{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "166e531b",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import datetime\n",
    "import glob\n",
    "import re"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "id": "c65a1428",
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.options.display.max_rows = 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "037e13ee",
   "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>Raison Sociale</th>\n",
       "      <th>SIREN</th>\n",
       "      <th>Année</th>\n",
       "      <th>Note</th>\n",
       "      <th>Structure</th>\n",
       "      <th>Nom UES</th>\n",
       "      <th>Entreprises UES (SIREN)</th>\n",
       "      <th>Région</th>\n",
       "      <th>Département</th>\n",
       "      <th>Pays</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>SAS HERNAS CARTONNAGE</td>\n",
       "      <td>5520176</td>\n",
       "      <td>2019</td>\n",
       "      <td>47.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Somme</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>SAS HERNAS CARTONNAGE</td>\n",
       "      <td>5520176</td>\n",
       "      <td>2020</td>\n",
       "      <td>88.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Somme</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>HERNAS CARTONNAGE</td>\n",
       "      <td>5520176</td>\n",
       "      <td>2021</td>\n",
       "      <td>88.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Somme</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>L'OISEAU BLEU</td>\n",
       "      <td>5620190</td>\n",
       "      <td>2020</td>\n",
       "      <td>94.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Somme</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>L'OISEAU BLEU</td>\n",
       "      <td>5620190</td>\n",
       "      <td>2021</td>\n",
       "      <td>88.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Somme</td>\n",
       "      <td>FRANCE</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16378</th>\n",
       "      <td>SANDVIK MINING AND CONSTRUCTION FRANCE</td>\n",
       "      <td>334537974</td>\n",
       "      <td>2020</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Auvergne-Rhône-Alpes</td>\n",
       "      <td>Rhône</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16379</th>\n",
       "      <td>SANDVIK MINING AND CONSTRUCTION FRANCE SAS</td>\n",
       "      <td>334537974</td>\n",
       "      <td>2021</td>\n",
       "      <td>54.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Auvergne-Rhône-Alpes</td>\n",
       "      <td>Rhône</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16380</th>\n",
       "      <td>pOLYCLINIQUE DE LA THIERACHE</td>\n",
       "      <td>334554623</td>\n",
       "      <td>2019</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Nord</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16381</th>\n",
       "      <td>POLYCLINIQUE DE LA THIERACHE</td>\n",
       "      <td>334554623</td>\n",
       "      <td>2020</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Nord</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16382</th>\n",
       "      <td>GEODIS CALBERSON AQUITAINE</td>\n",
       "      <td>334556537</td>\n",
       "      <td>2018</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Nouvelle-Aquitaine</td>\n",
       "      <td>Gironde</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>16383 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   Raison Sociale      SIREN  Année  Note  \\\n",
       "0                           SAS HERNAS CARTONNAGE    5520176    2019  47.0   \n",
       "1                           SAS HERNAS CARTONNAGE    5520176    2020  88.0   \n",
       "2                               HERNAS CARTONNAGE    5520176    2021  88.0   \n",
       "3                                   L'OISEAU BLEU    5620190    2020  94.0   \n",
       "4                                   L'OISEAU BLEU    5620190    2021  88.0   \n",
       "...                                           ...        ...     ...   ...   \n",
       "16378      SANDVIK MINING AND CONSTRUCTION FRANCE  334537974    2020   NaN   \n",
       "16379  SANDVIK MINING AND CONSTRUCTION FRANCE SAS  334537974    2021  54.0   \n",
       "16380                pOLYCLINIQUE DE LA THIERACHE  334554623    2019   NaN   \n",
       "16381                POLYCLINIQUE DE LA THIERACHE  334554623    2020   NaN   \n",
       "16382                  GEODIS CALBERSON AQUITAINE  334556537    2018   NaN   \n",
       "\n",
       "        Structure  Nom UES Entreprises UES (SIREN)                Région  \\\n",
       "0      Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "1      Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "2      Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "3      Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "4      Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "...           ...      ...                     ...                    ...   \n",
       "16378  Entreprise      NaN                     NaN  Auvergne-Rhône-Alpes   \n",
       "16379  Entreprise      NaN                     NaN  Auvergne-Rhône-Alpes   \n",
       "16380  Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "16381  Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "16382  Entreprise      NaN                     NaN     Nouvelle-Aquitaine   \n",
       "\n",
       "      Département    Pays  \n",
       "0            Somme  FRANCE  \n",
       "1            Somme  FRANCE  \n",
       "2            Somme  FRANCE  \n",
       "3            Somme  FRANCE  \n",
       "4            Somme  FRANCE  \n",
       "...            ...     ...  \n",
       "16378       Rhône  FRANCE  \n",
       "16379       Rhône  FRANCE  \n",
       "16380         Nord  FRANCE  \n",
       "16381         Nord  FRANCE  \n",
       "16382      Gironde  FRANCE  \n",
       "\n",
       "[16383 rows x 10 columns]"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "propre3 = pd.read_excel('index-egalite-fh.xls')\n",
    "propre3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "3a5e9ce5",
   "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>Raison Sociale</th>\n",
       "      <th>SIREN</th>\n",
       "      <th>Année</th>\n",
       "      <th>Note</th>\n",
       "      <th>Structure</th>\n",
       "      <th>Nom UES</th>\n",
       "      <th>Entreprises UES (SIREN)</th>\n",
       "      <th>Région</th>\n",
       "      <th>Département</th>\n",
       "      <th>Pays</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>SAS HERNAS CARTONNAGE</td>\n",
       "      <td>5520176</td>\n",
       "      <td>2019</td>\n",
       "      <td>47.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Somme</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>SAS HERNAS CARTONNAGE</td>\n",
       "      <td>5520176</td>\n",
       "      <td>2020</td>\n",
       "      <td>88.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Somme</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>HERNAS CARTONNAGE</td>\n",
       "      <td>5520176</td>\n",
       "      <td>2021</td>\n",
       "      <td>88.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Somme</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>L'OISEAU BLEU</td>\n",
       "      <td>5620190</td>\n",
       "      <td>2020</td>\n",
       "      <td>94.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Somme</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>L'OISEAU BLEU</td>\n",
       "      <td>5620190</td>\n",
       "      <td>2021</td>\n",
       "      <td>88.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Somme</td>\n",
       "      <td>FRANCE</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16378</th>\n",
       "      <td>SANDVIK MINING AND CONSTRUCTION FRANCE</td>\n",
       "      <td>334537974</td>\n",
       "      <td>2020</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Auvergne-Rhône-Alpes</td>\n",
       "      <td>Rhône</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16379</th>\n",
       "      <td>SANDVIK MINING AND CONSTRUCTION FRANCE SAS</td>\n",
       "      <td>334537974</td>\n",
       "      <td>2021</td>\n",
       "      <td>54.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Auvergne-Rhône-Alpes</td>\n",
       "      <td>Rhône</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16380</th>\n",
       "      <td>pOLYCLINIQUE DE LA THIERACHE</td>\n",
       "      <td>334554623</td>\n",
       "      <td>2019</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Nord</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16381</th>\n",
       "      <td>POLYCLINIQUE DE LA THIERACHE</td>\n",
       "      <td>334554623</td>\n",
       "      <td>2020</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Nord</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16382</th>\n",
       "      <td>GEODIS CALBERSON AQUITAINE</td>\n",
       "      <td>334556537</td>\n",
       "      <td>2018</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Nouvelle-Aquitaine</td>\n",
       "      <td>Gironde</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>16383 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   Raison Sociale      SIREN  Année  Note  \\\n",
       "0                           SAS HERNAS CARTONNAGE    5520176    2019  47.0   \n",
       "1                           SAS HERNAS CARTONNAGE    5520176    2020  88.0   \n",
       "2                               HERNAS CARTONNAGE    5520176    2021  88.0   \n",
       "3                                   L'OISEAU BLEU    5620190    2020  94.0   \n",
       "4                                   L'OISEAU BLEU    5620190    2021  88.0   \n",
       "...                                           ...        ...     ...   ...   \n",
       "16378      SANDVIK MINING AND CONSTRUCTION FRANCE  334537974    2020   NaN   \n",
       "16379  SANDVIK MINING AND CONSTRUCTION FRANCE SAS  334537974    2021  54.0   \n",
       "16380                pOLYCLINIQUE DE LA THIERACHE  334554623    2019   NaN   \n",
       "16381                POLYCLINIQUE DE LA THIERACHE  334554623    2020   NaN   \n",
       "16382                  GEODIS CALBERSON AQUITAINE  334556537    2018   NaN   \n",
       "\n",
       "        Structure  Nom UES Entreprises UES (SIREN)                Région  \\\n",
       "0      Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "1      Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "2      Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "3      Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "4      Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "...           ...      ...                     ...                    ...   \n",
       "16378  Entreprise      NaN                     NaN  Auvergne-Rhône-Alpes   \n",
       "16379  Entreprise      NaN                     NaN  Auvergne-Rhône-Alpes   \n",
       "16380  Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "16381  Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "16382  Entreprise      NaN                     NaN     Nouvelle-Aquitaine   \n",
       "\n",
       "      Département    Pays  \n",
       "0            Somme  FRANCE  \n",
       "1            Somme  FRANCE  \n",
       "2            Somme  FRANCE  \n",
       "3            Somme  FRANCE  \n",
       "4            Somme  FRANCE  \n",
       "...            ...     ...  \n",
       "16378       Rhône  FRANCE  \n",
       "16379       Rhône  FRANCE  \n",
       "16380         Nord  FRANCE  \n",
       "16381         Nord  FRANCE  \n",
       "16382      Gironde  FRANCE  \n",
       "\n",
       "[16383 rows x 10 columns]"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cols = [x for x in range (propre3.shape[1])]\n",
    "propre3 = pd.read_excel('index-egalite-fh.xls', usecols=cols)\n",
    "propre3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "4ab71567",
   "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>Raison Sociale</th>\n",
       "      <th>SIREN</th>\n",
       "      <th>Année</th>\n",
       "      <th>Note</th>\n",
       "      <th>Structure</th>\n",
       "      <th>Nom UES</th>\n",
       "      <th>Entreprises UES (SIREN)</th>\n",
       "      <th>Région</th>\n",
       "      <th>Département</th>\n",
       "      <th>Pays</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [Raison Sociale, SIREN, Année, Note, Structure, Nom UES, Entreprises UES (SIREN), Région, Département, Pays]\n",
       "Index: []"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "propre3.isnull()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "da27db27",
   "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>Raison Sociale</th>\n",
       "      <th>SIREN</th>\n",
       "      <th>Année</th>\n",
       "      <th>Note</th>\n",
       "      <th>Structure</th>\n",
       "      <th>Nom UES</th>\n",
       "      <th>Entreprises UES (SIREN)</th>\n",
       "      <th>Région</th>\n",
       "      <th>Département</th>\n",
       "      <th>Pays</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [Raison Sociale, SIREN, Année, Note, Structure, Nom UES, Entreprises UES (SIREN), Région, Département, Pays]\n",
       "Index: []"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "propre3.dropna(thresh=10, inplace=True)\n",
    "propre3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "9e1d69dd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0        False\n",
       "1        False\n",
       "2        False\n",
       "3        False\n",
       "4        False\n",
       "         ...  \n",
       "16378    False\n",
       "16379    False\n",
       "16380    False\n",
       "16381    False\n",
       "16382    False\n",
       "Length: 16383, dtype: bool"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "propre3.duplicated()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "id": "3eeb3630",
   "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>Raison Sociale</th>\n",
       "      <th>SIREN</th>\n",
       "      <th>Année</th>\n",
       "      <th>Note</th>\n",
       "      <th>Structure</th>\n",
       "      <th>Nom UES</th>\n",
       "      <th>Entreprises UES (SIREN)</th>\n",
       "      <th>Région</th>\n",
       "      <th>Département</th>\n",
       "      <th>Pays</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>SAS HERNAS CARTONNAGE</td>\n",
       "      <td>5520176</td>\n",
       "      <td>2019</td>\n",
       "      <td>47.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Somme</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>SAS HERNAS CARTONNAGE</td>\n",
       "      <td>5520176</td>\n",
       "      <td>2020</td>\n",
       "      <td>88.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Somme</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>HERNAS CARTONNAGE</td>\n",
       "      <td>5520176</td>\n",
       "      <td>2021</td>\n",
       "      <td>88.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Somme</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>L'OISEAU BLEU</td>\n",
       "      <td>5620190</td>\n",
       "      <td>2020</td>\n",
       "      <td>94.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Somme</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>L'OISEAU BLEU</td>\n",
       "      <td>5620190</td>\n",
       "      <td>2021</td>\n",
       "      <td>88.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Somme</td>\n",
       "      <td>FRANCE</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16378</th>\n",
       "      <td>SANDVIK MINING AND CONSTRUCTION FRANCE</td>\n",
       "      <td>334537974</td>\n",
       "      <td>2020</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Auvergne-Rhône-Alpes</td>\n",
       "      <td>Rhône</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16379</th>\n",
       "      <td>SANDVIK MINING AND CONSTRUCTION FRANCE SAS</td>\n",
       "      <td>334537974</td>\n",
       "      <td>2021</td>\n",
       "      <td>54.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Auvergne-Rhône-Alpes</td>\n",
       "      <td>Rhône</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16380</th>\n",
       "      <td>pOLYCLINIQUE DE LA THIERACHE</td>\n",
       "      <td>334554623</td>\n",
       "      <td>2019</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Nord</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16381</th>\n",
       "      <td>POLYCLINIQUE DE LA THIERACHE</td>\n",
       "      <td>334554623</td>\n",
       "      <td>2020</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Nord</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16382</th>\n",
       "      <td>GEODIS CALBERSON AQUITAINE</td>\n",
       "      <td>334556537</td>\n",
       "      <td>2018</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Nouvelle-Aquitaine</td>\n",
       "      <td>Gironde</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>16383 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   Raison Sociale      SIREN  Année  Note  \\\n",
       "0                           SAS HERNAS CARTONNAGE    5520176    2019  47.0   \n",
       "1                           SAS HERNAS CARTONNAGE    5520176    2020  88.0   \n",
       "2                               HERNAS CARTONNAGE    5520176    2021  88.0   \n",
       "3                                   L'OISEAU BLEU    5620190    2020  94.0   \n",
       "4                                   L'OISEAU BLEU    5620190    2021  88.0   \n",
       "...                                           ...        ...     ...   ...   \n",
       "16378      SANDVIK MINING AND CONSTRUCTION FRANCE  334537974    2020   NaN   \n",
       "16379  SANDVIK MINING AND CONSTRUCTION FRANCE SAS  334537974    2021  54.0   \n",
       "16380                pOLYCLINIQUE DE LA THIERACHE  334554623    2019   NaN   \n",
       "16381                POLYCLINIQUE DE LA THIERACHE  334554623    2020   NaN   \n",
       "16382                  GEODIS CALBERSON AQUITAINE  334556537    2018   NaN   \n",
       "\n",
       "        Structure  Nom UES Entreprises UES (SIREN)                Région  \\\n",
       "0      Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "1      Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "2      Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "3      Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "4      Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "...           ...      ...                     ...                    ...   \n",
       "16378  Entreprise      NaN                     NaN  Auvergne-Rhône-Alpes   \n",
       "16379  Entreprise      NaN                     NaN  Auvergne-Rhône-Alpes   \n",
       "16380  Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "16381  Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "16382  Entreprise      NaN                     NaN     Nouvelle-Aquitaine   \n",
       "\n",
       "      Département    Pays  \n",
       "0            Somme  FRANCE  \n",
       "1            Somme  FRANCE  \n",
       "2            Somme  FRANCE  \n",
       "3            Somme  FRANCE  \n",
       "4            Somme  FRANCE  \n",
       "...            ...     ...  \n",
       "16378       Rhône  FRANCE  \n",
       "16379       Rhône  FRANCE  \n",
       "16380         Nord  FRANCE  \n",
       "16381         Nord  FRANCE  \n",
       "16382      Gironde  FRANCE  \n",
       "\n",
       "[16383 rows x 10 columns]"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "propre3.drop_duplicates()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "98df451a",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/td/4cpgzdjd7fl57lzw14shn36m0000gn/T/ipykernel_20829/336442086.py:2: FutureWarning: The default value of regex will change from True to False in a future version. In addition, single character regular expressions will *not* be treated as literal strings when regex=True.\n",
      "  propre2[\"Code_Dechet\"] = propre2[\"Code_Dechet\"].str.replace(\"*\", \"\")\n"
     ]
    },
    {
     "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>Siren</th>\n",
       "      <th>Nom_Etablissement</th>\n",
       "      <th>Annee</th>\n",
       "      <th>Code_Operation_Eliminatio_Valorisation</th>\n",
       "      <th>Libelle_Operation_Eliminatio_Valorisation</th>\n",
       "      <th>Code_Departement</th>\n",
       "      <th>Code_Dechet</th>\n",
       "      <th>Libelle_Dechet</th>\n",
       "      <th>Quantite</th>\n",
       "      <th>Unite</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3012500</td>\n",
       "      <td>ESTIA RESEAUX DE CHALEUR</td>\n",
       "      <td>2020</td>\n",
       "      <td>D10</td>\n",
       "      <td>D10 Incinération à terre</td>\n",
       "      <td>57</td>\n",
       "      <td>11 01 11</td>\n",
       "      <td>Liquides aqueux de rinçage contenant des subs...</td>\n",
       "      <td>3.000</td>\n",
       "      <td>tonnes/an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3012500</td>\n",
       "      <td>ESTIA RESEAUX DE CHALEUR</td>\n",
       "      <td>2020</td>\n",
       "      <td>R1</td>\n",
       "      <td>R1 Utilisation principale comme combustible ou...</td>\n",
       "      <td>37</td>\n",
       "      <td>13 02 08</td>\n",
       "      <td>Autres huiles moteur, de boîte de vitesses et...</td>\n",
       "      <td>2.000</td>\n",
       "      <td>tonnes/an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3012500</td>\n",
       "      <td>ESTIA RESEAUX DE CHALEUR</td>\n",
       "      <td>2020</td>\n",
       "      <td>R13</td>\n",
       "      <td>R13 Stockage de déchets préalablement à l'u...</td>\n",
       "      <td>37</td>\n",
       "      <td>16 10 01</td>\n",
       "      <td>Déchets liquides aqueux contenant des substan...</td>\n",
       "      <td>0.494</td>\n",
       "      <td>tonnes/an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3012500</td>\n",
       "      <td>ESTIA RESEAUX DE CHALEUR</td>\n",
       "      <td>2020</td>\n",
       "      <td>R3</td>\n",
       "      <td>R3 Recyclage ou récupération des substances ...</td>\n",
       "      <td>45</td>\n",
       "      <td>15 01 10</td>\n",
       "      <td>Emballages contenant des résidus de substance...</td>\n",
       "      <td>0.035</td>\n",
       "      <td>tonnes/an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3012500</td>\n",
       "      <td>ESTIA RESEAUX DE CHALEUR</td>\n",
       "      <td>2020</td>\n",
       "      <td>R4</td>\n",
       "      <td>R4 Recyclage ou récupération des métaux et ...</td>\n",
       "      <td>57</td>\n",
       "      <td>16 06 03</td>\n",
       "      <td>Piles contenant du mercure</td>\n",
       "      <td>0.093</td>\n",
       "      <td>tonnes/an</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65530</th>\n",
       "      <td>9000029</td>\n",
       "      <td>BA 709 Cognac</td>\n",
       "      <td>2020</td>\n",
       "      <td>R13</td>\n",
       "      <td>R13 Stockage de déchets préalablement à l'u...</td>\n",
       "      <td>16</td>\n",
       "      <td>13 05 06</td>\n",
       "      <td>Hydrocarbures provenant de séparateurs eau/hy...</td>\n",
       "      <td>1.000</td>\n",
       "      <td>tonnes/an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65531</th>\n",
       "      <td>9000029</td>\n",
       "      <td>BA 709 Cognac</td>\n",
       "      <td>2020</td>\n",
       "      <td>R13</td>\n",
       "      <td>R13 Stockage de déchets préalablement à l'u...</td>\n",
       "      <td>16</td>\n",
       "      <td>15 01 10</td>\n",
       "      <td>Emballages contenant des résidus de substance...</td>\n",
       "      <td>0.360</td>\n",
       "      <td>tonnes/an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65532</th>\n",
       "      <td>9000029</td>\n",
       "      <td>BA 709 Cognac</td>\n",
       "      <td>2020</td>\n",
       "      <td>R13</td>\n",
       "      <td>R13 Stockage de déchets préalablement à l'u...</td>\n",
       "      <td>16</td>\n",
       "      <td>15 02 02</td>\n",
       "      <td>Absorbants, matériaux filtrants (y compris le...</td>\n",
       "      <td>0.510</td>\n",
       "      <td>tonnes/an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65533</th>\n",
       "      <td>9000029</td>\n",
       "      <td>BA 709 Cognac</td>\n",
       "      <td>2020</td>\n",
       "      <td>R13</td>\n",
       "      <td>R13 Stockage de déchets préalablement à l'u...</td>\n",
       "      <td>16</td>\n",
       "      <td>16 01 07</td>\n",
       "      <td>Filtres à huile</td>\n",
       "      <td>0.810</td>\n",
       "      <td>tonnes/an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65534</th>\n",
       "      <td>9000029</td>\n",
       "      <td>BA 709 Cognac</td>\n",
       "      <td>2020</td>\n",
       "      <td>R13</td>\n",
       "      <td>R13 Stockage de déchets préalablement à l'u...</td>\n",
       "      <td>16</td>\n",
       "      <td>16 01 14</td>\n",
       "      <td>Antigels contenant des substances dangereuses</td>\n",
       "      <td>1.000</td>\n",
       "      <td>tonnes/an</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>64100 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         Siren         Nom_Etablissement  Annee  \\\n",
       "0      3012500  ESTIA RESEAUX DE CHALEUR   2020   \n",
       "1      3012500  ESTIA RESEAUX DE CHALEUR   2020   \n",
       "2      3012500  ESTIA RESEAUX DE CHALEUR   2020   \n",
       "3      3012500  ESTIA RESEAUX DE CHALEUR   2020   \n",
       "4      3012500  ESTIA RESEAUX DE CHALEUR   2020   \n",
       "...        ...                       ...    ...   \n",
       "65530  9000029             BA 709 Cognac   2020   \n",
       "65531  9000029             BA 709 Cognac   2020   \n",
       "65532  9000029             BA 709 Cognac   2020   \n",
       "65533  9000029             BA 709 Cognac   2020   \n",
       "65534  9000029             BA 709 Cognac   2020   \n",
       "\n",
       "      Code_Operation_Eliminatio_Valorisation  \\\n",
       "0                                        D10   \n",
       "1                                        R1    \n",
       "2                                        R13   \n",
       "3                                        R3    \n",
       "4                                        R4    \n",
       "...                                      ...   \n",
       "65530                                    R13   \n",
       "65531                                    R13   \n",
       "65532                                    R13   \n",
       "65533                                    R13   \n",
       "65534                                    R13   \n",
       "\n",
       "               Libelle_Operation_Eliminatio_Valorisation Code_Departement  \\\n",
       "0                             D10 Incinération à terre               57   \n",
       "1      R1 Utilisation principale comme combustible ou...               37   \n",
       "2      R13 Stockage de déchets préalablement à l'u...               37   \n",
       "3      R3 Recyclage ou récupération des substances ...               45   \n",
       "4      R4 Recyclage ou récupération des métaux et ...               57   \n",
       "...                                                  ...              ...   \n",
       "65530  R13 Stockage de déchets préalablement à l'u...               16   \n",
       "65531  R13 Stockage de déchets préalablement à l'u...               16   \n",
       "65532  R13 Stockage de déchets préalablement à l'u...               16   \n",
       "65533  R13 Stockage de déchets préalablement à l'u...               16   \n",
       "65534  R13 Stockage de déchets préalablement à l'u...               16   \n",
       "\n",
       "      Code_Dechet                                     Libelle_Dechet  \\\n",
       "0        11 01 11  Liquides aqueux de rinçage contenant des subs...   \n",
       "1        13 02 08  Autres huiles moteur, de boîte de vitesses et...   \n",
       "2        16 10 01  Déchets liquides aqueux contenant des substan...   \n",
       "3        15 01 10  Emballages contenant des résidus de substance...   \n",
       "4        16 06 03                         Piles contenant du mercure   \n",
       "...           ...                                                ...   \n",
       "65530    13 05 06  Hydrocarbures provenant de séparateurs eau/hy...   \n",
       "65531    15 01 10  Emballages contenant des résidus de substance...   \n",
       "65532    15 02 02  Absorbants, matériaux filtrants (y compris le...   \n",
       "65533    16 01 07                                   Filtres à huile   \n",
       "65534    16 01 14      Antigels contenant des substances dangereuses   \n",
       "\n",
       "       Quantite      Unite  \n",
       "0         3.000  tonnes/an  \n",
       "1         2.000  tonnes/an  \n",
       "2         0.494  tonnes/an  \n",
       "3         0.035  tonnes/an  \n",
       "4         0.093  tonnes/an  \n",
       "...         ...        ...  \n",
       "65530     1.000  tonnes/an  \n",
       "65531     0.360  tonnes/an  \n",
       "65532     0.510  tonnes/an  \n",
       "65533     0.810  tonnes/an  \n",
       "65534     1.000  tonnes/an  \n",
       "\n",
       "[64100 rows x 10 columns]"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#propre2[\"Hommes 15-24\"] = chom[\"Hommes 15-24\"].astype(str)\n",
    "propre2[\"Code_Dechet\"] = propre2[\"Code_Dechet\"].str.replace(\"*\", \"\")\n",
    "propre2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "a0531825",
   "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>Raison Sociale</th>\n",
       "      <th>SIREN</th>\n",
       "      <th>Année</th>\n",
       "      <th>Note</th>\n",
       "      <th>Structure</th>\n",
       "      <th>Nom UES</th>\n",
       "      <th>Entreprises UES (SIREN)</th>\n",
       "      <th>Région</th>\n",
       "      <th>Département</th>\n",
       "      <th>Pays</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>SAS HERNAS CARTONNAGE</td>\n",
       "      <td>5520176</td>\n",
       "      <td>2019</td>\n",
       "      <td>47.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Somme</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>SAS HERNAS CARTONNAGE</td>\n",
       "      <td>5520176</td>\n",
       "      <td>2020</td>\n",
       "      <td>88.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Somme</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>HERNAS CARTONNAGE</td>\n",
       "      <td>5520176</td>\n",
       "      <td>2021</td>\n",
       "      <td>88.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Somme</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>L'OISEAU BLEU</td>\n",
       "      <td>5620190</td>\n",
       "      <td>2020</td>\n",
       "      <td>94.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Somme</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>L'OISEAU BLEU</td>\n",
       "      <td>5620190</td>\n",
       "      <td>2021</td>\n",
       "      <td>88.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Somme</td>\n",
       "      <td>FRANCE</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16378</th>\n",
       "      <td>SANDVIK MINING AND CONSTRUCTION FRANCE</td>\n",
       "      <td>334537974</td>\n",
       "      <td>2020</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Auvergne-Rhône-Alpes</td>\n",
       "      <td>Rhône</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16379</th>\n",
       "      <td>SANDVIK MINING AND CONSTRUCTION FRANCE SAS</td>\n",
       "      <td>334537974</td>\n",
       "      <td>2021</td>\n",
       "      <td>54.0</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Auvergne-Rhône-Alpes</td>\n",
       "      <td>Rhône</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16380</th>\n",
       "      <td>pOLYCLINIQUE DE LA THIERACHE</td>\n",
       "      <td>334554623</td>\n",
       "      <td>2019</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Nord</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16381</th>\n",
       "      <td>POLYCLINIQUE DE LA THIERACHE</td>\n",
       "      <td>334554623</td>\n",
       "      <td>2020</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hauts-de-France</td>\n",
       "      <td>Nord</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16382</th>\n",
       "      <td>GEODIS CALBERSON AQUITAINE</td>\n",
       "      <td>334556537</td>\n",
       "      <td>2018</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Entreprise</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Nouvelle-Aquitaine</td>\n",
       "      <td>Gironde</td>\n",
       "      <td>FRANCE</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>16383 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   Raison Sociale      SIREN  Année  Note  \\\n",
       "0                           SAS HERNAS CARTONNAGE    5520176    2019  47.0   \n",
       "1                           SAS HERNAS CARTONNAGE    5520176    2020  88.0   \n",
       "2                               HERNAS CARTONNAGE    5520176    2021  88.0   \n",
       "3                                   L'OISEAU BLEU    5620190    2020  94.0   \n",
       "4                                   L'OISEAU BLEU    5620190    2021  88.0   \n",
       "...                                           ...        ...     ...   ...   \n",
       "16378      SANDVIK MINING AND CONSTRUCTION FRANCE  334537974    2020   NaN   \n",
       "16379  SANDVIK MINING AND CONSTRUCTION FRANCE SAS  334537974    2021  54.0   \n",
       "16380                pOLYCLINIQUE DE LA THIERACHE  334554623    2019   NaN   \n",
       "16381                POLYCLINIQUE DE LA THIERACHE  334554623    2020   NaN   \n",
       "16382                  GEODIS CALBERSON AQUITAINE  334556537    2018   NaN   \n",
       "\n",
       "        Structure  Nom UES Entreprises UES (SIREN)                Région  \\\n",
       "0      Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "1      Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "2      Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "3      Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "4      Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "...           ...      ...                     ...                    ...   \n",
       "16378  Entreprise      NaN                     NaN  Auvergne-Rhône-Alpes   \n",
       "16379  Entreprise      NaN                     NaN  Auvergne-Rhône-Alpes   \n",
       "16380  Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "16381  Entreprise      NaN                     NaN        Hauts-de-France   \n",
       "16382  Entreprise      NaN                     NaN     Nouvelle-Aquitaine   \n",
       "\n",
       "      Département    Pays  \n",
       "0            Somme  FRANCE  \n",
       "1            Somme  FRANCE  \n",
       "2            Somme  FRANCE  \n",
       "3            Somme  FRANCE  \n",
       "4            Somme  FRANCE  \n",
       "...            ...     ...  \n",
       "16378       Rhône  FRANCE  \n",
       "16379       Rhône  FRANCE  \n",
       "16380         Nord  FRANCE  \n",
       "16381         Nord  FRANCE  \n",
       "16382      Gironde  FRANCE  \n",
       "\n",
       "[16383 rows x 10 columns]"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "propre3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "20b51a16",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/td/4cpgzdjd7fl57lzw14shn36m0000gn/T/ipykernel_20829/2049097812.py:2: FutureWarning: As the xlwt package is no longer maintained, the xlwt engine will be removed in a future version of pandas. This is the only engine in pandas that supports writing in the xls format. Install openpyxl and write to an xlsx file instead. You can set the option io.excel.xls.writer to 'xlwt' to silence this warning. While this option is deprecated and will also raise a warning, it can be globally set and the warning suppressed.\n",
      "  propre2.to_excel(file_name)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "DataFrame is written to Excel File successfully.\n"
     ]
    }
   ],
   "source": [
    "file_name = 'ProdProp_Dechets_Dangereux.xls'\n",
    "propre2.to_excel(file_name)\n",
    "print('DataFrame is written to Excel File successfully.')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "7bf33a3e",
   "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>Siren</th>\n",
       "      <th>Nom_Etablissement</th>\n",
       "      <th>Annee</th>\n",
       "      <th>Code_Operation_Eliminatio_Valorisation</th>\n",
       "      <th>Libelle_Operation_Eliminatio_Valorisation</th>\n",
       "      <th>Code_Departement</th>\n",
       "      <th>Code_Dechet</th>\n",
       "      <th>Libelle_Dechet</th>\n",
       "      <th>Quantite</th>\n",
       "      <th>Unite</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3012500</td>\n",
       "      <td>ESTIA RESEAUX DE CHALEUR</td>\n",
       "      <td>2020</td>\n",
       "      <td>D10</td>\n",
       "      <td>D10 Incinération à terre</td>\n",
       "      <td>57</td>\n",
       "      <td>11 01 11*</td>\n",
       "      <td>Liquides aqueux de rinçage contenant des subs...</td>\n",
       "      <td>3.000</td>\n",
       "      <td>tonnes/an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3012500</td>\n",
       "      <td>ESTIA RESEAUX DE CHALEUR</td>\n",
       "      <td>2020</td>\n",
       "      <td>R1</td>\n",
       "      <td>R1 Utilisation principale comme combustible ou...</td>\n",
       "      <td>37</td>\n",
       "      <td>13 02 08*</td>\n",
       "      <td>Autres huiles moteur, de boîte de vitesses et...</td>\n",
       "      <td>2.000</td>\n",
       "      <td>tonnes/an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3012500</td>\n",
       "      <td>ESTIA RESEAUX DE CHALEUR</td>\n",
       "      <td>2020</td>\n",
       "      <td>R13</td>\n",
       "      <td>R13 Stockage de déchets préalablement à l'u...</td>\n",
       "      <td>37</td>\n",
       "      <td>16 10 01*</td>\n",
       "      <td>Déchets liquides aqueux contenant des substan...</td>\n",
       "      <td>0.494</td>\n",
       "      <td>tonnes/an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3012500</td>\n",
       "      <td>ESTIA RESEAUX DE CHALEUR</td>\n",
       "      <td>2020</td>\n",
       "      <td>R3</td>\n",
       "      <td>R3 Recyclage ou récupération des substances ...</td>\n",
       "      <td>45</td>\n",
       "      <td>15 01 10*</td>\n",
       "      <td>Emballages contenant des résidus de substance...</td>\n",
       "      <td>0.035</td>\n",
       "      <td>tonnes/an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3012500</td>\n",
       "      <td>ESTIA RESEAUX DE CHALEUR</td>\n",
       "      <td>2020</td>\n",
       "      <td>R4</td>\n",
       "      <td>R4 Recyclage ou récupération des métaux et ...</td>\n",
       "      <td>57</td>\n",
       "      <td>16 06 03*</td>\n",
       "      <td>Piles contenant du mercure</td>\n",
       "      <td>0.093</td>\n",
       "      <td>tonnes/an</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65530</th>\n",
       "      <td>9000029</td>\n",
       "      <td>BA 709 Cognac</td>\n",
       "      <td>2020</td>\n",
       "      <td>R13</td>\n",
       "      <td>R13 Stockage de déchets préalablement à l'u...</td>\n",
       "      <td>16</td>\n",
       "      <td>13 05 06*</td>\n",
       "      <td>Hydrocarbures provenant de séparateurs eau/hy...</td>\n",
       "      <td>1.000</td>\n",
       "      <td>tonnes/an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65531</th>\n",
       "      <td>9000029</td>\n",
       "      <td>BA 709 Cognac</td>\n",
       "      <td>2020</td>\n",
       "      <td>R13</td>\n",
       "      <td>R13 Stockage de déchets préalablement à l'u...</td>\n",
       "      <td>16</td>\n",
       "      <td>15 01 10*</td>\n",
       "      <td>Emballages contenant des résidus de substance...</td>\n",
       "      <td>0.360</td>\n",
       "      <td>tonnes/an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65532</th>\n",
       "      <td>9000029</td>\n",
       "      <td>BA 709 Cognac</td>\n",
       "      <td>2020</td>\n",
       "      <td>R13</td>\n",
       "      <td>R13 Stockage de déchets préalablement à l'u...</td>\n",
       "      <td>16</td>\n",
       "      <td>15 02 02*</td>\n",
       "      <td>Absorbants, matériaux filtrants (y compris le...</td>\n",
       "      <td>0.510</td>\n",
       "      <td>tonnes/an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65533</th>\n",
       "      <td>9000029</td>\n",
       "      <td>BA 709 Cognac</td>\n",
       "      <td>2020</td>\n",
       "      <td>R13</td>\n",
       "      <td>R13 Stockage de déchets préalablement à l'u...</td>\n",
       "      <td>16</td>\n",
       "      <td>16 01 07*</td>\n",
       "      <td>Filtres à huile</td>\n",
       "      <td>0.810</td>\n",
       "      <td>tonnes/an</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65534</th>\n",
       "      <td>9000029</td>\n",
       "      <td>BA 709 Cognac</td>\n",
       "      <td>2020</td>\n",
       "      <td>R13</td>\n",
       "      <td>R13 Stockage de déchets préalablement à l'u...</td>\n",
       "      <td>16</td>\n",
       "      <td>16 01 14*</td>\n",
       "      <td>Antigels contenant des substances dangereuses</td>\n",
       "      <td>1.000</td>\n",
       "      <td>tonnes/an</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>64100 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         Siren         Nom_Etablissement  Annee  \\\n",
       "0      3012500  ESTIA RESEAUX DE CHALEUR   2020   \n",
       "1      3012500  ESTIA RESEAUX DE CHALEUR   2020   \n",
       "2      3012500  ESTIA RESEAUX DE CHALEUR   2020   \n",
       "3      3012500  ESTIA RESEAUX DE CHALEUR   2020   \n",
       "4      3012500  ESTIA RESEAUX DE CHALEUR   2020   \n",
       "...        ...                       ...    ...   \n",
       "65530  9000029             BA 709 Cognac   2020   \n",
       "65531  9000029             BA 709 Cognac   2020   \n",
       "65532  9000029             BA 709 Cognac   2020   \n",
       "65533  9000029             BA 709 Cognac   2020   \n",
       "65534  9000029             BA 709 Cognac   2020   \n",
       "\n",
       "      Code_Operation_Eliminatio_Valorisation  \\\n",
       "0                                        D10   \n",
       "1                                        R1    \n",
       "2                                        R13   \n",
       "3                                        R3    \n",
       "4                                        R4    \n",
       "...                                      ...   \n",
       "65530                                    R13   \n",
       "65531                                    R13   \n",
       "65532                                    R13   \n",
       "65533                                    R13   \n",
       "65534                                    R13   \n",
       "\n",
       "               Libelle_Operation_Eliminatio_Valorisation Code_Departement  \\\n",
       "0                             D10 Incinération à terre               57   \n",
       "1      R1 Utilisation principale comme combustible ou...               37   \n",
       "2      R13 Stockage de déchets préalablement à l'u...               37   \n",
       "3      R3 Recyclage ou récupération des substances ...               45   \n",
       "4      R4 Recyclage ou récupération des métaux et ...               57   \n",
       "...                                                  ...              ...   \n",
       "65530  R13 Stockage de déchets préalablement à l'u...               16   \n",
       "65531  R13 Stockage de déchets préalablement à l'u...               16   \n",
       "65532  R13 Stockage de déchets préalablement à l'u...               16   \n",
       "65533  R13 Stockage de déchets préalablement à l'u...               16   \n",
       "65534  R13 Stockage de déchets préalablement à l'u...               16   \n",
       "\n",
       "      Code_Dechet                                     Libelle_Dechet  \\\n",
       "0       11 01 11*  Liquides aqueux de rinçage contenant des subs...   \n",
       "1       13 02 08*  Autres huiles moteur, de boîte de vitesses et...   \n",
       "2       16 10 01*  Déchets liquides aqueux contenant des substan...   \n",
       "3       15 01 10*  Emballages contenant des résidus de substance...   \n",
       "4       16 06 03*                         Piles contenant du mercure   \n",
       "...           ...                                                ...   \n",
       "65530   13 05 06*  Hydrocarbures provenant de séparateurs eau/hy...   \n",
       "65531   15 01 10*  Emballages contenant des résidus de substance...   \n",
       "65532   15 02 02*  Absorbants, matériaux filtrants (y compris le...   \n",
       "65533   16 01 07*                                   Filtres à huile   \n",
       "65534   16 01 14*      Antigels contenant des substances dangereuses   \n",
       "\n",
       "       Quantite      Unite  \n",
       "0         3.000  tonnes/an  \n",
       "1         2.000  tonnes/an  \n",
       "2         0.494  tonnes/an  \n",
       "3         0.035  tonnes/an  \n",
       "4         0.093  tonnes/an  \n",
       "...         ...        ...  \n",
       "65530     1.000  tonnes/an  \n",
       "65531     0.360  tonnes/an  \n",
       "65532     0.510  tonnes/an  \n",
       "65533     0.810  tonnes/an  \n",
       "65534     1.000  tonnes/an  \n",
       "\n",
       "[64100 rows x 10 columns]"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "propre2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fbfa5f57",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}