régression linéaire

module3/exo3/exercice.ipynb

@@ -30,16 +30,20 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": 53,
    "metadata": {},
    "outputs": [],
    "source": [
     "import os\n",
+    "import numpy as np\n",
     "import pandas as pd\n",
     "import seaborn as sns\n",
     "import matplotlib.pyplot as plt\n",
+    "from datetime import datetime, timedelta\n",
     "\n",
-    "from statsmodels.tsa.seasonal import seasonal_decompose\n"
+    "from statsmodels.tsa.seasonal import seasonal_decompose\n",
+    "from sklearn.linear_model import LinearRegression\n",
+    "from sklearn.metrics import r2_score"
    ]
   },
   {
@@ -459,6 +463,140 @@
     "    axes[i].legend(loc='upper left');"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Apprentissage d'un modèle simple sur la tendance lente et prédictions"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "On cherche à apprendre un modèle linéaire sur la tendance lente."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 42,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "LinearRegression(copy_X=True, fit_intercept=True, n_jobs=1, normalize=False)"
+      ]
+     },
+     "execution_count": 42,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "X = np.arange(0, len(x)).reshape(-1, 1)\n",
+    "\n",
+    "# Model\n",
+    "model = LinearRegression()\n",
+    "model.fit(X, y)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 75,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "0.976691466648821"
+      ]
+     },
+     "execution_count": 75,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# Score du modèle sur les données d'entrâinement\n",
+    "r2_score(y, model.predict(X))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 51,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "78"
+      ]
+     },
+     "execution_count": 51,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# Nombre de semaines restant jusque 2025\n",
+    "def count_weeks_until_year(start_date, target_year):\n",
+    "    start_date = datetime.strptime(start_date, '%Y-%m-%d')\n",
+    "    target_date = datetime(target_year, 1, 1)\n",
+    "    delta = target_date - start_date\n",
+    "    weeks = delta.days // 7\n",
+    "    return weeks\n",
+    "\n",
+    "start_date_str = x[-1]\n",
+    "target_year = 2025\n",
+    "\n",
+    "weeks_until_target_year = count_weeks_until_year(start_date_str, target_year)\n",
+    "weeks_until_target_year"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 68,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Prédictions\n",
+    "weeks_untils_2025 = np.arange(int(X[[-1]]) + 1, int(X[[-1]]) + 1 + weeks_until_target_year)\n",
+    "preds = model.predict(weeks_untils_2025.reshape(-1, 1))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 69,
+   "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": [
+    "fig, ax = plt.subplots(1, 1)\n",
+    "ax.plot(res.trend, label='Tendance')\n",
+    "ax.plot(weeks_untils_2025, preds, color='r', label='prédictions')\n",
+    "plt.legend();"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Ce modèle n'est pas particulièrement adapté, il sous-estime très nettement la quantité de $CO_2$."
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,