>\n",
       "      <td>TV Show</td>\n",
       "      <td>Transformers: Robots in Disguise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Will Friedle, Darren Criss, Constance Zimmer, ...</td>\n",
       "      <td>United States</td>\n",
       "      <td>2018-09-08</td>\n",
       "      <td>2016.0</td>\n",
       "      <td>64.0</td>\n",
       "      <td>1 Season</td>\n",
       "      <td>Kids' TV</td>\n",
       "      <td>When a prison ship crash unleashes hundreds of...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>80125979</td>\n",
       "      <td>Movie</td>\n",
       "      <td>#realityhigh</td>\n",
       "      <td>Fernando Lebrija</td>\n",
       "      <td>Nesta Cooper, Kate Walsh, John Michael Higgins...</td>\n",
       "      <td>United States</td>\n",
       "      <td>2017-09-08</td>\n",
       "      <td>2017.0</td>\n",
       "      <td>57.0</td>\n",
       "      <td>99 min</td>\n",
       "      <td>Comedies</td>\n",
       "      <td>When nerdy high schooler Dani finally attracts...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    show_id     type                                    title  \\\n",
       "0  81145628    Movie  Norm of the North: King Sized Adventure   \n",
       "1  80117401    Movie               Jandino: Whatever it Takes   \n",
       "2  70234439  TV Show                       Transformers Prime   \n",
       "3  80058654  TV Show         Transformers: Robots in Disguise   \n",
       "4  80125979    Movie                             #realityhigh   \n",
       "\n",
       "                   director  \\\n",
       "0  Richard Finn, Tim Maltby   \n",
       "1                       NaN   \n",
       "2                       NaN   \n",
       "3                       NaN   \n",
       "4          Fernando Lebrija   \n",
       "\n",
       "                                                cast  \\\n",
       "0  Alan Marriott, Andrew Toth, Brian Dobson, Cole...   \n",
       "1                                   Jandino Asporaat   \n",
       "2  Peter Cullen, Sumalee Montano, Frank Welker, J...   \n",
       "3  Will Friedle, Darren Criss, Constance Zimmer, ...   \n",
       "4  Nesta Cooper, Kate Walsh, John Michael Higgins...   \n",
       "\n",
       "                                    country date_added  release_year  rating  \\\n",
       "0  United States, India, South Korea, China 2019-09-09        2019.0    41.0   \n",
       "1                            United Kingdom 2016-09-09        2016.0    52.0   \n",
       "2                             United States 2018-09-08        2013.0    82.0   \n",
       "3                             United States 2018-09-08        2016.0    64.0   \n",
       "4                             United States 2017-09-08        2017.0    57.0   \n",
       "\n",
       "   duration                           listed_in  \\\n",
       "0    90 min  Children & Family Movies, Comedies   \n",
       "1    94 min                     Stand-Up Comedy   \n",
       "2  1 Season                            Kids' TV   \n",
       "3  1 Season                            Kids' TV   \n",
       "4    99 min                            Comedies   \n",
       "\n",
       "                                         description  \n",
       "0  Before planning an awesome wedding for his gra...  \n",
       "1  Jandino Asporaat riffs on the challenges of ra...  \n",
       "2  With the help of three human allies, the Autob...  \n",
       "3  When a prison ship crash unleashes hundreds of...  \n",
       "4  When nerdy high schooler Dani finally attracts...  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Importação dos dados sem nenhum tratamento prévio\n",
    "# Import of data without any previous treatment\n",
    "\n",
    "data_raw = pd.read_excel('../../portfolio-data/Movie_Rating_Prediction/data/dataset_netflix.xlsx', engine='openpyxl')\n",
    "data_raw.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 23,
     "status": "ok",
     "timestamp": 1661397625047,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "ZfLpynWTgx0k",
    "outputId": "7abcc403-e0b2-4be2-a22e-ab7430c17c43"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 6234 entries, 0 to 6233\n",
      "Data columns (total 12 columns):\n",
      " #   Column        Non-Null Count  Dtype         \n",
      "---  ------        --------------  -----         \n",
      " 0   show_id       6234 non-null   int64         \n",
      " 1   type          6234 non-null   object        \n",
      " 2   title         6234 non-null   object        \n",
      " 3   director      4264 non-null   object        \n",
      " 4   cast          6233 non-null   object        \n",
      " 5   country       6233 non-null   object        \n",
      " 6   date_added    6222 non-null   datetime64[ns]\n",
      " 7   release_year  6233 non-null   float64       \n",
      " 8   rating        6233 non-null   float64       \n",
      " 9   duration      6233 non-null   object        \n",
      " 10  listed_in     6233 non-null   object        \n",
      " 11  description   6233 non-null   object        \n",
      "dtypes: datetime64[ns](1), float64(2), int64(1), object(8)\n",
      "memory usage: 584.6+ KB\n"
     ]
    }
   ],
   "source": [
    "# Verificação dos tipos e valores nulos dos dados\n",
    "# Checking data types and null values\n",
    "\n",
    "data_raw.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "eNZEvc2Vgx0l"
   },
   "source": [
    "## Tratamento dos dados / *Data processing*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 22,
     "status": "ok",
     "timestamp": 1661397625047,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "uVRggJkbgx0l",
    "outputId": "2c4e8a43-8685-4106-d92b-76cfb6d323f1"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 6233 entries, 0 to 6232\n",
      "Data columns (total 12 columns):\n",
      " #   Column        Non-Null Count  Dtype         \n",
      "---  ------        --------------  -----         \n",
      " 0   show_id       6233 non-null   int64         \n",
      " 1   type          6233 non-null   object        \n",
      " 2   title         6233 non-null   object        \n",
      " 3   director      4264 non-null   object        \n",
      " 4   cast          6233 non-null   object        \n",
      " 5   country       6233 non-null   object        \n",
      " 6   date_added    6222 non-null   datetime64[ns]\n",
      " 7   release_year  6233 non-null   float64       \n",
      " 8   rating        6233 non-null   float64       \n",
      " 9   duration      6233 non-null   object        \n",
      " 10  listed_in     6233 non-null   object        \n",
      " 11  description   6233 non-null   object        \n",
      "dtypes: datetime64[ns](1), float64(2), int64(1), object(8)\n",
      "memory usage: 584.5+ KB\n"
     ]
    }
   ],
   "source": [
    "# Eliminação dos filmes sem classificação, e verificação dos tipos e valores nulos dos dados\n",
    "# Elimination of unrated movies, and verification of data types and null values\n",
    "\n",
    "data = data_raw.copy()\n",
    "data = data.dropna(subset=['rating']).reset_index(drop=True)\n",
    "data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 21,
     "status": "ok",
     "timestamp": 1661397625048,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "y7s4KpxFgx0l",
    "outputId": "7d20898e-7da1-46b8-b146-2d99b9d8c793"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['Movie', 'TV Show'], dtype=object)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Verificação dos valores únicos da coluna type\n",
    "# Checking the unique values of the type column\n",
    "\n",
    "data['type'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 19,
     "status": "ok",
     "timestamp": 1661397625048,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "Rwhbv7yQgx0l",
    "outputId": "ec130fad-31d4-478e-c1e8-efb92269a494"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['90 min', '94 min', '1 Season', '99 min', '110 min', '60 min',\n",
       "       '78 min', '95 min', '58 min', '62 min', '65 min', '61 min',\n",
       "       '135 min', '98 min', '96 min', '79 min', '113 min', '80 min',\n",
       "       '77 min', '112 min', '106 min', '102 min', '114 min', '125 min',\n",
       "       '142 min', '133 min', '100 min', '86 min', '146 min', '10 min',\n",
       "       '122 min', '126 min', '116 min', '130 min', '70 min', '128 min',\n",
       "       '88 min', '108 min', '93 min', '64 min', '107 min', '92 min',\n",
       "       '111 min', '103 min', '40 min', '85 min', '72 min', '45 min',\n",
       "       '89 min', '24 min', '127 min', '163 min', '117 min', '42 min',\n",
       "       '104 min', '75 min', '82 min', '38 min', '97 min', '81 min',\n",
       "       '91 min', '152 min', '87 min', '121 min', '101 min', '119 min',\n",
       "       '83 min', '182 min', '124 min', '63 min', '139 min', '171 min',\n",
       "       '84 min', '69 min', '76 min', '41 min', '153 min', '141 min',\n",
       "       '52 min', '137 min', '132 min', '157 min', '109 min', '105 min',\n",
       "       '25 min', '71 min', '123 min', '67 min', '140 min', '138 min',\n",
       "       '149 min', '47 min', '68 min', '148 min', '54 min', '151 min',\n",
       "       '155 min', '162 min', '28 min', '160 min', '29 min', '57 min',\n",
       "       '46 min', '73 min', '74 min', '118 min', '66 min', '14 min',\n",
       "       '20 min', '115 min', '48 min', '168 min', '144 min', '161 min',\n",
       "       '55 min', '56 min', '51 min', '50 min', '23 min', '205 min',\n",
       "       '190 min', '131 min', '22 min', '59 min', '150 min', '159 min',\n",
       "       '145 min', '158 min', '143 min', '154 min', '147 min', '129 min',\n",
       "       '214 min', '136 min', '134 min', '44 min', '156 min', '30 min',\n",
       "       '120 min', '209 min', '166 min', '200 min', '185 min', '36 min',\n",
       "       '11 min', '53 min', '165 min', '164 min', '19 min', '177 min',\n",
       "       '32 min', '18 min', '170 min', '187 min', '173 min', '12 min',\n",
       "       '3 min', '176 min', '15 min', '49 min', '26 min', '37 min',\n",
       "       '189 min', '201 min', '179 min', '191 min', '193 min', '192 min',\n",
       "       '43 min', '172 min', '224 min', '27 min', '181 min', '178 min',\n",
       "       '228 min', '34 min', '312 min', '174 min', '169 min', '35 min',\n",
       "       '31 min', '180 min', '196 min', '167 min', '33 min', '203 min',\n",
       "       '195 min', '2 Seasons', '5 Seasons', '3 Seasons', '7 Seasons',\n",
       "       '4 Seasons', '8 Seasons', '6 Seasons', '9 Seasons', '14 Seasons',\n",
       "       '10 Seasons', '12 Seasons', '15 Seasons', '11 Seasons',\n",
       "       '13 Seasons'], dtype=object)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Verificação dos valores únicos da coluna duration\n",
    "# Checking the unique values of the duration column\n",
    "\n",
    "data['duration'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 300
    },
    "executionInfo": {
     "elapsed": 17,
     "status": "ok",
     "timestamp": 1661397625048,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "q1GOjZ4cgx0m",
    "outputId": "4c754f87-3578-433c-de6f-07ea044bc9df"
   },
   "outputs": [
    {
     "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>release_year</th>\n",
       "      <th>rating</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>6233.000000</td>\n",
       "      <td>6233.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>2013.370448</td>\n",
       "      <td>64.127547</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>8.768412</td>\n",
       "      <td>19.013894</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1925.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>2013.000000</td>\n",
       "      <td>59.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>2016.000000</td>\n",
       "      <td>68.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>2018.000000</td>\n",
       "      <td>76.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>2020.000000</td>\n",
       "      <td>97.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       release_year       rating\n",
       "count   6233.000000  6233.000000\n",
       "mean    2013.370448    64.127547\n",
       "std        8.768412    19.013894\n",
       "min     1925.000000     0.000000\n",
       "25%     2013.000000    59.000000\n",
       "50%     2016.000000    68.000000\n",
       "75%     2018.000000    76.000000\n",
       "max     2020.000000    97.000000"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Verificação dos mínimos e máximos das colunas release_year e rating\n",
    "# Verification of the minimum and maximum of the release_year and rating columns\n",
    "\n",
    "data[['release_year', 'rating']].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "executionInfo": {
     "elapsed": 15,
     "status": "ok",
     "timestamp": 1661397625048,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "BykVCSxzgx0m"
   },
   "outputs": [],
   "source": [
    "# Aplicação de encoder na coluna type\n",
    "# Encoder application on type column\n",
    "\n",
    "ohe_type = OneHotEncoder(sparse = False)\n",
    "type_encoded = ohe_type.fit_transform(data[['type']])\n",
    "data['type_e_' + ohe_type.categories_[0]] = type_encoded"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 16,
     "status": "ok",
     "timestamp": 1661397625049,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "dIQIl3Fugx0m",
    "outputId": "e97fca74-58a3-4c10-d364-803cb1c443c2"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array(['Movie', 'TV Show'], dtype=object)]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Verificação das categorias do encoder na coluna type\n",
    "# Checking the encoder categories in the type column\n",
    "\n",
    "ohe_type.categories_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 13,
     "status": "ok",
     "timestamp": 1661397625049,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "JarimSdGgx0m",
    "outputId": "d559e2bc-33b5-41d3-cfc4-266b1d4289ef"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['show_id', 'type', 'title', 'director', 'cast', 'country', 'date_added',\n",
       "       'release_year', 'rating', 'duration', 'listed_in', 'description',\n",
       "       'type_e_Movie', 'type_e_TV Show'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Relação das colunas atuais da tabela\n",
    "# List of current table columns\n",
    "\n",
    "data.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 206
    },
    "executionInfo": {
     "elapsed": 11,
     "status": "ok",
     "timestamp": 1661397625049,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "cP7VKVEmgx0m",
    "outputId": "e026e3da-9cb2-493b-bfb7-4e8cb4bf0d01"
   },
   "outputs": [
    {
     "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>type</th>\n",
       "      <th>type_e_Movie</th>\n",
       "      <th>type_e_TV Show</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Movie</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Movie</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>TV Show</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>TV Show</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Movie</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      type  type_e_Movie  type_e_TV Show\n",
       "0    Movie           1.0             0.0\n",
       "1    Movie           1.0             0.0\n",
       "2  TV Show           0.0             1.0\n",
       "3  TV Show           0.0             1.0\n",
       "4    Movie           1.0             0.0"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Verificação das colunas criadas a partir do encoder na coluna type\n",
    "# Check the columns created from the encoder in the type column\n",
    "\n",
    "data[['type', 'type_e_Movie', 'type_e_TV Show']].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "executionInfo": {
     "elapsed": 11,
     "status": "ok",
     "timestamp": 1661397625050,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "CxDNA7O7gx0n"
   },
   "outputs": [],
   "source": [
    "# Função para converter letras maiúsculas do texto em minúsculas\n",
    "# Function to convert uppercase letters of text to lowercase\n",
    "\n",
    "def lowercase(text): \n",
    "    lowercased = text.lower() \n",
    "    return lowercased"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "executionInfo": {
     "elapsed": 372,
     "status": "ok",
     "timestamp": 1661397625412,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "Z_xisAdHgx0n"
   },
   "outputs": [],
   "source": [
    "# Função para remover números do texto\n",
    "# Function to remove numbers from text\n",
    "\n",
    "def remove_numbers(text):\n",
    "    words_only = ''.join([i for i in text if not i.isdigit()])\n",
    "    return words_only"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "executionInfo": {
     "elapsed": 5,
     "status": "ok",
     "timestamp": 1661397625413,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "OlbjS09Agx0n"
   },
   "outputs": [],
   "source": [
    "# Função para remover pontuação do texto\n",
    "# Function to remove punctuation from text\n",
    "\n",
    "def remove_punctuation(text):\n",
    "    for punctuation in string.punctuation:\n",
    "        text = text.replace(punctuation, '')\n",
    "    return text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "executionInfo": {
     "elapsed": 4,
     "status": "ok",
     "timestamp": 1661397625413,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "VqvZ7H-0gx0n"
   },
   "outputs": [],
   "source": [
    "# Função para remover pontuação do texto exceto vírgula\n",
    "# Function to remove punctuation from text except comma\n",
    "\n",
    "punctuation_wo_comma = string.punctuation.replace(',', '')\n",
    "\n",
    "def remove_punctuation_except_comma(text):\n",
    "    for punctuation in punctuation_wo_comma:\n",
    "        text = text.replace(punctuation, '')\n",
    "    return text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "executionInfo": {
     "elapsed": 4,
     "status": "ok",
     "timestamp": 1661397625413,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "suWTSQnJgx0o"
   },
   "outputs": [],
   "source": [
    "# Função para extrair os números da coluna duration\n",
    "# Function to extract the numbers from the duration column\n",
    "\n",
    "def extract_duration_numbers(text, unit):\n",
    "    numbers_only = int(text[:text.find(' ')])\n",
    "    unit_only = text[(text.find(' ') + 1):]\n",
    "    if (unit == 'minutes' and unit_only == 'min') or (unit == 'seasons' and unit_only[:6] == 'Season'):\n",
    "        return numbers_only\n",
    "    else:\n",
    "        return 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "executionInfo": {
     "elapsed": 5,
     "status": "ok",
     "timestamp": 1661397625414,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "Fa7pvg-rgx0o"
   },
   "outputs": [],
   "source": [
    "# Função para remover stop words do texto\n",
    "# Function to remove stop words from text\n",
    "\n",
    "stop_words = set(stopwords.words('english')) \n",
    "\n",
    "def remove_stopwords(text):\n",
    "    tokenized = word_tokenize(text)\n",
    "    without_stopwords = [word for word in tokenized if not word in stop_words]\n",
    "    return without_stopwords"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "executionInfo": {
     "elapsed": 4,
     "status": "ok",
     "timestamp": 1661397625414,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "CkbG7VJlgx0o"
   },
   "outputs": [],
   "source": [
    "# Função para padronizar as palavras do texto\n",
    "# Function to standardize the words of the text\n",
    "\n",
    "def lemma(text):\n",
    "    lemmatizer = WordNetLemmatizer()\n",
    "    lemmatized = [lemmatizer.lemmatize(word) for word in text]\n",
    "    lemmatized_string = \" \".join(lemmatized)\n",
    "    return lemmatized_string"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "executionInfo": {
     "elapsed": 6145,
     "status": "ok",
     "timestamp": 1661397631555,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "z-YFMdaXgx0o",
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "# Aplicação dos tratamentos específicos para texto nas colunas de acordo com a necessidade\n",
    "# Application of specific treatments for text in columns as needed\n",
    "\n",
    "# Coluna director\n",
    "# Column director\n",
    "data['director'] = data['director'].dropna().apply(lambda x: lowercase(x))\n",
    "data['director'] = data['director'].dropna().apply(lambda x: remove_numbers(x))\n",
    "data['director'] = data['director'].dropna().apply(lambda x: remove_punctuation_except_comma(x))\n",
    "# Coluna cast\n",
    "# Column cast\n",
    "data['cast'] = data['cast'].apply(lambda x: '' if x == 'bam' else x)\n",
    "data['cast'] = data['cast'].dropna().apply(lambda x: lowercase(x))\n",
    "data['cast'] = data['cast'].dropna().apply(lambda x: remove_numbers(x))\n",
    "data['cast'] = data['cast'].dropna().apply(lambda x: remove_punctuation_except_comma(x))\n",
    "# Coluna country\n",
    "# Column country\n",
    "data['country'] = data['country'].apply(lambda x: '' if x == 'none' else x)\n",
    "data['country'] = data['country'].dropna().apply(lambda x: lowercase(x))\n",
    "data['country'] = data['country'].dropna().apply(lambda x: remove_numbers(x))\n",
    "data['country'] = data['country'].dropna().apply(lambda x: remove_punctuation_except_comma(x))\n",
    "# Coluna rating\n",
    "# Column rating\n",
    "data['rating'] = data['rating'].apply(lambda x: round(x / 10))\n",
    "# Coluna duration\n",
    "# Column duration\n",
    "data['duration'] = data['duration'].dropna().apply(lambda x: lowercase(x))\n",
    "data['duration'] = data['duration'].dropna().apply(lambda x: remove_punctuation(x))\n",
    "data['duration_minutes'] = data['duration'].dropna().apply(lambda x: extract_duration_numbers(x, 'minutes'))\n",
    "data['duration_seasons'] = data['duration'].dropna().apply(lambda x: extract_duration_numbers(x, 'seasons'))\n",
    "# Coluna listed_in\n",
    "# Column listed_in\n",
    "data['listed_in'] = data['listed_in'].dropna().apply(lambda x: lowercase(x))\n",
    "data['listed_in'] = data['listed_in'].dropna().apply(lambda x: remove_numbers(x))\n",
    "data['listed_in'] = data['listed_in'].dropna().apply(lambda x: remove_punctuation_except_comma(x))\n",
    "# Coluna description\n",
    "# Column description\n",
    "data['description'] = data['description'].dropna().apply(lambda x: lowercase(x))\n",
    "data['description'] = data['description'].dropna().apply(lambda x: remove_numbers(x))\n",
    "data['description'] = data['description'].dropna().apply(lambda x: remove_punctuation(x))\n",
    "data['description'] = data['description'].dropna().apply(lambda x: remove_stopwords(x))\n",
    "data['description'] = data['description'].dropna().apply(lambda x: lemma(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 424
    },
    "executionInfo": {
     "elapsed": 18,
     "status": "ok",
     "timestamp": 1661397631558,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "RRk5XGujgx0p",
    "outputId": "508ea41b-5d89-45b1-bf3b-4abf8e96cc49"
   },
   "outputs": [
    {
     "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>show_id</th>\n",
       "      <th>type</th>\n",
       "      <th>title</th>\n",
       "      <th>director</th>\n",
       "      <th>cast</th>\n",
       "      <th>country</th>\n",
       "      <th>date_added</th>\n",
       "      <th>release_year</th>\n",
       "      <th>rating</th>\n",
       "      <th>duration</th>\n",
       "      <th>listed_in</th>\n",
       "      <th>description</th>\n",
       "      <th>type_e_Movie</th>\n",
       "      <th>type_e_TV Show</th>\n",
       "      <th>duration_minutes</th>\n",
       "      <th>duration_seasons</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>81145628</td>\n",
       "      <td>Movie</td>\n",
       "      <td>Norm of the North: King Sized Adventure</td>\n",
       "      <td>richard finn, tim maltby</td>\n",
       "      <td>alan marriott, andrew toth, brian dobson, cole...</td>\n",
       "      <td>united states, india, south korea, china</td>\n",
       "      <td>2019-09-09</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>4</td>\n",
       "      <td>90 min</td>\n",
       "      <td>children  family movies, comedies</td>\n",
       "      <td>planning awesome wedding grandfather polar bea...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>90</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>80117401</td>\n",
       "      <td>Movie</td>\n",
       "      <td>Jandino: Whatever it Takes</td>\n",
       "      <td>NaN</td>\n",
       "      <td>jandino asporaat</td>\n",
       "      <td>united kingdom</td>\n",
       "      <td>2016-09-09</td>\n",
       "      <td>2016.0</td>\n",
       "      <td>5</td>\n",
       "      <td>94 min</td>\n",
       "      <td>standup comedy</td>\n",
       "      <td>jandino asporaat riff challenge raising kid se...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>94</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>70234439</td>\n",
       "      <td>TV Show</td>\n",
       "      <td>Transformers Prime</td>\n",
       "      <td>NaN</td>\n",
       "      <td>peter cullen, sumalee montano, frank welker, j...</td>\n",
       "      <td>united states</td>\n",
       "      <td>2018-09-08</td>\n",
       "      <td>2013.0</td>\n",
       "      <td>8</td>\n",
       "      <td>1 season</td>\n",
       "      <td>kids tv</td>\n",
       "      <td>help three human ally autobots protect earth o...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>80058654</td>\n",
       "      <td>TV Show</td>\n",
       "      <td>Transformers: Robots in Disguise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>will friedle, darren criss, constance zimmer, ...</td>\n",
       "      <td>united states</td>\n",
       "      <td>2018-09-08</td>\n",
       "      <td>2016.0</td>\n",
       "      <td>6</td>\n",
       "      <td>1 season</td>\n",
       "      <td>kids tv</td>\n",
       "      <td>prison ship crash unleashes hundred decepticon...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>80125979</td>\n",
       "      <td>Movie</td>\n",
       "      <td>#realityhigh</td>\n",
       "      <td>fernando lebrija</td>\n",
       "      <td>nesta cooper, kate walsh, john michael higgins...</td>\n",
       "      <td>united states</td>\n",
       "      <td>2017-09-08</td>\n",
       "      <td>2017.0</td>\n",
       "      <td>6</td>\n",
       "      <td>99 min</td>\n",
       "      <td>comedies</td>\n",
       "      <td>nerdy high schooler dani finally attracts inte...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>99</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    show_id     type                                    title  \\\n",
       "0  81145628    Movie  Norm of the North: King Sized Adventure   \n",
       "1  80117401    Movie               Jandino: Whatever it Takes   \n",
       "2  70234439  TV Show                       Transformers Prime   \n",
       "3  80058654  TV Show         Transformers: Robots in Disguise   \n",
       "4  80125979    Movie                             #realityhigh   \n",
       "\n",
       "                   director  \\\n",
       "0  richard finn, tim maltby   \n",
       "1                       NaN   \n",
       "2                       NaN   \n",
       "3                       NaN   \n",
       "4          fernando lebrija   \n",
       "\n",
       "                                                cast  \\\n",
       "0  alan marriott, andrew toth, brian dobson, cole...   \n",
       "1                                   jandino asporaat   \n",
       "2  peter cullen, sumalee montano, frank welker, j...   \n",
       "3  will friedle, darren criss, constance zimmer, ...   \n",
       "4  nesta cooper, kate walsh, john michael higgins...   \n",
       "\n",
       "                                    country date_added  release_year  rating  \\\n",
       "0  united states, india, south korea, china 2019-09-09        2019.0       4   \n",
       "1                            united kingdom 2016-09-09        2016.0       5   \n",
       "2                             united states 2018-09-08        2013.0       8   \n",
       "3                             united states 2018-09-08        2016.0       6   \n",
       "4                             united states 2017-09-08        2017.0       6   \n",
       "\n",
       "   duration                          listed_in  \\\n",
       "0    90 min  children  family movies, comedies   \n",
       "1    94 min                     standup comedy   \n",
       "2  1 season                            kids tv   \n",
       "3  1 season                            kids tv   \n",
       "4    99 min                           comedies   \n",
       "\n",
       "                                         description  type_e_Movie  \\\n",
       "0  planning awesome wedding grandfather polar bea...           1.0   \n",
       "1  jandino asporaat riff challenge raising kid se...           1.0   \n",
       "2  help three human ally autobots protect earth o...           0.0   \n",
       "3  prison ship crash unleashes hundred decepticon...           0.0   \n",
       "4  nerdy high schooler dani finally attracts inte...           1.0   \n",
       "\n",
       "   type_e_TV Show  duration_minutes  duration_seasons  \n",
       "0             0.0                90                 0  \n",
       "1             0.0                94                 0  \n",
       "2             1.0                 0                 0  \n",
       "3             1.0                 0                 0  \n",
       "4             0.0                99                 0  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Apresentação da tabela tratada\n",
    "# Presentation of the treated table\n",
    "\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "nYtgwqY2gx0p"
   },
   "source": [
    "## Feature engineering / *Feature engineering*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 297
    },
    "executionInfo": {
     "elapsed": 15,
     "status": "ok",
     "timestamp": 1661397631559,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "8-Fn_bLLgx0p",
    "outputId": "f07f39bb-2c70-40aa-d0df-a6624f127283",
    "scrolled": false
   },
   "outputs": [
    {
     "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>a l vijay</th>\n",
       "      <th>a salaam</th>\n",
       "      <th>aadish keluskar</th>\n",
       "      <th>aamir bashir</th>\n",
       "      <th>aamir khan</th>\n",
       "      <th>aanand rai</th>\n",
       "      <th>aaron burns</th>\n",
       "      <th>aaron hancox</th>\n",
       "      <th>aaron hann</th>\n",
       "      <th>aaron moorhead</th>\n",
       "      <th>...</th>\n",
       "      <th>zak hilditch</th>\n",
       "      <th>zakariya</th>\n",
       "      <th>zatella beatty</th>\n",
       "      <th>zeke norton</th>\n",
       "      <th>zhang yimou</th>\n",
       "      <th>ziga virc</th>\n",
       "      <th>zoe berriatua</th>\n",
       "      <th>zoe listerjones</th>\n",
       "      <th>zoya akhtar</th>\n",
       "      <th>zune kwok</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 3666 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   a l vijay  a salaam  aadish keluskar  aamir bashir  aamir khan  aanand rai  \\\n",
       "0        0.0       0.0              0.0           0.0         0.0         0.0   \n",
       "1        0.0       0.0              0.0           0.0         0.0         0.0   \n",
       "2        0.0       0.0              0.0           0.0         0.0         0.0   \n",
       "3        0.0       0.0              0.0           0.0         0.0         0.0   \n",
       "4        0.0       0.0              0.0           0.0         0.0         0.0   \n",
       "\n",
       "   aaron burns  aaron hancox  aaron hann  aaron moorhead  ...  zak hilditch  \\\n",
       "0          0.0           0.0         0.0             0.0  ...           0.0   \n",
       "1          0.0           0.0         0.0             0.0  ...           0.0   \n",
       "2          0.0           0.0         0.0             0.0  ...           0.0   \n",
       "3          0.0           0.0         0.0             0.0  ...           0.0   \n",
       "4          0.0           0.0         0.0             0.0  ...           0.0   \n",
       "\n",
       "   zakariya  zatella beatty  zeke norton  zhang yimou  ziga virc  \\\n",
       "0       0.0             0.0          0.0          0.0        0.0   \n",
       "1       0.0             0.0          0.0          0.0        0.0   \n",
       "2       0.0             0.0          0.0          0.0        0.0   \n",
       "3       0.0             0.0          0.0          0.0        0.0   \n",
       "4       0.0             0.0          0.0          0.0        0.0   \n",
       "\n",
       "   zoe berriatua  zoe listerjones  zoya akhtar  zune kwok  \n",
       "0            0.0              0.0          0.0        0.0  \n",
       "1            0.0              0.0          0.0        0.0  \n",
       "2            0.0              0.0          0.0        0.0  \n",
       "3            0.0              0.0          0.0        0.0  \n",
       "4            0.0              0.0          0.0        0.0  \n",
       "\n",
       "[5 rows x 3666 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Geração da matriz de importância por diretor a partir da coluna director\n",
    "# Generation of the importance matrix by director from the director column\n",
    "\n",
    "tf_idf_vectorizer = TfidfVectorizer(strip_accents='unicode', token_pattern=r\"(?u)[a-zA-Z][a-z ]+\")\n",
    "\n",
    "X = tf_idf_vectorizer.fit_transform(data['director'].fillna(''))\n",
    "\n",
    "director_vectorized = pd.DataFrame(X.toarray(), columns=tf_idf_vectorizer.get_feature_names())\n",
    "director_vectorized.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 314
    },
    "executionInfo": {
     "elapsed": 1173,
     "status": "ok",
     "timestamp": 1661397632718,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "KzxtIOuKgx0q",
    "outputId": "ce35db79-c601-4212-86c9-f00192db65a8"
   },
   "outputs": [
    {
     "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>a boogie wit tha hoodie</th>\n",
       "      <th>a murat ozgen</th>\n",
       "      <th>aabhas yadav</th>\n",
       "      <th>aachal munjal</th>\n",
       "      <th>aadarsh balakrishna</th>\n",
       "      <th>aadhi</th>\n",
       "      <th>aaditi pohankar</th>\n",
       "      <th>aaditya pratap singh</th>\n",
       "      <th>aadukalam naren</th>\n",
       "      <th>aadya bedi</th>\n",
       "      <th>...</th>\n",
       "      <th>zul vellani</th>\n",
       "      <th>zulay henao</th>\n",
       "      <th>zurawski</th>\n",
       "      <th>zuzanna szadkowski</th>\n",
       "      <th>zwelakhe mtsaka</th>\n",
       "      <th>zydrunas savickas</th>\n",
       "      <th>zygmunt hubner</th>\n",
       "      <th>zygmunt zintel</th>\n",
       "      <th>zynnell zuh</th>\n",
       "      <th>zyon allen</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 27452 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   a boogie wit tha hoodie  a murat ozgen  aabhas yadav  aachal munjal  \\\n",
       "0                      0.0            0.0           0.0            0.0   \n",
       "1                      0.0            0.0           0.0            0.0   \n",
       "2                      0.0            0.0           0.0            0.0   \n",
       "3                      0.0            0.0           0.0            0.0   \n",
       "4                      0.0            0.0           0.0            0.0   \n",
       "\n",
       "   aadarsh balakrishna  aadhi  aaditi pohankar  aaditya pratap singh  \\\n",
       "0                  0.0    0.0              0.0                   0.0   \n",
       "1                  0.0    0.0              0.0                   0.0   \n",
       "2                  0.0    0.0              0.0                   0.0   \n",
       "3                  0.0    0.0              0.0                   0.0   \n",
       "4                  0.0    0.0              0.0                   0.0   \n",
       "\n",
       "   aadukalam naren  aadya bedi  ...  zul vellani  zulay henao  zurawski  \\\n",
       "0              0.0         0.0  ...          0.0          0.0       0.0   \n",
       "1              0.0         0.0  ...          0.0          0.0       0.0   \n",
       "2              0.0         0.0  ...          0.0          0.0       0.0   \n",
       "3              0.0         0.0  ...          0.0          0.0       0.0   \n",
       "4              0.0         0.0  ...          0.0          0.0       0.0   \n",
       "\n",
       "   zuzanna szadkowski  zwelakhe mtsaka  zydrunas savickas  zygmunt hubner  \\\n",
       "0                 0.0              0.0                0.0             0.0   \n",
       "1                 0.0              0.0                0.0             0.0   \n",
       "2                 0.0              0.0                0.0             0.0   \n",
       "3                 0.0              0.0                0.0             0.0   \n",
       "4                 0.0              0.0                0.0             0.0   \n",
       "\n",
       "   zygmunt zintel  zynnell zuh  zyon allen  \n",
       "0             0.0          0.0         0.0  \n",
       "1             0.0          0.0         0.0  \n",
       "2             0.0          0.0         0.0  \n",
       "3             0.0          0.0         0.0  \n",
       "4             0.0          0.0         0.0  \n",
       "\n",
       "[5 rows x 27452 columns]"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Geração da matriz de importância por ator a partir da coluna cast\n",
    "# Generation of the importance matrix per actor from the cast column\n",
    "\n",
    "tf_idf_vectorizer = TfidfVectorizer(strip_accents='unicode', token_pattern=r\"(?u)[a-zA-Z][a-z ]+\")\n",
    "\n",
    "X = tf_idf_vectorizer.fit_transform(data['cast'].fillna(''))\n",
    "\n",
    "cast_vectorized = pd.DataFrame(X.toarray(), columns=tf_idf_vectorizer.get_feature_names())\n",
    "cast_vectorized.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 297
    },
    "executionInfo": {
     "elapsed": 10,
     "status": "ok",
     "timestamp": 1661397632718,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "ZgkAavaPgx0q",
    "outputId": "9f776535-b6a7-46d9-fe53-c5f0d377dbc8"
   },
   "outputs": [
    {
     "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>afghanistan</th>\n",
       "      <th>albania</th>\n",
       "      <th>argentina</th>\n",
       "      <th>armenia</th>\n",
       "      <th>australia</th>\n",
       "      <th>austria</th>\n",
       "      <th>azerbaijan</th>\n",
       "      <th>bangladesh</th>\n",
       "      <th>belgium</th>\n",
       "      <th>bermuda</th>\n",
       "      <th>...</th>\n",
       "      <th>ukraine</th>\n",
       "      <th>united arab emirates</th>\n",
       "      <th>united kingdom</th>\n",
       "      <th>united states</th>\n",
       "      <th>uruguay</th>\n",
       "      <th>vatican city</th>\n",
       "      <th>venezuela</th>\n",
       "      <th>vietnam</th>\n",
       "      <th>west germany</th>\n",
       "      <th>zimbabwe</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.244531</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 110 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   afghanistan  albania  argentina  armenia  australia  austria  azerbaijan  \\\n",
       "0          0.0      0.0        0.0      0.0        0.0      0.0         0.0   \n",
       "1          0.0      0.0        0.0      0.0        0.0      0.0         0.0   \n",
       "2          0.0      0.0        0.0      0.0        0.0      0.0         0.0   \n",
       "3          0.0      0.0        0.0      0.0        0.0      0.0         0.0   \n",
       "4          0.0      0.0        0.0      0.0        0.0      0.0         0.0   \n",
       "\n",
       "   bangladesh  belgium  bermuda  ...  ukraine  united arab emirates  \\\n",
       "0         0.0      0.0      0.0  ...      0.0                   0.0   \n",
       "1         0.0      0.0      0.0  ...      0.0                   0.0   \n",
       "2         0.0      0.0      0.0  ...      0.0                   0.0   \n",
       "3         0.0      0.0      0.0  ...      0.0                   0.0   \n",
       "4         0.0      0.0      0.0  ...      0.0                   0.0   \n",
       "\n",
       "   united kingdom  united states  uruguay  vatican city  venezuela  vietnam  \\\n",
       "0             0.0       0.244531      0.0           0.0        0.0      0.0   \n",
       "1             1.0       0.000000      0.0           0.0        0.0      0.0   \n",
       "2             0.0       1.000000      0.0           0.0        0.0      0.0   \n",
       "3             0.0       1.000000      0.0           0.0        0.0      0.0   \n",
       "4             0.0       1.000000      0.0           0.0        0.0      0.0   \n",
       "\n",
       "   west germany  zimbabwe  \n",
       "0           0.0       0.0  \n",
       "1           0.0       0.0  \n",
       "2           0.0       0.0  \n",
       "3           0.0       0.0  \n",
       "4           0.0       0.0  \n",
       "\n",
       "[5 rows x 110 columns]"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Geração da matriz de importância por país a partir da coluna country\n",
    "# Generation of the importance matrix by country from the country column\n",
    "\n",
    "tf_idf_vectorizer = TfidfVectorizer(strip_accents='unicode', token_pattern=r\"(?u)[a-zA-Z][a-z ]+\")\n",
    "\n",
    "X = tf_idf_vectorizer.fit_transform(data['country'].fillna(''))\n",
    "\n",
    "country_vectorized = pd.DataFrame(X.toarray(), columns=tf_idf_vectorizer.get_feature_names())\n",
    "country_vectorized.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 297
    },
    "executionInfo": {
     "elapsed": 7,
     "status": "ok",
     "timestamp": 1661397632719,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "IL5BmWoIgx0q",
    "outputId": "8e4f6a7d-acb0-480b-80b2-fc5eed29f48d"
   },
   "outputs": [
    {
     "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>action  adventure</th>\n",
       "      <th>anime features</th>\n",
       "      <th>anime series</th>\n",
       "      <th>british tv shows</th>\n",
       "      <th>children  family movies</th>\n",
       "      <th>classic  cult tv</th>\n",
       "      <th>classic movies</th>\n",
       "      <th>comedies</th>\n",
       "      <th>crime tv shows</th>\n",
       "      <th>cult movies</th>\n",
       "      <th>...</th>\n",
       "      <th>teen tv shows</th>\n",
       "      <th>thrillers</th>\n",
       "      <th>tv action  adventure</th>\n",
       "      <th>tv comedies</th>\n",
       "      <th>tv dramas</th>\n",
       "      <th>tv horror</th>\n",
       "      <th>tv mysteries</th>\n",
       "      <th>tv scifi  fantasy</th>\n",
       "      <th>tv shows</th>\n",
       "      <th>tv thrillers</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.812963</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.582315</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 42 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   action  adventure  anime features  anime series  british tv shows  \\\n",
       "0                0.0             0.0           0.0               0.0   \n",
       "1                0.0             0.0           0.0               0.0   \n",
       "2                0.0             0.0           0.0               0.0   \n",
       "3                0.0             0.0           0.0               0.0   \n",
       "4                0.0             0.0           0.0               0.0   \n",
       "\n",
       "   children  family movies  classic  cult tv  classic movies  comedies  \\\n",
       "0                 0.812963               0.0             0.0  0.582315   \n",
       "1                 0.000000               0.0             0.0  0.000000   \n",
       "2                 0.000000               0.0             0.0  0.000000   \n",
       "3                 0.000000               0.0             0.0  0.000000   \n",
       "4                 0.000000               0.0             0.0  1.000000   \n",
       "\n",
       "   crime tv shows  cult movies  ...  teen tv shows  thrillers  \\\n",
       "0             0.0          0.0  ...            0.0        0.0   \n",
       "1             0.0          0.0  ...            0.0        0.0   \n",
       "2             0.0          0.0  ...            0.0        0.0   \n",
       "3             0.0          0.0  ...            0.0        0.0   \n",
       "4             0.0          0.0  ...            0.0        0.0   \n",
       "\n",
       "   tv action  adventure  tv comedies  tv dramas  tv horror  tv mysteries  \\\n",
       "0                   0.0          0.0        0.0        0.0           0.0   \n",
       "1                   0.0          0.0        0.0        0.0           0.0   \n",
       "2                   0.0          0.0        0.0        0.0           0.0   \n",
       "3                   0.0          0.0        0.0        0.0           0.0   \n",
       "4                   0.0          0.0        0.0        0.0           0.0   \n",
       "\n",
       "   tv scifi  fantasy  tv shows  tv thrillers  \n",
       "0                0.0       0.0           0.0  \n",
       "1                0.0       0.0           0.0  \n",
       "2                0.0       0.0           0.0  \n",
       "3                0.0       0.0           0.0  \n",
       "4                0.0       0.0           0.0  \n",
       "\n",
       "[5 rows x 42 columns]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Geração da matriz de importância por categoria a partir da coluna listed_in\n",
    "# Generation of the importance matrix by category from the column listed_in\n",
    "\n",
    "tf_idf_vectorizer = TfidfVectorizer(strip_accents='unicode', token_pattern=r\"(?u)[a-zA-Z][a-z ]+\")\n",
    "\n",
    "X = tf_idf_vectorizer.fit_transform(data['listed_in'].fillna(''))\n",
    "\n",
    "listed_in_vectorized = pd.DataFrame(X.toarray(), columns=tf_idf_vectorizer.get_feature_names())\n",
    "listed_in_vectorized.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 236
    },
    "executionInfo": {
     "elapsed": 573,
     "status": "ok",
     "timestamp": 1661397633286,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "Cx4mhDgKgx0r",
    "outputId": "2c6bfc9b-9674-404d-fd0f-e6b59257977c"
   },
   "outputs": [
    {
     "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>aamir</th>\n",
       "      <th>aardman</th>\n",
       "      <th>aaron</th>\n",
       "      <th>abad</th>\n",
       "      <th>abagnale</th>\n",
       "      <th>abandon</th>\n",
       "      <th>abandoned</th>\n",
       "      <th>abandoning</th>\n",
       "      <th>abandonment</th>\n",
       "      <th>abargil</th>\n",
       "      <th>...</th>\n",
       "      <th>zone</th>\n",
       "      <th>zoologist</th>\n",
       "      <th>zordon</th>\n",
       "      <th>zorro</th>\n",
       "      <th>zoya</th>\n",
       "      <th>zuhu</th>\n",
       "      <th>zulu</th>\n",
       "      <th>zumbo</th>\n",
       "      <th>zurich</th>\n",
       "      <th>łukasz</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 15294 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   aamir  aardman  aaron  abad  abagnale  abandon  abandoned  abandoning  \\\n",
       "0    0.0      0.0    0.0   0.0       0.0      0.0        0.0         0.0   \n",
       "1    0.0      0.0    0.0   0.0       0.0      0.0        0.0         0.0   \n",
       "2    0.0      0.0    0.0   0.0       0.0      0.0        0.0         0.0   \n",
       "3    0.0      0.0    0.0   0.0       0.0      0.0        0.0         0.0   \n",
       "4    0.0      0.0    0.0   0.0       0.0      0.0        0.0         0.0   \n",
       "\n",
       "   abandonment  abargil  ...  zone  zoologist  zordon  zorro  zoya  zuhu  \\\n",
       "0          0.0      0.0  ...   0.0        0.0     0.0    0.0   0.0   0.0   \n",
       "1          0.0      0.0  ...   0.0        0.0     0.0    0.0   0.0   0.0   \n",
       "2          0.0      0.0  ...   0.0        0.0     0.0    0.0   0.0   0.0   \n",
       "3          0.0      0.0  ...   0.0        0.0     0.0    0.0   0.0   0.0   \n",
       "4          0.0      0.0  ...   0.0        0.0     0.0    0.0   0.0   0.0   \n",
       "\n",
       "   zulu  zumbo  zurich  łukasz  \n",
       "0   0.0    0.0     0.0     0.0  \n",
       "1   0.0    0.0     0.0     0.0  \n",
       "2   0.0    0.0     0.0     0.0  \n",
       "3   0.0    0.0     0.0     0.0  \n",
       "4   0.0    0.0     0.0     0.0  \n",
       "\n",
       "[5 rows x 15294 columns]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Geração da matriz de importância por palavra a partir da coluna description\n",
    "# Generation of the word importance matrix from the description column\n",
    "\n",
    "tf_idf_vectorizer = TfidfVectorizer(strip_accents='unicode')\n",
    "\n",
    "X = tf_idf_vectorizer.fit_transform(data['description'].fillna(''))\n",
    "\n",
    "description_vectorized = pd.DataFrame(X.toarray(), columns=tf_idf_vectorizer.get_feature_names())\n",
    "description_vectorized.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 5,
     "status": "ok",
     "timestamp": 1661397633286,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "8t0axheRgx0r",
    "outputId": "30e64ccf-5016-4517-cdb4-3920eea50255"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['show_id', 'type', 'title', 'director', 'cast', 'country', 'date_added',\n",
       "       'release_year', 'rating', 'duration', 'listed_in', 'description',\n",
       "       'type_e_Movie', 'type_e_TV Show', 'duration_minutes',\n",
       "       'duration_seasons'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Relação das colunas atuais da tabela\n",
    "# List of current table columns\n",
    "\n",
    "data.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "executionInfo": {
     "elapsed": 6946,
     "status": "ok",
     "timestamp": 1661397640229,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "ulTCH1GLgx0r"
   },
   "outputs": [],
   "source": [
    "# Montagem da matriz das features do modelo\n",
    "# Assembly of the model features matrix\n",
    "\n",
    "X = data[['release_year', 'type_e_Movie', 'type_e_TV Show', 'duration_minutes', 'duration_seasons']]\n",
    "X = pd.concat([X, director_vectorized], axis=1)\n",
    "X = pd.concat([X, cast_vectorized], axis=1)\n",
    "X = pd.concat([X, country_vectorized], axis=1)\n",
    "X = pd.concat([X, listed_in_vectorized], axis=1)\n",
    "X = pd.concat([X, description_vectorized], axis=1)\n",
    "X\n",
    "\n",
    "del director_vectorized\n",
    "del cast_vectorized\n",
    "del country_vectorized\n",
    "del listed_in_vectorized\n",
    "del description_vectorized"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 12,
     "status": "ok",
     "timestamp": 1661397640230,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "qXfrS-tmgx0r",
    "outputId": "e6d3db5e-fc4b-4de5-a154-d3da11c3ee5d"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       4\n",
       "1       5\n",
       "2       8\n",
       "3       6\n",
       "4       6\n",
       "       ..\n",
       "6228    9\n",
       "6229    8\n",
       "6230    6\n",
       "6231    8\n",
       "6232    9\n",
       "Name: rating, Length: 6233, dtype: int64"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Montagem do vetor do target do modelo\n",
    "# Assembly of the model's target vector\n",
    "\n",
    "y = data['rating']\n",
    "y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "executionInfo": {
     "elapsed": 11,
     "status": "ok",
     "timestamp": 1661397640230,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "mc9AyzKsAuxf"
   },
   "outputs": [],
   "source": [
    "del data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "qaBeyQLogx0s"
   },
   "source": [
    "## Divisão dos dados entre treino e teste / *Splitting the data between training and testing*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "executionInfo": {
     "elapsed": 4180,
     "status": "ok",
     "timestamp": 1661397644399,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "rB0dnRybgx0s"
   },
   "outputs": [],
   "source": [
    "# Separação dos dados em 70% para treino do modelo e 30% para teste\n",
    "# Data separation in 70% for training the model and 30% for testing\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=1)\n",
    "\n",
    "del X\n",
    "del y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 9851,
     "status": "ok",
     "timestamp": 1661397654223,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "4g_9FAoMgx0s",
    "outputId": "8e38cde2-e609-465b-ec20-b8720765a7c0"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.33554809420221315"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Cálculo da performance Accuracy do modelo sobre os dados de treino\n",
    "# Calculation of the model's Accuracy performance on the training data\n",
    "\n",
    "cv_nb = cross_validate(MultinomialNB(), X_train, y_train, scoring ='accuracy')\n",
    "\n",
    "cv_nb['test_score'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "executionInfo": {
     "elapsed": 2503,
     "status": "ok",
     "timestamp": 1661397656715,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "DbZKDy36gx0s"
   },
   "outputs": [],
   "source": [
    "# Treino do modelo e registro de suas predições sobre os dados de teste bem como das probabilidades destas predições\n",
    "# Training the model and recording its predictions on the test data as well as the probabilities of these predictions\n",
    "\n",
    "nb_model = MultinomialNB()\n",
    "nb_model.fit(X_train, y_train)\n",
    "y_pred = nb_model.predict(X_test)\n",
    "y_pred_proba = nb_model.predict_proba(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Y5usUfiNgx0s"
   },
   "source": [
    "## Matriz de Confusão e gráfico de Precision e Recall do seu modelo com os dados de teste / *Confusion Matrix and Precision and Recall plot of your model with test data*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 14,
     "status": "ok",
     "timestamp": 1661397656716,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "6yAry00Tgx0t",
    "outputId": "685cd7c9-bdae-4c1c-957e-ab1f9f0692e0"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix\n",
      "\n",
      "[[  0   0   0   0   0   6  42  62   0   0]\n",
      " [  0   0   0   0   0   0   2   0   0   0]\n",
      " [  0   0   0   0   0   1  15   2   0   0]\n",
      " [  0   0   0   0   0   5  68  11   0   0]\n",
      " [  0   0   0   0   0  18 110  30   0   0]\n",
      " [  0   0   0   0   0  45 315  90   0   0]\n",
      " [  0   0   0   0   0  63 293 164   0   0]\n",
      " [  0   0   0   0   0  70 156 241   0   0]\n",
      " [  0   0   0   0   0   6  11  42   0   0]\n",
      " [  0   0   0   0   0   0   0   2   0   0]]\n"
     ]
    }
   ],
   "source": [
    "# Apresentação da matriz de confusão solicitada dos dados de teste\n",
    "# Presentation of the requested confusion matrix of the test data\n",
    "\n",
    "confusion = confusion_matrix(y_test, y_pred)\n",
    "print('Confusion Matrix\\n')\n",
    "print(confusion)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 10,
     "status": "ok",
     "timestamp": 1661397656716,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "HyzOCV_Zgx0t",
    "outputId": "30abb366-978d-4aca-a8e0-d927dd351061"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Accuracy: 0.31\n",
      "\n",
      "Micro Precision: 0.31\n",
      "Micro Recall: 0.31\n",
      "Micro F1-score: 0.31\n",
      "\n",
      "Macro Precision: 0.09\n",
      "Macro Recall: 0.12\n",
      "Macro F1-score: 0.10\n",
      "\n",
      "Weighted Precision: 0.22\n",
      "Weighted Recall: 0.31\n",
      "Weighted F1-score: 0.25\n"
     ]
    }
   ],
   "source": [
    "# Cálculo das métricas de performance do modelo sobre os dados de teste\n",
    "# Calculation of model performance metrics on test data\n",
    "\n",
    "print('\\nAccuracy: {:.2f}\\n'.format(accuracy_score(y_test, y_pred)))\n",
    "\n",
    "print('Micro Precision: {:.2f}'.format(precision_score(y_test, y_pred, average='micro')))\n",
    "print('Micro Recall: {:.2f}'.format(recall_score(y_test, y_pred, average='micro')))\n",
    "print('Micro F1-score: {:.2f}\\n'.format(f1_score(y_test, y_pred, average='micro')))\n",
    "\n",
    "print('Macro Precision: {:.2f}'.format(precision_score(y_test, y_pred, average='macro')))\n",
    "print('Macro Recall: {:.2f}'.format(recall_score(y_test, y_pred, average='macro')))\n",
    "print('Macro F1-score: {:.2f}\\n'.format(f1_score(y_test, y_pred, average='macro')))\n",
    "\n",
    "print('Weighted Precision: {:.2f}'.format(precision_score(y_test, y_pred, average='weighted')))\n",
    "print('Weighted Recall: {:.2f}'.format(recall_score(y_test, y_pred, average='weighted')))\n",
    "print('Weighted F1-score: {:.2f}'.format(f1_score(y_test, y_pred, average='weighted')))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 8,
     "status": "ok",
     "timestamp": 1661397656716,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "vz6cUlR1gx0t",
    "outputId": "031f88ae-30dc-45a8-dc17-0e11b00f6235"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Classification Report\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "  Rating 0-5       0.00      0.00      0.00       110\n",
      " Rating 6-15       0.00      0.00      0.00         2\n",
      "Rating 16-25       0.00      0.00      0.00        18\n",
      "Rating 26-35       0.00      0.00      0.00        84\n",
      "Rating 36-45       0.00      0.00      0.00       158\n",
      "Rating 46-55       0.21      0.10      0.14       450\n",
      "Rating 56-65       0.29      0.56      0.38       520\n",
      "Rating 66-75       0.37      0.52      0.43       467\n",
      "Rating 76-85       0.00      0.00      0.00        59\n",
      "Rating 86-95       0.00      0.00      0.00         2\n",
      "\n",
      "    accuracy                           0.31      1870\n",
      "   macro avg       0.09      0.12      0.10      1870\n",
      "weighted avg       0.22      0.31      0.25      1870\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Cálculo das métricas de performance do modelo sobre os dados de teste\n",
    "# Calculation of model performance metrics on test data\n",
    "\n",
    "print('\\nClassification Report\\n')\n",
    "print(classification_report(y_test, y_pred, target_names=['Rating 0-5', 'Rating 6-15', 'Rating 16-25', 'Rating 26-35', 'Rating 36-45',\n",
    "                                                          'Rating 46-55', 'Rating 56-65', 'Rating 66-75', 'Rating 76-85', 'Rating 86-95']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 6,
     "status": "ok",
     "timestamp": 1661397656716,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "Gup9UfWrgx0t",
    "outputId": "ef1353e8-b136-4a8c-9616-4c7932498aae"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0., 0., 0., ..., 0., 1., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 1., 0.],\n",
       "       ...,\n",
       "       [0., 0., 0., ..., 0., 0., 1.],\n",
       "       [0., 0., 0., ..., 1., 0., 0.],\n",
       "       [0., 0., 0., ..., 1., 0., 0.]])"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Montagem da matriz de probabilidade dos targets dos dados de teste\n",
    "# Assembly of the probability matrix of test data targets\n",
    "\n",
    "y_test_proba = pd.DataFrame(y_test).reset_index(drop=True)\n",
    "\n",
    "n_classes = 10\n",
    "\n",
    "for i in range(n_classes):\n",
    "    y_test_proba.loc[y_test_proba['rating'] == i, i] = 1\n",
    "\n",
    "y_test_proba = y_test_proba.drop(columns='rating').fillna(0)\n",
    "y_test_proba = np.array(y_test_proba)\n",
    "y_test_proba"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 295
    },
    "executionInfo": {
     "elapsed": 775,
     "status": "ok",
     "timestamp": 1661397657487,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "vw2ojf57gx0u",
    "outputId": "95319ca9-f141-4aef-f06b-5f218f220fc5"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Montagem do gráfico de precision e recall solicitado\n",
    "# Assembly of the requested precision and recall graph\n",
    "\n",
    "n_classes = 10\n",
    "precision = dict()\n",
    "recall = dict()\n",
    "threshold = dict()\n",
    "\n",
    "for i in range(n_classes):\n",
    "    precision[i], recall[i], threshold[i] = precision_recall_curve(y_test_proba[:, i], y_pred_proba[:, i])\n",
    "    plt.plot(recall[i], precision[i], lw=2, label='class {}'.format(i))\n",
    "\n",
    "plt.xlabel(\"recall\")\n",
    "plt.ylabel(\"precision\")\n",
    "plt.legend(loc=\"best\")\n",
    "plt.title(\"Precision vs. Recall Curve\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 394
    },
    "executionInfo": {
     "elapsed": 12,
     "status": "ok",
     "timestamp": 1661397657488,
     "user": {
      "displayName": "Solano Jacon",
      "userId": "03193064938437592202"
     },
     "user_tz": 180
    },
    "id": "07AT_VT-gx0u",
    "outputId": "4a70ab9b-f427-4eb9-bb67-430a04e0f8aa"
   },
   "outputs": [
    {
     "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>actual</th>\n",
       "      <th>predicted</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>rating</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>110</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>18</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>84</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>158</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>450</td>\n",
       "      <td>214</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>520</td>\n",
       "      <td>1012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>467</td>\n",
       "      <td>644</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>59</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        actual  predicted\n",
       "rating                   \n",
       "0          110          0\n",
       "2            2          0\n",
       "3           18          0\n",
       "4           84          0\n",
       "5          158          0\n",
       "6          450        214\n",
       "7          520       1012\n",
       "8          467        644\n",
       "9           59          0\n",
       "10           2          0"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Montagem de comparativo entre quantidade de registros por target dos dados de teste reais e preditos\n",
    "# Comparison between number of records per target of actual and predicted test data\n",
    "\n",
    "test1 = pd.DataFrame(pd.DataFrame(y_test).value_counts())\n",
    "test1.columns = ['actual']\n",
    "test1 = test1.reset_index()\n",
    "test2 = pd.DataFrame(pd.DataFrame(y_pred).value_counts())\n",
    "test2.columns = ['predicted']\n",
    "test2 = test2.reset_index()\n",
    "test2.columns = ['rating', 'predicted']\n",
    "test = test1.merge(test2, how='outer', on='rating').fillna(0)\n",
    "test['predicted'] = test['predicted'].astype('int64')\n",
    "test = test.sort_values(by=['rating']).set_index('rating')\n",
    "test"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "XBf-PPiFgx0u"
   },
   "source": [
    "## Tendo em vista o resultado final, relacionar as oportunidades de melhoria para esse modelo / *Bearing in mind the final result, list the improvement opportunities for this model*"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "3rC_1o6tgx0u"
   },
   "source": [
    "- Balanceamento dos dados de treino do modelo com base na quantidade de filmes por classe de rating, de forma que cada classe possua a mesma quantidade de filmes para treino.\n",
    "- Agrupar ainda mais as classes do rating (atualmente em 10, para 5 por exemplo).\n",
    "- Verificar com o cliente a possibilidade de fornecer mais dados destes filmes, além dos já fornecidos (type, director, cast e country por exemplo).\n",
    "- Verificar com o cliente a possibilidade de fornecer dados de mais filmes, além dos já fornecidos.\n",
    "\n",
    "\n",
    "- *Balancing the training data of the model based on the number of films per rating class, so that each class has the same number of films for training.*\n",
    "- *Group rating classes even more (currently at 10, to 5 for example).*\n",
    "- *Check with the customer the possibility of providing more data on these films, in addition to those already provided (type, director, cast and country for example).*\n",
    "- *Check with the customer the possibility of providing data for more films, in addition to those already provided.*"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Y3yXEEg8gx0u"
   },
   "source": [
    "## Fim / *End*"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "Movie_Rating_Prediction.ipynb",
   "provenance": []
  },
  "hide_input": false,
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.6"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}