{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Sujet 1 : Concentration de CO2 dans l'atmosphère depuis 1958"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Charles David Keeling a lancé une campagne de mesure de la concentration de C02 dans l'atmosphère. Il a installé ces instrument à l'observatoire de Mauna Loa, Hawaii, Etats-Unis. Depuis 1958, nous avons continuellement des données.\n",
    "\n",
    "L'étude initiale devait étudier les variations saisonnière de la concentration, mais avec le réchauffement climatique, elle se tourne maintenant sur la croissance de la concentration.\n",
    "\n",
    "A partir des données hebdomadaires disponible sur le [site Web de l'institut Scripps](https://scrippsco2.ucsd.edu/data/atmospheric_co2/primary_mlo_co2_record.html), nous souhaitons reproduire l'analyse de l'évolution de la concentration de C02 dans l'atmosphère pour faire un modèle prédictif."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Environnement de travail"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nous définions quelques fonctions pour faciliter l'affichage des numéros de version associés à notre système et à nos modules. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "hidePrompt": false
   },
   "outputs": [],
   "source": [
    "def print_imported_modules():\n",
    "    import sys\n",
    "    print(\"Imported modules\")\n",
    "    for name, val in sorted(sys.modules.items()):\n",
    "        if(hasattr(val, '__version__')): \n",
    "            print(\"\\t\",val.__name__, val.__version__)\n",
    "            \n",
    "def print_sys_info():\n",
    "    import sys\n",
    "    import platform\n",
    "    print(\"System Info\")\n",
    "    print(\"\\t\",sys.version)\n",
    "    print(\"\\t\",platform.uname())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nous utilisons les modules usuels en traitement des données sous le langage python3 à la date du *6 Avril 2020* : numpy, pandas, seaborn, matplotlib, statsmodels, ...   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import statsmodels.api as sm\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ci-aprés un aperçu de notre environnement d'execution pour les personnes qui souhaiteraient reproduire ces travaux sur leur machine."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "System Info\n",
      "\t 3.6.4 |Anaconda, Inc.| (default, Mar 13 2018, 01:15:57) \n",
      "[GCC 7.2.0]\n",
      "\t uname_result(system='Linux', node='dc160b48d9b7', release='4.4.0-164-generic', version='#192-Ubuntu SMP Fri Sep 13 12:02:50 UTC 2019', machine='x86_64', processor='x86_64')\n",
      "Imported modules\n",
      "\t IPython 7.12.0\n",
      "\t IPython.core.release 7.12.0\n",
      "\t PIL 7.0.0\n",
      "\t PIL.Image 7.0.0\n",
      "\t PIL._version 7.0.0\n",
      "\t _csv 1.0\n",
      "\t _ctypes 1.1.0\n",
      "\t _curses b'2.2'\n",
      "\t decimal 1.70\n",
      "\t argparse 1.1\n",
      "\t backcall 0.1.0\n",
      "\t cffi 1.13.2\n",
      "\t csv 1.0\n",
      "\t ctypes 1.1.0\n",
      "\t cycler 0.10.0\n",
      "\t dateutil 2.8.1\n",
      "\t decimal 1.70\n",
      "\t decorator 4.4.1\n",
      "\t distutils 3.6.4\n",
      "\t ipaddress 1.0\n",
      "\t ipykernel 5.1.4\n",
      "\t ipykernel._version 5.1.4\n",
      "\t ipython_genutils 0.2.0\n",
      "\t ipython_genutils._version 0.2.0\n",
      "\t ipywidgets 7.2.1\n",
      "\t ipywidgets._version 7.2.1\n",
      "\t jedi 0.16.0\n",
      "\t json 2.0.9\n",
      "\t jupyter_client 6.0.0\n",
      "\t jupyter_client._version 6.0.0\n",
      "\t jupyter_core 4.6.3\n",
      "\t jupyter_core.version 4.6.3\n",
      "\t kiwisolver 1.1.0\n",
      "\t logging 0.5.1.2\n",
      "\t matplotlib 2.2.3\n",
      "\t matplotlib.backends.backend_agg 2.2.3\n",
      "\t numpy 1.15.2\n",
      "\t numpy.core 1.15.2\n",
      "\t numpy.core.multiarray 3.1\n",
      "\t numpy.lib 1.15.2\n",
      "\t numpy.linalg._umath_linalg b'0.1.5'\n",
      "\t numpy.matlib 1.15.2\n",
      "\t optparse 1.5.3\n",
      "\t pandas 0.22.0\n",
      "\t _libjson 1.33\n",
      "\t parso 0.6.0\n",
      "\t patsy 0.5.1\n",
      "\t patsy.version 0.5.1\n",
      "\t pexpect 4.8.0\n",
      "\t pickleshare 0.7.5\n",
      "\t platform 1.0.8\n",
      "\t prompt_toolkit 3.0.3\n",
      "\t ptyprocess 0.6.0\n",
      "\t pygments 2.5.2\n",
      "\t pyparsing 2.4.6\n",
      "\t pytz 2019.3\n",
      "\t re 2.2.1\n",
      "\t scipy 1.1.0\n",
      "\t scipy._lib.decorator 4.0.5\n",
      "\t scipy._lib.six 1.2.0\n",
      "\t scipy.fftpack._fftpack b'$Revision: $'\n",
      "\t scipy.fftpack.convolve b'$Revision: $'\n",
      "\t scipy.integrate._dop b'$Revision: $'\n",
      "\t scipy.integrate._ode $Id$\n",
      "\t scipy.integrate._odepack  1.9 \n",
      "\t scipy.integrate._quadpack  1.13 \n",
      "\t scipy.integrate.lsoda b'$Revision: $'\n",
      "\t scipy.integrate.vode b'$Revision: $'\n",
      "\t scipy.interpolate._fitpack  1.7 \n",
      "\t scipy.interpolate.dfitpack b'$Revision: $'\n",
      "\t scipy.linalg 0.4.9\n",
      "\t scipy.linalg._fblas b'$Revision: $'\n",
      "\t scipy.linalg._flapack b'$Revision: $'\n",
      "\t scipy.linalg._flinalg b'$Revision: $'\n",
      "\t scipy.ndimage 2.0\n",
      "\t scipy.optimize._cobyla b'$Revision: $'\n",
      "\t scipy.optimize._lbfgsb b'$Revision: $'\n",
      "\t scipy.optimize._minpack  1.10 \n",
      "\t scipy.optimize._nnls b'$Revision: $'\n",
      "\t scipy.optimize._slsqp b'$Revision: $'\n",
      "\t scipy.optimize.minpack2 b'$Revision: $'\n",
      "\t scipy.signal.spline 0.2\n",
      "\t scipy.sparse.linalg.eigen.arpack._arpack b'$Revision: $'\n",
      "\t scipy.sparse.linalg.isolve._iterative b'$Revision: $'\n",
      "\t scipy.special.specfun b'$Revision: $'\n",
      "\t scipy.stats.mvn b'$Revision: $'\n",
      "\t scipy.stats.statlib b'$Revision: $'\n",
      "\t seaborn 0.8.1\n",
      "\t seaborn.external.husl 2.1.0\n",
      "\t seaborn.external.six 1.10.0\n",
      "\t six 1.14.0\n",
      "\t statsmodels 0.9.0\n",
      "\t statsmodels.__init__ 0.9.0\n",
      "\t traitlets 4.3.3\n",
      "\t traitlets._version 4.3.3\n",
      "\t urllib.request 3.6\n",
      "\t zlib 1.0\n",
      "\t zmq 17.1.2\n",
      "\t zmq.sugar 17.1.2\n",
      "\t zmq.sugar.version 17.1.2\n"
     ]
    }
   ],
   "source": [
    "print_sys_info()\n",
    "print_imported_modules()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Chargement et inspection des données"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nous avons récupérer les données hebdomadaire le *6 Avril 2020* depuis le lien suivant : [https://scrippsco2.ucsd.edu/assets/data/atmospheric/stations/in_situ_co2/weekly/weekly_in_situ_co2_mlo.csv](https://scrippsco2.ucsd.edu/assets/data/atmospheric/stations/in_situ_co2/weekly/weekly_in_situ_co2_mlo.csv)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "filename = \"./weekly_in_situ_co2_mlo.csv\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nous affichons les premières lignes du fichier pour repérer d'éventuelle lignes à ignorer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ligne 0 : \"-------------------------------------------------------------------------------------------\"\n",
      "Ligne 1 : \" Atmospheric CO2 concentrations (ppm) derived from in situ air measurements                \"\n",
      "Ligne 2 : \" at Mauna Loa, Observatory, Hawaii: Latitude 19.5°N Longitude 155.6°W Elevation 3397m      \"\n",
      "Ligne 3 : \"                                                                                           \"\n",
      "Ligne 4 : \" Source: R. F. Keeling, S. J. Walker, S. C. Piper and A. F. Bollenbacher                   \"\n"
     ]
    }
   ],
   "source": [
    "def head(filename,n):\n",
    "    with open(filename,\"r\") as f:\n",
    "        lignes = f.readlines()\n",
    "    n = min(n,len(lignes))\n",
    "    for i,ligne in enumerate(lignes[:n]):\n",
    "        print(\"Ligne\",i,\":\",ligne,end=\"\")\n",
    "head(filename,5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Le fichier semble être correctement formaté :\n",
    "* Les lignes de commentaire/metadonnée commencent par \"\n",
    "* Les données ne commencent pas par \"\n",
    "\n",
    "Trouvons donc la première ligne de données."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "44"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def find_num_first_dataline(filename):\n",
    "    with open(filename,\"r\") as f:\n",
    "        lignes = f.readlines()\n",
    "    for i,ligne in enumerate(lignes):\n",
    "        if ligne[0] != '\"':\n",
    "            return i\n",
    "    raise Exception(\"No dataline found\")\n",
    "find_num_first_dataline(filename)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Aprés une inspection visuelle, nous avons aussi trouvé que les données commence ligne 44.\n",
    "Lors de cette inspection, nous avons pu relever les informations suivantes :\n",
    "1. La première colonne correspond aux dates d'acquisition\n",
    "2. Les données sont centrées sur 12h00 chaque jour\n",
    "3. La seconde colonne correspond aux concentrations mesurées\n",
    "4. La concentration est la concentration moyenne de C02 dans l'atomosphère de la journée"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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>Concentration</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1958-03-29</td>\n",
       "      <td>316.19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1958-04-05</td>\n",
       "      <td>317.31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1958-04-12</td>\n",
       "      <td>317.69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1958-04-19</td>\n",
       "      <td>317.58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1958-04-26</td>\n",
       "      <td>316.48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1958-05-03</td>\n",
       "      <td>316.95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1958-05-17</td>\n",
       "      <td>317.56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1958-05-24</td>\n",
       "      <td>317.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1958-07-05</td>\n",
       "      <td>315.85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1958-07-12</td>\n",
       "      <td>315.85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1958-07-19</td>\n",
       "      <td>315.46</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1958-07-26</td>\n",
       "      <td>315.59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>1958-08-02</td>\n",
       "      <td>315.64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>1958-08-09</td>\n",
       "      <td>315.10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>1958-08-16</td>\n",
       "      <td>315.09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>1958-08-30</td>\n",
       "      <td>314.14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1958-09-06</td>\n",
       "      <td>313.54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>1958-11-08</td>\n",
       "      <td>313.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>1958-11-15</td>\n",
       "      <td>313.26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>1958-11-22</td>\n",
       "      <td>313.57</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>1958-11-29</td>\n",
       "      <td>314.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>1958-12-06</td>\n",
       "      <td>314.56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>1958-12-13</td>\n",
       "      <td>314.41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>1958-12-20</td>\n",
       "      <td>314.77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>1958-12-27</td>\n",
       "      <td>315.21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>1959-01-03</td>\n",
       "      <td>315.24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>1959-01-10</td>\n",
       "      <td>315.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>1959-01-17</td>\n",
       "      <td>315.69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>1959-01-24</td>\n",
       "      <td>315.86</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>1959-01-31</td>\n",
       "      <td>315.42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3126</th>\n",
       "      <td>2019-07-06</td>\n",
       "      <td>412.69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3127</th>\n",
       "      <td>2019-07-13</td>\n",
       "      <td>412.30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3128</th>\n",
       "      <td>2019-07-20</td>\n",
       "      <td>411.76</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3129</th>\n",
       "      <td>2019-07-27</td>\n",
       "      <td>410.32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3130</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>410.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3131</th>\n",
       "      <td>2019-08-10</td>\n",
       "      <td>410.48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3132</th>\n",
       "      <td>2019-08-17</td>\n",
       "      <td>410.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3133</th>\n",
       "      <td>2019-08-24</td>\n",
       "      <td>409.52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3134</th>\n",
       "      <td>2019-08-31</td>\n",
       "      <td>409.32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3135</th>\n",
       "      <td>2019-09-07</td>\n",
       "      <td>408.80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3136</th>\n",
       "      <td>2019-09-14</td>\n",
       "      <td>408.61</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3137</th>\n",
       "      <td>2019-09-21</td>\n",
       "      <td>408.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3138</th>\n",
       "      <td>2019-09-28</td>\n",
       "      <td>408.28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3139</th>\n",
       "      <td>2019-10-05</td>\n",
       "      <td>407.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3140</th>\n",
       "      <td>2019-10-12</td>\n",
       "      <td>408.61</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3141</th>\n",
       "      <td>2019-10-19</td>\n",
       "      <td>408.77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3142</th>\n",
       "      <td>2019-10-26</td>\n",
       "      <td>408.68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3143</th>\n",
       "      <td>2019-11-02</td>\n",
       "      <td>409.86</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3144</th>\n",
       "      <td>2019-11-09</td>\n",
       "      <td>410.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3145</th>\n",
       "      <td>2019-11-16</td>\n",
       "      <td>410.22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3146</th>\n",
       "      <td>2019-11-23</td>\n",
       "      <td>410.48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3147</th>\n",
       "      <td>2019-11-30</td>\n",
       "      <td>410.92</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3148</th>\n",
       "      <td>2019-12-07</td>\n",
       "      <td>411.27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3149</th>\n",
       "      <td>2019-12-14</td>\n",
       "      <td>411.67</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3150</th>\n",
       "      <td>2019-12-21</td>\n",
       "      <td>412.30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3151</th>\n",
       "      <td>2019-12-28</td>\n",
       "      <td>412.59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3152</th>\n",
       "      <td>2020-01-04</td>\n",
       "      <td>413.19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3153</th>\n",
       "      <td>2020-01-11</td>\n",
       "      <td>413.39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3154</th>\n",
       "      <td>2020-01-25</td>\n",
       "      <td>413.36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3155</th>\n",
       "      <td>2020-02-01</td>\n",
       "      <td>413.99</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3156 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            Date  Concentration\n",
       "0     1958-03-29         316.19\n",
       "1     1958-04-05         317.31\n",
       "2     1958-04-12         317.69\n",
       "3     1958-04-19         317.58\n",
       "4     1958-04-26         316.48\n",
       "5     1958-05-03         316.95\n",
       "6     1958-05-17         317.56\n",
       "7     1958-05-24         317.99\n",
       "8     1958-07-05         315.85\n",
       "9     1958-07-12         315.85\n",
       "10    1958-07-19         315.46\n",
       "11    1958-07-26         315.59\n",
       "12    1958-08-02         315.64\n",
       "13    1958-08-09         315.10\n",
       "14    1958-08-16         315.09\n",
       "15    1958-08-30         314.14\n",
       "16    1958-09-06         313.54\n",
       "17    1958-11-08         313.05\n",
       "18    1958-11-15         313.26\n",
       "19    1958-11-22         313.57\n",
       "20    1958-11-29         314.01\n",
       "21    1958-12-06         314.56\n",
       "22    1958-12-13         314.41\n",
       "23    1958-12-20         314.77\n",
       "24    1958-12-27         315.21\n",
       "25    1959-01-03         315.24\n",
       "26    1959-01-10         315.50\n",
       "27    1959-01-17         315.69\n",
       "28    1959-01-24         315.86\n",
       "29    1959-01-31         315.42\n",
       "...          ...            ...\n",
       "3126  2019-07-06         412.69\n",
       "3127  2019-07-13         412.30\n",
       "3128  2019-07-20         411.76\n",
       "3129  2019-07-27         410.32\n",
       "3130  2019-08-03         410.50\n",
       "3131  2019-08-10         410.48\n",
       "3132  2019-08-17         410.05\n",
       "3133  2019-08-24         409.52\n",
       "3134  2019-08-31         409.32\n",
       "3135  2019-09-07         408.80\n",
       "3136  2019-09-14         408.61\n",
       "3137  2019-09-21         408.50\n",
       "3138  2019-09-28         408.28\n",
       "3139  2019-10-05         407.99\n",
       "3140  2019-10-12         408.61\n",
       "3141  2019-10-19         408.77\n",
       "3142  2019-10-26         408.68\n",
       "3143  2019-11-02         409.86\n",
       "3144  2019-11-09         410.15\n",
       "3145  2019-11-16         410.22\n",
       "3146  2019-11-23         410.48\n",
       "3147  2019-11-30         410.92\n",
       "3148  2019-12-07         411.27\n",
       "3149  2019-12-14         411.67\n",
       "3150  2019-12-21         412.30\n",
       "3151  2019-12-28         412.59\n",
       "3152  2020-01-04         413.19\n",
       "3153  2020-01-11         413.39\n",
       "3154  2020-01-25         413.36\n",
       "3155  2020-02-01         413.99\n",
       "\n",
       "[3156 rows x 2 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_csv(filename,skiprows=44,header=None,names=[\"Date\",\"Concentration\"])\n",
    "data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nous y somme presque. Il ne reste plus qu'à convertir les dates en Period Pandas."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "def convert_date(year_month_day):\n",
    "    year_month_day_str = str(year_month_day)\n",
    "    year = int(year_month_day[:4])\n",
    "    month = int(year_month_day[5:7])\n",
    "    day = int(year_month_day[8:])\n",
    "    return pd.Timestamp(year=year,month=month,day=day,hour=12,minute=0,second=0)\n",
    "\n",
    "data[\"Datetime\"] = [convert_date(ymd) for ymd in data[\"Date\"]]\n",
    "data[\"Timestamp\"] = [date.value/10**8 for date in data[\"Datetime\"]]\n",
    "data[\"Period\"] = [date.to_period('W') for date in data[\"Datetime\"]]\n",
    "data.set_index(\"Period\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hideCode": false
   },
   "source": [
    "Il faut retenir les élèments suivants :\n",
    "\n",
    "| Nom           | Signification                        |\n",
    "|:-------------:|:-------------------------------------|\n",
    "|**Date**       | La date donnée dans le fichier brute |\n",
    "|**Datetime**   | La date au format ISO-8601           |\n",
    "|**Timestamp**  | Temps unix                           |\n",
    "|**Period**     | Plage de temps                       |\n",
    "\n",
    "Et un petit graphique pour guider notre travail."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "data.plot(x=\"Period\",y=\"Concentration\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Décomposition de la concentration"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Sur le graphique, nous aperçevons une courbe croissante avec de faible oscilliations. Cette courbe se prête bien à une décompostion en saisonière."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from statsmodels.tsa.seasonal import seasonal_decompose\n",
    "# freq = 52 car 52 semaines par an\n",
    "result = seasonal_decompose(data['Concentration'],model=\"additive\", freq=52)\n",
    "result.plot()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "La tendance globale (Trend) de la concentration moyenne de C02 dans l'atmosphère est croissante.\n",
    "\n",
    "Les oscillations semblent bien être saisonières (Seasonal).\n",
    "\n",
    "**Nous noterons cependant que le bruit résiduel (Residual) est du même ordre que les oscillations saisonnières**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Modélisation du phénomène"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Au regard de la décomposition précédente, nous modéliserons la variation de la concentration comme la somme d'une fonction affine croissante et d'une fonction périodique."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Estimation de la fonction linéaire"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Pour estimer notre fonction affine, nous allons utiliser un modlèle linéaire génralisé : Concentration ~ 1 + Temps"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<caption>Generalized Linear Model Regression Results</caption>\n",
       "<tr>\n",
       "  <th>Dep. Variable:</th>    <td>Concentration</td>  <th>  No. Observations:  </th>  <td>  3156</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model:</th>                 <td>GLM</td>       <th>  Df Residuals:      </th>  <td>  3154</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model Family:</th>       <td>Gaussian</td>     <th>  Df Model:          </th>  <td>     1</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Link Function:</th>      <td>identity</td>     <th>  Scale:             </th> <td>  18.461</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Method:</th>               <td>IRLS</td>       <th>  Log-Likelihood:    </th> <td> -9078.1</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Date:</th>           <td>Tue, 07 Apr 2020</td> <th>  Deviance:          </th> <td>  58226.</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Time:</th>               <td>13:14:18</td>     <th>  Pearson chi2:      </th> <td>5.82e+04</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>No. Iterations:</th>         <td>3</td>        <th>  Covariance Type:   </th> <td>nonrobust</td>\n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "      <td></td>         <th>coef</th>     <th>std err</th>      <th>z</th>      <th>P>|z|</th>  <th>[0.025</th>    <th>0.975]</th>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Intercept</th> <td>  324.5375</td> <td>    0.114</td> <td> 2845.850</td> <td> 0.000</td> <td>  324.314</td> <td>  324.761</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Timestamp</th> <td>     5e-09</td> <td> 1.37e-11</td> <td>  364.652</td> <td> 0.000</td> <td> 4.97e-09</td> <td> 5.03e-09</td>\n",
       "</tr>\n",
       "</table>"
      ],
      "text/plain": [
       "<class 'statsmodels.iolib.summary.Summary'>\n",
       "\"\"\"\n",
       "                 Generalized Linear Model Regression Results                  \n",
       "==============================================================================\n",
       "Dep. Variable:          Concentration   No. Observations:                 3156\n",
       "Model:                            GLM   Df Residuals:                     3154\n",
       "Model Family:                Gaussian   Df Model:                            1\n",
       "Link Function:               identity   Scale:                          18.461\n",
       "Method:                          IRLS   Log-Likelihood:                -9078.1\n",
       "Date:                Tue, 07 Apr 2020   Deviance:                       58226.\n",
       "Time:                        13:14:18   Pearson chi2:                 5.82e+04\n",
       "No. Iterations:                     3   Covariance Type:             nonrobust\n",
       "==============================================================================\n",
       "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
       "------------------------------------------------------------------------------\n",
       "Intercept    324.5375      0.114   2845.850      0.000     324.314     324.761\n",
       "Timestamp       5e-09   1.37e-11    364.652      0.000    4.97e-09    5.03e-09\n",
       "==============================================================================\n",
       "\"\"\""
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data[\"Intercept\"]=1\n",
    "linmodel = sm.GLM(data['Concentration'],data[['Intercept','Timestamp']]).fit()\n",
    "linmodel.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "data_pred = pd.DataFrame({'Timestamp': np.linspace(start=data[\"Timestamp\"][0], stop=data[\"Timestamp\"][len(data[\"Timestamp\"])-1], num=50), 'Intercept': 1})\n",
    "data_pred['Concentration'] = linmodel.predict(data_pred)\n",
    "data_pred.plot(x=\"Timestamp\",y=\"Concentration\",kind=\"line\")\n",
    "plt.scatter(x=data[\"Timestamp\"],y=data[\"Concentration\"],color=\"red\")\n",
    "plt.grid(True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Prédiction en 2050"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nous allons contruire une fonction pour prédire la concentration à une date donnée."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2050-06-04 12:00:00    1.269077e+10\n",
       "dtype: float64"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def predict(year=2050,month=6,day=4,hour=12,minute=0,second=0):\n",
    "    datetime = pd.Timestamp(year=year,month=month,day=day,hour=hour,minute=minute,second=second)\n",
    "    timestamp = datetime.value\n",
    "    date_pred = pd.DataFrame({'Timestamp':[timestamp],'Intercept':1},index=[datetime])\n",
    "    return linmodel.predict(date_pred)\n",
    "predict()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "La concentration de C02 de l'atmosphère en 2050 sera de 10 Gppm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Conclusion"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "La droite que nous avons pris pour faire nos prédictions se trouve sous les mesures réelles. Cela signifie que la réalité pourrait être encore pire dans le future. Cependant, notre prédiction présentée ici est plein de biais : le plus important reste sans doute qu'il s'appuie sur un modèle linéaire."
   ]
  }
 ],
 "metadata": {
  "celltoolbar": "Aucun(e)",
  "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}