data_collection.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"img/saclay_center.png\" width=\"800\" height=\"600\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Datacamp Challenge : Prediction of daily transport validation in Paris public underground transports "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<i> Authors :Dina ABED , Siwar ABBES, Philippe DE SAINT CHAMAS, Mohamed Skander HELLAL, Gwendal HENGOAT, Jules ZACCARDI </i>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Collection"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Introduction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Basic data used are in our challenge are the open data available on the Ile-de-France Mobilités platform which provides validation data of the entire Ile-de-France network. The transport network seen by the validation data covers 725 train stops (RER, metro, Transilien). \n",
    "Source: https://data.iledefrance-mobilites.fr/explore/dataset/histo-validations/information/\n",
    "\n",
    "We have merged all the data available since the first semester of 2015 until 2019. To get a whole observations of the number of validations per day and per stop. \n",
    "\n",
    "Thinking about some informations that could impact the frequentation of stations, our brainstorming leads to try to find other databases that contain these significant features to our business model:\n",
    "\n",
    "- find to which metro/RER line corresponds each stop, Database exploited: https://data.iledefrance-mobilites.fr/explore/dataset/emplacement-des-gares-idf-data-generalisee/information/\n",
    "- Search for a database of public holidays (whether public holiday or not and if so, the type of this day: Weekend, Christmas, Toussaint,...), Database exploited: https://www.data.gouv.fr/fr/datasets/jours-feries-en-france/\n",
    "- Incident rate, incident data on all RATP lines per day, Database exploited: http://data.cquest.org/openeventdatabase/\n",
    "- Social movements, Data on SNCF social movements: https://ressources.data.sncf.com/explore/dataset/mouvements-sociaux-depuis-2002/table/?sort=date_de_debut\n",
    "- Referential of passenger stations, Database: https://data.sncf.com/explore/dataset/frequentation-gares/table/?sort=nom_gare\n",
    "\n",
    "In order to get the data for the challenge, we worked on engineering features from the different databases defined above and merging them into a challenge data csv file\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Importing the Libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "import sys\n",
    "import numpy as np\n",
    "import pandas as pd \n",
    "from datetime import timedelta, date\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import seaborn as sns\n",
    "sns.set_style('ticks')\n",
    "sns.set(rc={'figure.figsize':(13.7,9.27)})\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "import os\n",
    "import folium\n",
    "import pandas as pd\n",
    "from IPython.display import display\n",
    "from folium import Popup\n",
    "import calendar\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1 . Validation Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "validations-sur-le-reseau-ferre-nombre-de-validations-par-jour-1er-semestre-2019.csv\n",
      "2018S1_NB_FER.txt\n",
      "2017S2_NB_FER.txt\n",
      "2016S2_NB_FER.txt\n",
      "2015S2_NB_FER.csv\n",
      "2016S1_NB_FER.txt\n",
      "2017S1_NB_FER.txt\n",
      "2015S1_NB_FER.csv\n"
     ]
    }
   ],
   "source": [
    "allfiles = os.listdir('./data/data valid/')\n",
    "li = []\n",
    "\n",
    "for file in allfiles : \n",
    "    print(file)\n",
    "    file = './data/data valid/' + file \n",
    "    if file[-3:] == 'csv':\n",
    "        data = pd.read_csv(file, header=0, sep=';')\n",
    "        data['JOUR'] = pd.to_datetime(data['JOUR'])\n",
    "        li.append(data)\n",
    "    else:\n",
    "        data = pd.read_csv(file, header=0, sep='\\t')\n",
    "        data['JOUR'] = pd.to_datetime(data['JOUR'])\n",
    "        li.append(data)\n",
    "\n",
    "dt = pd.concat(li, axis=0, ignore_index=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Modifying the types"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "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>JOUR</th>\n",
       "      <th>CODE_STIF_TRNS</th>\n",
       "      <th>CODE_STIF_RES</th>\n",
       "      <th>CODE_STIF_ARRET</th>\n",
       "      <th>LIBELLE_ARRET</th>\n",
       "      <th>ID_REFA_LDA</th>\n",
       "      <th>CATEGORIE_TITRE</th>\n",
       "      <th>NB_VALD</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-04-29</td>\n",
       "      <td>810</td>\n",
       "      <td>801</td>\n",
       "      <td>582</td>\n",
       "      <td>NATION</td>\n",
       "      <td>71673</td>\n",
       "      <td>AMETHYSTE</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-04-29</td>\n",
       "      <td>810</td>\n",
       "      <td>801</td>\n",
       "      <td>582</td>\n",
       "      <td>NATION</td>\n",
       "      <td>71673</td>\n",
       "      <td>IMAGINE R</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-04-29</td>\n",
       "      <td>810</td>\n",
       "      <td>801</td>\n",
       "      <td>582</td>\n",
       "      <td>NATION</td>\n",
       "      <td>71673</td>\n",
       "      <td>TST</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-04-29</td>\n",
       "      <td>810</td>\n",
       "      <td>801</td>\n",
       "      <td>591</td>\n",
       "      <td>NEUVILLE UNIVERSITE</td>\n",
       "      <td>66436</td>\n",
       "      <td>?</td>\n",
       "      <td>73</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-04-29</td>\n",
       "      <td>810</td>\n",
       "      <td>801</td>\n",
       "      <td>591</td>\n",
       "      <td>NEUVILLE UNIVERSITE</td>\n",
       "      <td>66436</td>\n",
       "      <td>NAVIGO</td>\n",
       "      <td>2205</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        JOUR  CODE_STIF_TRNS CODE_STIF_RES CODE_STIF_ARRET  \\\n",
       "0 2019-04-29             810           801             582   \n",
       "1 2019-04-29             810           801             582   \n",
       "2 2019-04-29             810           801             582   \n",
       "3 2019-04-29             810           801             591   \n",
       "4 2019-04-29             810           801             591   \n",
       "\n",
       "         LIBELLE_ARRET ID_REFA_LDA CATEGORIE_TITRE NB_VALD  \n",
       "0               NATION       71673       AMETHYSTE       7  \n",
       "1               NATION       71673       IMAGINE R      19  \n",
       "2               NATION       71673             TST      12  \n",
       "3  NEUVILLE UNIVERSITE       66436               ?      73  \n",
       "4  NEUVILLE UNIVERSITE       66436          NAVIGO    2205  "
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dt.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9.450435321932229"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dt.loc[dt['NB_VALD'] == 'Moins de 5'].shape[0]/dt.shape[0]*100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "dt.loc[dt['NB_VALD'] == 'Moins de 5', 'NB_VALD'] = 5\n",
    "dt['NB_VALD'] = pd.to_numeric(dt['NB_VALD'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "dt['JOUR'] = dt['JOUR'].astype(str)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "JOUR               object\n",
       "CODE_STIF_TRNS      int64\n",
       "CODE_STIF_RES      object\n",
       "CODE_STIF_ARRET    object\n",
       "LIBELLE_ARRET      object\n",
       "ID_REFA_LDA        object\n",
       "CATEGORIE_TITRE    object\n",
       "NB_VALD             int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dt.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "dt_agg = pd.DataFrame(dt.groupby([\"JOUR\",\"LIBELLE_ARRET\"])['NB_VALD'].sum().reset_index())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fb01f65a650>"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 986.4x667.44 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dt_agg.groupby('JOUR').sum().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "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>JOUR</th>\n",
       "      <th>LIBELLE_ARRET</th>\n",
       "      <th>NB_VALD</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ABBESSES</td>\n",
       "      <td>750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ABLON</td>\n",
       "      <td>118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ACHERES-GRAND-CORMIER</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ACHERES-VILLE</td>\n",
       "      <td>482</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>AEROPORT CHARLES DE GAULLE 1</td>\n",
       "      <td>1257</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1026013</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>VOLTAIRE (LEON BLUM)</td>\n",
       "      <td>6306</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1026014</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>VOSVES</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1026015</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>VULAINES-SUR-SEINE-SAMOREAU</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1026016</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>WAGRAM</td>\n",
       "      <td>1840</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1026017</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>YERRES</td>\n",
       "      <td>1622</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1026018 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "               JOUR                 LIBELLE_ARRET  NB_VALD\n",
       "0        2015-01-01                      ABBESSES      750\n",
       "1        2015-01-01                         ABLON      118\n",
       "2        2015-01-01         ACHERES-GRAND-CORMIER       18\n",
       "3        2015-01-01                 ACHERES-VILLE      482\n",
       "4        2015-01-01  AEROPORT CHARLES DE GAULLE 1     1257\n",
       "...             ...                           ...      ...\n",
       "1026013  2019-06-30          VOLTAIRE (LEON BLUM)     6306\n",
       "1026014  2019-06-30                        VOSVES       10\n",
       "1026015  2019-06-30   VULAINES-SUR-SEINE-SAMOREAU       20\n",
       "1026016  2019-06-30                        WAGRAM     1840\n",
       "1026017  2019-06-30                        YERRES     1622\n",
       "\n",
       "[1026018 rows x 3 columns]"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dt_agg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1026018, 3)"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dt_agg.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "dt_agg.to_csv('./data/validation_data_agg_2015_2019.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2 . Positions Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "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>Name</th>\n",
       "      <th>lon</th>\n",
       "      <th>lat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>11novembre</td>\n",
       "      <td>48.849993</td>\n",
       "      <td>2.395659</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>11novembre1918</td>\n",
       "      <td>48.824212</td>\n",
       "      <td>2.516243</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>11novembrecourcellesuryvetterer</td>\n",
       "      <td>48.700031</td>\n",
       "      <td>2.099321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>12fevrier1934</td>\n",
       "      <td>48.817827</td>\n",
       "      <td>2.308055</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>17octobre1961</td>\n",
       "      <td>48.898062</td>\n",
       "      <td>2.216702</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                              Name        lon       lat\n",
       "0                       11novembre  48.849993  2.395659\n",
       "1                   11novembre1918  48.824212  2.516243\n",
       "2  11novembrecourcellesuryvetterer  48.700031  2.099321\n",
       "3                    12fevrier1934  48.817827  2.308055\n",
       "4                    17octobre1961  48.898062  2.216702"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pos = pd.read_csv(\"data/positions-geographiques-des-stations-du-reseau-ratp.csv\", sep=';')\n",
    "dt2 = pos[['Name','Coordinates']].copy()\n",
    "\n",
    "l = np.array(dt2[\"Coordinates\"].str.split(\",\").tolist())\n",
    "dt2[\"lon\"] = l[:,0]\n",
    "dt2[\"lat\"] = l[:,1]\n",
    "dt2.drop(\"Coordinates\",axis=1, inplace=True)\n",
    "\n",
    "dt2['Name'] = dt2['Name'].str.lower()\n",
    "dt2['Name'] = dt2['Name'].str.replace('[^\\w]','')\n",
    "dt2['Name'] = dt2['Name'].str.normalize('NFKD').str.encode('ascii', errors='ignore').str.decode('utf-8')\n",
    "\n",
    "dt2['lon'] = pd.to_numeric(dt2[\"lon\"])\n",
    "dt2['lat'] = pd.to_numeric(dt2[\"lat\"])\n",
    "\n",
    "dt2 = dt2.groupby('Name').mean().reset_index()\n",
    "dt2.drop_duplicates(inplace=True)\n",
    "\n",
    "dt2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "dt_agg['Name'] = dt_agg['LIBELLE_ARRET'].str.lower()\n",
    "dt_agg['Name'] = dt_agg['Name'].str.replace('[^\\w]','')\n",
    "dt_agg['Name'] = dt_agg['Name'].str.normalize('NFKD').str.encode('ascii', errors='ignore').str.decode('utf-8')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "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>JOUR</th>\n",
       "      <th>LIBELLE_ARRET</th>\n",
       "      <th>NB_VALD</th>\n",
       "      <th>lon</th>\n",
       "      <th>lat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ABBESSES</td>\n",
       "      <td>750</td>\n",
       "      <td>48.884505</td>\n",
       "      <td>2.338365</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ABLON</td>\n",
       "      <td>118</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ACHERES-GRAND-CORMIER</td>\n",
       "      <td>18</td>\n",
       "      <td>48.955388</td>\n",
       "      <td>2.092457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ACHERES-VILLE</td>\n",
       "      <td>482</td>\n",
       "      <td>48.970248</td>\n",
       "      <td>2.077672</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>AEROPORT CHARLES DE GAULLE 1</td>\n",
       "      <td>1257</td>\n",
       "      <td>49.010400</td>\n",
       "      <td>2.561357</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         JOUR                 LIBELLE_ARRET  NB_VALD        lon       lat\n",
       "0  2015-01-01                      ABBESSES      750  48.884505  2.338365\n",
       "1  2015-01-01                         ABLON      118        NaN       NaN\n",
       "2  2015-01-01         ACHERES-GRAND-CORMIER       18  48.955388  2.092457\n",
       "3  2015-01-01                 ACHERES-VILLE      482  48.970248  2.077672\n",
       "4  2015-01-01  AEROPORT CHARLES DE GAULLE 1     1257  49.010400  2.561357"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dt_agg_pos = dt_agg.merge(dt2, how='left', on='Name')\n",
    "dt_agg_pos.drop('Name',axis=1,inplace=True)\n",
    "dt_agg_pos.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3 . Stations Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "gare_desc = pd.read_csv('data/emplacement-des-gares-idf-data-generalisee.csv',delimiter = ';')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "gare_desc = pd.read_csv('data/emplacement-des-gares-idf-data-generalisee.csv',delimiter = ';')\n",
    "stations_desc = pd.read_csv('data/stations.csv',delimiter = ';')\n",
    "gare_desc = gare_desc[['nom_long','res_com','mode_','principal']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "dt_agg_pos = pd.merge(dt_agg_pos,gare_desc,left_on='LIBELLE_ARRET',right_on='nom_long',how = 'left') \n",
    "dt_agg_pos = dt_agg_pos.drop(columns = ['nom_long'],axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "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>nom_long</th>\n",
       "      <th>res_com</th>\n",
       "      <th>mode_</th>\n",
       "      <th>principal</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>REUILLY-DIDEROT</td>\n",
       "      <td>M1 / M8</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>LOUVRE-RIVOLI</td>\n",
       "      <td>M1</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>LES MILONS</td>\n",
       "      <td>T2</td>\n",
       "      <td>Tramway</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>LYCEE HENRI SELLIER</td>\n",
       "      <td>T4</td>\n",
       "      <td>Tramway</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>PARMENTIER</td>\n",
       "      <td>M3</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              nom_long  res_com    mode_  principal\n",
       "0      REUILLY-DIDEROT  M1 / M8    Metro          0\n",
       "1        LOUVRE-RIVOLI       M1    Metro          0\n",
       "2           LES MILONS       T2  Tramway          0\n",
       "3  LYCEE HENRI SELLIER       T4  Tramway          0\n",
       "4           PARMENTIER       M3    Metro          0"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gare_desc.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "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>nom_long</th>\n",
       "      <th>res_com</th>\n",
       "      <th>mode_</th>\n",
       "      <th>principal</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>419</th>\n",
       "      <td>NATIONALE</td>\n",
       "      <td>M6</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>522</th>\n",
       "      <td>NATION</td>\n",
       "      <td>RER A / M1 / M2 / M6 / M9</td>\n",
       "      <td>RER / Metro</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>573</th>\n",
       "      <td>ASSEMBLEE NATIONALE</td>\n",
       "      <td>M12</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                nom_long                    res_com        mode_  principal\n",
       "419            NATIONALE                         M6        Metro          0\n",
       "522               NATION  RER A / M1 / M2 / M6 / M9  RER / Metro          0\n",
       "573  ASSEMBLEE NATIONALE                        M12        Metro          0"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gare_desc[gare_desc['nom_long'].str.contains(\"NATION\")]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4 . Correspondences  Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "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>JOUR</th>\n",
       "      <th>LIBELLE_ARRET</th>\n",
       "      <th>NB_VALD</th>\n",
       "      <th>lon</th>\n",
       "      <th>lat</th>\n",
       "      <th>res_com</th>\n",
       "      <th>mode_</th>\n",
       "      <th>principal</th>\n",
       "      <th>nb_metro</th>\n",
       "      <th>nb_rer</th>\n",
       "      <th>line</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ABBESSES</td>\n",
       "      <td>750</td>\n",
       "      <td>48.884505</td>\n",
       "      <td>2.338365</td>\n",
       "      <td>M12</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ABLON</td>\n",
       "      <td>118</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RER C</td>\n",
       "      <td>RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ACHERES-GRAND-CORMIER</td>\n",
       "      <td>18</td>\n",
       "      <td>48.955388</td>\n",
       "      <td>2.092457</td>\n",
       "      <td>RER A</td>\n",
       "      <td>RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ACHERES-VILLE</td>\n",
       "      <td>482</td>\n",
       "      <td>48.970248</td>\n",
       "      <td>2.077672</td>\n",
       "      <td>LIGNE L / RER A</td>\n",
       "      <td>Train / RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>AEROPORT CHARLES DE GAULLE 1</td>\n",
       "      <td>1257</td>\n",
       "      <td>49.010400</td>\n",
       "      <td>2.561357</td>\n",
       "      <td>RER B / CDGVAL</td>\n",
       "      <td>RER / Navette</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         JOUR                 LIBELLE_ARRET  NB_VALD        lon       lat  \\\n",
       "0  2015-01-01                      ABBESSES      750  48.884505  2.338365   \n",
       "1  2015-01-01                         ABLON      118        NaN       NaN   \n",
       "2  2015-01-01         ACHERES-GRAND-CORMIER       18  48.955388  2.092457   \n",
       "3  2015-01-01                 ACHERES-VILLE      482  48.970248  2.077672   \n",
       "4  2015-01-01  AEROPORT CHARLES DE GAULLE 1     1257  49.010400  2.561357   \n",
       "\n",
       "           res_com          mode_  principal  nb_metro  nb_rer line  \n",
       "0              M12          Metro        0.0       1.0     0.0   12  \n",
       "1            RER C            RER        0.0       0.0     1.0  NaN  \n",
       "2            RER A            RER        0.0       0.0     1.0  NaN  \n",
       "3  LIGNE L / RER A    Train / RER        0.0       0.0     1.0  NaN  \n",
       "4   RER B / CDGVAL  RER / Navette        0.0       0.0     1.0  NaN  "
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr = pd.read_csv('data/correspondances.csv')\n",
    "corr['Name'] = corr['stop_name'].str.lower()\n",
    "corr['Name'] = corr['Name'].str.replace('[^\\w]','')\n",
    "corr['Name'] = corr['Name'].str.normalize('NFKD').str.encode('ascii', errors='ignore').str.decode('utf-8')\n",
    "\n",
    "dt_agg_pos['Name'] = dt_agg_pos['LIBELLE_ARRET'].str.lower()\n",
    "dt_agg_pos['Name'] = dt_agg_pos['Name'].str.replace('[^\\w]','')\n",
    "dt_agg_pos['Name'] = dt_agg_pos['Name'].str.normalize('NFKD').str.encode('ascii', errors='ignore').str.decode('utf-8')\n",
    "\n",
    "dt_agg_pos_corr = dt_agg_pos.merge(corr, how='left', on='Name')\n",
    "dt_agg_pos_corr.drop(['Name','stop_name'],axis=1,inplace=True)\n",
    "dt_agg_pos_corr.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "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>JOUR</th>\n",
       "      <th>LIBELLE_ARRET</th>\n",
       "      <th>NB_VALD</th>\n",
       "      <th>lon</th>\n",
       "      <th>lat</th>\n",
       "      <th>res_com</th>\n",
       "      <th>mode_</th>\n",
       "      <th>principal</th>\n",
       "      <th>nb_metro</th>\n",
       "      <th>nb_rer</th>\n",
       "      <th>line</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>139</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>CHATELET-LES HALLES</td>\n",
       "      <td>13278</td>\n",
       "      <td>48.861630</td>\n",
       "      <td>2.346824</td>\n",
       "      <td>RER A / RER B / RER D</td>\n",
       "      <td>RER</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>375</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>LES HALLES</td>\n",
       "      <td>8319</td>\n",
       "      <td>48.862012</td>\n",
       "      <td>2.346476</td>\n",
       "      <td>M4</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>915</th>\n",
       "      <td>2015-01-02</td>\n",
       "      <td>CHATELET-LES HALLES</td>\n",
       "      <td>22781</td>\n",
       "      <td>48.861630</td>\n",
       "      <td>2.346824</td>\n",
       "      <td>RER A / RER B / RER D</td>\n",
       "      <td>RER</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1152</th>\n",
       "      <td>2015-01-02</td>\n",
       "      <td>LES HALLES</td>\n",
       "      <td>14908</td>\n",
       "      <td>48.862012</td>\n",
       "      <td>2.346476</td>\n",
       "      <td>M4</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1696</th>\n",
       "      <td>2015-01-03</td>\n",
       "      <td>CHATELET-LES HALLES</td>\n",
       "      <td>24157</td>\n",
       "      <td>48.861630</td>\n",
       "      <td>2.346824</td>\n",
       "      <td>RER A / RER B / RER D</td>\n",
       "      <td>RER</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1109668</th>\n",
       "      <td>2019-06-28</td>\n",
       "      <td>LES HALLES</td>\n",
       "      <td>47956</td>\n",
       "      <td>48.862012</td>\n",
       "      <td>2.346476</td>\n",
       "      <td>M4</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1110173</th>\n",
       "      <td>2019-06-29</td>\n",
       "      <td>CHATELET-LES HALLES</td>\n",
       "      <td>28924</td>\n",
       "      <td>48.861630</td>\n",
       "      <td>2.346824</td>\n",
       "      <td>RER A / RER B / RER D</td>\n",
       "      <td>RER</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1110394</th>\n",
       "      <td>2019-06-29</td>\n",
       "      <td>LES HALLES</td>\n",
       "      <td>37463</td>\n",
       "      <td>48.862012</td>\n",
       "      <td>2.346476</td>\n",
       "      <td>M4</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1110896</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>CHATELET-LES HALLES</td>\n",
       "      <td>18839</td>\n",
       "      <td>48.861630</td>\n",
       "      <td>2.346824</td>\n",
       "      <td>RER A / RER B / RER D</td>\n",
       "      <td>RER</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1111118</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>LES HALLES</td>\n",
       "      <td>18762</td>\n",
       "      <td>48.862012</td>\n",
       "      <td>2.346476</td>\n",
       "      <td>M4</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2862 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "               JOUR        LIBELLE_ARRET  NB_VALD        lon       lat  \\\n",
       "139      2015-01-01  CHATELET-LES HALLES    13278  48.861630  2.346824   \n",
       "375      2015-01-01           LES HALLES     8319  48.862012  2.346476   \n",
       "915      2015-01-02  CHATELET-LES HALLES    22781  48.861630  2.346824   \n",
       "1152     2015-01-02           LES HALLES    14908  48.862012  2.346476   \n",
       "1696     2015-01-03  CHATELET-LES HALLES    24157  48.861630  2.346824   \n",
       "...             ...                  ...      ...        ...       ...   \n",
       "1109668  2019-06-28           LES HALLES    47956  48.862012  2.346476   \n",
       "1110173  2019-06-29  CHATELET-LES HALLES    28924  48.861630  2.346824   \n",
       "1110394  2019-06-29           LES HALLES    37463  48.862012  2.346476   \n",
       "1110896  2019-06-30  CHATELET-LES HALLES    18839  48.861630  2.346824   \n",
       "1111118  2019-06-30           LES HALLES    18762  48.862012  2.346476   \n",
       "\n",
       "                       res_com  mode_  principal  nb_metro  nb_rer line  \n",
       "139      RER A / RER B / RER D    RER        1.0       0.0     3.0  NaN  \n",
       "375                         M4  Metro        0.0       1.0     0.0    4  \n",
       "915      RER A / RER B / RER D    RER        1.0       0.0     3.0  NaN  \n",
       "1152                        M4  Metro        0.0       1.0     0.0    4  \n",
       "1696     RER A / RER B / RER D    RER        1.0       0.0     3.0  NaN  \n",
       "...                        ...    ...        ...       ...     ...  ...  \n",
       "1109668                     M4  Metro        0.0       1.0     0.0    4  \n",
       "1110173  RER A / RER B / RER D    RER        1.0       0.0     3.0  NaN  \n",
       "1110394                     M4  Metro        0.0       1.0     0.0    4  \n",
       "1110896  RER A / RER B / RER D    RER        1.0       0.0     3.0  NaN  \n",
       "1111118                     M4  Metro        0.0       1.0     0.0    4  \n",
       "\n",
       "[2862 rows x 11 columns]"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dt_agg_pos_corr[dt_agg_pos_corr['LIBELLE_ARRET'].str.contains(\"HALLES\")]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 5 . Stations Repository Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "ref = pd.read_csv('data/referentiel-gares-voyageurs.csv',delimiter = ';')\n",
    "ref  = ref[['Intitulé gare','Segment DRG','Nbre plateformes']]\n",
    "ref['Intitulé gare'] = ref['Intitulé gare'].apply(lambda x : x.upper())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "dt_agg_pos_corr = pd.merge(dt_agg_pos_corr,ref,left_on='LIBELLE_ARRET',right_on='Intitulé gare',how = 'left') \n",
    "dt_agg_pos_corr = dt_agg_pos_corr.drop(columns = ['Intitulé gare'],axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 6 . Incidents Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "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>duree</th>\n",
       "      <th>date</th>\n",
       "      <th>line_num</th>\n",
       "      <th>line_type</th>\n",
       "      <th>what</th>\n",
       "      <th>type_inc</th>\n",
       "      <th>day</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>00:00:00</td>\n",
       "      <td>2016-07-01 10:34:00</td>\n",
       "      <td>12</td>\n",
       "      <td>metro</td>\n",
       "      <td>pedestrian</td>\n",
       "      <td>unscheduled</td>\n",
       "      <td>2016-07-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>00:00:00</td>\n",
       "      <td>2016-07-01 10:42:00</td>\n",
       "      <td>12</td>\n",
       "      <td>metro</td>\n",
       "      <td>pedestrian</td>\n",
       "      <td>unscheduled</td>\n",
       "      <td>2016-07-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>00:00:00</td>\n",
       "      <td>2016-07-01 11:30:00</td>\n",
       "      <td>12</td>\n",
       "      <td>metro</td>\n",
       "      <td>pedestrian</td>\n",
       "      <td>unscheduled</td>\n",
       "      <td>2016-07-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>00:37:00</td>\n",
       "      <td>2016-07-01 12:57:00</td>\n",
       "      <td>6</td>\n",
       "      <td>metro</td>\n",
       "      <td>unattended_luggage</td>\n",
       "      <td>unscheduled</td>\n",
       "      <td>2016-07-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>00:10:00</td>\n",
       "      <td>2016-07-01 13:35:00</td>\n",
       "      <td>6</td>\n",
       "      <td>metro</td>\n",
       "      <td>unattended_luggage</td>\n",
       "      <td>unscheduled</td>\n",
       "      <td>2016-07-01</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     duree                 date line_num line_type                what  \\\n",
       "0 00:00:00  2016-07-01 10:34:00       12     metro          pedestrian   \n",
       "1 00:00:00  2016-07-01 10:42:00       12     metro          pedestrian   \n",
       "2 00:00:00  2016-07-01 11:30:00       12     metro          pedestrian   \n",
       "3 00:37:00  2016-07-01 12:57:00        6     metro  unattended_luggage   \n",
       "4 00:10:00  2016-07-01 13:35:00        6     metro  unattended_luggage   \n",
       "\n",
       "      type_inc         day  \n",
       "0  unscheduled  2016-07-01  \n",
       "1  unscheduled  2016-07-01  \n",
       "2  unscheduled  2016-07-01  \n",
       "3  unscheduled  2016-07-01  \n",
       "4  unscheduled  2016-07-01  "
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inc = pd.read_csv('./data/incidents_2016_2019_clean.csv')\n",
    "inc['day'] = inc['date'].str[:10]\n",
    "inc['duree'] = pd.to_timedelta(inc['duree'])\n",
    "inc.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "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>day</th>\n",
       "      <th>line_num</th>\n",
       "      <th>duree</th>\n",
       "      <th>type_inc</th>\n",
       "      <th>num_inc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2016-05-01</td>\n",
       "      <td>1</td>\n",
       "      <td>00:00:00</td>\n",
       "      <td>generic</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2016-05-01</td>\n",
       "      <td>11</td>\n",
       "      <td>00:00:00</td>\n",
       "      <td>smoke</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2016-05-01</td>\n",
       "      <td>13</td>\n",
       "      <td>00:26:00</td>\n",
       "      <td>breakdown</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2016-05-01</td>\n",
       "      <td>2</td>\n",
       "      <td>00:00:00</td>\n",
       "      <td>breakdown</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2016-05-01</td>\n",
       "      <td>7</td>\n",
       "      <td>00:00:00</td>\n",
       "      <td>breakdown</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         day line_num    duree   type_inc  num_inc\n",
       "0 2016-05-01        1 00:00:00    generic        2\n",
       "1 2016-05-01       11 00:00:00      smoke        1\n",
       "2 2016-05-01       13 00:26:00  breakdown        4\n",
       "3 2016-05-01        2 00:00:00  breakdown        3\n",
       "4 2016-05-01        7 00:00:00  breakdown        2"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inc_group = inc.groupby(['day','line_num']).agg({\n",
    "    'duree' : sum,\n",
    "    'what' : 'first',\n",
    "    'type_inc' : 'count'\n",
    "}).reset_index().rename(columns={'type_inc':'num_inc','what':'type_inc'})\n",
    "inc_group['day'] = pd.to_datetime(inc_group['day'])\n",
    "\n",
    "inc_group.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "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>day</th>\n",
       "      <th>line_num</th>\n",
       "      <th>duree</th>\n",
       "      <th>type_inc</th>\n",
       "      <th>num_inc</th>\n",
       "      <th>stop_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2016-05-01</td>\n",
       "      <td>1</td>\n",
       "      <td>0 days</td>\n",
       "      <td>generic</td>\n",
       "      <td>2</td>\n",
       "      <td>REUILLY-DIDEROT</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2016-05-01</td>\n",
       "      <td>1</td>\n",
       "      <td>0 days</td>\n",
       "      <td>generic</td>\n",
       "      <td>2</td>\n",
       "      <td>LOUVRE-RIVOLI</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2016-05-01</td>\n",
       "      <td>1</td>\n",
       "      <td>0 days</td>\n",
       "      <td>generic</td>\n",
       "      <td>2</td>\n",
       "      <td>CONCORDE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2016-05-01</td>\n",
       "      <td>1</td>\n",
       "      <td>0 days</td>\n",
       "      <td>generic</td>\n",
       "      <td>2</td>\n",
       "      <td>GEORGE V</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2016-05-01</td>\n",
       "      <td>1</td>\n",
       "      <td>0 days</td>\n",
       "      <td>generic</td>\n",
       "      <td>2</td>\n",
       "      <td>ARGENTINE</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         day line_num  duree type_inc  num_inc        stop_name\n",
       "0 2016-05-01        1 0 days  generic        2  REUILLY-DIDEROT\n",
       "1 2016-05-01        1 0 days  generic        2    LOUVRE-RIVOLI\n",
       "2 2016-05-01        1 0 days  generic        2         CONCORDE\n",
       "3 2016-05-01        1 0 days  generic        2         GEORGE V\n",
       "4 2016-05-01        1 0 days  generic        2        ARGENTINE"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inc_station = inc_group.merge(corr, how='left',left_on='line_num',right_on='line').drop(['line','Name','nb_metro','nb_rer'],axis=1)\n",
    "\n",
    "inc_station.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "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>JOUR</th>\n",
       "      <th>LIBELLE_ARRET</th>\n",
       "      <th>NB_VALD</th>\n",
       "      <th>lon</th>\n",
       "      <th>lat</th>\n",
       "      <th>res_com</th>\n",
       "      <th>mode_</th>\n",
       "      <th>principal</th>\n",
       "      <th>nb_metro</th>\n",
       "      <th>nb_rer</th>\n",
       "      <th>line</th>\n",
       "      <th>Segment DRG</th>\n",
       "      <th>Nbre plateformes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ABBESSES</td>\n",
       "      <td>750</td>\n",
       "      <td>48.884505</td>\n",
       "      <td>2.338365</td>\n",
       "      <td>M12</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ABLON</td>\n",
       "      <td>118</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RER C</td>\n",
       "      <td>RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>b</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ACHERES-GRAND-CORMIER</td>\n",
       "      <td>18</td>\n",
       "      <td>48.955388</td>\n",
       "      <td>2.092457</td>\n",
       "      <td>RER A</td>\n",
       "      <td>RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ACHERES-VILLE</td>\n",
       "      <td>482</td>\n",
       "      <td>48.970248</td>\n",
       "      <td>2.077672</td>\n",
       "      <td>LIGNE L / RER A</td>\n",
       "      <td>Train / RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>AEROPORT CHARLES DE GAULLE 1</td>\n",
       "      <td>1257</td>\n",
       "      <td>49.010400</td>\n",
       "      <td>2.561357</td>\n",
       "      <td>RER B / CDGVAL</td>\n",
       "      <td>RER / Navette</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         JOUR                 LIBELLE_ARRET  NB_VALD        lon       lat  \\\n",
       "0  2015-01-01                      ABBESSES      750  48.884505  2.338365   \n",
       "1  2015-01-01                         ABLON      118        NaN       NaN   \n",
       "2  2015-01-01         ACHERES-GRAND-CORMIER       18  48.955388  2.092457   \n",
       "3  2015-01-01                 ACHERES-VILLE      482  48.970248  2.077672   \n",
       "4  2015-01-01  AEROPORT CHARLES DE GAULLE 1     1257  49.010400  2.561357   \n",
       "\n",
       "           res_com          mode_  principal  nb_metro  nb_rer line  \\\n",
       "0              M12          Metro        0.0       1.0     0.0   12   \n",
       "1            RER C            RER        0.0       0.0     1.0  NaN   \n",
       "2            RER A            RER        0.0       0.0     1.0  NaN   \n",
       "3  LIGNE L / RER A    Train / RER        0.0       0.0     1.0  NaN   \n",
       "4   RER B / CDGVAL  RER / Navette        0.0       0.0     1.0  NaN   \n",
       "\n",
       "  Segment DRG  Nbre plateformes  \n",
       "0         NaN               NaN  \n",
       "1           b               1.0  \n",
       "2         NaN               NaN  \n",
       "3         NaN               NaN  \n",
       "4         NaN               NaN  "
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dt_agg_pos_corr.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "inc_station['Name'] = inc_station['stop_name'].str.lower()\n",
    "inc_station['Name'] = inc_station['Name'].str.replace('[^\\w]','')\n",
    "inc_station['Name'] = inc_station['Name'].str.normalize('NFKD').str.encode('ascii', errors='ignore').str.decode('utf-8')\n",
    "\n",
    "inc_station['day'] = inc_station['day'].astype(str)\n",
    "\n",
    "curr_dt = dt_agg_pos_corr.copy()\n",
    "\n",
    "curr_dt['Name'] = curr_dt['LIBELLE_ARRET'].str.lower()\n",
    "curr_dt['Name'] = curr_dt['Name'].str.replace('[^\\w]','')\n",
    "curr_dt['Name'] = curr_dt['Name'].str.normalize('NFKD').str.encode('ascii', errors='ignore').str.decode('utf-8')\n",
    "\n",
    "curr_dt['JOUR'] = curr_dt['JOUR'].astype(str)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "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>JOUR</th>\n",
       "      <th>LIBELLE_ARRET</th>\n",
       "      <th>NB_VALD</th>\n",
       "      <th>lon</th>\n",
       "      <th>lat</th>\n",
       "      <th>res_com</th>\n",
       "      <th>mode_</th>\n",
       "      <th>principal</th>\n",
       "      <th>nb_metro</th>\n",
       "      <th>nb_rer</th>\n",
       "      <th>line</th>\n",
       "      <th>Segment DRG</th>\n",
       "      <th>Nbre plateformes</th>\n",
       "      <th>duree</th>\n",
       "      <th>type_inc</th>\n",
       "      <th>num_inc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ABBESSES</td>\n",
       "      <td>750</td>\n",
       "      <td>48.884505</td>\n",
       "      <td>2.338365</td>\n",
       "      <td>M12</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ABLON</td>\n",
       "      <td>118</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RER C</td>\n",
       "      <td>RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>b</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ACHERES-GRAND-CORMIER</td>\n",
       "      <td>18</td>\n",
       "      <td>48.955388</td>\n",
       "      <td>2.092457</td>\n",
       "      <td>RER A</td>\n",
       "      <td>RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ACHERES-VILLE</td>\n",
       "      <td>482</td>\n",
       "      <td>48.970248</td>\n",
       "      <td>2.077672</td>\n",
       "      <td>LIGNE L / RER A</td>\n",
       "      <td>Train / RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>AEROPORT CHARLES DE GAULLE 1</td>\n",
       "      <td>1257</td>\n",
       "      <td>49.010400</td>\n",
       "      <td>2.561357</td>\n",
       "      <td>RER B / CDGVAL</td>\n",
       "      <td>RER / Navette</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         JOUR                 LIBELLE_ARRET  NB_VALD        lon       lat  \\\n",
       "0  2015-01-01                      ABBESSES      750  48.884505  2.338365   \n",
       "1  2015-01-01                         ABLON      118        NaN       NaN   \n",
       "2  2015-01-01         ACHERES-GRAND-CORMIER       18  48.955388  2.092457   \n",
       "3  2015-01-01                 ACHERES-VILLE      482  48.970248  2.077672   \n",
       "4  2015-01-01  AEROPORT CHARLES DE GAULLE 1     1257  49.010400  2.561357   \n",
       "\n",
       "           res_com          mode_  principal  nb_metro  nb_rer line  \\\n",
       "0              M12          Metro        0.0       1.0     0.0   12   \n",
       "1            RER C            RER        0.0       0.0     1.0  NaN   \n",
       "2            RER A            RER        0.0       0.0     1.0  NaN   \n",
       "3  LIGNE L / RER A    Train / RER        0.0       0.0     1.0  NaN   \n",
       "4   RER B / CDGVAL  RER / Navette        0.0       0.0     1.0  NaN   \n",
       "\n",
       "  Segment DRG  Nbre plateformes duree type_inc  num_inc  \n",
       "0         NaN               NaN   NaT      NaN      NaN  \n",
       "1           b               1.0   NaT      NaN      NaN  \n",
       "2         NaN               NaN   NaT      NaN      NaN  \n",
       "3         NaN               NaN   NaT      NaN      NaN  \n",
       "4         NaN               NaN   NaT      NaN      NaN  "
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dt_pos_cor_inc = curr_dt.merge(inc_station, how='left',left_on=['JOUR','Name'], right_on=['day','Name'])\n",
    "dt_pos_cor_inc.drop(['Name','day','line_num','stop_name'],axis=1,inplace=True)\n",
    "\n",
    "dt_pos_cor_inc.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "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>JOUR</th>\n",
       "      <th>LIBELLE_ARRET</th>\n",
       "      <th>NB_VALD</th>\n",
       "      <th>lon</th>\n",
       "      <th>lat</th>\n",
       "      <th>res_com</th>\n",
       "      <th>mode_</th>\n",
       "      <th>principal</th>\n",
       "      <th>nb_metro</th>\n",
       "      <th>nb_rer</th>\n",
       "      <th>line</th>\n",
       "      <th>Segment DRG</th>\n",
       "      <th>Nbre plateformes</th>\n",
       "      <th>duree</th>\n",
       "      <th>type_inc</th>\n",
       "      <th>num_inc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>375032</th>\n",
       "      <td>2016-05-01</td>\n",
       "      <td>ALEXANDRE DUMAS</td>\n",
       "      <td>10584</td>\n",
       "      <td>48.872031</td>\n",
       "      <td>2.416374</td>\n",
       "      <td>M2</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>00:00:00</td>\n",
       "      <td>breakdown</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>375039</th>\n",
       "      <td>2016-05-01</td>\n",
       "      <td>ANVERS</td>\n",
       "      <td>9382</td>\n",
       "      <td>48.882872</td>\n",
       "      <td>2.344164</td>\n",
       "      <td>M2</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>00:00:00</td>\n",
       "      <td>breakdown</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>375042</th>\n",
       "      <td>2016-05-01</td>\n",
       "      <td>ARGENTINE</td>\n",
       "      <td>8038</td>\n",
       "      <td>48.875693</td>\n",
       "      <td>2.289097</td>\n",
       "      <td>M1</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>00:00:00</td>\n",
       "      <td>generic</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>375058</th>\n",
       "      <td>2016-05-01</td>\n",
       "      <td>AVRON</td>\n",
       "      <td>5099</td>\n",
       "      <td>48.850355</td>\n",
       "      <td>2.465669</td>\n",
       "      <td>M2</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>00:00:00</td>\n",
       "      <td>breakdown</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>375061</th>\n",
       "      <td>2016-05-01</td>\n",
       "      <td>BALARD</td>\n",
       "      <td>13411</td>\n",
       "      <td>48.836384</td>\n",
       "      <td>2.278305</td>\n",
       "      <td>M8 / T3A</td>\n",
       "      <td>Metro / Tramway</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>01:55:00</td>\n",
       "      <td>breakdown</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1136466</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>QUAI DE LA GARE</td>\n",
       "      <td>4874</td>\n",
       "      <td>48.837381</td>\n",
       "      <td>2.373923</td>\n",
       "      <td>M6</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>05:45:00</td>\n",
       "      <td>passenger</td>\n",
       "      <td>12.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1136472</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>RASPAIL</td>\n",
       "      <td>1511</td>\n",
       "      <td>48.776791</td>\n",
       "      <td>2.349999</td>\n",
       "      <td>M4 / M6</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>05:45:00</td>\n",
       "      <td>passenger</td>\n",
       "      <td>12.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1136473</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>RASPAIL</td>\n",
       "      <td>1511</td>\n",
       "      <td>48.776791</td>\n",
       "      <td>2.349999</td>\n",
       "      <td>M4 / M6</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>05:45:00</td>\n",
       "      <td>passenger</td>\n",
       "      <td>12.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1136513</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>SAINT-JACQUES</td>\n",
       "      <td>1999</td>\n",
       "      <td>48.832992</td>\n",
       "      <td>2.336683</td>\n",
       "      <td>M6</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>05:45:00</td>\n",
       "      <td>passenger</td>\n",
       "      <td>12.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1136543</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>SARTROUVILLE</td>\n",
       "      <td>5800</td>\n",
       "      <td>48.937512</td>\n",
       "      <td>2.157311</td>\n",
       "      <td>LIGNE L / RER A</td>\n",
       "      <td>Train / RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>A</td>\n",
       "      <td>b</td>\n",
       "      <td>1.0</td>\n",
       "      <td>00:20:00</td>\n",
       "      <td>breakdown</td>\n",
       "      <td>16.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>141443 rows × 16 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "               JOUR    LIBELLE_ARRET  NB_VALD        lon       lat  \\\n",
       "375032   2016-05-01  ALEXANDRE DUMAS    10584  48.872031  2.416374   \n",
       "375039   2016-05-01           ANVERS     9382  48.882872  2.344164   \n",
       "375042   2016-05-01        ARGENTINE     8038  48.875693  2.289097   \n",
       "375058   2016-05-01            AVRON     5099  48.850355  2.465669   \n",
       "375061   2016-05-01           BALARD    13411  48.836384  2.278305   \n",
       "...             ...              ...      ...        ...       ...   \n",
       "1136466  2019-06-30  QUAI DE LA GARE     4874  48.837381  2.373923   \n",
       "1136472  2019-06-30          RASPAIL     1511  48.776791  2.349999   \n",
       "1136473  2019-06-30          RASPAIL     1511  48.776791  2.349999   \n",
       "1136513  2019-06-30    SAINT-JACQUES     1999  48.832992  2.336683   \n",
       "1136543  2019-06-30     SARTROUVILLE     5800  48.937512  2.157311   \n",
       "\n",
       "                 res_com            mode_  principal  nb_metro  nb_rer line  \\\n",
       "375032                M2            Metro        0.0       1.0     0.0    2   \n",
       "375039                M2            Metro        0.0       1.0     0.0    2   \n",
       "375042                M1            Metro        0.0       1.0     0.0    1   \n",
       "375058                M2            Metro        0.0       1.0     0.0    2   \n",
       "375061          M8 / T3A  Metro / Tramway        0.0       1.0     0.0    8   \n",
       "...                  ...              ...        ...       ...     ...  ...   \n",
       "1136466               M6            Metro        0.0       1.0     0.0    6   \n",
       "1136472          M4 / M6            Metro        0.0       2.0     0.0    4   \n",
       "1136473          M4 / M6            Metro        0.0       2.0     0.0    6   \n",
       "1136513               M6            Metro        0.0       1.0     0.0    6   \n",
       "1136543  LIGNE L / RER A      Train / RER        0.0       0.0     1.0    A   \n",
       "\n",
       "        Segment DRG  Nbre plateformes    duree   type_inc  num_inc  \n",
       "375032          NaN               NaN 00:00:00  breakdown      3.0  \n",
       "375039          NaN               NaN 00:00:00  breakdown      3.0  \n",
       "375042          NaN               NaN 00:00:00    generic      2.0  \n",
       "375058          NaN               NaN 00:00:00  breakdown      3.0  \n",
       "375061          NaN               NaN 01:55:00  breakdown      5.0  \n",
       "...             ...               ...      ...        ...      ...  \n",
       "1136466         NaN               NaN 05:45:00  passenger     12.0  \n",
       "1136472         NaN               NaN 05:45:00  passenger     12.0  \n",
       "1136473         NaN               NaN 05:45:00  passenger     12.0  \n",
       "1136513         NaN               NaN 05:45:00  passenger     12.0  \n",
       "1136543           b               1.0 00:20:00  breakdown     16.0  \n",
       "\n",
       "[141443 rows x 16 columns]"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dt_pos_cor_inc.loc[dt_pos_cor_inc['num_inc'].notna()]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 7 . Social movements Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "mvs = pd.read_csv('data/mouvements-sociaux.csv',delimiter=';')\n",
    "mvs['date_de_debut'] = pd.to_datetime(mvs['date_de_debut'],format=\"%Y/%m/%d\")\n",
    "mvs['date_de_fin'] = pd.to_datetime(mvs['date_de_fin'],format=\"%Y/%m/%d\")\n",
    "mvs  = mvs[mvs['date_de_debut'].isin(pd.date_range(start='20150101', end='20190630'))]\n",
    "mvs_dates = mvs[['date_de_debut','date_de_fin']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "list_dates = []\n",
    "for index in mvs_dates.index:\n",
    "    start_date = mvs_dates['date_de_debut'][index]\n",
    "    if  not pd.isna(mvs_dates['date_de_fin'][index]): \n",
    "        end_date = mvs_dates['date_de_fin'][index]\n",
    "        for n in range(int ((end_date - start_date).days)):\n",
    "            list_dates.append(start_date + timedelta(n))\n",
    "    else : \n",
    "        list_dates.append(start_date)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "mvs_dates = {'date': list_dates}\n",
    "mvs_dates = pd.DataFrame(data=mvs_dates)\n",
    "mvs_dates['est_greve'] = np.ones(mvs_dates.shape[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "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>JOUR</th>\n",
       "      <th>LIBELLE_ARRET</th>\n",
       "      <th>NB_VALD</th>\n",
       "      <th>lon</th>\n",
       "      <th>lat</th>\n",
       "      <th>res_com</th>\n",
       "      <th>mode_</th>\n",
       "      <th>principal</th>\n",
       "      <th>nb_metro</th>\n",
       "      <th>nb_rer</th>\n",
       "      <th>line</th>\n",
       "      <th>Segment DRG</th>\n",
       "      <th>Nbre plateformes</th>\n",
       "      <th>duree</th>\n",
       "      <th>type_inc</th>\n",
       "      <th>num_inc</th>\n",
       "      <th>est_greve</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ABBESSES</td>\n",
       "      <td>750</td>\n",
       "      <td>48.884505</td>\n",
       "      <td>2.338365</td>\n",
       "      <td>M12</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ABLON</td>\n",
       "      <td>118</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RER C</td>\n",
       "      <td>RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>b</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ACHERES-GRAND-CORMIER</td>\n",
       "      <td>18</td>\n",
       "      <td>48.955388</td>\n",
       "      <td>2.092457</td>\n",
       "      <td>RER A</td>\n",
       "      <td>RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ACHERES-VILLE</td>\n",
       "      <td>482</td>\n",
       "      <td>48.970248</td>\n",
       "      <td>2.077672</td>\n",
       "      <td>LIGNE L / RER A</td>\n",
       "      <td>Train / RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>AEROPORT CHARLES DE GAULLE 1</td>\n",
       "      <td>1257</td>\n",
       "      <td>49.010400</td>\n",
       "      <td>2.561357</td>\n",
       "      <td>RER B / CDGVAL</td>\n",
       "      <td>RER / Navette</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1138198</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>VOLTAIRE (LEON BLUM)</td>\n",
       "      <td>6306</td>\n",
       "      <td>48.857966</td>\n",
       "      <td>2.380567</td>\n",
       "      <td>M9</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1138199</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>VOSVES</td>\n",
       "      <td>10</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RER D</td>\n",
       "      <td>RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>c</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1138200</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>VULAINES-SUR-SEINE-SAMOREAU</td>\n",
       "      <td>20</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>LIGNE R</td>\n",
       "      <td>Train</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1138201</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>WAGRAM</td>\n",
       "      <td>1840</td>\n",
       "      <td>48.883712</td>\n",
       "      <td>2.305558</td>\n",
       "      <td>M3</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1138202</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>YERRES</td>\n",
       "      <td>1622</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RER D</td>\n",
       "      <td>RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>b</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1138203 rows × 17 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              JOUR                 LIBELLE_ARRET  NB_VALD        lon  \\\n",
       "0       2015-01-01                      ABBESSES      750  48.884505   \n",
       "1       2015-01-01                         ABLON      118        NaN   \n",
       "2       2015-01-01         ACHERES-GRAND-CORMIER       18  48.955388   \n",
       "3       2015-01-01                 ACHERES-VILLE      482  48.970248   \n",
       "4       2015-01-01  AEROPORT CHARLES DE GAULLE 1     1257  49.010400   \n",
       "...            ...                           ...      ...        ...   \n",
       "1138198 2019-06-30          VOLTAIRE (LEON BLUM)     6306  48.857966   \n",
       "1138199 2019-06-30                        VOSVES       10        NaN   \n",
       "1138200 2019-06-30   VULAINES-SUR-SEINE-SAMOREAU       20        NaN   \n",
       "1138201 2019-06-30                        WAGRAM     1840  48.883712   \n",
       "1138202 2019-06-30                        YERRES     1622        NaN   \n",
       "\n",
       "              lat          res_com          mode_  principal  nb_metro  \\\n",
       "0        2.338365              M12          Metro        0.0       1.0   \n",
       "1             NaN            RER C            RER        0.0       0.0   \n",
       "2        2.092457            RER A            RER        0.0       0.0   \n",
       "3        2.077672  LIGNE L / RER A    Train / RER        0.0       0.0   \n",
       "4        2.561357   RER B / CDGVAL  RER / Navette        0.0       0.0   \n",
       "...           ...              ...            ...        ...       ...   \n",
       "1138198  2.380567               M9          Metro        0.0       1.0   \n",
       "1138199       NaN            RER D            RER        0.0       0.0   \n",
       "1138200       NaN          LIGNE R          Train        0.0       0.0   \n",
       "1138201  2.305558               M3          Metro        0.0       1.0   \n",
       "1138202       NaN            RER D            RER        0.0       0.0   \n",
       "\n",
       "         nb_rer line Segment DRG  Nbre plateformes duree type_inc  num_inc  \\\n",
       "0           0.0   12         NaN               NaN   NaT      NaN      NaN   \n",
       "1           1.0  NaN           b               1.0   NaT      NaN      NaN   \n",
       "2           1.0  NaN         NaN               NaN   NaT      NaN      NaN   \n",
       "3           1.0  NaN         NaN               NaN   NaT      NaN      NaN   \n",
       "4           1.0  NaN         NaN               NaN   NaT      NaN      NaN   \n",
       "...         ...  ...         ...               ...   ...      ...      ...   \n",
       "1138198     0.0  NaN         NaN               NaN   NaT      NaN      NaN   \n",
       "1138199     1.0  NaN           c               1.0   NaT      NaN      NaN   \n",
       "1138200     0.0  NaN         NaN               NaN   NaT      NaN      NaN   \n",
       "1138201     0.0    3         NaN               NaN   NaT      NaN      NaN   \n",
       "1138202     1.0  NaN           b               1.0   NaT      NaN      NaN   \n",
       "\n",
       "         est_greve  \n",
       "0              0.0  \n",
       "1              0.0  \n",
       "2              0.0  \n",
       "3              0.0  \n",
       "4              0.0  \n",
       "...            ...  \n",
       "1138198        0.0  \n",
       "1138199        0.0  \n",
       "1138200        0.0  \n",
       "1138201        0.0  \n",
       "1138202        0.0  \n",
       "\n",
       "[1138203 rows x 17 columns]"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dt_pos_cor_inc['JOUR'] = pd.to_datetime(dt_pos_cor_inc['JOUR'])\n",
    "df_merged = pd.merge(dt_pos_cor_inc,mvs_dates, left_on='JOUR',right_on='date',how='left')\n",
    "df_merged = df_merged.drop(columns=['date'])\n",
    "df_merged[\"est_greve\"].fillna(0, inplace = True)\n",
    "df_merged"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###  8 . Stations timetable Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "horaires_gares = pd.read_csv('data/horaires-des-gares.csv',delimiter = ';')\n",
    "horaires_gares = horaires_gares[['Gare','Jour de la semaine','Horaire en jour normal', 'Horaire en jour férié']]\n",
    "horaires_gares['Gare']  = horaires_gares['Gare'].apply(lambda x : x.upper())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_merged['Jour de la semaine'] = df_merged['JOUR'].apply(lambda x : calendar.day_name [x.date().weekday()])\n",
    "jour_eng = ['Monday','Tuesday','Wednesday','Thursday','Friday','Saturday','Sunday']\n",
    "jour_fr = ['Lundi','Mardi','Mercredi','Jeudi','Vendredi','Samedi','Dimanche']\n",
    "horaires_gares['Jour de la semaine'] = horaires_gares['Jour de la semaine'].apply(lambda x : jour_eng[jour_fr.index(x)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "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>JOUR</th>\n",
       "      <th>LIBELLE_ARRET</th>\n",
       "      <th>NB_VALD</th>\n",
       "      <th>lon</th>\n",
       "      <th>lat</th>\n",
       "      <th>res_com</th>\n",
       "      <th>mode_</th>\n",
       "      <th>principal</th>\n",
       "      <th>nb_metro</th>\n",
       "      <th>nb_rer</th>\n",
       "      <th>line</th>\n",
       "      <th>Segment DRG</th>\n",
       "      <th>Nbre plateformes</th>\n",
       "      <th>duree</th>\n",
       "      <th>type_inc</th>\n",
       "      <th>num_inc</th>\n",
       "      <th>est_greve</th>\n",
       "      <th>Jour de la semaine</th>\n",
       "      <th>Horaire en jour normal</th>\n",
       "      <th>Horaire en jour férié</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ABBESSES</td>\n",
       "      <td>750</td>\n",
       "      <td>48.884505</td>\n",
       "      <td>2.338365</td>\n",
       "      <td>M12</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ABLON</td>\n",
       "      <td>118</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RER C</td>\n",
       "      <td>RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>b</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ACHERES-GRAND-CORMIER</td>\n",
       "      <td>18</td>\n",
       "      <td>48.955388</td>\n",
       "      <td>2.092457</td>\n",
       "      <td>RER A</td>\n",
       "      <td>RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ACHERES-VILLE</td>\n",
       "      <td>482</td>\n",
       "      <td>48.970248</td>\n",
       "      <td>2.077672</td>\n",
       "      <td>LIGNE L / RER A</td>\n",
       "      <td>Train / RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>AEROPORT CHARLES DE GAULLE 1</td>\n",
       "      <td>1257</td>\n",
       "      <td>49.010400</td>\n",
       "      <td>2.561357</td>\n",
       "      <td>RER B / CDGVAL</td>\n",
       "      <td>RER / Navette</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1143443</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>VOLTAIRE (LEON BLUM)</td>\n",
       "      <td>6306</td>\n",
       "      <td>48.857966</td>\n",
       "      <td>2.380567</td>\n",
       "      <td>M9</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Sunday</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1143444</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>VOSVES</td>\n",
       "      <td>10</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RER D</td>\n",
       "      <td>RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>c</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Sunday</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1143445</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>VULAINES-SUR-SEINE-SAMOREAU</td>\n",
       "      <td>20</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>LIGNE R</td>\n",
       "      <td>Train</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Sunday</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1143446</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>WAGRAM</td>\n",
       "      <td>1840</td>\n",
       "      <td>48.883712</td>\n",
       "      <td>2.305558</td>\n",
       "      <td>M3</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Sunday</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1143447</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>YERRES</td>\n",
       "      <td>1622</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RER D</td>\n",
       "      <td>RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>b</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Sunday</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1143448 rows × 20 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              JOUR                 LIBELLE_ARRET  NB_VALD        lon  \\\n",
       "0       2015-01-01                      ABBESSES      750  48.884505   \n",
       "1       2015-01-01                         ABLON      118        NaN   \n",
       "2       2015-01-01         ACHERES-GRAND-CORMIER       18  48.955388   \n",
       "3       2015-01-01                 ACHERES-VILLE      482  48.970248   \n",
       "4       2015-01-01  AEROPORT CHARLES DE GAULLE 1     1257  49.010400   \n",
       "...            ...                           ...      ...        ...   \n",
       "1143443 2019-06-30          VOLTAIRE (LEON BLUM)     6306  48.857966   \n",
       "1143444 2019-06-30                        VOSVES       10        NaN   \n",
       "1143445 2019-06-30   VULAINES-SUR-SEINE-SAMOREAU       20        NaN   \n",
       "1143446 2019-06-30                        WAGRAM     1840  48.883712   \n",
       "1143447 2019-06-30                        YERRES     1622        NaN   \n",
       "\n",
       "              lat          res_com          mode_  principal  nb_metro  \\\n",
       "0        2.338365              M12          Metro        0.0       1.0   \n",
       "1             NaN            RER C            RER        0.0       0.0   \n",
       "2        2.092457            RER A            RER        0.0       0.0   \n",
       "3        2.077672  LIGNE L / RER A    Train / RER        0.0       0.0   \n",
       "4        2.561357   RER B / CDGVAL  RER / Navette        0.0       0.0   \n",
       "...           ...              ...            ...        ...       ...   \n",
       "1143443  2.380567               M9          Metro        0.0       1.0   \n",
       "1143444       NaN            RER D            RER        0.0       0.0   \n",
       "1143445       NaN          LIGNE R          Train        0.0       0.0   \n",
       "1143446  2.305558               M3          Metro        0.0       1.0   \n",
       "1143447       NaN            RER D            RER        0.0       0.0   \n",
       "\n",
       "         nb_rer line Segment DRG  Nbre plateformes duree type_inc  num_inc  \\\n",
       "0           0.0   12         NaN               NaN   NaT      NaN      NaN   \n",
       "1           1.0  NaN           b               1.0   NaT      NaN      NaN   \n",
       "2           1.0  NaN         NaN               NaN   NaT      NaN      NaN   \n",
       "3           1.0  NaN         NaN               NaN   NaT      NaN      NaN   \n",
       "4           1.0  NaN         NaN               NaN   NaT      NaN      NaN   \n",
       "...         ...  ...         ...               ...   ...      ...      ...   \n",
       "1143443     0.0  NaN         NaN               NaN   NaT      NaN      NaN   \n",
       "1143444     1.0  NaN           c               1.0   NaT      NaN      NaN   \n",
       "1143445     0.0  NaN         NaN               NaN   NaT      NaN      NaN   \n",
       "1143446     0.0    3         NaN               NaN   NaT      NaN      NaN   \n",
       "1143447     1.0  NaN           b               1.0   NaT      NaN      NaN   \n",
       "\n",
       "         est_greve Jour de la semaine Horaire en jour normal  \\\n",
       "0              0.0           Thursday                    NaN   \n",
       "1              0.0           Thursday                    NaN   \n",
       "2              0.0           Thursday                    NaN   \n",
       "3              0.0           Thursday                    NaN   \n",
       "4              0.0           Thursday                    NaN   \n",
       "...            ...                ...                    ...   \n",
       "1143443        0.0             Sunday                    NaN   \n",
       "1143444        0.0             Sunday                    NaN   \n",
       "1143445        0.0             Sunday                    NaN   \n",
       "1143446        0.0             Sunday                    NaN   \n",
       "1143447        0.0             Sunday                    NaN   \n",
       "\n",
       "        Horaire en jour férié  \n",
       "0                         NaN  \n",
       "1                         NaN  \n",
       "2                         NaN  \n",
       "3                         NaN  \n",
       "4                         NaN  \n",
       "...                       ...  \n",
       "1143443                   NaN  \n",
       "1143444                   NaN  \n",
       "1143445                   NaN  \n",
       "1143446                   NaN  \n",
       "1143447                   NaN  \n",
       "\n",
       "[1143448 rows x 20 columns]"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_merged = pd.merge(df_merged,horaires_gares , left_on = ['LIBELLE_ARRET','Jour de la semaine'],right_on = ['Gare','Jour de la semaine'],how = 'left')\n",
    "df_merged = df_merged.drop(['Gare'],axis=1)\n",
    "df_merged"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 9 . Day off Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "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>date</th>\n",
       "      <th>est_jour_ferie</th>\n",
       "      <th>nom_jour_ferie</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1950-01-01</td>\n",
       "      <td>True</td>\n",
       "      <td>Jour de l'an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1950-01-02</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1950-01-03</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1950-01-04</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1950-01-05</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         date  est_jour_ferie nom_jour_ferie\n",
       "0  1950-01-01            True   Jour de l'an\n",
       "1  1950-01-02           False            NaN\n",
       "2  1950-01-03           False            NaN\n",
       "3  1950-01-04           False            NaN\n",
       "4  1950-01-05           False            NaN"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_ferie = pd.read_csv(\"./data/jours-feries.csv\")\n",
    "data_ferie.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Let's convert dates into datetime type\n",
    "data_ferie['date'] = pd.to_datetime(data_ferie['date'],format=\"%Y/%m/%d\")\n",
    "# Let's keep only dates between 2015-01-01 and 2019-06-30\n",
    "data_ferie=data_ferie[data_ferie['date'].isin(pd.date_range(start='20150101', end='20190630'))]\n",
    "# Let's replace NaN values by \"jour non ferie\"\n",
    "data_ferie[\"nom_jour_ferie\"].fillna(\"Jour non ferie\", inplace = True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "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>date</th>\n",
       "      <th>est_jour_ferie</th>\n",
       "      <th>type_jour</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>23741</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>True</td>\n",
       "      <td>Jour de l'an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23742</th>\n",
       "      <td>2015-01-02</td>\n",
       "      <td>False</td>\n",
       "      <td>Jour non ferie</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23743</th>\n",
       "      <td>2015-01-03</td>\n",
       "      <td>False</td>\n",
       "      <td>Jour non ferie</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23744</th>\n",
       "      <td>2015-01-04</td>\n",
       "      <td>False</td>\n",
       "      <td>Jour non ferie</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23745</th>\n",
       "      <td>2015-01-05</td>\n",
       "      <td>False</td>\n",
       "      <td>Jour non ferie</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            date  est_jour_ferie       type_jour\n",
       "23741 2015-01-01            True    Jour de l'an\n",
       "23742 2015-01-02           False  Jour non ferie\n",
       "23743 2015-01-03           False  Jour non ferie\n",
       "23744 2015-01-04           False  Jour non ferie\n",
       "23745 2015-01-05           False  Jour non ferie"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_ferie.columns = ['date','est_jour_ferie','type_jour']\n",
    "data_ferie.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "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>date</th>\n",
       "      <th>est_jour_ferie</th>\n",
       "      <th>type_jour</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>23741</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>1</td>\n",
       "      <td>Jour de l'an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23742</th>\n",
       "      <td>2015-01-02</td>\n",
       "      <td>0</td>\n",
       "      <td>Jour non ferie</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23743</th>\n",
       "      <td>2015-01-03</td>\n",
       "      <td>0</td>\n",
       "      <td>Jour non ferie</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23744</th>\n",
       "      <td>2015-01-04</td>\n",
       "      <td>0</td>\n",
       "      <td>Jour non ferie</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23745</th>\n",
       "      <td>2015-01-05</td>\n",
       "      <td>0</td>\n",
       "      <td>Jour non ferie</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            date  est_jour_ferie       type_jour\n",
       "23741 2015-01-01               1    Jour de l'an\n",
       "23742 2015-01-02               0  Jour non ferie\n",
       "23743 2015-01-03               0  Jour non ferie\n",
       "23744 2015-01-04               0  Jour non ferie\n",
       "23745 2015-01-05               0  Jour non ferie"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Encoder la colonne 'est_jour_ferié' : True => 1 et False => 0\n",
    "data_ferie.est_jour_ferie = data_ferie.est_jour_ferie.astype(int)\n",
    "data_ferie.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 10 .  Feature Engineering:  Week-End Days"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "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>date</th>\n",
       "      <th>est_jour_ferie</th>\n",
       "      <th>type_jour</th>\n",
       "      <th>est_weekend</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>23741</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>1</td>\n",
       "      <td>Jour de l'an</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23742</th>\n",
       "      <td>2015-01-02</td>\n",
       "      <td>0</td>\n",
       "      <td>Jour non ferie</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23743</th>\n",
       "      <td>2015-01-03</td>\n",
       "      <td>0</td>\n",
       "      <td>Jour non ferie</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23744</th>\n",
       "      <td>2015-01-04</td>\n",
       "      <td>0</td>\n",
       "      <td>Jour non ferie</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23745</th>\n",
       "      <td>2015-01-05</td>\n",
       "      <td>0</td>\n",
       "      <td>Jour non ferie</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            date  est_jour_ferie       type_jour  est_weekend\n",
       "23741 2015-01-01               1    Jour de l'an            0\n",
       "23742 2015-01-02               0  Jour non ferie            0\n",
       "23743 2015-01-03               0  Jour non ferie            1\n",
       "23744 2015-01-04               0  Jour non ferie            1\n",
       "23745 2015-01-05               0  Jour non ferie            0"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_ferie['est_weekend'] = data_ferie['date'].map(lambda x : 1 if x.weekday() >= 5 else 0 )\n",
    "data_ferie.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "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>date</th>\n",
       "      <th>est_jour_ferie</th>\n",
       "      <th>type_jour</th>\n",
       "      <th>est_weekend</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>23741</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>1</td>\n",
       "      <td>Jour de l'an</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23742</th>\n",
       "      <td>2015-01-02</td>\n",
       "      <td>0</td>\n",
       "      <td>Jour non ferie</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23743</th>\n",
       "      <td>2015-01-03</td>\n",
       "      <td>0</td>\n",
       "      <td>Weekend</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23744</th>\n",
       "      <td>2015-01-04</td>\n",
       "      <td>0</td>\n",
       "      <td>Weekend</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23745</th>\n",
       "      <td>2015-01-05</td>\n",
       "      <td>0</td>\n",
       "      <td>Jour non ferie</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            date  est_jour_ferie       type_jour  est_weekend\n",
       "23741 2015-01-01               1    Jour de l'an            0\n",
       "23742 2015-01-02               0  Jour non ferie            0\n",
       "23743 2015-01-03               0         Weekend            1\n",
       "23744 2015-01-04               0         Weekend            1\n",
       "23745 2015-01-05               0  Jour non ferie            0"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Ajouter le type \"Weekend\" pour la colonne \"type_jour\"\n",
    "mask = (data_ferie['est_weekend'] == 1) \n",
    "data_ferie['type_jour'][mask] = \"Weekend\"\n",
    "data_ferie.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "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>date</th>\n",
       "      <th>est_jour_ferie</th>\n",
       "      <th>type_jour</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>23741</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>1</td>\n",
       "      <td>Jour de l'an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23742</th>\n",
       "      <td>2015-01-02</td>\n",
       "      <td>0</td>\n",
       "      <td>Jour non ferie</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23743</th>\n",
       "      <td>2015-01-03</td>\n",
       "      <td>0</td>\n",
       "      <td>Weekend</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23744</th>\n",
       "      <td>2015-01-04</td>\n",
       "      <td>0</td>\n",
       "      <td>Weekend</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23745</th>\n",
       "      <td>2015-01-05</td>\n",
       "      <td>0</td>\n",
       "      <td>Jour non ferie</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            date  est_jour_ferie       type_jour\n",
       "23741 2015-01-01               1    Jour de l'an\n",
       "23742 2015-01-02               0  Jour non ferie\n",
       "23743 2015-01-03               0         Weekend\n",
       "23744 2015-01-04               0         Weekend\n",
       "23745 2015-01-05               0  Jour non ferie"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_ferie.drop('est_weekend',axis=1,inplace=True)\n",
    "data_ferie.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "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>JOUR</th>\n",
       "      <th>LIBELLE_ARRET</th>\n",
       "      <th>NB_VALD</th>\n",
       "      <th>lon</th>\n",
       "      <th>lat</th>\n",
       "      <th>res_com</th>\n",
       "      <th>mode_</th>\n",
       "      <th>principal</th>\n",
       "      <th>nb_metro</th>\n",
       "      <th>nb_rer</th>\n",
       "      <th>...</th>\n",
       "      <th>Nbre plateformes</th>\n",
       "      <th>duree</th>\n",
       "      <th>type_inc</th>\n",
       "      <th>num_inc</th>\n",
       "      <th>est_greve</th>\n",
       "      <th>Jour de la semaine</th>\n",
       "      <th>Horaire en jour normal</th>\n",
       "      <th>Horaire en jour férié</th>\n",
       "      <th>est_jour_ferie</th>\n",
       "      <th>type_jour</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ABBESSES</td>\n",
       "      <td>750</td>\n",
       "      <td>48.884505</td>\n",
       "      <td>2.338365</td>\n",
       "      <td>M12</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>Jour de l'an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ABLON</td>\n",
       "      <td>118</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RER C</td>\n",
       "      <td>RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>Jour de l'an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ACHERES-GRAND-CORMIER</td>\n",
       "      <td>18</td>\n",
       "      <td>48.955388</td>\n",
       "      <td>2.092457</td>\n",
       "      <td>RER A</td>\n",
       "      <td>RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>Jour de l'an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>ACHERES-VILLE</td>\n",
       "      <td>482</td>\n",
       "      <td>48.970248</td>\n",
       "      <td>2.077672</td>\n",
       "      <td>LIGNE L / RER A</td>\n",
       "      <td>Train / RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>Jour de l'an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>AEROPORT CHARLES DE GAULLE 1</td>\n",
       "      <td>1257</td>\n",
       "      <td>49.010400</td>\n",
       "      <td>2.561357</td>\n",
       "      <td>RER B / CDGVAL</td>\n",
       "      <td>RER / Navette</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>Jour de l'an</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 22 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        JOUR                 LIBELLE_ARRET  NB_VALD        lon       lat  \\\n",
       "0 2015-01-01                      ABBESSES      750  48.884505  2.338365   \n",
       "1 2015-01-01                         ABLON      118        NaN       NaN   \n",
       "2 2015-01-01         ACHERES-GRAND-CORMIER       18  48.955388  2.092457   \n",
       "3 2015-01-01                 ACHERES-VILLE      482  48.970248  2.077672   \n",
       "4 2015-01-01  AEROPORT CHARLES DE GAULLE 1     1257  49.010400  2.561357   \n",
       "\n",
       "           res_com          mode_  principal  nb_metro  nb_rer  ...  \\\n",
       "0              M12          Metro        0.0       1.0     0.0  ...   \n",
       "1            RER C            RER        0.0       0.0     1.0  ...   \n",
       "2            RER A            RER        0.0       0.0     1.0  ...   \n",
       "3  LIGNE L / RER A    Train / RER        0.0       0.0     1.0  ...   \n",
       "4   RER B / CDGVAL  RER / Navette        0.0       0.0     1.0  ...   \n",
       "\n",
       "  Nbre plateformes duree  type_inc num_inc est_greve  Jour de la semaine  \\\n",
       "0              NaN   NaT       NaN     NaN       0.0            Thursday   \n",
       "1              1.0   NaT       NaN     NaN       0.0            Thursday   \n",
       "2              NaN   NaT       NaN     NaN       0.0            Thursday   \n",
       "3              NaN   NaT       NaN     NaN       0.0            Thursday   \n",
       "4              NaN   NaT       NaN     NaN       0.0            Thursday   \n",
       "\n",
       "   Horaire en jour normal Horaire en jour férié est_jour_ferie     type_jour  \n",
       "0                     NaN                   NaN              1  Jour de l'an  \n",
       "1                     NaN                   NaN              1  Jour de l'an  \n",
       "2                     NaN                   NaN              1  Jour de l'an  \n",
       "3                     NaN                   NaN              1  Jour de l'an  \n",
       "4                     NaN                   NaN              1  Jour de l'an  \n",
       "\n",
       "[5 rows x 22 columns]"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_merged = pd.merge(df_merged, data_ferie, left_on='JOUR',right_on='date',how='left')\n",
    "df_merged.drop('date',axis=1,inplace=True)\n",
    "df_merged.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train,X_test = train_test_split(df_merged, test_size=0.33, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "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>JOUR</th>\n",
       "      <th>LIBELLE_ARRET</th>\n",
       "      <th>NB_VALD</th>\n",
       "      <th>lon</th>\n",
       "      <th>lat</th>\n",
       "      <th>res_com</th>\n",
       "      <th>mode_</th>\n",
       "      <th>principal</th>\n",
       "      <th>nb_metro</th>\n",
       "      <th>nb_rer</th>\n",
       "      <th>...</th>\n",
       "      <th>Nbre plateformes</th>\n",
       "      <th>duree</th>\n",
       "      <th>type_inc</th>\n",
       "      <th>num_inc</th>\n",
       "      <th>est_greve</th>\n",
       "      <th>Jour de la semaine</th>\n",
       "      <th>Horaire en jour normal</th>\n",
       "      <th>Horaire en jour férié</th>\n",
       "      <th>est_jour_ferie</th>\n",
       "      <th>type_jour</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>259473</th>\n",
       "      <td>2015-12-01</td>\n",
       "      <td>LES GRESILLONS</td>\n",
       "      <td>3008</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RER C</td>\n",
       "      <td>RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Tuesday</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>Jour non ferie</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>255030</th>\n",
       "      <td>2015-11-25</td>\n",
       "      <td>ERMONT-HALTE</td>\n",
       "      <td>544</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>LIGNE H</td>\n",
       "      <td>Train</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Wednesday</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>Jour non ferie</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>490696</th>\n",
       "      <td>2016-09-19</td>\n",
       "      <td>COURCELLE-SUR-YVETTE</td>\n",
       "      <td>1091</td>\n",
       "      <td>48.700934</td>\n",
       "      <td>2.099045</td>\n",
       "      <td>RER B</td>\n",
       "      <td>RER</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Monday</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>Jour non ferie</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1047961</th>\n",
       "      <td>2019-02-23</td>\n",
       "      <td>RAMBOUILLET</td>\n",
       "      <td>2718</td>\n",
       "      <td>48.844479</td>\n",
       "      <td>2.381176</td>\n",
       "      <td>LIGNE N</td>\n",
       "      <td>Train</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Saturday</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>Weekend</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1050591</th>\n",
       "      <td>2019-02-27</td>\n",
       "      <td>FR. POPULAIRE</td>\n",
       "      <td>7627</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Wednesday</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>Jour non ferie</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 22 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              JOUR         LIBELLE_ARRET  NB_VALD        lon       lat  \\\n",
       "259473  2015-12-01        LES GRESILLONS     3008        NaN       NaN   \n",
       "255030  2015-11-25          ERMONT-HALTE      544        NaN       NaN   \n",
       "490696  2016-09-19  COURCELLE-SUR-YVETTE     1091  48.700934  2.099045   \n",
       "1047961 2019-02-23           RAMBOUILLET     2718  48.844479  2.381176   \n",
       "1050591 2019-02-27         FR. POPULAIRE     7627        NaN       NaN   \n",
       "\n",
       "         res_com  mode_  principal  nb_metro  nb_rer  ... Nbre plateformes  \\\n",
       "259473     RER C    RER        0.0       0.0     1.0  ...              NaN   \n",
       "255030   LIGNE H  Train        0.0       0.0     0.0  ...              NaN   \n",
       "490696     RER B    RER        0.0       0.0     1.0  ...              NaN   \n",
       "1047961  LIGNE N  Train        1.0       0.0     0.0  ...              1.0   \n",
       "1050591      NaN    NaN        NaN       NaN     NaN  ...              NaN   \n",
       "\n",
       "        duree  type_inc num_inc est_greve  Jour de la semaine  \\\n",
       "259473    NaT       NaN     NaN       0.0             Tuesday   \n",
       "255030    NaT       NaN     NaN       0.0           Wednesday   \n",
       "490696    NaT       NaN     NaN       0.0              Monday   \n",
       "1047961   NaT       NaN     NaN       0.0            Saturday   \n",
       "1050591   NaT       NaN     NaN       0.0           Wednesday   \n",
       "\n",
       "         Horaire en jour normal Horaire en jour férié est_jour_ferie  \\\n",
       "259473                      NaN                   NaN              0   \n",
       "255030                      NaN                   NaN              0   \n",
       "490696                      NaN                   NaN              0   \n",
       "1047961                     NaN                   NaN              0   \n",
       "1050591                     NaN                   NaN              0   \n",
       "\n",
       "              type_jour  \n",
       "259473   Jour non ferie  \n",
       "255030   Jour non ferie  \n",
       "490696   Jour non ferie  \n",
       "1047961         Weekend  \n",
       "1050591  Jour non ferie  \n",
       "\n",
       "[5 rows x 22 columns]"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train.to_csv('./data/data challenge/data_challenge_TRAIN.csv',index=False)\n",
    "X_test.to_csv('./data/data challenge/data_challenge_TEST.csv',index=False)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}