01_Amazon_Prime_Video_Popularity_Prediction/Amazon Video Project (Github).ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "R88Ms0MTi0Ma"
   },
   "source": [
    "# Amazon Video Project"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "WA6lL1fni0Mb"
   },
   "source": [
    "**In this project, I use supervised learning models to predict whether a movie is going to perform well on Amazon Prime Video platform. **"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "bO94-bXZi0Md"
   },
   "source": [
    "## Contents"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "SIvRSRqAi0Md"
   },
   "source": [
    "<ul>\n",
    "<li>[Part 1: Data Exploration](#Part-1:-Data-Exploration)\n",
    "<li>[Part 2: Feature Preprocessing](#Part-2:-Feature-Preprocessing)\n",
    "<li>[Part 3: Model Training and Results Evaluation](#Part-3:-Model-Training-and-Result-Evaluation)\n",
    "<li>[Part 4: Feature Selection](#Part-4:-Feature-Selection)\n",
    "</ul>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "TUoI2S7Bi6iR"
   },
   "source": [
    "# Part 0: Setup Google Drive Environment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "1iLCfrlpe-pV"
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "file_id='#'\n",
    "link='https://drive.google.com/uc?export=download&id={FILE_ID}'\n",
    "csv_url=link.format(FILE_ID=file_id)\n",
    "my_data = pd.read_csv(csv_url)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4226, 16)"
      ]
     },
     "execution_count": 189,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "my_data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 224
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 4996,
     "status": "ok",
     "timestamp": 1546590668026,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "KbCamNSoLcuX",
    "outputId": "8c4cf166-4ca6-4e74-9f0d-5d1aa6a28dfe",
    "scrolled": true
   },
   "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>video_id</th>\n",
       "      <th>cvt_per_day</th>\n",
       "      <th>weighted_categorical_position</th>\n",
       "      <th>weighted_horizontal_poition</th>\n",
       "      <th>import_id</th>\n",
       "      <th>release_year</th>\n",
       "      <th>genres</th>\n",
       "      <th>imdb_votes</th>\n",
       "      <th>budget</th>\n",
       "      <th>boxoffice</th>\n",
       "      <th>imdb_rating</th>\n",
       "      <th>duration_in_mins</th>\n",
       "      <th>metacritic_score</th>\n",
       "      <th>awards</th>\n",
       "      <th>mpaa</th>\n",
       "      <th>star_category</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>385504</td>\n",
       "      <td>307127.605608</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013</td>\n",
       "      <td>Action,Thriller,Drama</td>\n",
       "      <td>69614</td>\n",
       "      <td>15000000</td>\n",
       "      <td>42930462</td>\n",
       "      <td>6.5</td>\n",
       "      <td>112.301017</td>\n",
       "      <td>51</td>\n",
       "      <td>other award</td>\n",
       "      <td>PG-13</td>\n",
       "      <td>1.710000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>300175</td>\n",
       "      <td>270338.426375</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013</td>\n",
       "      <td>Comedy,Crime,Thriller</td>\n",
       "      <td>46705</td>\n",
       "      <td>15000000</td>\n",
       "      <td>3301046</td>\n",
       "      <td>6.5</td>\n",
       "      <td>94.983250</td>\n",
       "      <td>41</td>\n",
       "      <td>no award</td>\n",
       "      <td>R</td>\n",
       "      <td>3.250000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>361899</td>\n",
       "      <td>256165.867446</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>other</td>\n",
       "      <td>2012</td>\n",
       "      <td>Crime,Drama</td>\n",
       "      <td>197596</td>\n",
       "      <td>26000000</td>\n",
       "      <td>37397291</td>\n",
       "      <td>7.3</td>\n",
       "      <td>115.763675</td>\n",
       "      <td>58</td>\n",
       "      <td>other award</td>\n",
       "      <td>R</td>\n",
       "      <td>2.646667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>308314</td>\n",
       "      <td>196622.720996</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2008</td>\n",
       "      <td>Thriller,Drama,War,Documentary,Mystery,Action</td>\n",
       "      <td>356339</td>\n",
       "      <td>15000000</td>\n",
       "      <td>15700000</td>\n",
       "      <td>7.6</td>\n",
       "      <td>130.703583</td>\n",
       "      <td>94</td>\n",
       "      <td>Oscar</td>\n",
       "      <td>R</td>\n",
       "      <td>1.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>307201</td>\n",
       "      <td>159841.652064</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013</td>\n",
       "      <td>Crime,Thriller,Mystery,Documentary</td>\n",
       "      <td>46720</td>\n",
       "      <td>27220000</td>\n",
       "      <td>8551228</td>\n",
       "      <td>6.4</td>\n",
       "      <td>105.545533</td>\n",
       "      <td>37</td>\n",
       "      <td>other award</td>\n",
       "      <td>R</td>\n",
       "      <td>3.066667</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   video_id    cvt_per_day  weighted_categorical_position  \\\n",
       "0    385504  307127.605608                              1   \n",
       "1    300175  270338.426375                              1   \n",
       "2    361899  256165.867446                              1   \n",
       "3    308314  196622.720996                              3   \n",
       "4    307201  159841.652064                              1   \n",
       "\n",
       "   weighted_horizontal_poition  import_id  release_year  \\\n",
       "0                            3  lionsgate          2013   \n",
       "1                            3  lionsgate          2013   \n",
       "2                            3      other          2012   \n",
       "3                            4  lionsgate          2008   \n",
       "4                            3  lionsgate          2013   \n",
       "\n",
       "                                          genres  imdb_votes    budget  \\\n",
       "0                          Action,Thriller,Drama       69614  15000000   \n",
       "1                          Comedy,Crime,Thriller       46705  15000000   \n",
       "2                                    Crime,Drama      197596  26000000   \n",
       "3  Thriller,Drama,War,Documentary,Mystery,Action      356339  15000000   \n",
       "4             Crime,Thriller,Mystery,Documentary       46720  27220000   \n",
       "\n",
       "   boxoffice  imdb_rating  duration_in_mins  metacritic_score       awards  \\\n",
       "0   42930462          6.5        112.301017                51  other award   \n",
       "1    3301046          6.5         94.983250                41     no award   \n",
       "2   37397291          7.3        115.763675                58  other award   \n",
       "3   15700000          7.6        130.703583                94        Oscar   \n",
       "4    8551228          6.4        105.545533                37  other award   \n",
       "\n",
       "    mpaa  star_category  \n",
       "0  PG-13       1.710000  \n",
       "1      R       3.250000  \n",
       "2      R       2.646667  \n",
       "3      R       1.666667  \n",
       "4      R       3.066667  "
      ]
     },
     "execution_count": 190,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "my_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**cumulative view time per day is an average value, this is an engagement metric**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Num of rows: 4226\n",
      "Num of columns: 16\n"
     ]
    }
   ],
   "source": [
    "print (\"Num of rows: \" + str(my_data.shape[0])) # row count\n",
    "print (\"Num of columns: \" + str(my_data.shape[1])) # col count"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "a6bG_gAPi0Me"
   },
   "source": [
    "# Part 1: Data Exploration"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "bspx2K6fi0Me"
   },
   "source": [
    "### Part 1.1: Understand the Raw Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "kuTHKjk-i0Mf"
   },
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "# imbalanced-learn is a python package offering a number of re-sampling techniques \n",
    "# commonly used in datasets showing strong between-class imbalance. \n",
    "# It is compatible with scikit-learn and is part of scikit-learn-contrib projects.\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import imblearn\n",
    "# will show all the columns\n",
    "pd.set_option('display.max_columns', None)\n",
    "video_df = my_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "video_id                           int64\n",
       "cvt_per_day                      float64\n",
       "weighted_categorical_position      int64\n",
       "weighted_horizontal_poition        int64\n",
       "import_id                         object\n",
       "release_year                       int64\n",
       "genres                            object\n",
       "imdb_votes                         int64\n",
       "budget                             int64\n",
       "boxoffice                          int64\n",
       "imdb_rating                      float64\n",
       "duration_in_mins                 float64\n",
       "metacritic_score                   int64\n",
       "awards                            object\n",
       "mpaa                              object\n",
       "star_category                    float64\n",
       "dtype: object"
      ]
     },
     "execution_count": 193,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "video_df.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 51
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 4943,
     "status": "ok",
     "timestamp": 1546590668030,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "C99Z9b7ai0Mm",
    "outputId": "8b0901e5-5fe5-436f-a5e2-10cc77b14340"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Num of rows: 4226\n",
      "Num of columns: 16\n"
     ]
    }
   ],
   "source": [
    "print (\"Num of rows: \" + str(video_df.shape[0])) # row count\n",
    "print (\"Num of columns: \" + str(video_df.shape[1])) # col count"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {},
   "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>video_id</th>\n",
       "      <th>cvt_per_day</th>\n",
       "      <th>weighted_categorical_position</th>\n",
       "      <th>weighted_horizontal_poition</th>\n",
       "      <th>import_id</th>\n",
       "      <th>release_year</th>\n",
       "      <th>genres</th>\n",
       "      <th>imdb_votes</th>\n",
       "      <th>budget</th>\n",
       "      <th>boxoffice</th>\n",
       "      <th>imdb_rating</th>\n",
       "      <th>duration_in_mins</th>\n",
       "      <th>metacritic_score</th>\n",
       "      <th>awards</th>\n",
       "      <th>mpaa</th>\n",
       "      <th>star_category</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>385504</td>\n",
       "      <td>307127.605608</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013</td>\n",
       "      <td>Action,Thriller,Drama</td>\n",
       "      <td>69614</td>\n",
       "      <td>15000000</td>\n",
       "      <td>42930462</td>\n",
       "      <td>6.5</td>\n",
       "      <td>112.301017</td>\n",
       "      <td>51</td>\n",
       "      <td>other award</td>\n",
       "      <td>PG-13</td>\n",
       "      <td>1.710000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>300175</td>\n",
       "      <td>270338.426375</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013</td>\n",
       "      <td>Comedy,Crime,Thriller</td>\n",
       "      <td>46705</td>\n",
       "      <td>15000000</td>\n",
       "      <td>3301046</td>\n",
       "      <td>6.5</td>\n",
       "      <td>94.983250</td>\n",
       "      <td>41</td>\n",
       "      <td>no award</td>\n",
       "      <td>R</td>\n",
       "      <td>3.250000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>361899</td>\n",
       "      <td>256165.867446</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>other</td>\n",
       "      <td>2012</td>\n",
       "      <td>Crime,Drama</td>\n",
       "      <td>197596</td>\n",
       "      <td>26000000</td>\n",
       "      <td>37397291</td>\n",
       "      <td>7.3</td>\n",
       "      <td>115.763675</td>\n",
       "      <td>58</td>\n",
       "      <td>other award</td>\n",
       "      <td>R</td>\n",
       "      <td>2.646667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>308314</td>\n",
       "      <td>196622.720996</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2008</td>\n",
       "      <td>Thriller,Drama,War,Documentary,Mystery,Action</td>\n",
       "      <td>356339</td>\n",
       "      <td>15000000</td>\n",
       "      <td>15700000</td>\n",
       "      <td>7.6</td>\n",
       "      <td>130.703583</td>\n",
       "      <td>94</td>\n",
       "      <td>Oscar</td>\n",
       "      <td>R</td>\n",
       "      <td>1.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>307201</td>\n",
       "      <td>159841.652064</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013</td>\n",
       "      <td>Crime,Thriller,Mystery,Documentary</td>\n",
       "      <td>46720</td>\n",
       "      <td>27220000</td>\n",
       "      <td>8551228</td>\n",
       "      <td>6.4</td>\n",
       "      <td>105.545533</td>\n",
       "      <td>37</td>\n",
       "      <td>other award</td>\n",
       "      <td>R</td>\n",
       "      <td>3.066667</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   video_id    cvt_per_day  weighted_categorical_position  \\\n",
       "0    385504  307127.605608                              1   \n",
       "1    300175  270338.426375                              1   \n",
       "2    361899  256165.867446                              1   \n",
       "3    308314  196622.720996                              3   \n",
       "4    307201  159841.652064                              1   \n",
       "\n",
       "   weighted_horizontal_poition  import_id  release_year  \\\n",
       "0                            3  lionsgate          2013   \n",
       "1                            3  lionsgate          2013   \n",
       "2                            3      other          2012   \n",
       "3                            4  lionsgate          2008   \n",
       "4                            3  lionsgate          2013   \n",
       "\n",
       "                                          genres  imdb_votes    budget  \\\n",
       "0                          Action,Thriller,Drama       69614  15000000   \n",
       "1                          Comedy,Crime,Thriller       46705  15000000   \n",
       "2                                    Crime,Drama      197596  26000000   \n",
       "3  Thriller,Drama,War,Documentary,Mystery,Action      356339  15000000   \n",
       "4             Crime,Thriller,Mystery,Documentary       46720  27220000   \n",
       "\n",
       "   boxoffice  imdb_rating  duration_in_mins  metacritic_score       awards  \\\n",
       "0   42930462          6.5        112.301017                51  other award   \n",
       "1    3301046          6.5         94.983250                41     no award   \n",
       "2   37397291          7.3        115.763675                58  other award   \n",
       "3   15700000          7.6        130.703583                94        Oscar   \n",
       "4    8551228          6.4        105.545533                37  other award   \n",
       "\n",
       "    mpaa  star_category  \n",
       "0  PG-13       1.710000  \n",
       "1      R       3.250000  \n",
       "2      R       2.646667  \n",
       "3      R       1.666667  \n",
       "4      R       3.066667  "
      ]
     },
     "execution_count": 195,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "video_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "metadata": {},
   "outputs": [],
   "source": [
    "# check duplication"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "OCglmJ9Oi0Mo"
   },
   "source": [
    "### Part 1.2: Data cleaning"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "JxlrXRG3i0Mp"
   },
   "source": [
    "Remove Extra Whitespace"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 4934,
     "status": "ok",
     "timestamp": 1546590668031,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "2Vf8iYmWi0Mq",
    "outputId": "c3741432-6b27-4692-88e5-ebf126ef1024",
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'PG-13'"
      ]
     },
     "execution_count": 197,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check categorical feature\n",
    "video_df['mpaa'][0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "SsAbAjhvi0Mx"
   },
   "source": [
    "### Part 1.3:  Understand the features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 296
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 5279,
     "status": "ok",
     "timestamp": 1546590668405,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "rJ0AdxwLi0Mz",
    "outputId": "efe4edd7-5acb-4210-c94f-10e4e8124d82",
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x11ac51668>"
      ]
     },
     "execution_count": 198,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# check the feature distribution\n",
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "# sns.distplot: Flexibly plot a univariate distribution of observations. \n",
    "# kde: kernel density estimate\n",
    "sns.distplot(video_df['cvt_per_day'], kde=True) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['video_id', 'cvt_per_day', 'weighted_categorical_position',\n",
       "       'weighted_horizontal_poition', 'import_id', 'release_year', 'genres',\n",
       "       'imdb_votes', 'budget', 'boxoffice', 'imdb_rating', 'duration_in_mins',\n",
       "       'metacritic_score', 'awards', 'mpaa', 'star_category'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 199,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "video_df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 386
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 5679,
     "status": "ok",
     "timestamp": 1546590668822,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "4DKTTdB6i0M2",
    "outputId": "27dda677-7a01-4a77-cce7-920b585f40a8"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x11ae56470>"
      ]
     },
     "execution_count": 200,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# correlations between all the features\n",
    "corr = video_df[video_df.columns].corr()\n",
    "\n",
    "# show heapmap of correlations\n",
    "sns.heatmap(corr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 472
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 5670,
     "status": "ok",
     "timestamp": 1546590668825,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "1qfEnNW_i0M5",
    "outputId": "70b4558b-8562-4e1b-d4a4-d2062490e56f"
   },
   "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>video_id</th>\n",
       "      <th>cvt_per_day</th>\n",
       "      <th>weighted_categorical_position</th>\n",
       "      <th>weighted_horizontal_poition</th>\n",
       "      <th>release_year</th>\n",
       "      <th>imdb_votes</th>\n",
       "      <th>budget</th>\n",
       "      <th>boxoffice</th>\n",
       "      <th>imdb_rating</th>\n",
       "      <th>duration_in_mins</th>\n",
       "      <th>metacritic_score</th>\n",
       "      <th>star_category</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>video_id</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.122352</td>\n",
       "      <td>0.050145</td>\n",
       "      <td>-0.067163</td>\n",
       "      <td>0.231145</td>\n",
       "      <td>0.047766</td>\n",
       "      <td>0.088285</td>\n",
       "      <td>0.071539</td>\n",
       "      <td>-0.099815</td>\n",
       "      <td>0.023838</td>\n",
       "      <td>0.063930</td>\n",
       "      <td>0.170079</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cvt_per_day</th>\n",
       "      <td>0.122352</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.148000</td>\n",
       "      <td>-0.301663</td>\n",
       "      <td>0.045859</td>\n",
       "      <td>0.297697</td>\n",
       "      <td>0.315886</td>\n",
       "      <td>0.312209</td>\n",
       "      <td>0.058894</td>\n",
       "      <td>0.152217</td>\n",
       "      <td>0.249159</td>\n",
       "      <td>0.246956</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>weighted_categorical_position</th>\n",
       "      <td>0.050145</td>\n",
       "      <td>-0.148000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.084066</td>\n",
       "      <td>-0.068902</td>\n",
       "      <td>-0.063700</td>\n",
       "      <td>-0.089899</td>\n",
       "      <td>-0.073628</td>\n",
       "      <td>-0.116443</td>\n",
       "      <td>-0.174461</td>\n",
       "      <td>-0.044317</td>\n",
       "      <td>-0.122851</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>weighted_horizontal_poition</th>\n",
       "      <td>-0.067163</td>\n",
       "      <td>-0.301663</td>\n",
       "      <td>-0.084066</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.026634</td>\n",
       "      <td>-0.220749</td>\n",
       "      <td>-0.211356</td>\n",
       "      <td>-0.244709</td>\n",
       "      <td>-0.162021</td>\n",
       "      <td>-0.102344</td>\n",
       "      <td>-0.255015</td>\n",
       "      <td>-0.167801</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>release_year</th>\n",
       "      <td>0.231145</td>\n",
       "      <td>0.045859</td>\n",
       "      <td>-0.068902</td>\n",
       "      <td>0.026634</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.082661</td>\n",
       "      <td>-0.040121</td>\n",
       "      <td>-0.128091</td>\n",
       "      <td>-0.198856</td>\n",
       "      <td>-0.096811</td>\n",
       "      <td>-0.167038</td>\n",
       "      <td>-0.198454</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>imdb_votes</th>\n",
       "      <td>0.047766</td>\n",
       "      <td>0.297697</td>\n",
       "      <td>-0.063700</td>\n",
       "      <td>-0.220749</td>\n",
       "      <td>-0.082661</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.272852</td>\n",
       "      <td>0.531467</td>\n",
       "      <td>0.163264</td>\n",
       "      <td>0.155670</td>\n",
       "      <td>0.321860</td>\n",
       "      <td>0.225192</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>budget</th>\n",
       "      <td>0.088285</td>\n",
       "      <td>0.315886</td>\n",
       "      <td>-0.089899</td>\n",
       "      <td>-0.211356</td>\n",
       "      <td>-0.040121</td>\n",
       "      <td>0.272852</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.504618</td>\n",
       "      <td>0.105671</td>\n",
       "      <td>0.251342</td>\n",
       "      <td>0.314608</td>\n",
       "      <td>0.321721</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>boxoffice</th>\n",
       "      <td>0.071539</td>\n",
       "      <td>0.312209</td>\n",
       "      <td>-0.073628</td>\n",
       "      <td>-0.244709</td>\n",
       "      <td>-0.128091</td>\n",
       "      <td>0.531467</td>\n",
       "      <td>0.504618</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.163673</td>\n",
       "      <td>0.226184</td>\n",
       "      <td>0.464965</td>\n",
       "      <td>0.322329</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>imdb_rating</th>\n",
       "      <td>-0.099815</td>\n",
       "      <td>0.058894</td>\n",
       "      <td>-0.116443</td>\n",
       "      <td>-0.162021</td>\n",
       "      <td>-0.198856</td>\n",
       "      <td>0.163264</td>\n",
       "      <td>0.105671</td>\n",
       "      <td>0.163673</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.344055</td>\n",
       "      <td>0.346935</td>\n",
       "      <td>0.191768</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>duration_in_mins</th>\n",
       "      <td>0.023838</td>\n",
       "      <td>0.152217</td>\n",
       "      <td>-0.174461</td>\n",
       "      <td>-0.102344</td>\n",
       "      <td>-0.096811</td>\n",
       "      <td>0.155670</td>\n",
       "      <td>0.251342</td>\n",
       "      <td>0.226184</td>\n",
       "      <td>0.344055</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.327050</td>\n",
       "      <td>0.309110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>metacritic_score</th>\n",
       "      <td>0.063930</td>\n",
       "      <td>0.249159</td>\n",
       "      <td>-0.044317</td>\n",
       "      <td>-0.255015</td>\n",
       "      <td>-0.167038</td>\n",
       "      <td>0.321860</td>\n",
       "      <td>0.314608</td>\n",
       "      <td>0.464965</td>\n",
       "      <td>0.346935</td>\n",
       "      <td>0.327050</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.452445</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>star_category</th>\n",
       "      <td>0.170079</td>\n",
       "      <td>0.246956</td>\n",
       "      <td>-0.122851</td>\n",
       "      <td>-0.167801</td>\n",
       "      <td>-0.198454</td>\n",
       "      <td>0.225192</td>\n",
       "      <td>0.321721</td>\n",
       "      <td>0.322329</td>\n",
       "      <td>0.191768</td>\n",
       "      <td>0.309110</td>\n",
       "      <td>0.452445</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                               video_id  cvt_per_day  \\\n",
       "video_id                       1.000000     0.122352   \n",
       "cvt_per_day                    0.122352     1.000000   \n",
       "weighted_categorical_position  0.050145    -0.148000   \n",
       "weighted_horizontal_poition   -0.067163    -0.301663   \n",
       "release_year                   0.231145     0.045859   \n",
       "imdb_votes                     0.047766     0.297697   \n",
       "budget                         0.088285     0.315886   \n",
       "boxoffice                      0.071539     0.312209   \n",
       "imdb_rating                   -0.099815     0.058894   \n",
       "duration_in_mins               0.023838     0.152217   \n",
       "metacritic_score               0.063930     0.249159   \n",
       "star_category                  0.170079     0.246956   \n",
       "\n",
       "                               weighted_categorical_position  \\\n",
       "video_id                                            0.050145   \n",
       "cvt_per_day                                        -0.148000   \n",
       "weighted_categorical_position                       1.000000   \n",
       "weighted_horizontal_poition                        -0.084066   \n",
       "release_year                                       -0.068902   \n",
       "imdb_votes                                         -0.063700   \n",
       "budget                                             -0.089899   \n",
       "boxoffice                                          -0.073628   \n",
       "imdb_rating                                        -0.116443   \n",
       "duration_in_mins                                   -0.174461   \n",
       "metacritic_score                                   -0.044317   \n",
       "star_category                                      -0.122851   \n",
       "\n",
       "                               weighted_horizontal_poition  release_year  \\\n",
       "video_id                                         -0.067163      0.231145   \n",
       "cvt_per_day                                      -0.301663      0.045859   \n",
       "weighted_categorical_position                    -0.084066     -0.068902   \n",
       "weighted_horizontal_poition                       1.000000      0.026634   \n",
       "release_year                                      0.026634      1.000000   \n",
       "imdb_votes                                       -0.220749     -0.082661   \n",
       "budget                                           -0.211356     -0.040121   \n",
       "boxoffice                                        -0.244709     -0.128091   \n",
       "imdb_rating                                      -0.162021     -0.198856   \n",
       "duration_in_mins                                 -0.102344     -0.096811   \n",
       "metacritic_score                                 -0.255015     -0.167038   \n",
       "star_category                                    -0.167801     -0.198454   \n",
       "\n",
       "                               imdb_votes    budget  boxoffice  imdb_rating  \\\n",
       "video_id                         0.047766  0.088285   0.071539    -0.099815   \n",
       "cvt_per_day                      0.297697  0.315886   0.312209     0.058894   \n",
       "weighted_categorical_position   -0.063700 -0.089899  -0.073628    -0.116443   \n",
       "weighted_horizontal_poition     -0.220749 -0.211356  -0.244709    -0.162021   \n",
       "release_year                    -0.082661 -0.040121  -0.128091    -0.198856   \n",
       "imdb_votes                       1.000000  0.272852   0.531467     0.163264   \n",
       "budget                           0.272852  1.000000   0.504618     0.105671   \n",
       "boxoffice                        0.531467  0.504618   1.000000     0.163673   \n",
       "imdb_rating                      0.163264  0.105671   0.163673     1.000000   \n",
       "duration_in_mins                 0.155670  0.251342   0.226184     0.344055   \n",
       "metacritic_score                 0.321860  0.314608   0.464965     0.346935   \n",
       "star_category                    0.225192  0.321721   0.322329     0.191768   \n",
       "\n",
       "                               duration_in_mins  metacritic_score  \\\n",
       "video_id                               0.023838          0.063930   \n",
       "cvt_per_day                            0.152217          0.249159   \n",
       "weighted_categorical_position         -0.174461         -0.044317   \n",
       "weighted_horizontal_poition           -0.102344         -0.255015   \n",
       "release_year                          -0.096811         -0.167038   \n",
       "imdb_votes                             0.155670          0.321860   \n",
       "budget                                 0.251342          0.314608   \n",
       "boxoffice                              0.226184          0.464965   \n",
       "imdb_rating                            0.344055          0.346935   \n",
       "duration_in_mins                       1.000000          0.327050   \n",
       "metacritic_score                       0.327050          1.000000   \n",
       "star_category                          0.309110          0.452445   \n",
       "\n",
       "                               star_category  \n",
       "video_id                            0.170079  \n",
       "cvt_per_day                         0.246956  \n",
       "weighted_categorical_position      -0.122851  \n",
       "weighted_horizontal_poition        -0.167801  \n",
       "release_year                       -0.198454  \n",
       "imdb_votes                          0.225192  \n",
       "budget                              0.321721  \n",
       "boxoffice                           0.322329  \n",
       "imdb_rating                         0.191768  \n",
       "duration_in_mins                    0.309110  \n",
       "metacritic_score                    0.452445  \n",
       "star_category                       1.000000  "
      ]
     },
     "execution_count": 201,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check the actual values of correlations\n",
    "corr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 5660,
     "status": "ok",
     "timestamp": 1546590668826,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "wtjI61m6i0M8",
    "outputId": "2101f665-017c-4910-ecd4-2e864669925a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.31220891751693813\n",
      "0.31588570143658673\n"
     ]
    }
   ],
   "source": [
    "# calculate two features pearson coefficient of correlation\n",
    "# how to interpret: https://www.dummies.com/education/math/statistics/how-to-interpret-a-correlation-coefficient-r/\n",
    "from scipy.stats import pearsonr\n",
    "print (pearsonr(video_df['cvt_per_day'], video_df['boxoffice'])[0])\n",
    "print (pearsonr(video_df['cvt_per_day'], video_df['budget'])[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "aFa4d6t3i0NH"
   },
   "source": [
    "# Part 2: Feature Preprocessing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Part 2.1 Processing Missing Values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "metadata": {},
   "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>video_id</th>\n",
       "      <th>cvt_per_day</th>\n",
       "      <th>weighted_categorical_position</th>\n",
       "      <th>weighted_horizontal_poition</th>\n",
       "      <th>import_id</th>\n",
       "      <th>release_year</th>\n",
       "      <th>genres</th>\n",
       "      <th>imdb_votes</th>\n",
       "      <th>budget</th>\n",
       "      <th>boxoffice</th>\n",
       "      <th>imdb_rating</th>\n",
       "      <th>duration_in_mins</th>\n",
       "      <th>metacritic_score</th>\n",
       "      <th>awards</th>\n",
       "      <th>mpaa</th>\n",
       "      <th>star_category</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>385504</td>\n",
       "      <td>307127.605608</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013</td>\n",
       "      <td>Action,Thriller,Drama</td>\n",
       "      <td>69614</td>\n",
       "      <td>15000000</td>\n",
       "      <td>42930462</td>\n",
       "      <td>6.5</td>\n",
       "      <td>112.301017</td>\n",
       "      <td>51</td>\n",
       "      <td>other award</td>\n",
       "      <td>PG-13</td>\n",
       "      <td>1.710000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>300175</td>\n",
       "      <td>270338.426375</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013</td>\n",
       "      <td>Comedy,Crime,Thriller</td>\n",
       "      <td>46705</td>\n",
       "      <td>15000000</td>\n",
       "      <td>3301046</td>\n",
       "      <td>6.5</td>\n",
       "      <td>94.983250</td>\n",
       "      <td>41</td>\n",
       "      <td>no award</td>\n",
       "      <td>R</td>\n",
       "      <td>3.250000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>361899</td>\n",
       "      <td>256165.867446</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>other</td>\n",
       "      <td>2012</td>\n",
       "      <td>Crime,Drama</td>\n",
       "      <td>197596</td>\n",
       "      <td>26000000</td>\n",
       "      <td>37397291</td>\n",
       "      <td>7.3</td>\n",
       "      <td>115.763675</td>\n",
       "      <td>58</td>\n",
       "      <td>other award</td>\n",
       "      <td>R</td>\n",
       "      <td>2.646667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>308314</td>\n",
       "      <td>196622.720996</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2008</td>\n",
       "      <td>Thriller,Drama,War,Documentary,Mystery,Action</td>\n",
       "      <td>356339</td>\n",
       "      <td>15000000</td>\n",
       "      <td>15700000</td>\n",
       "      <td>7.6</td>\n",
       "      <td>130.703583</td>\n",
       "      <td>94</td>\n",
       "      <td>Oscar</td>\n",
       "      <td>R</td>\n",
       "      <td>1.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>307201</td>\n",
       "      <td>159841.652064</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013</td>\n",
       "      <td>Crime,Thriller,Mystery,Documentary</td>\n",
       "      <td>46720</td>\n",
       "      <td>27220000</td>\n",
       "      <td>8551228</td>\n",
       "      <td>6.4</td>\n",
       "      <td>105.545533</td>\n",
       "      <td>37</td>\n",
       "      <td>other award</td>\n",
       "      <td>R</td>\n",
       "      <td>3.066667</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   video_id    cvt_per_day  weighted_categorical_position  \\\n",
       "0    385504  307127.605608                              1   \n",
       "1    300175  270338.426375                              1   \n",
       "2    361899  256165.867446                              1   \n",
       "3    308314  196622.720996                              3   \n",
       "4    307201  159841.652064                              1   \n",
       "\n",
       "   weighted_horizontal_poition  import_id  release_year  \\\n",
       "0                            3  lionsgate          2013   \n",
       "1                            3  lionsgate          2013   \n",
       "2                            3      other          2012   \n",
       "3                            4  lionsgate          2008   \n",
       "4                            3  lionsgate          2013   \n",
       "\n",
       "                                          genres  imdb_votes    budget  \\\n",
       "0                          Action,Thriller,Drama       69614  15000000   \n",
       "1                          Comedy,Crime,Thriller       46705  15000000   \n",
       "2                                    Crime,Drama      197596  26000000   \n",
       "3  Thriller,Drama,War,Documentary,Mystery,Action      356339  15000000   \n",
       "4             Crime,Thriller,Mystery,Documentary       46720  27220000   \n",
       "\n",
       "   boxoffice  imdb_rating  duration_in_mins  metacritic_score       awards  \\\n",
       "0   42930462          6.5        112.301017                51  other award   \n",
       "1    3301046          6.5         94.983250                41     no award   \n",
       "2   37397291          7.3        115.763675                58  other award   \n",
       "3   15700000          7.6        130.703583                94        Oscar   \n",
       "4    8551228          6.4        105.545533                37  other award   \n",
       "\n",
       "    mpaa  star_category  \n",
       "0  PG-13       1.710000  \n",
       "1      R       3.250000  \n",
       "2      R       2.646667  \n",
       "3      R       1.666667  \n",
       "4      R       3.066667  "
      ]
     },
     "execution_count": 203,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "video_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # visualize the missing values\n",
    "# import missingno as msno\n",
    "# msno.matrix(video_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "video_id                         0\n",
       "cvt_per_day                      0\n",
       "weighted_categorical_position    0\n",
       "weighted_horizontal_poition      0\n",
       "import_id                        0\n",
       "release_year                     0\n",
       "genres                           0\n",
       "imdb_votes                       0\n",
       "budget                           0\n",
       "boxoffice                        0\n",
       "imdb_rating                      0\n",
       "duration_in_mins                 0\n",
       "metacritic_score                 0\n",
       "awards                           0\n",
       "mpaa                             0\n",
       "star_category                    0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 205,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "video_df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "metadata": {},
   "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>video_id</th>\n",
       "      <th>cvt_per_day</th>\n",
       "      <th>weighted_categorical_position</th>\n",
       "      <th>weighted_horizontal_poition</th>\n",
       "      <th>release_year</th>\n",
       "      <th>imdb_votes</th>\n",
       "      <th>budget</th>\n",
       "      <th>boxoffice</th>\n",
       "      <th>imdb_rating</th>\n",
       "      <th>duration_in_mins</th>\n",
       "      <th>metacritic_score</th>\n",
       "      <th>star_category</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>4226.000000</td>\n",
       "      <td>4226.000000</td>\n",
       "      <td>4226.000000</td>\n",
       "      <td>4226.000000</td>\n",
       "      <td>4226.000000</td>\n",
       "      <td>4226.000000</td>\n",
       "      <td>4.226000e+03</td>\n",
       "      <td>4.226000e+03</td>\n",
       "      <td>4226.000000</td>\n",
       "      <td>4226.000000</td>\n",
       "      <td>4226.000000</td>\n",
       "      <td>4226.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>280371.162565</td>\n",
       "      <td>4218.630239</td>\n",
       "      <td>7.782537</td>\n",
       "      <td>28.103644</td>\n",
       "      <td>2001.056791</td>\n",
       "      <td>6462.924042</td>\n",
       "      <td>2.150743e+06</td>\n",
       "      <td>2.536338e+06</td>\n",
       "      <td>5.257099</td>\n",
       "      <td>89.556123</td>\n",
       "      <td>15.973734</td>\n",
       "      <td>0.954651</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>112640.127822</td>\n",
       "      <td>13036.079964</td>\n",
       "      <td>6.134183</td>\n",
       "      <td>11.863649</td>\n",
       "      <td>17.496849</td>\n",
       "      <td>31596.006790</td>\n",
       "      <td>7.176604e+06</td>\n",
       "      <td>8.243516e+06</td>\n",
       "      <td>2.122810</td>\n",
       "      <td>21.086183</td>\n",
       "      <td>26.205217</td>\n",
       "      <td>0.955045</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>7909.000000</td>\n",
       "      <td>2.187625</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1916.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>4.037250</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>285104.250000</td>\n",
       "      <td>351.168776</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>1998.000000</td>\n",
       "      <td>81.000000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>4.300000</td>\n",
       "      <td>82.601712</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>313891.500000</td>\n",
       "      <td>1193.499989</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>2008.000000</td>\n",
       "      <td>535.000000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>5.800000</td>\n",
       "      <td>90.730308</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>349345.750000</td>\n",
       "      <td>3356.788816</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>36.000000</td>\n",
       "      <td>2012.000000</td>\n",
       "      <td>3053.000000</td>\n",
       "      <td>1.500000e+06</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>6.800000</td>\n",
       "      <td>99.500312</td>\n",
       "      <td>41.000000</td>\n",
       "      <td>1.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>394880.000000</td>\n",
       "      <td>307127.605608</td>\n",
       "      <td>41.000000</td>\n",
       "      <td>70.000000</td>\n",
       "      <td>2017.000000</td>\n",
       "      <td>948630.000000</td>\n",
       "      <td>1.070000e+08</td>\n",
       "      <td>1.842088e+08</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>246.016767</td>\n",
       "      <td>100.000000</td>\n",
       "      <td>4.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            video_id    cvt_per_day  weighted_categorical_position  \\\n",
       "count    4226.000000    4226.000000                    4226.000000   \n",
       "mean   280371.162565    4218.630239                       7.782537   \n",
       "std    112640.127822   13036.079964                       6.134183   \n",
       "min      7909.000000       2.187625                       1.000000   \n",
       "25%    285104.250000     351.168776                       4.000000   \n",
       "50%    313891.500000    1193.499989                       6.000000   \n",
       "75%    349345.750000    3356.788816                       9.000000   \n",
       "max    394880.000000  307127.605608                      41.000000   \n",
       "\n",
       "       weighted_horizontal_poition  release_year     imdb_votes        budget  \\\n",
       "count                  4226.000000   4226.000000    4226.000000  4.226000e+03   \n",
       "mean                     28.103644   2001.056791    6462.924042  2.150743e+06   \n",
       "std                      11.863649     17.496849   31596.006790  7.176604e+06   \n",
       "min                       1.000000   1916.000000       0.000000  0.000000e+00   \n",
       "25%                      20.000000   1998.000000      81.000000  0.000000e+00   \n",
       "50%                      28.000000   2008.000000     535.000000  0.000000e+00   \n",
       "75%                      36.000000   2012.000000    3053.000000  1.500000e+06   \n",
       "max                      70.000000   2017.000000  948630.000000  1.070000e+08   \n",
       "\n",
       "          boxoffice  imdb_rating  duration_in_mins  metacritic_score  \\\n",
       "count  4.226000e+03  4226.000000       4226.000000       4226.000000   \n",
       "mean   2.536338e+06     5.257099         89.556123         15.973734   \n",
       "std    8.243516e+06     2.122810         21.086183         26.205217   \n",
       "min    0.000000e+00     0.000000          4.037250          0.000000   \n",
       "25%    0.000000e+00     4.300000         82.601712          0.000000   \n",
       "50%    0.000000e+00     5.800000         90.730308          0.000000   \n",
       "75%    0.000000e+00     6.800000         99.500312         41.000000   \n",
       "max    1.842088e+08    10.000000        246.016767        100.000000   \n",
       "\n",
       "       star_category  \n",
       "count    4226.000000  \n",
       "mean        0.954651  \n",
       "std         0.955045  \n",
       "min         0.000000  \n",
       "25%         0.000000  \n",
       "50%         1.000000  \n",
       "75%         1.666667  \n",
       "max         4.000000  "
      ]
     },
     "execution_count": 206,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#严谨一点应该drop掉ID和year\n",
    "video_df.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**there are columns that have a minimum value of zero (0). On some columns, a value of zero does not make sense and indicates an invalid or missing value.** \n",
    "-  imdb votes\n",
    "-  budget\n",
    "-  boxoffice\n",
    "-  imdb_rating\n",
    "-  metacritic_score\n",
    "-  star_category"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 224
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 6019,
     "status": "ok",
     "timestamp": 1546590669197,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "pxtf6XoJi0NI",
    "outputId": "9463a048-e23b-4e7f-bd5d-47708b430043",
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "imdb_votes           344\n",
      "budget              2454\n",
      "boxoffice           3194\n",
      "imdb_rating          344\n",
      "metacritic_score    3012\n",
      "star_category       1846\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print((video_df[['imdb_votes','budget','boxoffice','imdb_rating','metacritic_score','star_category']] == 0).sum())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**imdb votes and rating might not be invalid zeros, since many movies are not known by the public, indicating their 'unpopularness'**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "invalid_zeros = ['budget','boxoffice','metacritic_score','star_category','imdb_rating','imdb_votes']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "video_id                            0\n",
      "cvt_per_day                         0\n",
      "weighted_categorical_position       0\n",
      "weighted_horizontal_poition         0\n",
      "import_id                           0\n",
      "release_year                        0\n",
      "genres                              0\n",
      "imdb_votes                          0\n",
      "budget                           2454\n",
      "boxoffice                        3194\n",
      "imdb_rating                         0\n",
      "duration_in_mins                    0\n",
      "metacritic_score                 3012\n",
      "awards                              0\n",
      "mpaa                                0\n",
      "star_category                    1846\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# mark zero values as missing or NaN\n",
    "video_df[invalid_zeros]= video_df[invalid_zeros].replace(0, np.NaN)\n",
    "# count the number of NaN values in each column\n",
    "print(video_df.isnull().sum())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***Based on my understanding:***\n",
    "-  **uninformative missing: budget, boxoffice,metacritic_score,star_category**\n",
    "-  **infrmative missing: imdb_votes,imdb_rating**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['video_id', 'cvt_per_day', 'weighted_categorical_position',\n",
       "       'weighted_horizontal_poition', 'import_id', 'release_year', 'genres',\n",
       "       'imdb_votes', 'budget', 'boxoffice', 'imdb_rating', 'duration_in_mins',\n",
       "       'awards', 'mpaa', 'star_category'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Solution: \n",
    "# drop metacritic_score column\n",
    "original = video_df\n",
    "video_df = video_df.drop('metacritic_score',axis = 1)\n",
    "video_df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "# impute with median for budget and boxoffice column\n",
    "video_df[\"budget\"].fillna(video_df[\"budget\"].median(), inplace=True)\n",
    "video_df[\"boxoffice\"].fillna(video_df[\"boxoffice\"].median(), inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "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>video_id</th>\n",
       "      <th>cvt_per_day</th>\n",
       "      <th>weighted_categorical_position</th>\n",
       "      <th>weighted_horizontal_poition</th>\n",
       "      <th>release_year</th>\n",
       "      <th>imdb_votes</th>\n",
       "      <th>budget</th>\n",
       "      <th>boxoffice</th>\n",
       "      <th>imdb_rating</th>\n",
       "      <th>duration_in_mins</th>\n",
       "      <th>star_category</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>4226.000000</td>\n",
       "      <td>4226.000000</td>\n",
       "      <td>4226.000000</td>\n",
       "      <td>4226.000000</td>\n",
       "      <td>4226.000000</td>\n",
       "      <td>4226.000000</td>\n",
       "      <td>4.226000e+03</td>\n",
       "      <td>4.226000e+03</td>\n",
       "      <td>4226.000000</td>\n",
       "      <td>4226.000000</td>\n",
       "      <td>2380.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>280371.162565</td>\n",
       "      <td>4218.630239</td>\n",
       "      <td>7.782537</td>\n",
       "      <td>28.103644</td>\n",
       "      <td>2001.056791</td>\n",
       "      <td>6462.924042</td>\n",
       "      <td>3.021780e+06</td>\n",
       "      <td>8.999335e+06</td>\n",
       "      <td>5.257099</td>\n",
       "      <td>89.556123</td>\n",
       "      <td>1.695107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>112640.127822</td>\n",
       "      <td>13036.079964</td>\n",
       "      <td>6.134183</td>\n",
       "      <td>11.863649</td>\n",
       "      <td>17.496849</td>\n",
       "      <td>31596.006790</td>\n",
       "      <td>6.950107e+06</td>\n",
       "      <td>6.975854e+06</td>\n",
       "      <td>2.122810</td>\n",
       "      <td>21.086183</td>\n",
       "      <td>0.603482</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>7909.000000</td>\n",
       "      <td>2.187625</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1916.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.316000e+03</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>4.037250</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>285104.250000</td>\n",
       "      <td>351.168776</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>1998.000000</td>\n",
       "      <td>81.000000</td>\n",
       "      <td>1.500000e+06</td>\n",
       "      <td>8.551228e+06</td>\n",
       "      <td>4.300000</td>\n",
       "      <td>82.601712</td>\n",
       "      <td>1.112778</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>313891.500000</td>\n",
       "      <td>1193.499989</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>2008.000000</td>\n",
       "      <td>535.000000</td>\n",
       "      <td>1.500000e+06</td>\n",
       "      <td>8.551228e+06</td>\n",
       "      <td>5.800000</td>\n",
       "      <td>90.730308</td>\n",
       "      <td>1.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>349345.750000</td>\n",
       "      <td>3356.788816</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>36.000000</td>\n",
       "      <td>2012.000000</td>\n",
       "      <td>3053.000000</td>\n",
       "      <td>1.500000e+06</td>\n",
       "      <td>8.551228e+06</td>\n",
       "      <td>6.800000</td>\n",
       "      <td>99.500312</td>\n",
       "      <td>2.018750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>394880.000000</td>\n",
       "      <td>307127.605608</td>\n",
       "      <td>41.000000</td>\n",
       "      <td>70.000000</td>\n",
       "      <td>2017.000000</td>\n",
       "      <td>948630.000000</td>\n",
       "      <td>1.070000e+08</td>\n",
       "      <td>1.842088e+08</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>246.016767</td>\n",
       "      <td>4.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            video_id    cvt_per_day  weighted_categorical_position  \\\n",
       "count    4226.000000    4226.000000                    4226.000000   \n",
       "mean   280371.162565    4218.630239                       7.782537   \n",
       "std    112640.127822   13036.079964                       6.134183   \n",
       "min      7909.000000       2.187625                       1.000000   \n",
       "25%    285104.250000     351.168776                       4.000000   \n",
       "50%    313891.500000    1193.499989                       6.000000   \n",
       "75%    349345.750000    3356.788816                       9.000000   \n",
       "max    394880.000000  307127.605608                      41.000000   \n",
       "\n",
       "       weighted_horizontal_poition  release_year     imdb_votes        budget  \\\n",
       "count                  4226.000000   4226.000000    4226.000000  4.226000e+03   \n",
       "mean                     28.103644   2001.056791    6462.924042  3.021780e+06   \n",
       "std                      11.863649     17.496849   31596.006790  6.950107e+06   \n",
       "min                       1.000000   1916.000000       0.000000  1.000000e+00   \n",
       "25%                      20.000000   1998.000000      81.000000  1.500000e+06   \n",
       "50%                      28.000000   2008.000000     535.000000  1.500000e+06   \n",
       "75%                      36.000000   2012.000000    3053.000000  1.500000e+06   \n",
       "max                      70.000000   2017.000000  948630.000000  1.070000e+08   \n",
       "\n",
       "          boxoffice  imdb_rating  duration_in_mins  star_category  \n",
       "count  4.226000e+03  4226.000000       4226.000000    2380.000000  \n",
       "mean   8.999335e+06     5.257099         89.556123       1.695107  \n",
       "std    6.975854e+06     2.122810         21.086183       0.603482  \n",
       "min    1.316000e+03     0.000000          4.037250       1.000000  \n",
       "25%    8.551228e+06     4.300000         82.601712       1.112778  \n",
       "50%    8.551228e+06     5.800000         90.730308       1.666667  \n",
       "75%    8.551228e+06     6.800000         99.500312       2.018750  \n",
       "max    1.842088e+08    10.000000        246.016767       4.000000  "
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "video_df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "video_df[\"star_category\"].fillna(video_df[\"star_category\"].median(), inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "video_id                         0\n",
      "cvt_per_day                      0\n",
      "weighted_categorical_position    0\n",
      "weighted_horizontal_poition      0\n",
      "import_id                        0\n",
      "release_year                     0\n",
      "genres                           0\n",
      "imdb_votes                       0\n",
      "budget                           0\n",
      "boxoffice                        0\n",
      "imdb_rating                      0\n",
      "duration_in_mins                 0\n",
      "awards                           0\n",
      "mpaa                             0\n",
      "star_category                    0\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# double check there's no untreated null value\n",
    "print(video_df.isnull().sum())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Part 2.2 Catagorical Feature -- Genre"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [],
   "source": [
    "# make a copy of the curret state of video_df data\n",
    "video_before_genre = video_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# catagorical feature"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "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>video_id</th>\n",
       "      <th>cvt_per_day</th>\n",
       "      <th>weighted_categorical_position</th>\n",
       "      <th>weighted_horizontal_poition</th>\n",
       "      <th>import_id</th>\n",
       "      <th>release_year</th>\n",
       "      <th>genres</th>\n",
       "      <th>imdb_votes</th>\n",
       "      <th>budget</th>\n",
       "      <th>boxoffice</th>\n",
       "      <th>imdb_rating</th>\n",
       "      <th>duration_in_mins</th>\n",
       "      <th>awards</th>\n",
       "      <th>mpaa</th>\n",
       "      <th>star_category</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>385504</td>\n",
       "      <td>307127.605608</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013</td>\n",
       "      <td>Action,Thriller,Drama</td>\n",
       "      <td>69614</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>42930462.0</td>\n",
       "      <td>6.5</td>\n",
       "      <td>112.301017</td>\n",
       "      <td>other award</td>\n",
       "      <td>PG-13</td>\n",
       "      <td>1.710000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>300175</td>\n",
       "      <td>270338.426375</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013</td>\n",
       "      <td>Comedy,Crime,Thriller</td>\n",
       "      <td>46705</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>3301046.0</td>\n",
       "      <td>6.5</td>\n",
       "      <td>94.983250</td>\n",
       "      <td>no award</td>\n",
       "      <td>R</td>\n",
       "      <td>3.250000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>361899</td>\n",
       "      <td>256165.867446</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>other</td>\n",
       "      <td>2012</td>\n",
       "      <td>Crime,Drama</td>\n",
       "      <td>197596</td>\n",
       "      <td>26000000.0</td>\n",
       "      <td>37397291.0</td>\n",
       "      <td>7.3</td>\n",
       "      <td>115.763675</td>\n",
       "      <td>other award</td>\n",
       "      <td>R</td>\n",
       "      <td>2.646667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>308314</td>\n",
       "      <td>196622.720996</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2008</td>\n",
       "      <td>Thriller,Drama,War,Documentary,Mystery,Action</td>\n",
       "      <td>356339</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>15700000.0</td>\n",
       "      <td>7.6</td>\n",
       "      <td>130.703583</td>\n",
       "      <td>Oscar</td>\n",
       "      <td>R</td>\n",
       "      <td>1.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>307201</td>\n",
       "      <td>159841.652064</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013</td>\n",
       "      <td>Crime,Thriller,Mystery,Documentary</td>\n",
       "      <td>46720</td>\n",
       "      <td>27220000.0</td>\n",
       "      <td>8551228.0</td>\n",
       "      <td>6.4</td>\n",
       "      <td>105.545533</td>\n",
       "      <td>other award</td>\n",
       "      <td>R</td>\n",
       "      <td>3.066667</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   video_id    cvt_per_day  weighted_categorical_position  \\\n",
       "0    385504  307127.605608                              1   \n",
       "1    300175  270338.426375                              1   \n",
       "2    361899  256165.867446                              1   \n",
       "3    308314  196622.720996                              3   \n",
       "4    307201  159841.652064                              1   \n",
       "\n",
       "   weighted_horizontal_poition  import_id  release_year  \\\n",
       "0                            3  lionsgate          2013   \n",
       "1                            3  lionsgate          2013   \n",
       "2                            3      other          2012   \n",
       "3                            4  lionsgate          2008   \n",
       "4                            3  lionsgate          2013   \n",
       "\n",
       "                                          genres  imdb_votes      budget  \\\n",
       "0                          Action,Thriller,Drama       69614  15000000.0   \n",
       "1                          Comedy,Crime,Thriller       46705  15000000.0   \n",
       "2                                    Crime,Drama      197596  26000000.0   \n",
       "3  Thriller,Drama,War,Documentary,Mystery,Action      356339  15000000.0   \n",
       "4             Crime,Thriller,Mystery,Documentary       46720  27220000.0   \n",
       "\n",
       "    boxoffice  imdb_rating  duration_in_mins       awards   mpaa  \\\n",
       "0  42930462.0          6.5        112.301017  other award  PG-13   \n",
       "1   3301046.0          6.5         94.983250     no award      R   \n",
       "2  37397291.0          7.3        115.763675  other award      R   \n",
       "3  15700000.0          7.6        130.703583        Oscar      R   \n",
       "4   8551228.0          6.4        105.545533  other award      R   \n",
       "\n",
       "   star_category  \n",
       "0       1.710000  \n",
       "1       3.250000  \n",
       "2       2.646667  \n",
       "3       1.666667  \n",
       "4       3.066667  "
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "video_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [],
   "source": [
    "genre_set = set()\n",
    "for i,row in video_df.iterrows():\n",
    "    this_row = row['genres'].split(',')\n",
    "    for elem in this_row:\n",
    "        if elem not in genre_set:\n",
    "            genre_set.add(elem)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "27"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(genre_set)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "video_id   \n",
       "385504    0      Action\n",
       "          1    Thriller\n",
       "          2       Drama\n",
       "300175    0      Comedy\n",
       "          1       Crime\n",
       "dtype: object"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cleaned = video_df.set_index('video_id').genres.str.split(',', expand=True).stack()\n",
    "cleaned.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "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>video_id</th>\n",
       "      <th>g_Action</th>\n",
       "      <th>g_Adult</th>\n",
       "      <th>g_Adventure</th>\n",
       "      <th>g_Animation</th>\n",
       "      <th>g_Anime</th>\n",
       "      <th>g_Comedy</th>\n",
       "      <th>g_Crime</th>\n",
       "      <th>g_Documentary</th>\n",
       "      <th>g_Drama</th>\n",
       "      <th>g_Fantasy</th>\n",
       "      <th>g_Foreign/International</th>\n",
       "      <th>g_Holiday</th>\n",
       "      <th>g_Horror</th>\n",
       "      <th>g_Independent</th>\n",
       "      <th>g_Kids &amp; Family</th>\n",
       "      <th>g_LGBT</th>\n",
       "      <th>g_Lifestyle</th>\n",
       "      <th>g_Music</th>\n",
       "      <th>g_Musicals</th>\n",
       "      <th>g_Mystery</th>\n",
       "      <th>g_Reality</th>\n",
       "      <th>g_Romance</th>\n",
       "      <th>g_Sci-Fi</th>\n",
       "      <th>g_Sport</th>\n",
       "      <th>g_Thriller</th>\n",
       "      <th>g_War</th>\n",
       "      <th>g_Western</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>7909</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>7910</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>7911</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>7942</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>7943</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   video_id  g_Action  g_Adult  g_Adventure  g_Animation  g_Anime  g_Comedy  \\\n",
       "0      7909         0        0            0            0        0         1   \n",
       "1      7910         0        0            0            0        0         1   \n",
       "2      7911         0        0            0            0        0         0   \n",
       "3      7942         0        0            0            0        0         0   \n",
       "4      7943         0        0            0            0        0         0   \n",
       "\n",
       "   g_Crime  g_Documentary  g_Drama  g_Fantasy  g_Foreign/International  \\\n",
       "0        0              0        1          0                        0   \n",
       "1        1              0        1          0                        0   \n",
       "2        0              0        0          1                        0   \n",
       "3        0              0        1          0                        0   \n",
       "4        1              0        1          0                        0   \n",
       "\n",
       "   g_Holiday  g_Horror  g_Independent  g_Kids & Family  g_LGBT  g_Lifestyle  \\\n",
       "0          0         0              0                0       0            0   \n",
       "1          0         1              0                0       0            0   \n",
       "2          0         1              0                0       0            0   \n",
       "3          0         0              0                0       0            0   \n",
       "4          0         0              0                0       0            0   \n",
       "\n",
       "   g_Music  g_Musicals  g_Mystery  g_Reality  g_Romance  g_Sci-Fi  g_Sport  \\\n",
       "0        0           0          1          0          0         0        0   \n",
       "1        0           0          1          0          0         0        0   \n",
       "2        0           0          1          0          0         1        0   \n",
       "3        0           0          0          0          0         0        0   \n",
       "4        0           0          0          0          0         0        0   \n",
       "\n",
       "   g_Thriller  g_War  g_Western  \n",
       "0           0      0          0  \n",
       "1           0      0          0  \n",
       "2           1      0          0  \n",
       "3           0      0          0  \n",
       "4           0      0          0  "
      ]
     },
     "execution_count": 114,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "genres_subset = pd.get_dummies(cleaned, prefix='g').groupby(level=0).sum().reset_index()\n",
    "genres_subset.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "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>video_id</th>\n",
       "      <th>cvt_per_day</th>\n",
       "      <th>weighted_categorical_position</th>\n",
       "      <th>weighted_horizontal_poition</th>\n",
       "      <th>import_id</th>\n",
       "      <th>release_year</th>\n",
       "      <th>genres</th>\n",
       "      <th>imdb_votes</th>\n",
       "      <th>budget</th>\n",
       "      <th>boxoffice</th>\n",
       "      <th>imdb_rating</th>\n",
       "      <th>duration_in_mins</th>\n",
       "      <th>awards</th>\n",
       "      <th>mpaa</th>\n",
       "      <th>star_category</th>\n",
       "      <th>g_Action</th>\n",
       "      <th>g_Adult</th>\n",
       "      <th>g_Adventure</th>\n",
       "      <th>g_Animation</th>\n",
       "      <th>g_Anime</th>\n",
       "      <th>g_Comedy</th>\n",
       "      <th>g_Crime</th>\n",
       "      <th>g_Documentary</th>\n",
       "      <th>g_Drama</th>\n",
       "      <th>g_Fantasy</th>\n",
       "      <th>g_Foreign/International</th>\n",
       "      <th>g_Holiday</th>\n",
       "      <th>g_Horror</th>\n",
       "      <th>g_Independent</th>\n",
       "      <th>g_Kids &amp; Family</th>\n",
       "      <th>g_LGBT</th>\n",
       "      <th>g_Lifestyle</th>\n",
       "      <th>g_Music</th>\n",
       "      <th>g_Musicals</th>\n",
       "      <th>g_Mystery</th>\n",
       "      <th>g_Reality</th>\n",
       "      <th>g_Romance</th>\n",
       "      <th>g_Sci-Fi</th>\n",
       "      <th>g_Sport</th>\n",
       "      <th>g_Thriller</th>\n",
       "      <th>g_War</th>\n",
       "      <th>g_Western</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>385504</td>\n",
       "      <td>307127.605608</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013</td>\n",
       "      <td>Action,Thriller,Drama</td>\n",
       "      <td>69614</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>42930462.0</td>\n",
       "      <td>6.5</td>\n",
       "      <td>112.301017</td>\n",
       "      <td>other award</td>\n",
       "      <td>PG-13</td>\n",
       "      <td>1.710000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>300175</td>\n",
       "      <td>270338.426375</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013</td>\n",
       "      <td>Comedy,Crime,Thriller</td>\n",
       "      <td>46705</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>3301046.0</td>\n",
       "      <td>6.5</td>\n",
       "      <td>94.983250</td>\n",
       "      <td>no award</td>\n",
       "      <td>R</td>\n",
       "      <td>3.250000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>361899</td>\n",
       "      <td>256165.867446</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>other</td>\n",
       "      <td>2012</td>\n",
       "      <td>Crime,Drama</td>\n",
       "      <td>197596</td>\n",
       "      <td>26000000.0</td>\n",
       "      <td>37397291.0</td>\n",
       "      <td>7.3</td>\n",
       "      <td>115.763675</td>\n",
       "      <td>other award</td>\n",
       "      <td>R</td>\n",
       "      <td>2.646667</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>308314</td>\n",
       "      <td>196622.720996</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2008</td>\n",
       "      <td>Thriller,Drama,War,Documentary,Mystery,Action</td>\n",
       "      <td>356339</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>15700000.0</td>\n",
       "      <td>7.6</td>\n",
       "      <td>130.703583</td>\n",
       "      <td>Oscar</td>\n",
       "      <td>R</td>\n",
       "      <td>1.666667</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>307201</td>\n",
       "      <td>159841.652064</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013</td>\n",
       "      <td>Crime,Thriller,Mystery,Documentary</td>\n",
       "      <td>46720</td>\n",
       "      <td>27220000.0</td>\n",
       "      <td>8551228.0</td>\n",
       "      <td>6.4</td>\n",
       "      <td>105.545533</td>\n",
       "      <td>other award</td>\n",
       "      <td>R</td>\n",
       "      <td>3.066667</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   video_id    cvt_per_day  weighted_categorical_position  \\\n",
       "0    385504  307127.605608                              1   \n",
       "1    300175  270338.426375                              1   \n",
       "2    361899  256165.867446                              1   \n",
       "3    308314  196622.720996                              3   \n",
       "4    307201  159841.652064                              1   \n",
       "\n",
       "   weighted_horizontal_poition  import_id  release_year  \\\n",
       "0                            3  lionsgate          2013   \n",
       "1                            3  lionsgate          2013   \n",
       "2                            3      other          2012   \n",
       "3                            4  lionsgate          2008   \n",
       "4                            3  lionsgate          2013   \n",
       "\n",
       "                                          genres  imdb_votes      budget  \\\n",
       "0                          Action,Thriller,Drama       69614  15000000.0   \n",
       "1                          Comedy,Crime,Thriller       46705  15000000.0   \n",
       "2                                    Crime,Drama      197596  26000000.0   \n",
       "3  Thriller,Drama,War,Documentary,Mystery,Action      356339  15000000.0   \n",
       "4             Crime,Thriller,Mystery,Documentary       46720  27220000.0   \n",
       "\n",
       "    boxoffice  imdb_rating  duration_in_mins       awards   mpaa  \\\n",
       "0  42930462.0          6.5        112.301017  other award  PG-13   \n",
       "1   3301046.0          6.5         94.983250     no award      R   \n",
       "2  37397291.0          7.3        115.763675  other award      R   \n",
       "3  15700000.0          7.6        130.703583        Oscar      R   \n",
       "4   8551228.0          6.4        105.545533  other award      R   \n",
       "\n",
       "   star_category  g_Action  g_Adult  g_Adventure  g_Animation  g_Anime  \\\n",
       "0       1.710000         1        0            0            0        0   \n",
       "1       3.250000         0        0            0            0        0   \n",
       "2       2.646667         0        0            0            0        0   \n",
       "3       1.666667         1        0            0            0        0   \n",
       "4       3.066667         0        0            0            0        0   \n",
       "\n",
       "   g_Comedy  g_Crime  g_Documentary  g_Drama  g_Fantasy  \\\n",
       "0         0        0              0        1          0   \n",
       "1         1        1              0        0          0   \n",
       "2         0        1              0        1          0   \n",
       "3         0        0              1        1          0   \n",
       "4         0        1              1        0          0   \n",
       "\n",
       "   g_Foreign/International  g_Holiday  g_Horror  g_Independent  \\\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",
       "   g_Kids & Family  g_LGBT  g_Lifestyle  g_Music  g_Musicals  g_Mystery  \\\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          1   \n",
       "4                0       0            0        0           0          1   \n",
       "\n",
       "   g_Reality  g_Romance  g_Sci-Fi  g_Sport  g_Thriller  g_War  g_Western  \n",
       "0          0          0         0        0           1      0          0  \n",
       "1          0          0         0        0           1      0          0  \n",
       "2          0          0         0        0           0      0          0  \n",
       "3          0          0         0        0           1      1          0  \n",
       "4          0          0         0        0           1      0          0  "
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "video_updated = pd.merge(video_df,genres_subset,on='video_id')\n",
    "video_updated.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4226, 42)"
      ]
     },
     "execution_count": 119,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "video_updated.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Part2.3 Catagorical feature --- mpaa"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['PG-13', 'R', 'PG', 'NC-17', 'NotRated', 'G'], dtype=object)"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "video_updated['mpaa'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [],
   "source": [
    "#make a copy again\n",
    "video_before_mpaa = video_updated"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [],
   "source": [
    "# recode answers on a 1,2,3,4,5,6,scale\n",
    "video_updated = video_updated.replace({'G':1, \n",
    "                                       'PG':2, \n",
    "                                       'PG-13':3,\n",
    "                                       'R':4, \n",
    "                                       'NC-17':5,\n",
    "                                       'NotRated':0})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "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>video_id</th>\n",
       "      <th>cvt_per_day</th>\n",
       "      <th>weighted_categorical_position</th>\n",
       "      <th>weighted_horizontal_poition</th>\n",
       "      <th>import_id</th>\n",
       "      <th>release_year</th>\n",
       "      <th>genres</th>\n",
       "      <th>imdb_votes</th>\n",
       "      <th>budget</th>\n",
       "      <th>boxoffice</th>\n",
       "      <th>imdb_rating</th>\n",
       "      <th>duration_in_mins</th>\n",
       "      <th>awards</th>\n",
       "      <th>mpaa</th>\n",
       "      <th>star_category</th>\n",
       "      <th>g_Action</th>\n",
       "      <th>g_Adult</th>\n",
       "      <th>g_Adventure</th>\n",
       "      <th>g_Animation</th>\n",
       "      <th>g_Anime</th>\n",
       "      <th>g_Comedy</th>\n",
       "      <th>g_Crime</th>\n",
       "      <th>g_Documentary</th>\n",
       "      <th>g_Drama</th>\n",
       "      <th>g_Fantasy</th>\n",
       "      <th>g_Foreign/International</th>\n",
       "      <th>g_Holiday</th>\n",
       "      <th>g_Horror</th>\n",
       "      <th>g_Independent</th>\n",
       "      <th>g_Kids &amp; Family</th>\n",
       "      <th>g_LGBT</th>\n",
       "      <th>g_Lifestyle</th>\n",
       "      <th>g_Music</th>\n",
       "      <th>g_Musicals</th>\n",
       "      <th>g_Mystery</th>\n",
       "      <th>g_Reality</th>\n",
       "      <th>g_Romance</th>\n",
       "      <th>g_Sci-Fi</th>\n",
       "      <th>g_Sport</th>\n",
       "      <th>g_Thriller</th>\n",
       "      <th>g_War</th>\n",
       "      <th>g_Western</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>385504</td>\n",
       "      <td>307127.605608</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013</td>\n",
       "      <td>Action,Thriller,Drama</td>\n",
       "      <td>69614</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>42930462.0</td>\n",
       "      <td>6.5</td>\n",
       "      <td>112.301017</td>\n",
       "      <td>other award</td>\n",
       "      <td>3</td>\n",
       "      <td>1.710000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>300175</td>\n",
       "      <td>270338.426375</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013</td>\n",
       "      <td>Comedy,Crime,Thriller</td>\n",
       "      <td>46705</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>3301046.0</td>\n",
       "      <td>6.5</td>\n",
       "      <td>94.983250</td>\n",
       "      <td>no award</td>\n",
       "      <td>4</td>\n",
       "      <td>3.250000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>361899</td>\n",
       "      <td>256165.867446</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>other</td>\n",
       "      <td>2012</td>\n",
       "      <td>Crime,Drama</td>\n",
       "      <td>197596</td>\n",
       "      <td>26000000.0</td>\n",
       "      <td>37397291.0</td>\n",
       "      <td>7.3</td>\n",
       "      <td>115.763675</td>\n",
       "      <td>other award</td>\n",
       "      <td>4</td>\n",
       "      <td>2.646667</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>308314</td>\n",
       "      <td>196622.720996</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2008</td>\n",
       "      <td>Thriller,Drama,War,Documentary,Mystery,Action</td>\n",
       "      <td>356339</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>15700000.0</td>\n",
       "      <td>7.6</td>\n",
       "      <td>130.703583</td>\n",
       "      <td>Oscar</td>\n",
       "      <td>4</td>\n",
       "      <td>1.666667</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>307201</td>\n",
       "      <td>159841.652064</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013</td>\n",
       "      <td>Crime,Thriller,Mystery,Documentary</td>\n",
       "      <td>46720</td>\n",
       "      <td>27220000.0</td>\n",
       "      <td>8551228.0</td>\n",
       "      <td>6.4</td>\n",
       "      <td>105.545533</td>\n",
       "      <td>other award</td>\n",
       "      <td>4</td>\n",
       "      <td>3.066667</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   video_id    cvt_per_day  weighted_categorical_position  \\\n",
       "0    385504  307127.605608                              1   \n",
       "1    300175  270338.426375                              1   \n",
       "2    361899  256165.867446                              1   \n",
       "3    308314  196622.720996                              3   \n",
       "4    307201  159841.652064                              1   \n",
       "\n",
       "   weighted_horizontal_poition  import_id  release_year  \\\n",
       "0                            3  lionsgate          2013   \n",
       "1                            3  lionsgate          2013   \n",
       "2                            3      other          2012   \n",
       "3                            4  lionsgate          2008   \n",
       "4                            3  lionsgate          2013   \n",
       "\n",
       "                                          genres  imdb_votes      budget  \\\n",
       "0                          Action,Thriller,Drama       69614  15000000.0   \n",
       "1                          Comedy,Crime,Thriller       46705  15000000.0   \n",
       "2                                    Crime,Drama      197596  26000000.0   \n",
       "3  Thriller,Drama,War,Documentary,Mystery,Action      356339  15000000.0   \n",
       "4             Crime,Thriller,Mystery,Documentary       46720  27220000.0   \n",
       "\n",
       "    boxoffice  imdb_rating  duration_in_mins       awards  mpaa  \\\n",
       "0  42930462.0          6.5        112.301017  other award     3   \n",
       "1   3301046.0          6.5         94.983250     no award     4   \n",
       "2  37397291.0          7.3        115.763675  other award     4   \n",
       "3  15700000.0          7.6        130.703583        Oscar     4   \n",
       "4   8551228.0          6.4        105.545533  other award     4   \n",
       "\n",
       "   star_category  g_Action  g_Adult  g_Adventure  g_Animation  g_Anime  \\\n",
       "0       1.710000         1        0            0            0        0   \n",
       "1       3.250000         0        0            0            0        0   \n",
       "2       2.646667         0        0            0            0        0   \n",
       "3       1.666667         1        0            0            0        0   \n",
       "4       3.066667         0        0            0            0        0   \n",
       "\n",
       "   g_Comedy  g_Crime  g_Documentary  g_Drama  g_Fantasy  \\\n",
       "0         0        0              0        1          0   \n",
       "1         1        1              0        0          0   \n",
       "2         0        1              0        1          0   \n",
       "3         0        0              1        1          0   \n",
       "4         0        1              1        0          0   \n",
       "\n",
       "   g_Foreign/International  g_Holiday  g_Horror  g_Independent  \\\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",
       "   g_Kids & Family  g_LGBT  g_Lifestyle  g_Music  g_Musicals  g_Mystery  \\\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          1   \n",
       "4                0       0            0        0           0          1   \n",
       "\n",
       "   g_Reality  g_Romance  g_Sci-Fi  g_Sport  g_Thriller  g_War  g_Western  \n",
       "0          0          0         0        0           1      0          0  \n",
       "1          0          0         0        0           1      0          0  \n",
       "2          0          0         0        0           0      0          0  \n",
       "3          0          0         0        0           1      1          0  \n",
       "4          0          0         0        0           1      0          0  "
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "video_updated.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Part2.4 Datetime: release year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [],
   "source": [
    "import datetime"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [],
   "source": [
    "video_updated['release_year'] = pd.to_datetime(video_updated['release_year'],format = '%Y')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "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>video_id</th>\n",
       "      <th>cvt_per_day</th>\n",
       "      <th>weighted_categorical_position</th>\n",
       "      <th>weighted_horizontal_poition</th>\n",
       "      <th>import_id</th>\n",
       "      <th>release_year</th>\n",
       "      <th>genres</th>\n",
       "      <th>imdb_votes</th>\n",
       "      <th>budget</th>\n",
       "      <th>boxoffice</th>\n",
       "      <th>imdb_rating</th>\n",
       "      <th>duration_in_mins</th>\n",
       "      <th>awards</th>\n",
       "      <th>mpaa</th>\n",
       "      <th>star_category</th>\n",
       "      <th>g_Action</th>\n",
       "      <th>g_Adult</th>\n",
       "      <th>g_Adventure</th>\n",
       "      <th>g_Animation</th>\n",
       "      <th>g_Anime</th>\n",
       "      <th>g_Comedy</th>\n",
       "      <th>g_Crime</th>\n",
       "      <th>g_Documentary</th>\n",
       "      <th>g_Drama</th>\n",
       "      <th>g_Fantasy</th>\n",
       "      <th>g_Foreign/International</th>\n",
       "      <th>g_Holiday</th>\n",
       "      <th>g_Horror</th>\n",
       "      <th>g_Independent</th>\n",
       "      <th>g_Kids &amp; Family</th>\n",
       "      <th>g_LGBT</th>\n",
       "      <th>g_Lifestyle</th>\n",
       "      <th>g_Music</th>\n",
       "      <th>g_Musicals</th>\n",
       "      <th>g_Mystery</th>\n",
       "      <th>g_Reality</th>\n",
       "      <th>g_Romance</th>\n",
       "      <th>g_Sci-Fi</th>\n",
       "      <th>g_Sport</th>\n",
       "      <th>g_Thriller</th>\n",
       "      <th>g_War</th>\n",
       "      <th>g_Western</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>385504</td>\n",
       "      <td>307127.605608</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013-01-01</td>\n",
       "      <td>Action,Thriller,Drama</td>\n",
       "      <td>69614</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>42930462.0</td>\n",
       "      <td>6.5</td>\n",
       "      <td>112.301017</td>\n",
       "      <td>other award</td>\n",
       "      <td>3</td>\n",
       "      <td>1.710000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>300175</td>\n",
       "      <td>270338.426375</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013-01-01</td>\n",
       "      <td>Comedy,Crime,Thriller</td>\n",
       "      <td>46705</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>3301046.0</td>\n",
       "      <td>6.5</td>\n",
       "      <td>94.983250</td>\n",
       "      <td>no award</td>\n",
       "      <td>4</td>\n",
       "      <td>3.250000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>361899</td>\n",
       "      <td>256165.867446</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>other</td>\n",
       "      <td>2012-01-01</td>\n",
       "      <td>Crime,Drama</td>\n",
       "      <td>197596</td>\n",
       "      <td>26000000.0</td>\n",
       "      <td>37397291.0</td>\n",
       "      <td>7.3</td>\n",
       "      <td>115.763675</td>\n",
       "      <td>other award</td>\n",
       "      <td>4</td>\n",
       "      <td>2.646667</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>308314</td>\n",
       "      <td>196622.720996</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2008-01-01</td>\n",
       "      <td>Thriller,Drama,War,Documentary,Mystery,Action</td>\n",
       "      <td>356339</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>15700000.0</td>\n",
       "      <td>7.6</td>\n",
       "      <td>130.703583</td>\n",
       "      <td>Oscar</td>\n",
       "      <td>4</td>\n",
       "      <td>1.666667</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>307201</td>\n",
       "      <td>159841.652064</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>lionsgate</td>\n",
       "      <td>2013-01-01</td>\n",
       "      <td>Crime,Thriller,Mystery,Documentary</td>\n",
       "      <td>46720</td>\n",
       "      <td>27220000.0</td>\n",
       "      <td>8551228.0</td>\n",
       "      <td>6.4</td>\n",
       "      <td>105.545533</td>\n",
       "      <td>other award</td>\n",
       "      <td>4</td>\n",
       "      <td>3.066667</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   video_id    cvt_per_day  weighted_categorical_position  \\\n",
       "0    385504  307127.605608                              1   \n",
       "1    300175  270338.426375                              1   \n",
       "2    361899  256165.867446                              1   \n",
       "3    308314  196622.720996                              3   \n",
       "4    307201  159841.652064                              1   \n",
       "\n",
       "   weighted_horizontal_poition  import_id release_year  \\\n",
       "0                            3  lionsgate   2013-01-01   \n",
       "1                            3  lionsgate   2013-01-01   \n",
       "2                            3      other   2012-01-01   \n",
       "3                            4  lionsgate   2008-01-01   \n",
       "4                            3  lionsgate   2013-01-01   \n",
       "\n",
       "                                          genres  imdb_votes      budget  \\\n",
       "0                          Action,Thriller,Drama       69614  15000000.0   \n",
       "1                          Comedy,Crime,Thriller       46705  15000000.0   \n",
       "2                                    Crime,Drama      197596  26000000.0   \n",
       "3  Thriller,Drama,War,Documentary,Mystery,Action      356339  15000000.0   \n",
       "4             Crime,Thriller,Mystery,Documentary       46720  27220000.0   \n",
       "\n",
       "    boxoffice  imdb_rating  duration_in_mins       awards  mpaa  \\\n",
       "0  42930462.0          6.5        112.301017  other award     3   \n",
       "1   3301046.0          6.5         94.983250     no award     4   \n",
       "2  37397291.0          7.3        115.763675  other award     4   \n",
       "3  15700000.0          7.6        130.703583        Oscar     4   \n",
       "4   8551228.0          6.4        105.545533  other award     4   \n",
       "\n",
       "   star_category  g_Action  g_Adult  g_Adventure  g_Animation  g_Anime  \\\n",
       "0       1.710000         1        0            0            0        0   \n",
       "1       3.250000         0        0            0            0        0   \n",
       "2       2.646667         0        0            0            0        0   \n",
       "3       1.666667         1        0            0            0        0   \n",
       "4       3.066667         0        0            0            0        0   \n",
       "\n",
       "   g_Comedy  g_Crime  g_Documentary  g_Drama  g_Fantasy  \\\n",
       "0         0        0              0        1          0   \n",
       "1         1        1              0        0          0   \n",
       "2         0        1              0        1          0   \n",
       "3         0        0              1        1          0   \n",
       "4         0        1              1        0          0   \n",
       "\n",
       "   g_Foreign/International  g_Holiday  g_Horror  g_Independent  \\\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",
       "   g_Kids & Family  g_LGBT  g_Lifestyle  g_Music  g_Musicals  g_Mystery  \\\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          1   \n",
       "4                0       0            0        0           0          1   \n",
       "\n",
       "   g_Reality  g_Romance  g_Sci-Fi  g_Sport  g_Thriller  g_War  g_Western  \n",
       "0          0          0         0        0           1      0          0  \n",
       "1          0          0         0        0           1      0          0  \n",
       "2          0          0         0        0           0      0          0  \n",
       "3          0          0         0        0           1      1          0  \n",
       "4          0          0         0        0           1      0          0  "
      ]
     },
     "execution_count": 129,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "video_updated.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "JMTIEpY7IfPp"
   },
   "source": [
    "Read more for handling [categorical feature](https://github.com/scikit-learn-contrib/categorical-encoding)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [],
   "source": [
    "video_updated1 = pd.get_dummies(video_updated, columns=['import_id'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [],
   "source": [
    "video_updated2 = pd.get_dummies(video_updated1, columns=['awards'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {},
   "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>video_id</th>\n",
       "      <th>cvt_per_day</th>\n",
       "      <th>weighted_categorical_position</th>\n",
       "      <th>weighted_horizontal_poition</th>\n",
       "      <th>release_year</th>\n",
       "      <th>genres</th>\n",
       "      <th>imdb_votes</th>\n",
       "      <th>budget</th>\n",
       "      <th>boxoffice</th>\n",
       "      <th>imdb_rating</th>\n",
       "      <th>duration_in_mins</th>\n",
       "      <th>mpaa</th>\n",
       "      <th>star_category</th>\n",
       "      <th>g_Action</th>\n",
       "      <th>g_Adult</th>\n",
       "      <th>g_Adventure</th>\n",
       "      <th>g_Animation</th>\n",
       "      <th>g_Anime</th>\n",
       "      <th>g_Comedy</th>\n",
       "      <th>g_Crime</th>\n",
       "      <th>g_Documentary</th>\n",
       "      <th>g_Drama</th>\n",
       "      <th>g_Fantasy</th>\n",
       "      <th>g_Foreign/International</th>\n",
       "      <th>g_Holiday</th>\n",
       "      <th>g_Horror</th>\n",
       "      <th>g_Independent</th>\n",
       "      <th>g_Kids &amp; Family</th>\n",
       "      <th>g_LGBT</th>\n",
       "      <th>g_Lifestyle</th>\n",
       "      <th>g_Music</th>\n",
       "      <th>g_Musicals</th>\n",
       "      <th>g_Mystery</th>\n",
       "      <th>g_Reality</th>\n",
       "      <th>g_Romance</th>\n",
       "      <th>g_Sci-Fi</th>\n",
       "      <th>g_Sport</th>\n",
       "      <th>g_Thriller</th>\n",
       "      <th>g_War</th>\n",
       "      <th>g_Western</th>\n",
       "      <th>import_id_lionsgate</th>\n",
       "      <th>import_id_mgm</th>\n",
       "      <th>import_id_other</th>\n",
       "      <th>import_id_paramount</th>\n",
       "      <th>awards_BAFTA</th>\n",
       "      <th>awards_Golden Globe</th>\n",
       "      <th>awards_Oscar</th>\n",
       "      <th>awards_no award</th>\n",
       "      <th>awards_other award</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>385504</td>\n",
       "      <td>307127.605608</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2013-01-01</td>\n",
       "      <td>Action,Thriller,Drama</td>\n",
       "      <td>69614</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>42930462.0</td>\n",
       "      <td>6.5</td>\n",
       "      <td>112.301017</td>\n",
       "      <td>3</td>\n",
       "      <td>1.710000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>300175</td>\n",
       "      <td>270338.426375</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2013-01-01</td>\n",
       "      <td>Comedy,Crime,Thriller</td>\n",
       "      <td>46705</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>3301046.0</td>\n",
       "      <td>6.5</td>\n",
       "      <td>94.983250</td>\n",
       "      <td>4</td>\n",
       "      <td>3.250000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>361899</td>\n",
       "      <td>256165.867446</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2012-01-01</td>\n",
       "      <td>Crime,Drama</td>\n",
       "      <td>197596</td>\n",
       "      <td>26000000.0</td>\n",
       "      <td>37397291.0</td>\n",
       "      <td>7.3</td>\n",
       "      <td>115.763675</td>\n",
       "      <td>4</td>\n",
       "      <td>2.646667</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>308314</td>\n",
       "      <td>196622.720996</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>2008-01-01</td>\n",
       "      <td>Thriller,Drama,War,Documentary,Mystery,Action</td>\n",
       "      <td>356339</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>15700000.0</td>\n",
       "      <td>7.6</td>\n",
       "      <td>130.703583</td>\n",
       "      <td>4</td>\n",
       "      <td>1.666667</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>307201</td>\n",
       "      <td>159841.652064</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2013-01-01</td>\n",
       "      <td>Crime,Thriller,Mystery,Documentary</td>\n",
       "      <td>46720</td>\n",
       "      <td>27220000.0</td>\n",
       "      <td>8551228.0</td>\n",
       "      <td>6.4</td>\n",
       "      <td>105.545533</td>\n",
       "      <td>4</td>\n",
       "      <td>3.066667</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   video_id    cvt_per_day  weighted_categorical_position  \\\n",
       "0    385504  307127.605608                              1   \n",
       "1    300175  270338.426375                              1   \n",
       "2    361899  256165.867446                              1   \n",
       "3    308314  196622.720996                              3   \n",
       "4    307201  159841.652064                              1   \n",
       "\n",
       "   weighted_horizontal_poition release_year  \\\n",
       "0                            3   2013-01-01   \n",
       "1                            3   2013-01-01   \n",
       "2                            3   2012-01-01   \n",
       "3                            4   2008-01-01   \n",
       "4                            3   2013-01-01   \n",
       "\n",
       "                                          genres  imdb_votes      budget  \\\n",
       "0                          Action,Thriller,Drama       69614  15000000.0   \n",
       "1                          Comedy,Crime,Thriller       46705  15000000.0   \n",
       "2                                    Crime,Drama      197596  26000000.0   \n",
       "3  Thriller,Drama,War,Documentary,Mystery,Action      356339  15000000.0   \n",
       "4             Crime,Thriller,Mystery,Documentary       46720  27220000.0   \n",
       "\n",
       "    boxoffice  imdb_rating  duration_in_mins  mpaa  star_category  g_Action  \\\n",
       "0  42930462.0          6.5        112.301017     3       1.710000         1   \n",
       "1   3301046.0          6.5         94.983250     4       3.250000         0   \n",
       "2  37397291.0          7.3        115.763675     4       2.646667         0   \n",
       "3  15700000.0          7.6        130.703583     4       1.666667         1   \n",
       "4   8551228.0          6.4        105.545533     4       3.066667         0   \n",
       "\n",
       "   g_Adult  g_Adventure  g_Animation  g_Anime  g_Comedy  g_Crime  \\\n",
       "0        0            0            0        0         0        0   \n",
       "1        0            0            0        0         1        1   \n",
       "2        0            0            0        0         0        1   \n",
       "3        0            0            0        0         0        0   \n",
       "4        0            0            0        0         0        1   \n",
       "\n",
       "   g_Documentary  g_Drama  g_Fantasy  g_Foreign/International  g_Holiday  \\\n",
       "0              0        1          0                        0          0   \n",
       "1              0        0          0                        0          0   \n",
       "2              0        1          0                        0          0   \n",
       "3              1        1          0                        0          0   \n",
       "4              1        0          0                        0          0   \n",
       "\n",
       "   g_Horror  g_Independent  g_Kids & Family  g_LGBT  g_Lifestyle  g_Music  \\\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",
       "   g_Musicals  g_Mystery  g_Reality  g_Romance  g_Sci-Fi  g_Sport  g_Thriller  \\\n",
       "0           0          0          0          0         0        0           1   \n",
       "1           0          0          0          0         0        0           1   \n",
       "2           0          0          0          0         0        0           0   \n",
       "3           0          1          0          0         0        0           1   \n",
       "4           0          1          0          0         0        0           1   \n",
       "\n",
       "   g_War  g_Western  import_id_lionsgate  import_id_mgm  import_id_other  \\\n",
       "0      0          0                    1              0                0   \n",
       "1      0          0                    1              0                0   \n",
       "2      0          0                    0              0                1   \n",
       "3      1          0                    1              0                0   \n",
       "4      0          0                    1              0                0   \n",
       "\n",
       "   import_id_paramount  awards_BAFTA  awards_Golden Globe  awards_Oscar  \\\n",
       "0                    0             0                    0             0   \n",
       "1                    0             0                    0             0   \n",
       "2                    0             0                    0             0   \n",
       "3                    0             0                    0             1   \n",
       "4                    0             0                    0             0   \n",
       "\n",
       "   awards_no award  awards_other award  \n",
       "0                0                   1  \n",
       "1                1                   0  \n",
       "2                0                   1  \n",
       "3                0                   0  \n",
       "4                0                   1  "
      ]
     },
     "execution_count": 143,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "video_updated2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 224
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 6299,
     "status": "ok",
     "timestamp": 1546590669500,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "MrsjiqTwi0NQ",
    "outputId": "d355a36a-2d78-42b1-d88e-ad491626a60c",
    "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>video_id</th>\n",
       "      <th>weighted_categorical_position</th>\n",
       "      <th>weighted_horizontal_poition</th>\n",
       "      <th>release_year</th>\n",
       "      <th>imdb_votes</th>\n",
       "      <th>budget</th>\n",
       "      <th>boxoffice</th>\n",
       "      <th>imdb_rating</th>\n",
       "      <th>duration_in_mins</th>\n",
       "      <th>mpaa</th>\n",
       "      <th>star_category</th>\n",
       "      <th>g_Action</th>\n",
       "      <th>g_Adult</th>\n",
       "      <th>g_Adventure</th>\n",
       "      <th>g_Animation</th>\n",
       "      <th>g_Anime</th>\n",
       "      <th>g_Comedy</th>\n",
       "      <th>g_Crime</th>\n",
       "      <th>g_Documentary</th>\n",
       "      <th>g_Drama</th>\n",
       "      <th>g_Fantasy</th>\n",
       "      <th>g_Foreign/International</th>\n",
       "      <th>g_Holiday</th>\n",
       "      <th>g_Horror</th>\n",
       "      <th>g_Independent</th>\n",
       "      <th>g_Kids &amp; Family</th>\n",
       "      <th>g_LGBT</th>\n",
       "      <th>g_Lifestyle</th>\n",
       "      <th>g_Music</th>\n",
       "      <th>g_Musicals</th>\n",
       "      <th>g_Mystery</th>\n",
       "      <th>g_Reality</th>\n",
       "      <th>g_Romance</th>\n",
       "      <th>g_Sci-Fi</th>\n",
       "      <th>g_Sport</th>\n",
       "      <th>g_Thriller</th>\n",
       "      <th>g_War</th>\n",
       "      <th>g_Western</th>\n",
       "      <th>import_id_lionsgate</th>\n",
       "      <th>import_id_mgm</th>\n",
       "      <th>import_id_other</th>\n",
       "      <th>import_id_paramount</th>\n",
       "      <th>awards_BAFTA</th>\n",
       "      <th>awards_Golden Globe</th>\n",
       "      <th>awards_Oscar</th>\n",
       "      <th>awards_no award</th>\n",
       "      <th>awards_other award</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>385504</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2013-01-01</td>\n",
       "      <td>69614</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>42930462.0</td>\n",
       "      <td>6.5</td>\n",
       "      <td>112.301017</td>\n",
       "      <td>3</td>\n",
       "      <td>1.710000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>300175</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2013-01-01</td>\n",
       "      <td>46705</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>3301046.0</td>\n",
       "      <td>6.5</td>\n",
       "      <td>94.983250</td>\n",
       "      <td>4</td>\n",
       "      <td>3.250000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>361899</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2012-01-01</td>\n",
       "      <td>197596</td>\n",
       "      <td>26000000.0</td>\n",
       "      <td>37397291.0</td>\n",
       "      <td>7.3</td>\n",
       "      <td>115.763675</td>\n",
       "      <td>4</td>\n",
       "      <td>2.646667</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>308314</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>2008-01-01</td>\n",
       "      <td>356339</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>15700000.0</td>\n",
       "      <td>7.6</td>\n",
       "      <td>130.703583</td>\n",
       "      <td>4</td>\n",
       "      <td>1.666667</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>307201</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2013-01-01</td>\n",
       "      <td>46720</td>\n",
       "      <td>27220000.0</td>\n",
       "      <td>8551228.0</td>\n",
       "      <td>6.4</td>\n",
       "      <td>105.545533</td>\n",
       "      <td>4</td>\n",
       "      <td>3.066667</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   video_id  weighted_categorical_position  weighted_horizontal_poition  \\\n",
       "0    385504                              1                            3   \n",
       "1    300175                              1                            3   \n",
       "2    361899                              1                            3   \n",
       "3    308314                              3                            4   \n",
       "4    307201                              1                            3   \n",
       "\n",
       "  release_year  imdb_votes      budget   boxoffice  imdb_rating  \\\n",
       "0   2013-01-01       69614  15000000.0  42930462.0          6.5   \n",
       "1   2013-01-01       46705  15000000.0   3301046.0          6.5   \n",
       "2   2012-01-01      197596  26000000.0  37397291.0          7.3   \n",
       "3   2008-01-01      356339  15000000.0  15700000.0          7.6   \n",
       "4   2013-01-01       46720  27220000.0   8551228.0          6.4   \n",
       "\n",
       "   duration_in_mins  mpaa  star_category  g_Action  g_Adult  g_Adventure  \\\n",
       "0        112.301017     3       1.710000         1        0            0   \n",
       "1         94.983250     4       3.250000         0        0            0   \n",
       "2        115.763675     4       2.646667         0        0            0   \n",
       "3        130.703583     4       1.666667         1        0            0   \n",
       "4        105.545533     4       3.066667         0        0            0   \n",
       "\n",
       "   g_Animation  g_Anime  g_Comedy  g_Crime  g_Documentary  g_Drama  g_Fantasy  \\\n",
       "0            0        0         0        0              0        1          0   \n",
       "1            0        0         1        1              0        0          0   \n",
       "2            0        0         0        1              0        1          0   \n",
       "3            0        0         0        0              1        1          0   \n",
       "4            0        0         0        1              1        0          0   \n",
       "\n",
       "   g_Foreign/International  g_Holiday  g_Horror  g_Independent  \\\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",
       "   g_Kids & Family  g_LGBT  g_Lifestyle  g_Music  g_Musicals  g_Mystery  \\\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          1   \n",
       "4                0       0            0        0           0          1   \n",
       "\n",
       "   g_Reality  g_Romance  g_Sci-Fi  g_Sport  g_Thriller  g_War  g_Western  \\\n",
       "0          0          0         0        0           1      0          0   \n",
       "1          0          0         0        0           1      0          0   \n",
       "2          0          0         0        0           0      0          0   \n",
       "3          0          0         0        0           1      1          0   \n",
       "4          0          0         0        0           1      0          0   \n",
       "\n",
       "   import_id_lionsgate  import_id_mgm  import_id_other  import_id_paramount  \\\n",
       "0                    1              0                0                    0   \n",
       "1                    1              0                0                    0   \n",
       "2                    0              0                1                    0   \n",
       "3                    1              0                0                    0   \n",
       "4                    1              0                0                    0   \n",
       "\n",
       "   awards_BAFTA  awards_Golden Globe  awards_Oscar  awards_no award  \\\n",
       "0             0                    0             0                0   \n",
       "1             0                    0             0                1   \n",
       "2             0                    0             0                0   \n",
       "3             0                    0             1                0   \n",
       "4             0                    0             0                0   \n",
       "\n",
       "   awards_other award  \n",
       "0                   1  \n",
       "1                   0  \n",
       "2                   1  \n",
       "3                   0  \n",
       "4                   1  "
      ]
     },
     "execution_count": 144,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "to_drop= ['cvt_per_day','genres']\n",
    "video_updated2= video_updated2.drop(to_drop, axis=1)\n",
    "video_updated2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = video_updated2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {},
   "outputs": [],
   "source": [
    "X['release_year'] = video_df['release_year']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = video_df['cvt_per_day']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 391
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 6292,
     "status": "ok",
     "timestamp": 1546590669501,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "JuPhtUkJi0NW",
    "outputId": "23671147-880b-4337-8adc-64740609ceda"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Feature space holds 4226 observations and 47 features\n",
      "Unique target labels: [2.18762475e+00 2.86661397e+00 2.93569112e+00 ... 2.56165867e+05\n",
      " 2.70338426e+05 3.07127606e+05]\n",
      "         0         1         2         3          4         5         6   \\\n",
      "0  0.933462 -1.105826 -2.116264  0.682673   1.998941  1.723662  4.864658   \n",
      "1  0.175836 -1.105826 -2.116264  0.682673   1.273795  1.723662 -0.816956   \n",
      "2  0.723876 -1.105826 -2.116264  0.625513   6.049995  3.306559  4.071375   \n",
      "3  0.248101 -0.779746 -2.031963  0.396873  11.074737  1.723662  0.960665   \n",
      "4  0.238219 -1.105826 -2.116264  0.682673   1.274270  3.482116 -0.064244   \n",
      "\n",
      "         7         8         9         10        11        12        13  \\\n",
      "0  0.585567  1.078791  0.790878  0.060300  2.145923 -0.026653 -0.306543   \n",
      "1  0.585567  0.257409  1.356690  3.459786 -0.458520 -0.026653 -0.306543   \n",
      "2  0.962471  1.243025  1.356690  2.127953 -0.458520 -0.026653 -0.306543   \n",
      "3  1.103810  1.951625  1.356690 -0.035357  2.145923 -0.026653 -0.306543   \n",
      "4  0.538454  0.758378  1.356690  3.055086 -0.458520 -0.026653 -0.306543   \n",
      "\n",
      "         14        15        16        17        18        19        20  \\\n",
      "0 -0.177444 -0.051085 -0.623873 -0.339608 -0.434452  1.232873 -0.247001   \n",
      "1 -0.177444 -0.051085  1.602890  2.944568 -0.434452 -0.811113 -0.247001   \n",
      "2 -0.177444 -0.051085 -0.623873  2.944568 -0.434452  1.232873 -0.247001   \n",
      "3 -0.177444 -0.051085 -0.623873 -0.339608  2.301752  1.232873 -0.247001   \n",
      "4 -0.177444 -0.051085 -0.623873  2.944568  2.301752 -0.811113 -0.247001   \n",
      "\n",
      "         21        22        23        24        25       26        27  \\\n",
      "0 -0.124005 -0.015385 -0.469017 -0.320204 -0.266379 -0.02176 -0.040733   \n",
      "1 -0.124005 -0.015385 -0.469017 -0.320204 -0.266379 -0.02176 -0.040733   \n",
      "2 -0.124005 -0.015385 -0.469017 -0.320204 -0.266379 -0.02176 -0.040733   \n",
      "3 -0.124005 -0.015385 -0.469017 -0.320204 -0.266379 -0.02176 -0.040733   \n",
      "4 -0.124005 -0.015385 -0.469017 -0.320204 -0.266379 -0.02176 -0.040733   \n",
      "\n",
      "         28        29        30        31       32        33       34  \\\n",
      "0 -0.205354 -0.127883 -0.312053 -0.046198 -0.40322 -0.306543 -0.13623   \n",
      "1 -0.205354 -0.127883 -0.312053 -0.046198 -0.40322 -0.306543 -0.13623   \n",
      "2 -0.205354 -0.127883 -0.312053 -0.046198 -0.40322 -0.306543 -0.13623   \n",
      "3 -0.205354 -0.127883  3.204580 -0.046198 -0.40322 -0.306543 -0.13623   \n",
      "4 -0.205354 -0.127883  3.204580 -0.046198 -0.40322 -0.306543 -0.13623   \n",
      "\n",
      "         35        36        37        38        39        40        41  \\\n",
      "0  1.951342 -0.157268 -0.157268  2.289595 -0.343065 -1.531666 -0.185786   \n",
      "1  1.951342 -0.157268 -0.157268  2.289595 -0.343065 -1.531666 -0.185786   \n",
      "2 -0.512468 -0.157268 -0.157268 -0.436758 -0.343065  0.652884 -0.185786   \n",
      "3  1.951342  6.358567 -0.157268  2.289595 -0.343065 -1.531666 -0.185786   \n",
      "4  1.951342 -0.157268 -0.157268  2.289595 -0.343065 -1.531666 -0.185786   \n",
      "\n",
      "         42        43        44        45        46  \n",
      "0 -0.067203 -0.102573 -0.185786 -1.245147  1.381472  \n",
      "1 -0.067203 -0.102573 -0.185786  0.803118 -0.723865  \n",
      "2 -0.067203 -0.102573 -0.185786 -1.245147  1.381472  \n",
      "3 -0.067203 -0.102573  5.382530 -1.245147 -0.723865  \n",
      "4 -0.067203 -0.102573 -0.185786 -1.245147  1.381472  \n"
     ]
    }
   ],
   "source": [
    "# Scale the data, using standardization\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "scaler = StandardScaler()\n",
    "X = scaler.fit_transform(X)\n",
    "\n",
    "print (\"Feature space holds %d observations and %d features\" % X.shape)\n",
    "print (\"Unique target labels:\", np.unique(y))\n",
    "print (pd.DataFrame(X).head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 238
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 6290,
     "status": "ok",
     "timestamp": 1546590669501,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "Psvc7vM8i0Nc",
    "outputId": "38ff8502-ba6c-4765-e496-19096e629690"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.933462  , -1.10582603, -2.11626404, ..., -0.18578623,\n",
       "        -1.24514749,  1.38147213],\n",
       "       [ 0.17583592, -1.10582603, -2.11626404, ..., -0.18578623,\n",
       "         0.80311771, -0.72386549],\n",
       "       [ 0.723876  , -1.10582603, -2.11626404, ..., -0.18578623,\n",
       "        -1.24514749,  1.38147213],\n",
       "       ...,\n",
       "       [-2.38625409, -0.77974566, -0.09303838, ..., -0.18578623,\n",
       "        -1.24514749,  1.38147213],\n",
       "       [-2.25529059, -0.61670548, -1.10465121, ..., -0.18578623,\n",
       "         0.80311771, -0.72386549],\n",
       "       [-2.38968134, -0.12758492,  0.58137018, ..., -0.18578623,\n",
       "         0.80311771, -0.72386549]])"
      ]
     },
     "execution_count": 153,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "q3x9ySX_i0Nd"
   },
   "source": [
    "# Part 3: Model Training and Result Evaluation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "77OjmSl9i0Nf"
   },
   "source": [
    "### Part 3.1: Split dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 51
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 6276,
     "status": "ok",
     "timestamp": 1546590669501,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "Uay8Md5li0Nh",
    "outputId": "4b0495cb-e8a5-4568-ba50-a56eec228293"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "training data has 3380 observation with 47 features\n",
      "test data has 846 observation with 47 features\n"
     ]
    }
   ],
   "source": [
    "# Split data into training and testing\n",
    "from sklearn import model_selection\n",
    "\n",
    "# Reserve 20% for testing\n",
    "X_train, X_test, y_train, y_test = model_selection.train_test_split(X, y, test_size=0.2)\n",
    "\n",
    "print('training data has %d observation with %d features'% X_train.shape)\n",
    "print('test data has %d observation with %d features'% X_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "c4UTtCQTi0Nl"
   },
   "source": [
    "### Part 3.2: Model Training and Selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {
    "cellView": "code",
    "colab": {},
    "colab_type": "code",
    "id": "EAhSxINLi0Nl"
   },
   "outputs": [],
   "source": [
    "#@title build models\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "# from sklearn.neighbors import KNeighborsClassifier \n",
    "# from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "# Linear Regression\n",
    "lm = LinearRegression() \n",
    "\n",
    "# # K Nearest Neighbors\n",
    "# classifier_KNN = KNeighborsClassifier()\n",
    "\n",
    "# # Random Forest\n",
    "# classifier_RF = RandomForestClassifier()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 85
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 6261,
     "status": "ok",
     "timestamp": 1546590669504,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "Av0IRSoBQ3pe",
    "outputId": "7ae3da57-3703-4e57-9eb0-4df7315b54cf"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LinearRegression(copy_X=True, fit_intercept=True, n_jobs=None,\n",
       "         normalize=False)"
      ]
     },
     "execution_count": 182,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Train the model\n",
    "lm.fit(X_train, y_train)\n",
    "# classifier_logistic.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 4.21196471e+02, -1.28950069e+03, -3.02086328e+03,  7.57617508e+02,\n",
       "        2.03668726e+03,  1.96450458e+03,  3.89829768e+02, -4.90119032e+02,\n",
       "        3.26610777e+02,  4.63561212e+02,  1.11259243e+03,  5.12401661e+02,\n",
       "       -4.39087115e+01, -1.56371643e+02, -1.28095681e+02, -1.05144982e+02,\n",
       "        5.68497745e+02,  4.70296249e+02, -2.22100765e+02, -1.83680813e+02,\n",
       "        2.57562204e+02, -2.83224768e+02,  1.00929155e+07,  2.90660503e+02,\n",
       "        2.84252618e+01,  4.30443290e+02,  6.76854237e+01, -1.10066734e+02,\n",
       "        1.64871819e+02,  2.33621776e+02, -1.35609266e+02, -1.44615224e+02,\n",
       "       -2.00176198e+02, -6.60378188e+01,  4.39337004e+01,  9.92587856e+02,\n",
       "        4.15314492e+02,  3.71834193e+02,  1.12769940e+16,  9.43690011e+15,\n",
       "        1.40738693e+16,  5.52143038e+15,  9.71609948e+14,  1.47416891e+15,\n",
       "        2.60816152e+15,  7.09042394e+15,  6.89821360e+15])"
      ]
     },
     "execution_count": 183,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "linear_model.coef_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 1.24066971e+04,  5.91869715e+03, -1.63730285e+03, -5.30802851e+02,\n",
       "        1.33806971e+04,  3.04269715e+03,  1.24119715e+03,  5.16669715e+03,\n",
       "        9.25619715e+03,  5.69697149e+02,  6.77869715e+03,  7.06669715e+03,\n",
       "        3.68869715e+03,  3.91969715e+03,  2.21669715e+03,  2.93719715e+03,\n",
       "       -3.81830285e+03,  1.17169715e+03,  8.46869715e+03,  9.17697149e+02,\n",
       "        1.31346971e+04, -1.06568029e+04,  4.96219715e+03,  2.44469715e+03,\n",
       "        6.12519715e+03,  1.22301971e+04,  7.70569715e+03,  1.52611971e+04,\n",
       "        1.13626971e+04,  1.80419715e+03, -3.15380285e+03, -4.90330285e+03,\n",
       "        6.45169715e+03,  9.98697149e+02, -9.32802851e+02, -4.56302851e+02,\n",
       "        5.62069715e+03, -3.73180285e+03,  2.39969715e+03,  1.06196971e+04,\n",
       "        2.05881971e+04,  3.28019715e+03, -8.78802851e+02, -2.41030285e+03,\n",
       "        1.97669715e+03, -4.95280285e+03, -1.84330285e+03, -3.65630285e+03,\n",
       "        3.85069715e+03, -2.62680285e+03, -1.92430285e+03,  5.62269715e+03,\n",
       "        9.63069715e+03, -9.34802851e+02,  1.42796971e+04,  9.63769715e+03,\n",
       "        1.69769715e+03, -8.80302851e+02, -9.84802851e+02,  1.65069715e+03,\n",
       "        9.41769715e+03,  7.12269715e+03,  3.01869715e+03,  2.54781971e+04,\n",
       "        4.29697149e+02, -5.67480285e+03,  1.06216971e+04,  1.45586971e+04,\n",
       "       -1.30880285e+03,  3.59619715e+03,  1.35269715e+03,  5.03819715e+03,\n",
       "        2.02469715e+03,  8.65769715e+03,  2.91119715e+03, -1.70180285e+03,\n",
       "        1.59986971e+04,  1.20186971e+04,  1.60769715e+03,  2.34669715e+03,\n",
       "       -1.30285108e+00, -2.08580285e+03,  7.55769715e+03,  1.14119715e+03,\n",
       "       -4.79330285e+03,  1.69486971e+04, -3.68302851e+02, -1.55330285e+03,\n",
       "        6.58519715e+03,  2.78719715e+03,  4.56569715e+03,  7.61419715e+03,\n",
       "        3.89569715e+03,  9.29269715e+03,  4.45869715e+03,  8.83769715e+03,\n",
       "       -6.72802851e+02,  1.01476971e+04, -3.67730285e+03,  1.02319715e+03,\n",
       "        2.38891971e+04,  4.64869715e+03, -8.34130285e+03,  4.21469715e+03,\n",
       "        5.43619715e+03,  1.63019715e+03, -2.07230285e+03,  1.16951971e+04,\n",
       "       -3.33630285e+03, -2.33630285e+03,  1.42876971e+04,  1.36576971e+04,\n",
       "        1.33819715e+03,  2.08669715e+03,  1.44666971e+04, -9.49630285e+03,\n",
       "        8.88669715e+03,  7.91469715e+03, -1.78180285e+03,  3.30169715e+03,\n",
       "        9.13519715e+03, -4.76530285e+03, -3.96330285e+03,  8.88669715e+03,\n",
       "        1.81971489e+01,  1.76269715e+03,  7.87069715e+03, -1.66130285e+03,\n",
       "        4.37469715e+03,  3.98169715e+03,  2.69469715e+03, -4.13028511e+01,\n",
       "        1.82019715e+03,  1.26306971e+04,  1.54219715e+03,  3.20669715e+03,\n",
       "        3.29296971e+04, -1.31130285e+03, -1.23630285e+03,  6.22919715e+03,\n",
       "        4.22069715e+03,  5.20369715e+03,  1.01219715e+03,  2.60469715e+03,\n",
       "        6.43669715e+03,  1.13781971e+04,  6.03019715e+03,  6.55469715e+03,\n",
       "        3.20346971e+04, -6.43330285e+03, -2.63530285e+03,  8.61019715e+03,\n",
       "        1.26969715e+03, -3.98028511e+01,  6.66697149e+02,  9.30569715e+03,\n",
       "       -2.13802851e+02,  6.04869715e+03,  1.08271971e+04,  5.04019715e+03,\n",
       "        1.16719715e+03, -5.32802851e+02,  3.57069715e+03,  3.57069715e+03,\n",
       "        1.18569715e+03,  8.97069715e+03, -7.27030285e+03,  2.17869715e+03,\n",
       "       -1.14180285e+03,  2.19219715e+03, -3.61630285e+03,  2.67669715e+03,\n",
       "        7.72669715e+03, -1.78780285e+03,  2.84069715e+03,  1.07819715e+03,\n",
       "        9.94619715e+03,  8.58697149e+02,  1.61069715e+03,  3.66519715e+03,\n",
       "       -3.90530285e+03,  3.35769715e+03, -2.89930285e+03,  1.24769715e+03,\n",
       "        1.15926971e+04, -3.49302851e+02,  6.16269715e+03,  3.65519715e+03,\n",
       "        3.51197149e+02,  3.57919715e+03,  8.94419715e+03,  7.95197149e+02,\n",
       "        1.43169715e+03,  6.51469715e+03,  1.39096971e+04, -3.20030285e+03,\n",
       "        1.55219715e+03, -1.44930285e+03, -4.80530285e+03, -1.74580285e+03,\n",
       "        7.95069715e+03,  4.38069715e+03,  6.61869715e+03,  2.64269715e+03,\n",
       "        2.27219715e+03,  1.25161971e+04,  3.24419715e+03,  4.32197149e+02,\n",
       "        4.30946971e+04, -3.45630285e+03,  1.06236971e+04,  5.17669715e+03,\n",
       "        1.17869715e+03,  5.85869715e+03,  6.62669715e+03,  3.75269715e+03,\n",
       "        6.35519715e+03,  1.00691971e+04,  7.55169715e+03,  4.64519715e+03,\n",
       "       -2.63480285e+03,  8.58269715e+03,  7.91819715e+03,  1.27419715e+03,\n",
       "        3.87169715e+03,  3.42819715e+03,  6.35569715e+03,  1.74219715e+03,\n",
       "        8.21569715e+03,  9.69714892e+00,  6.49619715e+03, -4.87302851e+02,\n",
       "        2.33691971e+04,  1.32156971e+04,  4.46319715e+03,  1.20269715e+03,\n",
       "       -4.82330285e+03,  1.60019715e+03,  2.01419715e+03, -1.52980285e+03,\n",
       "        1.01826971e+04, -2.26030285e+03,  1.41697149e+02,  5.12569715e+03,\n",
       "        3.20269715e+03, -6.13230285e+03, -1.78680285e+03,  1.89697149e+02,\n",
       "        7.20869715e+03,  1.31197149e+02,  2.95819715e+03,  2.64046971e+04,\n",
       "        5.48669715e+03,  3.70469715e+03, -7.32302851e+02, -3.56302851e+02,\n",
       "        2.64386971e+04,  1.11519715e+03,  6.58369715e+03,  1.94119715e+03,\n",
       "        3.00697149e+02,  6.19719715e+03, -7.53302851e+02,  3.61669715e+03,\n",
       "        9.41469715e+03,  3.21119715e+03,  3.16569715e+03,  4.27469715e+03,\n",
       "        4.17469715e+03,  1.07746971e+04,  8.92819715e+03, -4.23930285e+03,\n",
       "        7.69419715e+03,  9.64519715e+03,  1.03096971e+04, -8.62302851e+02,\n",
       "        2.10019715e+03,  3.87619715e+03,  4.51069715e+03,  3.35469715e+03,\n",
       "        8.24969715e+03,  1.51431971e+04, -4.40480285e+03,  3.47697149e+02,\n",
       "       -9.04302851e+02,  3.21869715e+03,  8.86969715e+03, -2.00980285e+03,\n",
       "       -5.68680285e+03, -2.13530285e+03,  7.39819715e+03, -3.94630285e+03,\n",
       "       -4.11730285e+03,  1.09619715e+03, -1.00530285e+03,  5.26669715e+03,\n",
       "        2.35697149e+02,  1.40069715e+03,  3.67819715e+03,  6.91069715e+03,\n",
       "        6.96269715e+03,  4.07669715e+03,  7.26669715e+03, -2.41302851e+02,\n",
       "       -6.27430285e+03, -1.01280285e+03, -4.61080285e+03,  1.94469715e+03,\n",
       "        1.12069715e+03, -4.48080285e+03, -7.23028511e+01,  6.52269715e+03,\n",
       "        6.40519715e+03, -3.13730285e+03,  5.89019715e+03,  5.50197149e+02,\n",
       "        1.33391971e+04,  3.25569715e+03, -4.32130285e+03,  6.88697149e+02,\n",
       "       -3.80430285e+03,  8.32469715e+03, -3.62802851e+02,  6.12669715e+03,\n",
       "        5.08169715e+03,  4.76569715e+03, -1.75830285e+03,  8.65069715e+03,\n",
       "        1.26756971e+04,  4.57819715e+03,  8.39919715e+03,  8.54969715e+03,\n",
       "        5.64569715e+03,  6.32969715e+03,  5.56569715e+03,  7.96519715e+03,\n",
       "        5.48669715e+03,  4.83469715e+03,  2.19086971e+04,  1.31769715e+03,\n",
       "        1.78697149e+02,  3.85969715e+03,  1.50569715e+03, -1.21880285e+03,\n",
       "        6.43769715e+03,  1.81119715e+03, -1.21730285e+03,  6.46969715e+03,\n",
       "        1.08086971e+04,  1.33719715e+03, -1.85730285e+03,  6.94697149e+02,\n",
       "       -1.41280285e+03,  5.72769715e+03,  6.56196228e+08, -6.95030285e+03,\n",
       "        8.26697149e+02,  1.17036971e+04,  1.80069715e+03,  1.40176971e+04,\n",
       "        6.46269715e+03,  3.55469715e+03,  8.03197149e+02, -4.00630285e+03,\n",
       "        9.50269715e+03, -2.60780285e+03, -1.47302851e+02, -1.11330285e+03,\n",
       "        4.12069715e+03,  7.09519715e+03,  2.35419715e+03,  3.30697149e+02,\n",
       "        1.30671971e+04,  1.50169715e+03,  3.91369715e+03,  6.14969715e+03,\n",
       "       -1.58802851e+02,  7.12619715e+03,  5.57469715e+03,  6.66069715e+03,\n",
       "        1.04069715e+03, -1.16080285e+03, -4.43302851e+02, -3.53430285e+03,\n",
       "        3.27069715e+03,  3.93169715e+03,  8.36971489e+01,  2.10531971e+04,\n",
       "        4.45919715e+03,  1.26901971e+04,  1.61626971e+04,  6.20669715e+03,\n",
       "        6.09197149e+02, -1.83930285e+03,  1.64096971e+04,  7.75069715e+03,\n",
       "        3.72519715e+03,  3.34519715e+03,  9.60569715e+03,  8.85869715e+03,\n",
       "        9.15369715e+03,  5.17869715e+03,  3.39419715e+03, -7.65302851e+02,\n",
       "        2.14569715e+03, -2.19680285e+03,  1.45419715e+03,  7.32169715e+03,\n",
       "        1.66819715e+03,  8.92669715e+03, -2.56930285e+03, -1.65930285e+03,\n",
       "       -4.27930285e+03,  9.69769715e+03,  9.06669715e+03,  2.64519715e+03,\n",
       "        1.78856971e+04,  1.21871971e+04, -3.07930285e+03,  1.00366971e+04,\n",
       "        9.88319715e+03, -1.67980285e+03,  1.17206971e+04,  1.03766971e+04,\n",
       "        6.14419715e+03, -9.48028511e+01,  9.12119715e+03,  1.10269715e+03,\n",
       "       -5.94180285e+03,  4.97369715e+03,  4.04769715e+03,  9.39697149e+02,\n",
       "        1.08906971e+04,  2.90269715e+03,  2.86769715e+03,  1.90419715e+03,\n",
       "        1.31726971e+04,  8.30269715e+03, -2.19730285e+03,  1.53236971e+04,\n",
       "       -4.73802851e+02,  3.23319715e+03,  3.85069715e+03,  2.56919715e+03,\n",
       "       -2.46130285e+03, -3.05730285e+03,  4.63669715e+03,  1.04756971e+04,\n",
       "       -6.82302851e+02, -3.09780285e+03,  1.44719715e+03,  4.90619715e+03,\n",
       "        8.93669715e+03, -3.30630285e+03,  1.20486971e+04,  1.04846971e+04,\n",
       "       -3.66930285e+03, -2.79480285e+03,  1.09369715e+03,  3.20869715e+03,\n",
       "        2.66697149e+02,  5.35569715e+03,  2.94119715e+03,  2.25419715e+03,\n",
       "        2.80419715e+03, -1.03530285e+03,  2.92119715e+03, -7.53302851e+02,\n",
       "        4.55219715e+03,  1.17016971e+04,  3.75119715e+03,  2.12106971e+04,\n",
       "        6.48269715e+03, -6.59302851e+02,  3.71519715e+03,  3.46969715e+03,\n",
       "        1.63669715e+03, -3.35580285e+03,  3.17919715e+03,  1.38869715e+03,\n",
       "        2.06697149e+02, -3.30530285e+03,  6.92669715e+03, -1.74230285e+03,\n",
       "        1.56869715e+03, -4.83028511e+01,  1.37869715e+03,  5.00069715e+03,\n",
       "        1.16886971e+04,  1.10769715e+03,  1.26369715e+03,  5.74697149e+02,\n",
       "       -5.00802851e+02,  5.01769715e+03,  1.58669715e+03, -5.27430285e+03,\n",
       "        4.83469715e+03,  3.96669715e+03,  1.32069715e+03, -5.47302851e+02,\n",
       "        4.78569715e+03, -1.70730285e+03,  1.31626971e+04, -1.20680285e+03,\n",
       "       -1.99030285e+03,  4.48869715e+03,  8.91869715e+03,  9.49697149e+02,\n",
       "        8.33769715e+03,  3.55719715e+03,  6.04869715e+03,  1.42421971e+04,\n",
       "       -2.21480285e+03,  2.75269715e+03,  5.23969715e+03,  4.26219715e+03,\n",
       "       -2.89302851e+02,  2.85819715e+03, -1.89880285e+03,  1.09369715e+03,\n",
       "        2.68369715e+03, -2.04930285e+03,  6.34619715e+03,  4.62069715e+03,\n",
       "        4.54469715e+03,  7.14069715e+03,  2.11697149e+02,  3.50619715e+03,\n",
       "        8.12369715e+03,  6.38669715e+03,  1.07001971e+04,  1.22776971e+04,\n",
       "        1.73269715e+03,  4.39069715e+03,  3.78697149e+02,  1.15096971e+04,\n",
       "        3.21969715e+03,  1.30551971e+04,  6.40169715e+03,  1.09026971e+04,\n",
       "       -1.03680285e+03,  4.62269715e+03,  2.40786971e+04,  1.49969715e+03,\n",
       "        3.05769715e+03,  1.00386971e+04,  4.69269715e+03,  6.00269715e+03,\n",
       "        3.84119715e+03,  2.68697149e+02,  1.73471971e+04,  5.64569715e+03,\n",
       "       -3.69730285e+03,  1.01569715e+03,  7.33697149e+02,  3.76219715e+03,\n",
       "        1.23197149e+02, -1.18630285e+03,  2.84269715e+03,  2.31769715e+03,\n",
       "       -1.26980285e+03, -1.35430285e+03,  1.50546971e+04,  5.99869715e+03,\n",
       "        2.02697149e+02,  1.25069715e+03,  2.86419715e+03,  1.79319715e+03,\n",
       "       -7.18028511e+01,  8.93869715e+03,  1.87319715e+03,  2.30569715e+03,\n",
       "        8.98369715e+03,  4.54819715e+03,  1.46971971e+04,  1.68119715e+03,\n",
       "       -1.13180285e+03,  3.66369715e+03, -3.43302851e+02,  1.08719715e+03,\n",
       "        1.45281971e+04,  6.27619715e+03,  4.26369715e+03,  1.00761971e+04,\n",
       "        1.39341971e+04, -1.85730285e+03,  4.43669715e+03,  2.53769715e+03,\n",
       "        5.63697149e+02,  1.82269715e+03, -2.73830285e+03,  1.00666971e+04,\n",
       "        5.19319715e+03,  6.21069715e+03,  1.33151971e+04,  4.61419715e+03,\n",
       "       -6.28080285e+03,  2.04919715e+03,  4.54319715e+03, -6.06230285e+03,\n",
       "        6.03697149e+02, -8.35302851e+02,  1.09746971e+04,  4.59819715e+03,\n",
       "        5.40719715e+03,  4.16697149e+02,  4.08819715e+03,  5.65197149e+02,\n",
       "        3.31119715e+03,  2.72869715e+03,  2.82719715e+03,  8.54697149e+02,\n",
       "        1.05769715e+03, -3.17302851e+02,  5.85869715e+03,  8.60269715e+03,\n",
       "        7.25119715e+03,  7.51369715e+03, -3.41230285e+03, -1.05343029e+04,\n",
       "       -6.76302851e+02, -1.76302851e+02,  2.56319715e+03, -6.45302851e+02,\n",
       "       -8.94802851e+02,  8.61669715e+03,  7.36319715e+03,  1.75076971e+04,\n",
       "        5.77369715e+03,  2.40569715e+03, -1.86802851e+02, -3.27930285e+03,\n",
       "        7.64819715e+03,  1.54116971e+04, -2.42030285e+03,  2.30169715e+03,\n",
       "        5.07869715e+03,  5.79319715e+03, -3.13530285e+03,  1.74819715e+03,\n",
       "       -1.10930285e+03,  1.78669715e+03,  7.66697149e+02, -4.53530285e+03,\n",
       "       -8.82480285e+03,  6.99469715e+03,  1.03169715e+03, -3.88028511e+01,\n",
       "        3.83469715e+03,  3.38869715e+03,  5.70697149e+02,  1.05436971e+04,\n",
       "        1.01146971e+04,  3.16469715e+03,  7.82619715e+03,  2.51971489e+01,\n",
       "        4.19069715e+03,  1.18346971e+04,  1.09266971e+04,  3.87069715e+03,\n",
       "        2.13869715e+03,  1.57366971e+04,  4.65319715e+03,  2.13319715e+03,\n",
       "        2.36269715e+03,  3.66919715e+03,  6.06069715e+03,  6.22419715e+03,\n",
       "        4.02269715e+03,  1.80519715e+03,  9.11069715e+03,  3.45469715e+03,\n",
       "        1.96069715e+03,  2.14169715e+03,  7.57969715e+03,  8.06697149e+02,\n",
       "        2.34819715e+03,  5.05069715e+03,  7.00669715e+03, -4.06802851e+02,\n",
       "        1.34316971e+04,  7.89169715e+03,  5.25469715e+03,  4.58197149e+02,\n",
       "       -3.21380285e+03,  1.12536971e+04, -1.66130285e+03,  2.60569715e+03,\n",
       "        3.39269715e+03,  3.94669715e+03,  1.62919715e+03,  5.16319715e+03,\n",
       "        1.87321971e+04, -4.80180285e+03, -3.90330285e+03,  2.35266971e+04,\n",
       "       -1.26030285e+03,  5.18197149e+02,  3.24669715e+03,  9.87569715e+03,\n",
       "        2.01819715e+03,  1.19896971e+04, -3.29230285e+03,  6.08169715e+03,\n",
       "        4.28169715e+03, -2.46630285e+03,  2.28969715e+03,  6.42519715e+03,\n",
       "       -1.52130285e+03,  1.32469715e+03,  3.45419715e+03, -3.78930285e+03,\n",
       "        4.84919715e+03, -3.73230285e+03,  1.96419715e+03, -2.68802851e+02,\n",
       "       -8.61302851e+02,  2.08269715e+03,  1.08136971e+04,  1.22619715e+03,\n",
       "        5.59469715e+03, -1.13480285e+03,  2.76819715e+03,  2.76996971e+04,\n",
       "        7.53869715e+03, -4.39430285e+03,  4.96969715e+03,  5.98869715e+03,\n",
       "        8.53869715e+03,  6.81669715e+03,  1.61876971e+04,  2.24161971e+04,\n",
       "        7.98969715e+03,  1.25469715e+03,  8.87669715e+03,  1.73986971e+04,\n",
       "       -2.33302851e+02,  6.48769715e+03,  3.67069715e+03,  1.39811971e+04,\n",
       "        9.81369715e+03,  7.60197149e+02,  2.00269715e+03,  2.51669715e+03,\n",
       "        9.22019715e+03,  7.24869715e+03,  7.14769715e+03,  3.22469715e+03,\n",
       "        7.32119715e+03, -1.62630285e+03,  2.52269715e+03,  1.02076971e+04,\n",
       "        8.55469715e+03,  8.18697149e+02,  5.28697149e+02,  1.57746971e+04,\n",
       "        1.82269715e+03, -1.12880285e+03,  2.34619715e+03,  8.91697149e+02,\n",
       "        1.68506971e+04,  4.29569715e+03, -3.19802851e+02,  2.36869715e+03,\n",
       "        1.70169715e+03,  1.45669715e+03, -6.56180285e+03, -1.30230285e+03,\n",
       "       -6.76302851e+02, -3.93028511e+01,  4.34219715e+03,  7.42697149e+02,\n",
       "        8.83697149e+02, -2.79930285e+03, -4.25030285e+03,  3.41819715e+03,\n",
       "        1.06426971e+04,  2.51069715e+03, -2.70430285e+03,  3.54319715e+03,\n",
       "        9.07569715e+03,  2.71919715e+03,  7.22369715e+03,  8.02269715e+03,\n",
       "        1.21456971e+04,  2.39269715e+03,  5.43119715e+03,  8.42269715e+03,\n",
       "        1.40719715e+03,  1.62336971e+04, -2.99380285e+03,  1.06536971e+04,\n",
       "        3.09269715e+03,  1.47819715e+03,  8.78419715e+03,  2.65219715e+03,\n",
       "        1.62486971e+04,  4.02697149e+02,  2.06451971e+04,  4.76769715e+03,\n",
       "        1.46881971e+04,  5.93869715e+03,  3.18369715e+03,  7.20669715e+03,\n",
       "       -2.02530285e+03,  1.66696971e+04,  1.01241971e+04,  6.69719715e+03,\n",
       "        1.03166971e+04,  1.01341971e+04,  7.73569715e+03,  6.33869715e+03,\n",
       "        2.28869715e+03,  5.52169715e+03, -8.34802851e+02,  1.35869715e+03,\n",
       "       -2.21302851e+02,  2.84569715e+03,  1.37919715e+03, -2.62780285e+03,\n",
       "        3.05769715e+03,  9.47769715e+03, -7.08802851e+02,  6.26169715e+03,\n",
       "       -3.44680285e+03,  1.50631971e+04,  1.13066971e+04,  7.46719715e+03,\n",
       "       -2.03130285e+03,  1.19186971e+04,  7.78319715e+03,  3.13569715e+03,\n",
       "       -2.21302851e+02,  4.22119715e+03, -6.41380285e+03,  8.63169715e+03,\n",
       "        1.34351971e+04,  2.64669715e+03,  4.63197149e+02,  5.52619715e+03,\n",
       "        5.53969715e+03,  7.50669715e+03,  4.61019715e+03,  9.29919715e+03,\n",
       "        2.27919715e+03,  6.06269715e+03])"
      ]
     },
     "execution_count": 184,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# prediction of test data\n",
    "lm.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "508969334482409.25"
      ]
     },
     "execution_count": 185,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import mean_squared_error\n",
    "mse = mean_squared_error(y_test,lm.predict(X_test))\n",
    "mse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-6124417.041929917"
      ]
     },
     "execution_count": 186,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Accuracy of test data\n",
    "lm.score(X_test,y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 68
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 7008,
     "status": "ok",
     "timestamp": 1546590670272,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "1OCgNSNri0Nn",
    "outputId": "725804d5-f202-4cfd-edbf-7e4a42fe7a20"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model accuracy of Logistic Regression is: 0.866\n",
      "Model accuracy of KNN is: 0.900\n",
      "Model accuracy of Random Forest is: 0.950\n"
     ]
    }
   ],
   "source": [
    "# Use 5-fold Cross Validation to get the accuracy for different models\n",
    "model_names = ['Logistic Regression','KNN','Random Forest']\n",
    "model_list = [classifier_logistic, classifier_KNN, classifier_RF]\n",
    "count = 0\n",
    "\n",
    "for classifier in model_list:\n",
    "    cv_score = model_selection.cross_val_score(classifier, X_train, y_train, cv=5)\n",
    "    # cprint(cv_score)\n",
    "    print('Model accuracy of %s is: %.3f'%(model_names[count],cv_score.mean()))\n",
    "    count += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 68
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 7008,
     "status": "ok",
     "timestamp": 1546590670272,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "1OCgNSNri0Nn",
    "outputId": "725804d5-f202-4cfd-edbf-7e4a42fe7a20"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model accuracy of Logistic Regression is: 0.866\n",
      "Model accuracy of KNN is: 0.900\n",
      "Model accuracy of Random Forest is: 0.951\n"
     ]
    }
   ],
   "source": [
    "# Use 10-fold Cross Validation to get the accuracy for different models\n",
    "model_names = ['Logistic Regression','KNN','Random Forest']\n",
    "model_list = [classifier_logistic, classifier_KNN, classifier_RF]\n",
    "count = 0\n",
    "\n",
    "for classifier in model_list:\n",
    "    cv_score = model_selection.cross_val_score(classifier, X_train, y_train, cv=10)\n",
    "    # cprint(cv_score)\n",
    "    print('Model accuracy of %s is: %.3f'%(model_names[count],cv_score.mean()))\n",
    "    count += 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "https://scikit-learn.org/stable/modules/model_evaluation.html"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "G1GCzaPxi0Np"
   },
   "source": [
    "#### Homework 2: Can you do prediction with SVM model?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 8367,
     "status": "ok",
     "timestamp": 1546590671639,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "yxmKTdahi0Nq",
    "outputId": "96726650-8664-47cd-a045-e9b27ca58ded"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model accuracy of SVM is: 0.926\n"
     ]
    }
   ],
   "source": [
    "####\n",
    "## Your Code\n",
    "####\n",
    "# SVC\n",
    "from sklearn.svm import SVC \n",
    "\n",
    "classifier_SVC = SVC()\n",
    "\n",
    "cv_score = model_selection.cross_val_score(classifier_SVC, X_train, y_train, cv=5)\n",
    "print('Model accuracy of SVM is: %.3f'%(cv_score.mean()))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "7J-23z78i0Ns"
   },
   "source": [
    "### (Optional) Part 3.3: Use Grid Search to Find Optimal Hyperparameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Hpe9PEAAi0Nt"
   },
   "outputs": [],
   "source": [
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "# helper function for printing out grid search results \n",
    "def print_grid_search_metrics(gs):\n",
    "    print (\"Best score: %0.3f\" % gs.best_score_)\n",
    "    print (\"Best parameters set:\")\n",
    "    best_parameters = gs.best_params_\n",
    "    for param_name in sorted(parameters.keys()):\n",
    "        print(\"\\t%s: %r\" % (param_name, best_parameters[param_name]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "qvYo9I5Ti0Nv"
   },
   "source": [
    "#### Part 3.3.1: Find Optimal Hyperparameters - LogisticRegression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 170
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 13697,
     "status": "ok",
     "timestamp": 1546590676979,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "wOc48syxi0Nx",
    "outputId": "0d807a12-998b-4e05-8244-7376c435440f",
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=5, error_score='raise-deprecating',\n",
       "       estimator=LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
       "          intercept_scaling=1, max_iter=100, multi_class='warn',\n",
       "          n_jobs=None, penalty='l2', random_state=None, solver='warn',\n",
       "          tol=0.0001, verbose=0, warm_start=False),\n",
       "       fit_params=None, iid='warn', n_jobs=None,\n",
       "       param_grid={'penalty': ('l1', 'l2'), 'C': (1, 5, 10)},\n",
       "       pre_dispatch='2*n_jobs', refit=True, return_train_score='warn',\n",
       "       scoring=None, verbose=0)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Possible hyperparamter options for Logistic Regression Regularization\n",
    "# Penalty is choosed from L1 or L2\n",
    "# C is the lambda value(weight) for L1 and L2\n",
    "parameters = {\n",
    "    'penalty':('l1', 'l2'), \n",
    "    'C':(1, 5, 10)\n",
    "}\n",
    "Grid_LR = GridSearchCV(LogisticRegression(),parameters, cv=5)\n",
    "Grid_LR.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 85
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 13695,
     "status": "ok",
     "timestamp": 1546590676983,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "nN5rU0e-i0N1",
    "outputId": "821b843a-066e-4139-c77c-8656ad49056e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best score: 0.867\n",
      "Best parameters set:\n",
      "\tC: 1\n",
      "\tpenalty: 'l1'\n"
     ]
    }
   ],
   "source": [
    "# the best hyperparameter combination\n",
    "print_grid_search_metrics(Grid_LR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "TtkDsXgui0N3"
   },
   "outputs": [],
   "source": [
    "# best model\n",
    "best_LR_model = Grid_LR.best_estimator_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "9u9YFedOi0N6"
   },
   "source": [
    "#### Part 3.3.2: Find Optimal Hyperparameters: KNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 136
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 7505,
     "status": "ok",
     "timestamp": 1546590907448,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "o78422XVi0N6",
    "outputId": "bc5c8daa-d335-428e-bcae-6956f1913a1e"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=5, error_score='raise-deprecating',\n",
       "       estimator=KNeighborsClassifier(algorithm='auto', leaf_size=30, metric='minkowski',\n",
       "           metric_params=None, n_jobs=None, n_neighbors=5, p=2,\n",
       "           weights='uniform'),\n",
       "       fit_params=None, iid='warn', n_jobs=None,\n",
       "       param_grid={'n_neighbors': [3, 5, 7, 10]}, pre_dispatch='2*n_jobs',\n",
       "       refit=True, return_train_score='warn', scoring=None, verbose=0)"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Possible hyperparamter options for KNN\n",
    "# Choose k\n",
    "parameters = {\n",
    "    'n_neighbors':[3,5,7,10] \n",
    "}\n",
    "Grid_KNN = GridSearchCV(KNeighborsClassifier(),parameters, cv=5)\n",
    "Grid_KNN.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 68
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 21017,
     "status": "ok",
     "timestamp": 1546590684319,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "ydaRZVAIi0N_",
    "outputId": "25ae3e92-3fed-4e4c-dc86-1a3384aab6d1",
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best score: 0.900\n",
      "Best parameters set:\n",
      "\tn_neighbors: 5\n"
     ]
    }
   ],
   "source": [
    "# best k\n",
    "print_grid_search_metrics(Grid_KNN)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "nKn_oKLSi0OB"
   },
   "source": [
    "#### Part 3.3.3: Find Optimal Hyperparameters: Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 204
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 29616,
     "status": "ok",
     "timestamp": 1546590692925,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "NniAZIPfi0OC",
    "outputId": "65d13e27-b964-4b39-b550-69df3679a154"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=5, error_score='raise-deprecating',\n",
       "       estimator=RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "            max_depth=None, max_features='auto', max_leaf_nodes=None,\n",
       "            min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "            min_samples_leaf=1, min_samples_split=2,\n",
       "            min_weight_fraction_leaf=0.0, n_estimators='warn', n_jobs=None,\n",
       "            oob_score=False, random_state=None, verbose=0,\n",
       "            warm_start=False),\n",
       "       fit_params=None, iid='warn', n_jobs=None,\n",
       "       param_grid={'n_estimators': [40, 60, 80]}, pre_dispatch='2*n_jobs',\n",
       "       refit=True, return_train_score='warn', scoring=None, verbose=0)"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Possible hyperparamter options for Random Forest\n",
    "# Choose the number of trees\n",
    "parameters = {\n",
    "    'n_estimators' : [40,60,80]\n",
    "}\n",
    "Grid_RF = GridSearchCV(RandomForestClassifier(),parameters, cv=5)\n",
    "Grid_RF.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 68
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 29612,
     "status": "ok",
     "timestamp": 1546590692927,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "ScPiI-Bfi0OE",
    "outputId": "40f70dab-ac62-4d5c-8535-821ce8e15d0c",
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best score: 0.958\n",
      "Best parameters set:\n",
      "\tn_estimators: 80\n"
     ]
    }
   ],
   "source": [
    "# best number of tress\n",
    "print_grid_search_metrics(Grid_RF)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "xJgfri_Mi0OG"
   },
   "outputs": [],
   "source": [
    "# best random forest\n",
    "best_RF_model = Grid_RF.best_estimator_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "# best_RF_model.fit(X_train,y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "xxDAOrGIi0OI"
   },
   "source": [
    "### Part 3.4: Model Evaluation - Confusion Matrix (Precision, Recall, Accuracy)\n",
    "\n",
    "class of interest as positive\n",
    "\n",
    "TP: correctly labeled real churn\n",
    "\n",
    "Precision(PPV, positive predictive value): tp / (tp + fp);\n",
    "Total number of true predictive churn divided by the total number of predictive churn;\n",
    "High Precision means low fp, not many return users were predicted as churn users. \n",
    "在所有预估的churn user 中，是真的churn user的概率是多少\n",
    "\n",
    "\n",
    "Recall(sensitivity, hit rate, true positive rate): tp / (tp + fn)\n",
    "Predict most postive or churn user correctly. High recall means low fn, not many churn users were predicted as return users.\n",
    "在所有真的churn user 中，我们判断出真的churn user概率是多少"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "o-tP94iFi0OI"
   },
   "outputs": [],
   "source": [
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import classification_report\n",
    "from sklearn.metrics import precision_score\n",
    "from sklearn.metrics import recall_score\n",
    "\n",
    "# calculate accuracy, precision and recall\n",
    "def cal_evaluation(classifier, cm):\n",
    "    tn = cm[0][0]\n",
    "    fp = cm[0][1]\n",
    "    fn = cm[1][0]\n",
    "    tp = cm[1][1]\n",
    "    accuracy  = (tp + tn) / (tp + fp + fn + tn + 0.0)\n",
    "    precision = tp / (tp + fp + 0.0)\n",
    "    recall = tp / (tp + fn + 0.0)\n",
    "    print (classifier)\n",
    "    print (\"Accuracy is: %0.3f\" % accuracy)\n",
    "    print (\"precision is: %0.3f\" % precision)\n",
    "    print (\"recall is: %0.3f\" % recall)\n",
    "\n",
    "# print out confusion matrices\n",
    "def draw_confusion_matrices(confusion_matricies):\n",
    "    class_names = ['Not','Churn']\n",
    "    for cm in confusion_matrices:\n",
    "        classifier, cm = cm[0], cm[1]\n",
    "        cal_evaluation(classifier, cm)\n",
    "        fig = plt.figure()\n",
    "        ax = fig.add_subplot(111)\n",
    "        cax = ax.matshow(cm, interpolation='nearest',cmap=plt.get_cmap('Reds'))\n",
    "        plt.title('Confusion matrix for %s' % classifier)\n",
    "        fig.colorbar(cax)\n",
    "        ax.set_xticklabels([''] + class_names)\n",
    "        ax.set_yticklabels([''] + class_names)\n",
    "        plt.xlabel('Predicted')\n",
    "        plt.ylabel('True')\n",
    "        plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 695
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 29963,
     "status": "ok",
     "timestamp": 1546590693285,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "OpSGaN49i0OL",
    "outputId": "e62d4f84-feaf-4e1a-b98c-968c435e1b9d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random Forest\n",
      "Accuracy is: 0.945\n",
      "precision is: 0.895\n",
      "recall is: 0.703\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAVAAAAEQCAYAAADmsCy1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAH8JJREFUeJzt3XucHlWd5/HPtzvcEwnQGCEJhIGIG1kJmGGDqIMEHWAZEkUQhpcENjOZcRhGRUR0fK3RcVcUlYujaBhmCYhcZEUyyqAYyMpdA4Rwl4DEJORCAgl3MPjbP+o8pGj6eZ7q6qe76+n+vvOqV1edOnXqPLdfzqlTF0UEZmbWex2DXQEzs3blAGpmVpIDqJlZSQ6gZmYlOYCamZXkAGpmVpIDaItI2kbSf0jaKOnHfSjnBEm/bGXdBouk90l6pOS2e0taLOk5Sf/U6rr1N0lPSDp0sOth/WvYBVBJfy1pkaTnJa2S9J+S3tuCoj8KjAF2iohjyhYSEZdFxIdaUJ9+JSkk7dUoT0TcHBF7l9zFGcBNETEqIs4vWcbrJM2R9Mf0uW+QdJukA/ta7mCTdLGkV9Prqk0fG+A6DNv/LIZVAJV0GnAu8L/Jgt1uwPeA6S0ofnfgdxGxqQVltT1JI/pYxO7AAy3e95URMRLoAm4CSvcUKuYbETEyN13Z2wIkdfZHxYa8iBgWE7A98DxwTIM8W5EF2CfTdC6wVVp3MLAC+AywFlgFnJzWfRl4Ffhj2scsYA7ww1zZE4AARqTlk4DHgeeA3wMn5NJvyW33HuC3wMb09z25dQuBfwFuTeX8Euiq89pq9T8jV/8ZwBHA74CngS/k8h8A3A5sSHn/Fdgyrft1ei0vpNf7sVz5nwNWA5fW0tI2e6Z97J+WdwWeAg7uoa43Aq8BL6fy354+v0vSNsuALwIduffsVuAcYD3w1R7K7P55TEqvYee0vAPws1T+M2l+XNH3Gvh4qtd64J+BJ4BDe/G9KvS59PC6Lu7p9aZ1/yXVewPZf0ZHddvuAuC69Dkemur5TeAPwBrg+8A2KX9Xek82pDrdTNYAuxT4E/BS+qzOGOzf+oDGlcGuwIC9UDgM2EQKYHXyfAW4A3grsDNwG/Avad3BafuvAFukL/iLwA5pffcfaPflCekHOwLYDngW2Dut2wV4Z5o/iRRAgR3Tj/njabvj0/JOaf1C4DGyALNNWj6rzmur1f9/pvr/LVmw+BEwCnhn+hHskfK/G5ia9jsBeAj4VK68APbqofyvpx/iNuQCaMrzt8CDwLbAL4BvNvgsFgJ/k1u+BLg21XUCWXCZlXvPNgGnpvpu00N5r38ewJbAWcA6Nv+HthNwdKrbKLLW6U+71afH95osGD8PvD+99m+n+tQCaJHvVaHPpYfXdTE9/4exBbAU+EJ6vYeQBf69c9ttBA4iC4Rbk/0HNJ/sezcK+A/gayn/18gC6hZpeh+gtO6J2msdbtOgV2DAXiicAKxukucx4Ijc8l8CT6T5g9MXeURu/Vpgapp//QdaZ3kCbwygG9IPdptudTiJzQH048Bvuq2/HTgpzS8Evphb9w/A9XVeW63+nWl5VKrPf8vluQuYUWf7TwHX5JZ7CqCvAlt3S1vRrZz5wH3AElIrrM7+FpICKNCZyp6UW/93wMLce/aHJp/tnFTGBrLW7Xp6aP3m8k8GnulWnx7fa7Lgd0Vu3XZpX7UAWuR7VfZzuZispb4hTetS+vvIegIdubyXA3Ny212SWyeyluieubQDgd+n+a+Q/Qe2Vw91eIJhGkCH0zHQ9UBXk2Nzu5J1w2qWpbTXy4g3HuN8ERjZ24pExAtk3d6/B1ZJ+rmkdxSoT61OY3PLq3tRn/UR8Vqafyn9XZNb/1Jte0lvl/QzSaslPUt23LirQdkAT0XEy03yXAjsA3wnIl5pkremi6zV0/2zyb8PywuUc1VEjCY7/n0/WSsbAEnbSvqBpGXp9f4aGN3t2GC993rX/P7T57s+l7fI96rQ51LHNyNidJpqn9GuwPKI+FO3/dZ7z3Yma33flQbZNgDXp3SAs8latL+U9LikMxvUZ9gYTgH0duAVsuNL9TxJNnhRs1tKK+MFsi9kzdvyKyPiFxHxQbLu+8NkgaVZfWp1WlmyTr1xAVm9JkbEW8i6gmqyTcNbe0kaSXb87yJgjqQdC9ZlHdnx5e6fTf59KHxbsYhYB8xOddglJX8G2Jus5fcWsu44NH/NkB23HF9bkLQt2SGBmlZ+r4p6EhgvKf8bb/SerSML1O/MBePtIxt0IyKei4jPRMSfAUcBp0ma1kM5w8qwCaARsZGsq/VdSTNSi2MLSYdL+kbKdjnwRUk7S+pK+X9YcpeLgfdL2k3S9sDnayskjZE0XdJ2ZEH9ebID8d1dB7w9nXo1Ip2eMonsYH5/G0V2nPb51Dr+RLf1a4A/62WZ5wGLIuJvgJ+THVNrKrXOrgL+l6RRknYHTqP8Z0NEPEJ2HPaMlDSKLIBsSIH9S70o7mrgSEnvlbQlWXc3/9tq5feqqDvJWslnpO/5wcBfAVf0lDm1VC8EzpH0VgBJYyX9ZZo/UtJekkR27PQ1Nn9ny3wXhoRhE0ABIuJbZD+8L5IdqF8O/CPw05Tlq8AisuNz9wF3p7Qy+7oBuDKVdRdvDHodqR5Pko1o/gVvDlBExHrgSLLW0XqyH/uRqQXV304H/pps4OFCsteSNweYl7p7xzYrTNJ0soG82us8Ddhf0gkF63MqWav+ceAWskGWfy+4bT1nA7NTwDiXbHBoHdmAz/VFC4mIB4BTUp1WkQ30rchladn3qhd1epUsYB5O9pq+B5wYEQ832OxzZN30O9JhjF+RtcoBJqbl58l6c9+LiJvSuq+R/QexQdLpLX8xFVYbRbMKkxTAtyPiM2n5dGBkRMxpsM0MsvNSHxyYWg5Pkt5GFnz/nGwQZw3Zf8hHRcSRg1k363/DqgXaxl4BPpK6f0XNIOvuWz9J3dlryM4G2DMi3k12qGZMH8vt60UINkAcQNvDJmAu8OnuKyRNkHSjpCWSFqRjru8hO9B/drqefM+BrvAw8QHgjxHx+rHciLiX7CTzkZKulvSwpMtSsK1d9tiV5qdIWpjm50i6VNKtwKWSTpL0E0nXS3o0d5zeKsQBtH18FzghDUjlfQeYFxHvAi4Dzo+I28jOt/xsREyOiMcGuK7DxT5kx7d7sh/ZubOTyAZYDipQ3iSy8ymPT8uTyU53+6/AxySNr7ulDQoH0DYREc+SXY3T/c5EB5INXkB2WV0rboxiffebiFiRRrcXk11I0cz8iHgpt7wgIjamc2sf5M2ntNkgcwBtL+eSXWe/3WBXxIDs+vJ311mXv0jgNbIr0CA7HFP73W3dbZsXCpZhFeEA2kYi4mmy8yFn5ZJvA45L8yeQHX+D7PSjUQNXu2HpRmArSbNrCZLeRXYZZT1PsDnoHt1/VbOB4ADafr7FGy+pPBU4WdISsmvnP5nSrwA+K+keDyL1j8jOAfwwcKikxyQ9QHZO5OoGm30ZOE/SIrJWpbUxnwdqZlaSW6BmZiU5gJqZleQAamZWkgOomVlJDqBDQP40Gqs+f15DhwPo0OAfZHvx5zVEOICamZU0bM4D3VqKUUP0/4uXCbYu9OSJ9rL7fu8a7Cr0i6fWrWfnrp2aZ2xDd92zeF1E7Nw8Z33jNSJeLvCUkHX86RcRcVhf9tVXw+ba2lF0cPQbHlFkVff9WxYOdhWsl7Td6O4PQey1lwmOLnC7hx/wXG/uj9svhk0ANbP2INrn2KIDqJlVioARKnBIqgJHHx1AzaxyOooc0ncANTN7M3fhzcxKEKKjSBe+AhxAzaxy2qUF2i71NLNhQmTHQJtNhcqSPi3pAUn3S7pc0taS9pB0p6Slkq6UtGXKu1VaXprWT2hWvgOomVWLoFNqOjUtRhpL9hDGKRGxD9BJ9vibrwPnRMRewDNsfkTOLOCZlH5OyteQA6iZVUrtPNBmU0EjgG0kjQC2BVYBhwBXp/XzgBlpfnpaJq2fJjWO1A6gZlY5rejCR8RK4JvAH8gC50bgLmBDRGxK2VYAY9P8WGB52nZTyt/wmlsHUDOrnIIt0C5Ji3LTG+5yJWkHslblHsCuZI8Db+m18x6FN7NKyQaRCo0SrYuIKQ3WHwr8PiKeApD0E+AgYLSkEamVOQ5YmfKvBMYDK1KXf3tgfaMKuAVqZpWSXcrZfCrgD8BUSdumY5nTgAeBm4CPpjwzgWvT/Py0TFp/YzS5XZ1boGZWOa1o2UXEnZKuBu4GNgH3AHOBnwNXSPpqSrsobXIRcKmkpcDTZCP2DTmAmlnldLTo/rYR8SXgS92SHwcO6CHvy8AxvSnfAdTMKqV2In07cAA1s8ppl8EZB1AzqxT14lLNweYAamaVU+iGyhXgAGpmleJHepiZ9YG78GZmJQi17DSm/uYAamaV4xaomVkJAjodQM3MynEX3sysBJ8HambWBz6NycyspDZpgDqAmlm19OKGyoPOAdTMKsddeDOzktqj/ekAamYV1ORpwpXRLi1lMxsmVHBqWo60t6TFuelZSZ+StKOkGyQ9mv7ukPJL0vmSlkpaImn/ZvtwADWzyin4WOOGIuKRiJgcEZOBdwMvAtcAZwILImIisCAtAxwOTEzTbOCCIvU0M6sUqfnUS9OAxyJiGdmz4uel9HnAjDQ/HbgkMneQPf54l0aFOoCaWaVk9wNV06mXjgMuT/NjImJVml8NjEnzY4HluW1WpLS6HEDNrHIKHgPtkrQoN83usSxpS+Ao4Mfd16Xnvjd89nsjHoU3s8opeC38uoiYUiDf4cDdEbEmLa+RtEtErEpd9LUpfSUwPrfduJRWv56FqmlmNmBU6F8vHM/m7jvAfGBmmp8JXJtLPzGNxk8FNua6+j1yC9TMKqXoaUqFypK2Az4I/F0u+SzgKkmzgGXAsSn9OuAIYCnZiP3Jzcp3ADWzamnh7ewi4gVgp25p68lG5bvnDeCU3pTvAGpmleMbKpuZldDKLnx/cwA1s8ppk0vhHUDNrHraJH46gJpZ9fTyNKVB4wBqZpXixxqbmfVBm8RPB1Azq5526cJX6lJOSSHpW7nl0yXNabLNDEmT+r1yZjZg+uF2dv2iUgEUeAX4iKSuXmwzA3AANRsistvZ9f2GygOhKvWo2QTMBT7dfYWkCZJuTLfaXyBpN0nvIbtN1dnplv17DnSFzaz1WvFIj4FQtQAK8F3gBEnbd0v/DjAvIt4FXAacHxG3kd1B5bPp1v2PDXBdzawfdEhNpyqoXACNiGeBS4B/6rbqQOBHaf5S4L3NypI0u3az1ZfL3zPVzAZQqx4qNxAqF0CTc4FZwHZ9KSQi5kbElIiYsnVl3nIza0hCBaYqqGQAjYingavIgmjNbWTPNQE4Abg5zT8HjBq42plZf+tQ86kKKhlAk28B+dH4U4GTJS0BPg58MqVfAXxW0j0eRDIbGtShplMVVOpE+ogYmZtfA2ybW14GHNLDNrfi05jMhgwJOqrctMupVAA1MwMqc4yzmTaJ82Y2nLTqSiRJoyVdLelhSQ9JOlDSjpJukPRo+rtDyitJ50tams43379Z+Q6gZlY5LRyFPw+4PiLeAewLPAScCSyIiInAgrQM2eOPJ6ZpNnBBs8IdQM2sUkRrWqDpYpz3AxcBRMSrEbEBmA7MS9nmkV0OTkq/JDJ3AKPTc+PrcgA1s2pR4SuRumoXyqRpdreS9gCeAv5POkvn39Jjjsfknve+GhiT5scCy3Pbr0hpdXkQycwqRnQUO01pXURMabB+BLA/cGpE3CnpPDZ314HsUcaSSl+m6BaomVWKAHU0nwpYAayIiDvT8tVkAXVNrWue/q5N61cC43Pbj0tpdTmAmlm1qDWDSBGxGlguae+UNA14kOwGRDNT2kzg2jQ/HzgxjcZPBTbmuvo9chfezCqnhaeBngpcJmlL4HHgZLKG41WSZgHLgGNT3uuAI4ClwIspb0MOoGZWOa06kT4iFgM9HSed1kPeAE7pTfkOoGZWOW1yIZIDqJlViwSdFblZSDMOoGZWOe1yLbwDqJlVTpvETwdQM6uW2qWc7cAB1MyqRdW5YXIzDqBmVjkeRDIzK8FdeDOzPvAovJlZGb244/xgcwA1s8pxC9TMrKQ2iZ8OoGZWLRJ0dLZHBHUANbOK6dVD4waVA6iZVY/PAzUzK8ktUDOzEtQ+o/B+JpKZVYygs6P5VKQk6QlJ90laLGlRSttR0g2SHk1/d0jpknS+pKWSlkjav1n5DqBmVikSqENNp174QERMzj0C+UxgQURMBBaw+VHHhwMT0zQbuKBZwQ6gZlY9UvOpvOnAvDQ/D5iRS78kMncAo2uPP67HAdTMKqdgC7RL0qLcNLuHogL4paS7cuvH5B5XvBoYk+bHAstz265IaXV5EMnMqqdYC3Ndrltez3sjYqWktwI3SHo4vzIiQlKUraZboGZWLVJ2HmizqYCIWJn+rgWuAQ4A1tS65unv2pR9JTA+t/m4lFaXA6iZVY46O5pOTcuQtpM0qjYPfAi4H5gPzEzZZgLXpvn5wIlpNH4qsDHX1e+Ru/BmVi2tu6PyGOCadE7pCOBHEXG9pN8CV0maBSwDjk35rwOOAJYCLwInN9uBA6iZVY5a0DeOiMeBfXtIXw9M6yE9gFN6sw8HUDOrnja5EskB1MyqxU/lNDPrg6HWApW0VUS80p+VMTOTKDTKXgVNaynpAEn3AY+m5X0lfaffa2Zmw1eLzgPtb0XC/PnAkcB6gIi4F/hAf1bKzIazAtfBV6SLX6QL3xERy7rdn++1fqqPmVnb3A+0SABdLukAICR1AqcCv+vfapnZsCUq00VvpkgA/QRZN343YA3wq5RmZtYv2mUQqWkATRfhHzcAdTEzq9QxzmaaBlBJF5LdU+8NIqKne++ZmfXZUDqR/le5+a2BD/PGm46ambXWUGmBRsSV+WVJlwK39FuNzGx4G2KDSN3tweZb4LeN3SfvwwU3/WKwq2G9EE8/OdhVsEEyZE5jkvQMm4+BdgBPs/kpdmZmLabCjy0ebA0DqLL/BvZl823t/5TumWdm1j9ad0PlftcwzKdgeV1EvJYmB08z639tcilnkXbyYkn79XtNzMwAEHR0NJ+KliZ1SrpH0s/S8h6S7pS0VNKVkrZM6Vul5aVp/YRmZdethaRa934/4LeSHpF0d6rI3YVrb2bWW61tgX4SeCi3/HXgnIjYC3gGmJXSZwHPpPRzUr6GGoXx36S/RwF7kz1s6Rjgo+mvmVnr1Y6BtiCAShoH/Hfg39KygEOAq1OWecCMND89LZPWT1OT0wEaDSIJICIeK1RTM7OWEHR2FsnYJWlRbnluRMztludc4AxgVFreCdgQEZvS8gpgbJofS7pIKCI2SdqY8q+rV4FGAXRnSafVWxkR326wrZlZecVamOsiYkr9InQksDYi7pJ0cKuqltcogHYCI0ktUTOzAdG605gOAo6SdATZZehvAc4DRksakVqh49h8muZKYDywIo0BbU+6kXw9jQLoqoj4Sh9fgJlZ77UggEbE54HPZ8XpYOD0iDhB0o/JxnKuAGYC16ZN5qfl29P6G5udutloEMktTzMbBK09jakHnwNOk7SU7BjnRSn9ImCnlH4aBa64bNQCndaXGpqZlSL6GiDfJCIWAgvT/OPAAT3keZlenmFUN4BGxNO9qqGZWatU5EqjZsrcjcnMrN8IoRa3QPuLA6iZVY9boGZmJbTR3ZgcQM2sehxAzczKKHwp56BzADWzanEX3sysDxxAzczKUMtPpO8vDqBmVj1ugZqZleBjoGZmZXkU3sysPLdAzcxKcBfezKwsj8KbmZXnFqiZWQkCOtpjEKk92slmNowIOgpMzUqRtpb0G0n3SnpA0pdT+h6S7pS0VNKVkrZM6Vul5aVp/YRm+3AANbPqUUfzqblXgEMiYl9gMnCYpKnA14FzImIv4BlgVso/C3gmpZ+T8jXkAGpm1SM1n5qIzPNpcYs0BXAIcHVKnwfMSPPT0zJp/TSp8Y4cQM2sWlT4qZxdkhblptlvLkqdkhYDa4EbgMeADemZ8AArgLFpfiywHCCt30j21M66PIhkZtVTbBR+XURMaZQhIl4DJksaDVwDvKMFtXudA6iZVU+LR+EjYoOkm4ADgdGSRqRW5jhgZcq2EhgPrJA0AtgeWN+wmi2tpZlZXxXvwjcpRjunlieStgE+CDwE3AR8NGWbCVyb5uenZdL6GyMiGu3DLVAzq57WnEi/CzBPUidZY/GqiPiZpAeBKyR9FbgHuCjlvwi4VNJS4GnguGY7cAA1s+opdppSQxGxBNivh/THgQN6SH8ZOKY3+3AANbNqUbET5avAAdTMqqdNLuV0ADWzilFLuvADwQHUzKpFuAtvZlaab2dnZlaSu/BmZiV4FN7MrA88Cm9mVoZH4c3MyvEovJlZH7gFamZWkk9jAklvA84F/hzYAKwBfgocFRFH9ue+zaxd+bnwpGeJXAPMi4jjUtq+wFF9LHdE7nb8ZjbU+LHGAHwA+GNEfL+WEBH3AjcDIyVdLelhSZfVHtwk6QlJXWl+iqSFaX6OpEsl3Up2v76TJP1E0vWSHpX0jX58HWY2oAo8UK4iXfz+7MLvA9xVZ91+wDuBJ4FbgYOAW5qUNwl4b0S8JOkksseU7kf26NJHJH0nIpbnN0gPmZoNsNu4sZhZm2iTLvxg1fI3EbEiIv4ELAYmFNhmfkS8lFteEBEb001QHwR2775BRMyNiCkRMWXnroYP1zOzqhBt0wLtzwD6APDuOuteyc2/xuaW8KZcnbbuts0LBcsws7aWTqRvNjUrRRov6SZJD0p6QNInU/qOkm5Ih/9ukLRDSpek8yUtlbRE0v7N9tGfAfRGYKv8s5olvQt4X4NtnmBz0D26/6pmZpXW0dl8am4T8JmImARMBU6RNAk4k6wHOxFYkJYBDgcmpmk2cEHTavb+lRWTnmb3YeBQSY9JegD4GrC6wWZfBs6TtIisVWlmw03tZiLNpiYiYlVE3J3mnyN7IudYYDowL2WbB8xI89OBSyJzB9njj3dptI9+7fZGxJPAsT2sujCX5x9z8zcDb++hnDndli8GLs4t+5xSs6GkxVciSZpANuh8JzAmIlalVauBMWl+LJAfiF6R0lZRh48bmln1FBsk6kq91Zq5ETH3zUVpJPB/gU9FxLPKlR0RIanhs98bcQA1s4opfDemdRExpWFJ0hZkwfOyiPhJSl4jaZeIWJW66GtT+kpgfG7zcSmtrvY42crMhhVJTacCZQi4CHgoIr6dWzUfmJnmZwLX5tJPTKPxU4GNua5+j9wCNbNqkaCjJaHpIODjwH2SFqe0LwBnAVdJmgUsY/M4zXXAEcBS4EXg5GY7cAA1s+ppwf1AI+IWstPyezKth/wBnNKbfTiAmln1+H6gZmYl1C7lbAMOoGZWMX4mkplZeW6BmpmVIEFne9xQ2QHUzKrHXXgzs5LchTczK8ODSGZm5bkFamZWggSd7RGa2qOWZjasFLlZSBU4gJpZ9fgYqJlZCb6U08ysLI/Cm5mV5xaomVkJvpTTzKwP2qQL3x61NLPhRWo+FSpG/y5praT7c2k7SrpB0qPp7w4pXZLOl7RU0hJJ+zcr3wHUzCpIBaZCLgYO65Z2JrAgIiYCC9IywOHAxDTNBi5oVrgDqJlVTIHWZ8EWaET8Gni6W/J0YF6anwfMyKVfEpk7gNHpscd1OYCaWfW0KIDWMSb3uOLVwJg0PxZYnsu3IqXV5UEkM6sWUXQQqUvSotzy3IiY25tdRURIit5sk+cAambVU6yBuS4ippQofY2kXSJiVeqir03pK4HxuXzjUlpd7sKbWQW1bBCpJ/OBmWl+JnBtLv3ENBo/FdiY6+r3yC1QM6uYPh/j3FySdDlwMFl3fwXwJeAs4CpJs4BlwLEp+3XAEcBS4EXg5GblO4CaWfW0KIBGxPF1Vk3rIW8Ap/SmfAdQM6ueNrkSyQHUzCrINxMxM+u9vp/nOWAcQM2sehxAzczKcgA1MyvFD5UzMyvFj/QwMyvPLVAzsxL8VE4zs75wADUzK8ctUDOzktojfjqAmlnVeBTezKwcDyKZmfWFA6iZWTlugZqZleG7MZmZldcmg0jK7mI/9El6iuz5J0NRF7BusCthhQ3lz2v3iNi5LwVIup7sPWpmXUQc1pd99dWwCaBDmaRFJR/vaoPAn9fQ0R7tZDOzCnIANTMryQF0aJg72BWwXvHnNUQ4gA4BETFoP0hJr0laLOl+ST+WtG0fyjpY0s/S/FGSzmyQd7SkfyixjzmSTi9bx1YYzM/LWssB1PrqpYiYHBH7AK8Cf59fqUyvv2cRMT8izmqQZTTQ6wBq1koOoNZKNwN7SZog6RFJlwD3A+MlfUjS7ZLuTi3VkQCSDpP0sKS7gY/UCpJ0kqR/TfNjJF0j6d40vQc4C9gztX7PTvk+K+m3kpZI+nKurH+W9DtJtwB7D9i7YUOeT6S3lpA0AjgcuD4lTQRmRsQdkrqALwKHRsQLkj4HnCbpG8CFwCHAUuDKOsWfD/y/iPiwpE5gJHAmsE9ETE77/1Da5wFkF1LPl/R+4AXgOGAy2ff9buCu1r56G64cQK2vtpG0OM3fDFwE7Aosi4g7UvpUYBJwa3ra4pbA7cA7gN9HxKMAkn4IzO5hH4cAJwJExGvARkk7dMvzoTTdk5ZHkgXUUcA1EfFi2sf8Pr1asxwHUOurl2qtwJoUJF/IJwE3RMTx3fK9Ybs+EvC1iPhBt318qoX7MHsDHwO1gXAHcJCkvQAkbSfp7cDDwARJe6Z8x9fZfgHwibRtp6TtgefIWpc1vwD+R+7Y6lhJbwV+DcyQtI2kUcBftfi12TDmAGr9LiKeAk4CLpe0hNR9j4iXybrsP0+DSGvrFPFJ4AOS7iM7fjkpItaTHRK4X9LZEfFL4EfA7Snf1cCoiLib7NjqvcB/Ar/ttxdqw46vhTczK8ktUDOzkhxAzcxKcgA1MyvJAdTMrCQHUDOzkhxAzcxKcgA1Myvp/wNZqxm5lLPyHAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Logistic Regression\n",
      "Accuracy is: 0.869\n",
      "precision is: 0.640\n",
      "recall is: 0.221\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "# Confusion matrix, accuracy, precison and recall for random forest and logistic regression\n",
    "confusion_matrices = [\n",
    "    (\"Random Forest\", confusion_matrix(y_test,best_RF_model.predict(X_test))),\n",
    "    (\"Logistic Regression\", confusion_matrix(y_test,best_LR_model.predict(X_test))),\n",
    "]\n",
    "\n",
    "draw_confusion_matrices(confusion_matrices)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "OvHlyhPBi0OT"
   },
   "source": [
    "### Part 3.4: Model Evaluation - ROC & AUC"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "jx_3XkgKi0OW"
   },
   "source": [
    "RandomForestClassifier, KNeighborsClassifier and LogisticRegression have predict_prob() function "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "-Os_ZLTvi0OX"
   },
   "source": [
    "#### Part 3.4.1: ROC of RF Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "UypvQMVBi0OY"
   },
   "outputs": [],
   "source": [
    "from sklearn.metrics import roc_curve\n",
    "from sklearn import metrics\n",
    "\n",
    "# Use predict_proba to get the probability results of Random Forest\n",
    "y_pred_rf = best_RF_model.predict_proba(X_test)[:, 1]\n",
    "fpr_rf, tpr_rf, _ = roc_curve(y_test, y_pred_rf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 294
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 30110,
     "status": "ok",
     "timestamp": 1546590693440,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "s3PR-PdPi0Ob",
    "outputId": "2c536e85-d92c-47ed-8250-b5a149a660df"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# ROC curve of Random Forest result\n",
    "plt.figure(1)\n",
    "plt.plot([0, 1], [0, 1], 'k--')\n",
    "plt.plot(fpr_rf, tpr_rf, label='RF')\n",
    "plt.xlabel('False positive rate')\n",
    "plt.ylabel('True positive rate')\n",
    "plt.title('ROC curve - RF model')\n",
    "plt.legend(loc='best')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 30106,
     "status": "ok",
     "timestamp": 1546590693441,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "R89IUMYDi0Oe",
    "outputId": "a0c12454-4a22-4520-bfc1-ed849ba0d0de"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.924573502722323"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn import metrics\n",
    "\n",
    "# AUC score\n",
    "metrics.auc(fpr_rf,tpr_rf)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "-1DVqnJVi0Oh"
   },
   "source": [
    "#### Part 3.4.1: ROC of LR Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "t-q5XJPoi0Oi"
   },
   "outputs": [],
   "source": [
    "# Use predict_proba to get the probability results of Logistic Regression\n",
    "y_pred_lr = best_LR_model.predict_proba(X_test)[:, 1]\n",
    "fpr_lr, tpr_lr, _ = roc_curve(y_test, y_pred_lr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 294
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 30437,
     "status": "ok",
     "timestamp": 1546590693779,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "KZSrN-1Mi0Ok",
    "outputId": "9c18fcd0-1793-4f64-f3aa-0dcfd0c0d69e"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# ROC Curve\n",
    "plt.figure(1)\n",
    "plt.plot([0, 1], [0, 1], 'k--')\n",
    "plt.plot(fpr_lr, tpr_lr, label='LR')\n",
    "plt.xlabel('False positive rate')\n",
    "plt.ylabel('True positive rate')\n",
    "plt.title('ROC curve - LR Model')\n",
    "plt.legend(loc='best')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 30433,
     "status": "ok",
     "timestamp": 1546590693780,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "LHAyxishi0On",
    "outputId": "4b458f3d-1046-48ed-c10a-195bdb00f882"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8251260334744908"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# AUC score\n",
    "metrics.auc(fpr_lr,tpr_lr)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "gHHurD8Ii0Oq"
   },
   "source": [
    "# Part 4: Feature Selection"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "dSx4TPO-i0Or"
   },
   "source": [
    "### Part 4.1:  Logistic Regression Model - Feature Selection Discussion "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "BtLHUixoi0Ot"
   },
   "source": [
    "The corelated features that we are interested in: (total_day_minutes, total_day_charge), (total_eve_minutes, total_eve_charge), (total_intl_minutes, total_intl_charge)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 323
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 30427,
     "status": "ok",
     "timestamp": 1546590693780,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "cQaXOIsUi0Ou",
    "outputId": "259e6840-e8be-4cd7-84f9-672c979f82ef",
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Logistic Regression (L1) Coefficients\n",
      "voice_mail_plan: -0.8657\n",
      "number_customer_service_calls: 0.673\n",
      "total_day_charge: 0.6428\n",
      "intl_plan: 0.613\n",
      "number_vmail_messages: 0.396\n",
      "total_eve_minutes: 0.3151\n",
      "total_intl_calls: -0.1743\n",
      "total_intl_charge: 0.1381\n",
      "total_night_charge: 0.1117\n",
      "total_intl_minutes: 0.1052\n",
      "total_day_minutes: 0.0916\n",
      "total_night_minutes: 0.09\n",
      "account_length: 0.0534\n",
      "total_day_calls: 0.0495\n",
      "total_eve_charge: 0.0427\n",
      "total_night_calls: -0.0288\n",
      "total_eve_calls: -0.0284\n"
     ]
    }
   ],
   "source": [
    "# add L1 regularization to logistic regression\n",
    "# check the coef for feature selection\n",
    "LRmodel_l1 = LogisticRegression(penalty=\"l1\")\n",
    "LRmodel_l1.fit(X, y)\n",
    "LRmodel_l1.coef_[0]\n",
    "print (\"Logistic Regression (L1) Coefficients\")\n",
    "for k,v in sorted(zip(map(lambda x: round(x, 4), LRmodel_l1.coef_[0]), \\\n",
    "                      churn_feat_space.columns), key=lambda k_v:(-abs(k_v[0]),k_v[1])):\n",
    "    print (v + \": \" + str(k))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 323
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 30626,
     "status": "ok",
     "timestamp": 1546590693986,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "majifZZqi0O9",
    "outputId": "166228d1-ab5d-4ee5-bf28-58cc26750af1"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Logistic Regression (L2) Coefficients\n",
      "voice_mail_plan: -0.8929\n",
      "number_customer_service_calls: 0.6731\n",
      "intl_plan: 0.6132\n",
      "number_vmail_messages: 0.4234\n",
      "total_day_minutes: 0.3687\n",
      "total_day_charge: 0.3664\n",
      "total_eve_charge: 0.18\n",
      "total_eve_minutes: 0.1794\n",
      "total_intl_calls: -0.1761\n",
      "total_intl_charge: 0.1235\n",
      "total_intl_minutes: 0.1217\n",
      "total_night_minutes: 0.1028\n",
      "total_night_charge: 0.1008\n",
      "account_length: 0.0552\n",
      "total_day_calls: 0.0513\n",
      "total_night_calls: -0.0307\n",
      "total_eve_calls: -0.0304\n"
     ]
    }
   ],
   "source": [
    "# add L2 regularization to logistic regression\n",
    "# check the coef for feature selection\n",
    "LRmodel_l2 = LogisticRegression(penalty=\"l2\")\n",
    "LRmodel_l2.fit(X, y)\n",
    "LRmodel_l2.coef_[0]\n",
    "print (\"Logistic Regression (L2) Coefficients\")\n",
    "for k,v in sorted(zip(map(lambda x: round(x, 4), LRmodel_l2.coef_[0]), \\\n",
    "                      churn_feat_space.columns), key=lambda k_v:(-abs(k_v[0]),k_v[1])):\n",
    "    print (v + \": \" + str(k))\n",
    "  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "uqs41ydLi0O_"
   },
   "source": [
    "### Part 4.2:  Random Forest Model - Feature Importance Discussion"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 323
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 30621,
     "status": "ok",
     "timestamp": 1546590693987,
     "user": {
      "displayName": "Arthur M",
      "photoUrl": "",
      "userId": "16377123886564016720"
     },
     "user_tz": 480
    },
    "id": "MPxUM2lei0PA",
    "outputId": "f528803b-5200-450e-e5df-97d4c4b3dff4"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Feature importance ranking by Random Forest Model:\n",
      "total_day_minutes: 0.1681\n",
      "total_day_charge: 0.1369\n",
      "number_customer_service_calls: 0.1071\n",
      "intl_plan: 0.0888\n",
      "total_eve_minutes: 0.0716\n",
      "total_eve_charge: 0.0653\n",
      "total_intl_calls: 0.047\n",
      "total_intl_charge: 0.0436\n",
      "total_night_minutes: 0.0422\n",
      "total_night_charge: 0.0367\n",
      "total_intl_minutes: 0.0347\n",
      "voice_mail_plan: 0.0325\n",
      "total_day_calls: 0.0277\n",
      "account_length: 0.0276\n",
      "number_vmail_messages: 0.0249\n",
      "total_eve_calls: 0.0233\n",
      "total_night_calls: 0.0223\n"
     ]
    }
   ],
   "source": [
    "# check feature importance of random forest for feature selection\n",
    "forest = RandomForestClassifier()\n",
    "forest.fit(X, y)\n",
    "\n",
    "importances = forest.feature_importances_\n",
    "\n",
    "# Print the feature ranking\n",
    "print(\"Feature importance ranking by Random Forest Model:\")\n",
    "for k,v in sorted(zip(map(lambda x: round(x, 4), importances), churn_feat_space.columns), reverse=True):\n",
    "    print (v + \": \" + str(k))"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "Copy of Supervised Learning Project.ipynb",
   "provenance": [
    {
     "file_id": "1BtyDGHRJSmlyi29CPUH8ZNmzh-j0eg3d",
     "timestamp": 1545526424684
    }
   ],
   "version": "0.3.2"
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}