anomalies sur les semaines manquantes

module3/exo3/exercice.ipynb

@@ -30,7 +30,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 1,
+   "execution_count": 11,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -49,7 +49,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": 12,
    "metadata": {},
    "outputs": [
     {
@@ -80,7 +80,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": 13,
    "metadata": {},
    "outputs": [
     {
@@ -141,7 +141,7 @@
        "1958-04-26  316.48"
       ]
      },
-     "execution_count": 3,
+     "execution_count": 13,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -154,7 +154,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": 14,
    "metadata": {},
    "outputs": [
     {
@@ -183,7 +183,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 15,
    "metadata": {},
    "outputs": [
     {
@@ -228,7 +228,7 @@
        "Index: []"
       ]
      },
-     "execution_count": 5,
+     "execution_count": 15,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -240,7 +240,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": 17,
    "metadata": {},
    "outputs": [
     {
@@ -264,28 +264,34 @@
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th></th>\n",
+       "      <th>index</th>\n",
        "      <th>ppm</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
-       "      <th>1958-03-29</th>\n",
+       "      <th>0</th>\n",
+       "      <td>1958-03-29</td>\n",
        "      <td>316.19</td>\n",
        "    </tr>\n",
        "    <tr>\n",
-       "      <th>1958-04-05</th>\n",
+       "      <th>1</th>\n",
+       "      <td>1958-04-05</td>\n",
        "      <td>317.31</td>\n",
        "    </tr>\n",
        "    <tr>\n",
-       "      <th>1958-04-12</th>\n",
+       "      <th>2</th>\n",
+       "      <td>1958-04-12</td>\n",
        "      <td>317.69</td>\n",
        "    </tr>\n",
        "    <tr>\n",
-       "      <th>1958-04-19</th>\n",
+       "      <th>3</th>\n",
+       "      <td>1958-04-19</td>\n",
        "      <td>317.58</td>\n",
        "    </tr>\n",
        "    <tr>\n",
-       "      <th>1958-04-26</th>\n",
+       "      <th>4</th>\n",
+       "      <td>1958-04-26</td>\n",
        "      <td>316.48</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
@@ -293,24 +299,65 @@
        "</div>"
       ],
       "text/plain": [
-       "               ppm\n",
-       "1958-03-29  316.19\n",
-       "1958-04-05  317.31\n",
-       "1958-04-12  317.69\n",
-       "1958-04-19  317.58\n",
-       "1958-04-26  316.48"
+       "        index     ppm\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"
       ]
      },
-     "execution_count": 6,
+     "execution_count": 17,
      "metadata": {},
      "output_type": "execute_result"
     }
    ],
    "source": [
-    "data = raw_data.copy()\n",
+    "data = raw_data.copy().reset_index()\n",
     "data.head()"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "On vérifie si toutes les semaines sont présentes dans les données : ce n'est pas le cas pour les années ci-dessous. Deux cas de figure se présentent :\n",
+    "- pour l'année 1958 (première date de la série) et l'année 2023 (dernière date de la série) $\\rightarrow$ pas de problème\n",
+    "- pour les autres années $\\rightarrow$ quelques semaines sont manquantes, on décide de continuer l'analyse tel quel mais on garde l'information en tête"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 38,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "1958 25\n",
+      "1959 48\n",
+      "1962 48\n",
+      "1963 49\n",
+      "1964 31\n",
+      "1966 49\n",
+      "1984 48\n",
+      "2003 49\n",
+      "2005 49\n",
+      "2023 26\n"
+     ]
+    }
+   ],
+   "source": [
+    "# calcul du nombre de semaines présentes par année\n",
+    "data['year'] = data['index'].apply(lambda x: x[:4])\n",
+    "years = data.groupby(['year'])['year'].count().index\n",
+    "nb_weeks = data.groupby(['year'])['year'].count().values\n",
+    "for x, y in zip(years, nb_weeks):\n",
+    "    if y < 50:\n",
+    "        print(x, y)"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},