diff --git a/.ipynb_checkpoints/Experiment100PadiWeb-checkpoint.ipynb b/.ipynb_checkpoints/Experiment100PadiWeb-checkpoint.ipynb
new file mode 100644
index 0000000000000000000000000000000000000000..c9742bae6db3e6ab5212fd93908027e41cb25609
--- /dev/null
+++ b/.ipynb_checkpoints/Experiment100PadiWeb-checkpoint.ipynb
@@ -0,0 +1,890 @@
+{
+ "cells": [
+ {
+ "cell_type": "code",
+ "execution_count": 1,
+ "metadata": {
+ "ExecuteTime": {
+ "end_time": "2018-07-16T13:57:15.309377Z",
+ "start_time": "2018-07-16T13:57:14.152795Z"
+ }
+ },
+ "outputs": [],
+ "source": [
+ "import pandas as pd\n",
+ "import networkx as nx\n",
+ "import bqplot.pyplot as plt\n",
+ "%matplotlib inline"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 2,
+ "metadata": {
+ "ExecuteTime": {
+ "end_time": "2018-07-16T13:57:15.320691Z",
+ "start_time": "2018-07-16T13:57:15.311390Z"
+ }
+ },
+ "outputs": [],
+ "source": [
+ "data_bilan=pd.read_csv(\"is_bilan.csv\",sep=\";\")"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 3,
+ "metadata": {
+ "ExecuteTime": {
+ "end_time": "2018-07-16T13:57:15.326164Z",
+ "start_time": "2018-07-16T13:57:15.322564Z"
+ }
+ },
+ "outputs": [],
+ "source": [
+ "data_bilan[\"IS_BILAN\"]=data_bilan[\"IS_BILAN\"].apply(lambda x: \"BILAN\" if x ==1 else \"EPIDEMIE\")"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# Analyse de la structure des STRs avec un cas d'étude : Bilan/Récapitulatif d'une épidémie\n",
+ "\n",
+ "**La spatialité s'exprime-t-elle de la même manière dans certaines classes ou types de document ?** Dans le domaine de surveillance d'épidémies animales utilisant Google News, les chercheurs ont besoin de différencier un récapitulatif/bilan de la situation concernant une épidémie et la déclaration de celle-ci. Dans cette expérimentation, nous allons essayer de voir si ces deux classes de documents possèdent des caractéristiques spécifiques au travers de la STR.\n",
+ "\n",
+ "\n",
+ "## Définition des deux classes\n",
+ "\n",
+ "À l'aide du corpus de PadiWeb, on selectionne un échantillon de 100 documents que l'on divise en deux classes:\n",
+ "\n",
+ " * **Bilan**. Un récapitulatif d'un événement terminé ou en cours.\n",
+ " * **Épidémie**. Son but est d'annoncer le déclenchment d'une épidémie (le point de départ).\n",
+ "\n",
+ "L'effectif de chacune des classes est indiqué ci-dessous."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 4,
+ "metadata": {
+ "ExecuteTime": {
+ "end_time": "2018-07-16T13:57:15.447421Z",
+ "start_time": "2018-07-16T13:57:15.329307Z"
+ }
+ },
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "<matplotlib.axes._subplots.AxesSubplot at 0x113adceb8>"
+ ]
+ },
+ "execution_count": 4,
+ "metadata": {},
+ "output_type": "execute_result"
+ },
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<matplotlib.figure.Figure at 0x10d85f240>"
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "data_bilan.groupby(\"IS_BILAN\").count().plot.pie(\"ID_TEXT\")"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 5,
+ "metadata": {
+ "ExecuteTime": {
+ "end_time": "2018-07-16T13:57:15.457004Z",
+ "start_time": "2018-07-16T13:57:15.450071Z"
+ }
+ },
+ "outputs": [],
+ "source": [
+ "import numpy as np\n",
+ "def number_of_edges(x,color=None):\n",
+ " \"\"\"\n",
+ " Dedicated function to count edges based on their color\n",
+ " \"\"\"\n",
+ " if not color:\n",
+ " return len(x.number_of_edges())\n",
+ " edges=list(x.edges(data=True))\n",
+ " cp=0\n",
+ " for ed in edges:\n",
+ " if ed[-1][\"color\"] == color:\n",
+ " cp+=1\n",
+ " return cp\n",
+ "\n",
+ "\n",
+ "def flattern(A):\n",
+ " rt = []\n",
+ " for i in A:\n",
+ " if isinstance(i, list):\n",
+ " rt.extend(flattern(i))\n",
+ " elif isinstance(i, np.ndarray):\n",
+ " rt.extend(flattern(i.tolist()))\n",
+ " else:\n",
+ " rt.append(i)\n",
+ " return rt\n",
+ "def most_common(lst):\n",
+ " if not lst:\n",
+ " return \"P-PPL\"\n",
+ " if len(list(set(lst))) >1 and \"P-PPL\" in set(lst):\n",
+ " lst=[x for x in lst if x != \"PPL\"]\n",
+ " return max(set(lst), key=lst.count)\n",
+ "\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 6,
+ "metadata": {
+ "ExecuteTime": {
+ "end_time": "2018-07-16T13:57:15.544376Z",
+ "start_time": "2018-07-16T13:57:15.459343Z"
+ }
+ },
+ "outputs": [],
+ "source": [
+ "import nxpd\n",
+ "nxpd.nxpdParams[\"show\"]=\"ipynb\"\n",
+ "from strpython.helpers.gazeteer_helpers import get_data\n",
+ "def class_graph(g):\n",
+ " mapping={}\n",
+ " g2=g.copy()\n",
+ " for n in g2:\n",
+ " c=get_data(n)[\"class\"]\n",
+ " g2.nodes[n][\"label\"]=most_common(c)\n",
+ " return g2"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "Pour faire une comparaison entre les STRs générées dans chaque classe de document, on utilise plusieurs indicateurs :\n",
+ "\n",
+ " * **Granularité** La granularité est définie le niveau dans l'échelle spatiale ($village$ < $ville < pays$) d'une entité. Ici, elle nous indique à quel niveau la spatialité est utilisé pour décrire la situation.\n",
+ " * **Densité** La densité est définie par le nombre d'arêtes moyen pour un noeud dans un graphe. Un graphe d'une STR avec une forte densité, indique une forte cohésion entre les entités spatiales.\n",
+ " * **Ratio $Relation_i/Relation_j$** Dans la STR, chaque entité peut-être reliée à une autre par deux type de relations : inclusion et adjacence. Avec ce ratio, on souhaite savoir combien il existe de $relation_j$ pour une $relation_i$. Par exemple, pour une relation d'inclusion, combien de relations d'adjacence ?\n",
+ " * **Nombre de noeuds(entités spatiales)** Indique si des textes sont fortement spatialisés.\n",
+ "\n",
+ "\n",
+ "### Calcul de la granularité d'une STR\n",
+ "\n",
+ "On récupére les **classes associées** aux différentes **entités de la STR**, puis on récupére **la classe la plus fréquente**. Par exemple:\n",
+ "\n",
+ "$STR_1$ --> France, Montpellier, Clapiers, Caen --> [A-PCLI], [P-PPL, A-ADM4], [A-ADM4], [A-ADM4]\n",
+ "\n",
+ "On a donc pour granularité : **A-ADM4**\n",
+ "\n",
+ "### Calcul de la densité d'une STR\n",
+ "Le calcul de la densité d'une STR (ici son graphe) se calcule à l'aide de la formule suivante : $$\\frac{2\\times|E|}{|V|\\times(|V|-1)}$$\n",
+ "\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 7,
+ "metadata": {
+ "ExecuteTime": {
+ "end_time": "2018-07-16T13:57:16.300927Z",
+ "start_time": "2018-07-16T13:57:15.546668Z"
+ }
+ },
+ "outputs": [],
+ "source": [
+ "data_bilan[\"GRAPH\"]=data_bilan[\"ID_TEXT\"].apply(lambda x:nx.read_gexf(\"str_PADI100/{0}.gexf\".format(x)))\n",
+ "data_bilan[\"GRAPH_C\"]=data_bilan[\"GRAPH\"].apply(lambda x:class_graph(x))"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 24,
+ "metadata": {
+ "ExecuteTime": {
+ "end_time": "2018-07-16T13:58:25.430534Z",
+ "start_time": "2018-07-16T13:58:25.419515Z"
+ }
+ },
+ "outputs": [],
+ "source": [
+ "data_bilan[\"DENSITY\"]=data_bilan[\"GRAPH\"].apply(lambda x: (2*x.number_of_edges())/(x.number_of_nodes()*(x.number_of_nodes()-1)) if len(x) >1 else 0)\n",
+ "data_bilan[\"NB_NODE\"]=data_bilan[\"GRAPH\"].apply(lambda x: len(x))\n",
+ "data_bilan[\"NB_ED_ADJ\"]=data_bilan[\"GRAPH\"].apply(lambda x: number_of_edges(x,color=\"green\")/2)\n",
+ "data_bilan[\"NB_ED_INC\"]=data_bilan[\"GRAPH\"].apply(lambda x: number_of_edges(x,color=\"red\"))\n",
+ "data_bilan[\"R_ADJ_INC\"]=((data_bilan[\"NB_ED_ADJ\"])/data_bilan[\"NB_ED_INC\"]).replace([np.inf, -np.inf], np.nan).fillna(0)\n",
+ "data_bilan[\"R_INC_ADJ\"]=(data_bilan[\"NB_ED_INC\"]/(data_bilan[\"NB_ED_ADJ\"])).replace([np.inf, -np.inf], np.nan).fillna(0)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 25,
+ "metadata": {
+ "ExecuteTime": {
+ "end_time": "2018-07-16T13:58:26.522241Z",
+ "start_time": "2018-07-16T13:58:25.908441Z"
+ }
+ },
+ "outputs": [],
+ "source": [
+ "data_bilan[\"CLASS\"]=data_bilan[\"GRAPH\"].apply(lambda x: flattern([get_data(n)[\"class\"] for n in list(x.nodes())]))\n",
+ "data_bilan[\"MEAN_LVL\"]=data_bilan[\"CLASS\"].apply(lambda x: most_common(x) if len(x)>0 else \"\")"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 26,
+ "metadata": {
+ "ExecuteTime": {
+ "end_time": "2018-07-16T13:58:26.782966Z",
+ "start_time": "2018-07-16T13:58:26.759801Z"
+ },
+ "scrolled": true
+ },
+ "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>ID_TEXT</th>\n",
+ " <th>IS_BILAN</th>\n",
+ " <th>MIXED</th>\n",
+ " <th>GRAPH</th>\n",
+ " <th>GRAPH_C</th>\n",
+ " <th>DENSITY</th>\n",
+ " <th>NB_NODE</th>\n",
+ " <th>NB_ED_ADJ</th>\n",
+ " <th>NB_ED_INC</th>\n",
+ " <th>R_ADJ_INC</th>\n",
+ " <th>R_INC_ADJ</th>\n",
+ " <th>CLASS</th>\n",
+ " <th>MEAN_LVL</th>\n",
+ " </tr>\n",
+ " </thead>\n",
+ " <tbody>\n",
+ " <tr>\n",
+ " <th>0</th>\n",
+ " <td>0</td>\n",
+ " <td>BILAN</td>\n",
+ " <td>0</td>\n",
+ " <td>(GD4103071, GD4468122, GD95073, GD791183)</td>\n",
+ " <td>(GD4103071, GD4468122, GD95073, GD791183)</td>\n",
+ " <td>0.000000</td>\n",
+ " <td>4</td>\n",
+ " <td>0.0</td>\n",
+ " <td>0</td>\n",
+ " <td>0.0</td>\n",
+ " <td>0.0</td>\n",
+ " <td>[P-PPLA, P-PPL, P-PPLA, A-ADM1, P-PPLA]</td>\n",
+ " <td>P-PPLA</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>1</th>\n",
+ " <td>1</td>\n",
+ " <td>EPIDEMIE</td>\n",
+ " <td>0</td>\n",
+ " <td>(GD1685421)</td>\n",
+ " <td>(GD1685421)</td>\n",
+ " <td>0.000000</td>\n",
+ " <td>1</td>\n",
+ " <td>0.0</td>\n",
+ " <td>0</td>\n",
+ " <td>0.0</td>\n",
+ " <td>0.0</td>\n",
+ " <td>[P-PPL]</td>\n",
+ " <td>P-PPL</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>2</th>\n",
+ " <td>2</td>\n",
+ " <td>BILAN</td>\n",
+ " <td>1</td>\n",
+ " <td>(GD2032795)</td>\n",
+ " <td>(GD2032795)</td>\n",
+ " <td>0.000000</td>\n",
+ " <td>1</td>\n",
+ " <td>0.0</td>\n",
+ " <td>0</td>\n",
+ " <td>0.0</td>\n",
+ " <td>0.0</td>\n",
+ " <td>[A-PCLI]</td>\n",
+ " <td>A-PCLI</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>3</th>\n",
+ " <td>3</td>\n",
+ " <td>BILAN</td>\n",
+ " <td>0</td>\n",
+ " <td>(GD1626932, GD3274230)</td>\n",
+ " <td>(GD1626932, GD3274230)</td>\n",
+ " <td>0.000000</td>\n",
+ " <td>2</td>\n",
+ " <td>0.0</td>\n",
+ " <td>0</td>\n",
+ " <td>0.0</td>\n",
+ " <td>0.0</td>\n",
+ " <td>[A-PCLI, P-PPL]</td>\n",
+ " <td>A-PCLI</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>4</th>\n",
+ " <td>4</td>\n",
+ " <td>EPIDEMIE</td>\n",
+ " <td>0</td>\n",
+ " <td>(GD639917, GD3789919, GD1316637, GD2055944)</td>\n",
+ " <td>(GD639917, GD3789919, GD1316637, GD2055944)</td>\n",
+ " <td>0.166667</td>\n",
+ " <td>4</td>\n",
+ " <td>0.0</td>\n",
+ " <td>1</td>\n",
+ " <td>0.0</td>\n",
+ " <td>0.0</td>\n",
+ " <td>[A-PCLI, A-ADM1, P-PPLA, P-PPLC, A-PCLI, A-PCLI]</td>\n",
+ " <td>A-PCLI</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>5</th>\n",
+ " <td>5</td>\n",
+ " <td>EPIDEMIE</td>\n",
+ " <td>1</td>\n",
+ " <td>(GD639917, GD3995806, GD3789919, GD1316637, GD...</td>\n",
+ " <td>(GD639917, GD3995806, GD3789919, GD1316637, GD...</td>\n",
+ " <td>0.200000</td>\n",
+ " <td>5</td>\n",
+ " <td>0.0</td>\n",
+ " <td>2</td>\n",
+ " <td>0.0</td>\n",
+ " <td>0.0</td>\n",
+ " <td>[A-PCLI, P-PPL, A-ADM1, P-PPLA, P-PPLC, A-PCLI...</td>\n",
+ " <td>A-PCLI</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>6</th>\n",
+ " <td>6</td>\n",
+ " <td>EPIDEMIE</td>\n",
+ " <td>0</td>\n",
+ " <td>(GD639917, GD3789919, GD2055944)</td>\n",
+ " <td>(GD639917, GD3789919, GD2055944)</td>\n",
+ " <td>0.000000</td>\n",
+ " <td>3</td>\n",
+ " <td>0.0</td>\n",
+ " <td>0</td>\n",
+ " <td>0.0</td>\n",
+ " <td>0.0</td>\n",
+ " <td>[A-PCLI, A-ADM1, P-PPLA, P-PPLC, A-PCLI]</td>\n",
+ " <td>A-PCLI</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>7</th>\n",
+ " <td>7</td>\n",
+ " <td>BILAN</td>\n",
+ " <td>0</td>\n",
+ " <td>(GD5526704, GD976842, GD1316637, GD2055944)</td>\n",
+ " <td>(GD5526704, GD976842, GD1316637, GD2055944)</td>\n",
+ " <td>0.333333</td>\n",
+ " <td>4</td>\n",
+ " <td>1.0</td>\n",
+ " <td>0</td>\n",
+ " <td>0.0</td>\n",
+ " <td>0.0</td>\n",
+ " <td>[A-PCLI, A-PCLI, A-PCLI, A-PCLI]</td>\n",
+ " <td>A-PCLI</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>8</th>\n",
+ " <td>8</td>\n",
+ " <td>BILAN</td>\n",
+ " <td>0</td>\n",
+ " <td>(GD2908705, GD1404948, GD9642903, GD3995806, G...</td>\n",
+ " <td>(GD2908705, GD1404948, GD9642903, GD3995806, G...</td>\n",
+ " <td>0.285714</td>\n",
+ " <td>7</td>\n",
+ " <td>2.0</td>\n",
+ " <td>2</td>\n",
+ " <td>1.0</td>\n",
+ " <td>1.0</td>\n",
+ " <td>[A-ADM1, P-PPL, P-PPL, P-PPL, A-ADM1, A-ADM1, ...</td>\n",
+ " <td>P-PPL</td>\n",
+ " </tr>\n",
+ " </tbody>\n",
+ "</table>\n",
+ "</div>"
+ ],
+ "text/plain": [
+ " ID_TEXT IS_BILAN MIXED \\\n",
+ "0 0 BILAN 0 \n",
+ "1 1 EPIDEMIE 0 \n",
+ "2 2 BILAN 1 \n",
+ "3 3 BILAN 0 \n",
+ "4 4 EPIDEMIE 0 \n",
+ "5 5 EPIDEMIE 1 \n",
+ "6 6 EPIDEMIE 0 \n",
+ "7 7 BILAN 0 \n",
+ "8 8 BILAN 0 \n",
+ "\n",
+ " GRAPH \\\n",
+ "0 (GD4103071, GD4468122, GD95073, GD791183) \n",
+ "1 (GD1685421) \n",
+ "2 (GD2032795) \n",
+ "3 (GD1626932, GD3274230) \n",
+ "4 (GD639917, GD3789919, GD1316637, GD2055944) \n",
+ "5 (GD639917, GD3995806, GD3789919, GD1316637, GD... \n",
+ "6 (GD639917, GD3789919, GD2055944) \n",
+ "7 (GD5526704, GD976842, GD1316637, GD2055944) \n",
+ "8 (GD2908705, GD1404948, GD9642903, GD3995806, G... \n",
+ "\n",
+ " GRAPH_C DENSITY NB_NODE \\\n",
+ "0 (GD4103071, GD4468122, GD95073, GD791183) 0.000000 4 \n",
+ "1 (GD1685421) 0.000000 1 \n",
+ "2 (GD2032795) 0.000000 1 \n",
+ "3 (GD1626932, GD3274230) 0.000000 2 \n",
+ "4 (GD639917, GD3789919, GD1316637, GD2055944) 0.166667 4 \n",
+ "5 (GD639917, GD3995806, GD3789919, GD1316637, GD... 0.200000 5 \n",
+ "6 (GD639917, GD3789919, GD2055944) 0.000000 3 \n",
+ "7 (GD5526704, GD976842, GD1316637, GD2055944) 0.333333 4 \n",
+ "8 (GD2908705, GD1404948, GD9642903, GD3995806, G... 0.285714 7 \n",
+ "\n",
+ " NB_ED_ADJ NB_ED_INC R_ADJ_INC R_INC_ADJ \\\n",
+ "0 0.0 0 0.0 0.0 \n",
+ "1 0.0 0 0.0 0.0 \n",
+ "2 0.0 0 0.0 0.0 \n",
+ "3 0.0 0 0.0 0.0 \n",
+ "4 0.0 1 0.0 0.0 \n",
+ "5 0.0 2 0.0 0.0 \n",
+ "6 0.0 0 0.0 0.0 \n",
+ "7 1.0 0 0.0 0.0 \n",
+ "8 2.0 2 1.0 1.0 \n",
+ "\n",
+ " CLASS MEAN_LVL \n",
+ "0 [P-PPLA, P-PPL, P-PPLA, A-ADM1, P-PPLA] P-PPLA \n",
+ "1 [P-PPL] P-PPL \n",
+ "2 [A-PCLI] A-PCLI \n",
+ "3 [A-PCLI, P-PPL] A-PCLI \n",
+ "4 [A-PCLI, A-ADM1, P-PPLA, P-PPLC, A-PCLI, A-PCLI] A-PCLI \n",
+ "5 [A-PCLI, P-PPL, A-ADM1, P-PPLA, P-PPLC, A-PCLI... A-PCLI \n",
+ "6 [A-PCLI, A-ADM1, P-PPLA, P-PPLC, A-PCLI] A-PCLI \n",
+ "7 [A-PCLI, A-PCLI, A-PCLI, A-PCLI] A-PCLI \n",
+ "8 [A-ADM1, P-PPL, P-PPL, P-PPL, A-ADM1, A-ADM1, ... P-PPL "
+ ]
+ },
+ "execution_count": 26,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "data_bilan.head(9)"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# Résultats\n",
+ "\n",
+ "### Granularité sur les documents de classe **BILAN**"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 27,
+ "metadata": {
+ "ExecuteTime": {
+ "end_time": "2018-07-16T13:58:27.641675Z",
+ "start_time": "2018-07-16T13:58:27.457994Z"
+ }
+ },
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "<matplotlib.axes._subplots.AxesSubplot at 0x11409d5c0>"
+ ]
+ },
+ "execution_count": 27,
+ "metadata": {},
+ "output_type": "execute_result"
+ },
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<matplotlib.figure.Figure at 0x114971860>"
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "data_bilan[data_bilan[\"IS_BILAN\"] == \"BILAN\"].groupby(\"MEAN_LVL\").count().plot.pie(\"ID_TEXT\")"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "### Granularité sur les documents de classe **EPIDEMIE**"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 28,
+ "metadata": {
+ "ExecuteTime": {
+ "end_time": "2018-07-16T13:58:28.848371Z",
+ "start_time": "2018-07-16T13:58:28.698404Z"
+ }
+ },
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "<matplotlib.axes._subplots.AxesSubplot at 0x114c629b0>"
+ ]
+ },
+ "execution_count": 28,
+ "metadata": {},
+ "output_type": "execute_result"
+ },
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<matplotlib.figure.Figure at 0x114c6bcc0>"
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "data_bilan[data_bilan[\"IS_BILAN\"] == \"EPIDEMIE\"].groupby(\"MEAN_LVL\").count().plot.pie(\"ID_TEXT\")"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "### Valeurs moyennes obtenues pour chaque indicateur"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 23,
+ "metadata": {
+ "ExecuteTime": {
+ "end_time": "2018-07-16T13:57:51.583786Z",
+ "start_time": "2018-07-16T13:57:51.570256Z"
+ }
+ },
+ "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>ID_TEXT</th>\n",
+ " <th>MIXED</th>\n",
+ " <th>DENSITY</th>\n",
+ " <th>NB_NODE</th>\n",
+ " <th>NB_ED_ADJ</th>\n",
+ " <th>NB_ED_INC</th>\n",
+ " <th>R_ADJ_INC</th>\n",
+ " <th>R_INC_ADJ</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>IS_BILAN</th>\n",
+ " <th></th>\n",
+ " <th></th>\n",
+ " <th></th>\n",
+ " <th></th>\n",
+ " <th></th>\n",
+ " <th></th>\n",
+ " <th></th>\n",
+ " <th></th>\n",
+ " </tr>\n",
+ " </thead>\n",
+ " <tbody>\n",
+ " <tr>\n",
+ " <th>BILAN</th>\n",
+ " <td>51.588235</td>\n",
+ " <td>0.014706</td>\n",
+ " <td>0.293545</td>\n",
+ " <td>6.455882</td>\n",
+ " <td>2.058824</td>\n",
+ " <td>2.705882</td>\n",
+ " <td>0.279597</td>\n",
+ " <td>1.049048</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>EPIDEMIE</th>\n",
+ " <td>46.727273</td>\n",
+ " <td>0.030303</td>\n",
+ " <td>0.379501</td>\n",
+ " <td>4.636364</td>\n",
+ " <td>1.484848</td>\n",
+ " <td>1.303030</td>\n",
+ " <td>0.199495</td>\n",
+ " <td>0.240657</td>\n",
+ " </tr>\n",
+ " </tbody>\n",
+ "</table>\n",
+ "</div>"
+ ],
+ "text/plain": [
+ " ID_TEXT MIXED DENSITY NB_NODE NB_ED_ADJ NB_ED_INC \\\n",
+ "IS_BILAN \n",
+ "BILAN 51.588235 0.014706 0.293545 6.455882 2.058824 2.705882 \n",
+ "EPIDEMIE 46.727273 0.030303 0.379501 4.636364 1.484848 1.303030 \n",
+ "\n",
+ " R_ADJ_INC R_INC_ADJ \n",
+ "IS_BILAN \n",
+ "BILAN 0.279597 1.049048 \n",
+ "EPIDEMIE 0.199495 0.240657 "
+ ]
+ },
+ "execution_count": 23,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "data_bilan.groupby(\"IS_BILAN\").mean()"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "## Analyse des résultats\n",
+ "\n",
+ "\n",
+ "### Granularité\n",
+ "\n",
+ "En regardant les deux camemberts ci-dessus, on remarque que la granularité observé dans les STRs différe selon le type de texte. Les textes de classe **épidémie** sont généralement plus \"haut\" dans la hiérarchie spatiale, de part la forte présence de classe telles que: A-PCLI ($\\approx$Pays), A-ADM1(premier découpage administratif d'un pays *equiv* région en France, état aux Etats-Unis, *etc.*). Ceux de la classe **BILAN**, ont une granularité un peu plus fine avec un spectre de classe plus étendue : T-ISL (ile), S-BLDG (batiment).\n",
+ "\n",
+ "En se basant sur la classification proposé, on conclue que les documents de type **bilan** sont plus \"fin\" spatialement que ceux de la classe **épidémie**. \n",
+ "\n",
+ "### Densité/ Nombre de noeuds/ Nombre d'arrêtes\n",
+ "\n",
+ "Malheuresement la densité moyenne ne permet de faire aucune conclusion.\n",
+ "\n",
+ "On observe que le nombre de noeuds dans les documents de classes Bilan est plus élevé. Ce qui indique que le nombre d'entités spatiales dans ces documents est plus élevés. Ce qui semble tout à fait normal car contrairement à une déclaration d'épidémie, le bilan fait un récapitulatif de la propagation d'une maladie sur un laps de temps et une spatialité (souvent) plus importante.\n",
+ "\n",
+ "Pour le nombre de relations d'ajacence et d'inclusion, on observe un même rapport de \"force\" : Il y a plus d'arêtes d'inclusion que d'arêtes d'adjacence.\n",
+ "\n",
+ "### Ratio Adjacence/Inclusion VS Inclusion/Adjacence\n",
+ "\n",
+ "| CLASSE | ADJ/INC | INC/ADJ |\n",
+ "|----------|----------|----------|\n",
+ "| BILAN | 0.559194 | 1.04905 |\n",
+ "| EPIDEMIE | 0.39899 | 0.240657 |\n",
+ "\n",
+ "On reprend les résultats concernat les rapports ADJ/INC (combien de relations d'inclusion pour une relation d'adjacence ?) et INC/ADJ (le contraire de ADJ/INC). A partir de ces résultats, on observe que les rapports sont inversés ! Pour les documents de classe EPIDEMIE, on va favoriser plus les relations d'inclusion, contrairement aux documents de classe BILAN qui favorisent les relations d'adjacences.\n",
+ "\n",
+ "Est-ce que parce que les relations d'inclusions sont favorisés (ratio ADJ/INC élevé), on se retrouve sur des zones limitées, donc plus local ? Ca rentre bien dans le cadre de la classe épidémie.\n",
+ "\n",
+ "Est-ce qu'un ratio élevé INC/ADJ traduit une information concernant la dispertion d'une maladie ? \n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 14,
+ "metadata": {
+ "ExecuteTime": {
+ "end_time": "2018-07-16T13:57:17.407531Z",
+ "start_time": "2018-07-16T13:57:17.405043Z"
+ }
+ },
+ "outputs": [],
+ "source": [
+ "from ipywidgets import interact, interactive, fixed, interact_manual\n",
+ "import ipywidgets as widgets"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 15,
+ "metadata": {
+ "ExecuteTime": {
+ "end_time": "2018-07-16T13:57:17.491943Z",
+ "start_time": "2018-07-16T13:57:17.410121Z"
+ }
+ },
+ "outputs": [
+ {
+ "data": {
+ "application/vnd.jupyter.widget-view+json": {
+ "model_id": "e5118491edd14dd3aefcd65285a312a5",
+ "version_major": 2,
+ "version_minor": 0
+ },
+ "text/plain": [
+ "interactive(children=(IntSlider(value=0, description='x'), Output()), _dom_classes=('widget-interact',))"
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "from nxpd import draw\n",
+ "\n",
+ "def f(x):\n",
+ " global data_bilan\n",
+ " return draw(data_bilan[data_bilan[\"IS_BILAN\"]==\"BILAN\"].iloc[x][\"GRAPH_C\"],show=\"ipynb\")\n",
+ "interact(f, x=widgets.IntSlider(min=0,max=100,step=1));"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 16,
+ "metadata": {
+ "ExecuteTime": {
+ "end_time": "2018-07-16T13:57:17.500755Z",
+ "start_time": "2018-07-16T13:57:17.494176Z"
+ }
+ },
+ "outputs": [],
+ "source": [
+ "dd=data_bilan.groupby(\"IS_BILAN\").mean()"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 17,
+ "metadata": {
+ "ExecuteTime": {
+ "end_time": "2018-07-16T13:57:17.511979Z",
+ "start_time": "2018-07-16T13:57:17.502663Z"
+ }
+ },
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "|:---------|---------:|---------:|\n",
+ "| BILAN | 0.279597 | 2.0981 |\n",
+ "| EPIDEMIE | 0.199495 | 0.481313 |\n"
+ ]
+ }
+ ],
+ "source": [
+ "from tabulate import tabulate\n",
+ "print(tabulate(dd[[\"R_ADJ_INC\", \"R_INC_ADJ\"]],tablefmt=\"pipe\"))"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": []
+ }
+ ],
+ "metadata": {
+ "kernelspec": {
+ "display_name": "Python 3",
+ "language": "python",
+ "name": "python3"
+ },
+ "language_info": {
+ "codemirror_mode": {
+ "name": "ipython",
+ "version": 3
+ },
+ "file_extension": ".py",
+ "mimetype": "text/x-python",
+ "name": "python",
+ "nbconvert_exporter": "python",
+ "pygments_lexer": "ipython3",
+ "version": "3.6.5"
+ },
+ "toc": {
+ "nav_menu": {},
+ "number_sections": true,
+ "sideBar": true,
+ "skip_h1_title": false,
+ "toc_cell": false,
+ "toc_position": {},
+ "toc_section_display": "block",
+ "toc_window_display": false
+ },
+ "varInspector": {
+ "cols": {
+ "lenName": 16,
+ "lenType": 16,
+ "lenVar": 40
+ },
+ "kernels_config": {
+ "python": {
+ "delete_cmd_postfix": "",
+ "delete_cmd_prefix": "del ",
+ "library": "var_list.py",
+ "varRefreshCmd": "print(var_dic_list())"
+ },
+ "r": {
+ "delete_cmd_postfix": ") ",
+ "delete_cmd_prefix": "rm(",
+ "library": "var_list.r",
+ "varRefreshCmd": "cat(var_dic_list()) "
+ }
+ },
+ "types_to_exclude": [
+ "module",
+ "function",
+ "builtin_function_or_method",
+ "instance",
+ "_Feature"
+ ],
+ "window_display": false
+ }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}
diff --git a/Experiment100PadiWeb.ipynb b/Experiment100PadiWeb.ipynb
index 33318b64c01ae39954a0f8fb44e2051e8b0f3444..0b0a37e39cfcc519d5d145565b35107b6b430906 100644
--- a/Experiment100PadiWeb.ipynb
+++ b/Experiment100PadiWeb.ipynb
@@ -5,8 +5,8 @@
"execution_count": 1,
"metadata": {
"ExecuteTime": {
- "end_time": "2018-07-16T13:19:10.843088Z",
- "start_time": "2018-07-16T13:19:09.478873Z"
+ "end_time": "2018-07-16T13:57:15.309377Z",
+ "start_time": "2018-07-16T13:57:14.152795Z"
}
},
"outputs": [],
@@ -22,8 +22,8 @@
"execution_count": 2,
"metadata": {
"ExecuteTime": {
- "end_time": "2018-07-16T13:19:10.860729Z",
- "start_time": "2018-07-16T13:19:10.845346Z"
+ "end_time": "2018-07-16T13:57:15.320691Z",
+ "start_time": "2018-07-16T13:57:15.311390Z"
}
},
"outputs": [],
@@ -36,8 +36,8 @@
"execution_count": 3,
"metadata": {
"ExecuteTime": {
- "end_time": "2018-07-16T13:19:10.866942Z",
- "start_time": "2018-07-16T13:19:10.862464Z"
+ "end_time": "2018-07-16T13:57:15.326164Z",
+ "start_time": "2018-07-16T13:57:15.322564Z"
}
},
"outputs": [],
@@ -69,15 +69,15 @@
"execution_count": 4,
"metadata": {
"ExecuteTime": {
- "end_time": "2018-07-16T13:19:10.998510Z",
- "start_time": "2018-07-16T13:19:10.868556Z"
+ "end_time": "2018-07-16T13:57:15.447421Z",
+ "start_time": "2018-07-16T13:57:15.329307Z"
}
},
"outputs": [
{
"data": {
"text/plain": [
- "<matplotlib.axes._subplots.AxesSubplot at 0x109c1af98>"
+ "<matplotlib.axes._subplots.AxesSubplot at 0x113adceb8>"
]
},
"execution_count": 4,
@@ -88,7 +88,7 @@
"data": {
"image/png": "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\n",
"text/plain": [
- "<matplotlib.figure.Figure at 0x10214ef60>"
+ "<matplotlib.figure.Figure at 0x10d85f240>"
]
},
"metadata": {},
@@ -104,8 +104,8 @@
"execution_count": 5,
"metadata": {
"ExecuteTime": {
- "end_time": "2018-07-16T13:19:11.006615Z",
- "start_time": "2018-07-16T13:19:11.000324Z"
+ "end_time": "2018-07-16T13:57:15.457004Z",
+ "start_time": "2018-07-16T13:57:15.450071Z"
}
},
"outputs": [],
@@ -149,8 +149,8 @@
"execution_count": 6,
"metadata": {
"ExecuteTime": {
- "end_time": "2018-07-16T13:19:11.103841Z",
- "start_time": "2018-07-16T13:19:11.008238Z"
+ "end_time": "2018-07-16T13:57:15.544376Z",
+ "start_time": "2018-07-16T13:57:15.459343Z"
}
},
"outputs": [],
@@ -197,8 +197,8 @@
"execution_count": 7,
"metadata": {
"ExecuteTime": {
- "end_time": "2018-07-16T13:19:12.360291Z",
- "start_time": "2018-07-16T13:19:11.105857Z"
+ "end_time": "2018-07-16T13:57:16.300927Z",
+ "start_time": "2018-07-16T13:57:15.546668Z"
}
},
"outputs": [],
@@ -209,30 +209,30 @@
},
{
"cell_type": "code",
- "execution_count": 8,
+ "execution_count": 24,
"metadata": {
"ExecuteTime": {
- "end_time": "2018-07-16T13:19:12.378025Z",
- "start_time": "2018-07-16T13:19:12.362254Z"
+ "end_time": "2018-07-16T13:58:25.430534Z",
+ "start_time": "2018-07-16T13:58:25.419515Z"
}
},
"outputs": [],
"source": [
"data_bilan[\"DENSITY\"]=data_bilan[\"GRAPH\"].apply(lambda x: (2*x.number_of_edges())/(x.number_of_nodes()*(x.number_of_nodes()-1)) if len(x) >1 else 0)\n",
"data_bilan[\"NB_NODE\"]=data_bilan[\"GRAPH\"].apply(lambda x: len(x))\n",
- "data_bilan[\"NB_ED_ADJ\"]=data_bilan[\"GRAPH\"].apply(lambda x: number_of_edges(x,color=\"green\"))\n",
+ "data_bilan[\"NB_ED_ADJ\"]=data_bilan[\"GRAPH\"].apply(lambda x: number_of_edges(x,color=\"green\")/2)\n",
"data_bilan[\"NB_ED_INC\"]=data_bilan[\"GRAPH\"].apply(lambda x: number_of_edges(x,color=\"red\"))\n",
- "data_bilan[\"R_ADJ_INC\"]=((data_bilan[\"NB_ED_ADJ\"]/2)/data_bilan[\"NB_ED_INC\"]).replace([np.inf, -np.inf], np.nan).fillna(0)\n",
- "data_bilan[\"R_INC_ADJ\"]=(data_bilan[\"NB_ED_INC\"]/(data_bilan[\"NB_ED_ADJ\"]/2)).replace([np.inf, -np.inf], np.nan).fillna(0)"
+ "data_bilan[\"R_ADJ_INC\"]=((data_bilan[\"NB_ED_ADJ\"])/data_bilan[\"NB_ED_INC\"]).replace([np.inf, -np.inf], np.nan).fillna(0)\n",
+ "data_bilan[\"R_INC_ADJ\"]=(data_bilan[\"NB_ED_INC\"]/(data_bilan[\"NB_ED_ADJ\"])).replace([np.inf, -np.inf], np.nan).fillna(0)"
]
},
{
"cell_type": "code",
- "execution_count": 9,
+ "execution_count": 25,
"metadata": {
"ExecuteTime": {
- "end_time": "2018-07-16T13:19:13.124807Z",
- "start_time": "2018-07-16T13:19:12.380975Z"
+ "end_time": "2018-07-16T13:58:26.522241Z",
+ "start_time": "2018-07-16T13:58:25.908441Z"
}
},
"outputs": [],
@@ -243,11 +243,11 @@
},
{
"cell_type": "code",
- "execution_count": 10,
+ "execution_count": 26,
"metadata": {
"ExecuteTime": {
- "end_time": "2018-07-16T13:19:13.155512Z",
- "start_time": "2018-07-16T13:19:13.127069Z"
+ "end_time": "2018-07-16T13:58:26.782966Z",
+ "start_time": "2018-07-16T13:58:26.759801Z"
},
"scrolled": true
},
@@ -298,7 +298,7 @@
" <td>(GD4103071, GD4468122, GD95073, GD791183)</td>\n",
" <td>0.000000</td>\n",
" <td>4</td>\n",
- " <td>0</td>\n",
+ " <td>0.0</td>\n",
" <td>0</td>\n",
" <td>0.0</td>\n",
" <td>0.0</td>\n",
@@ -314,7 +314,7 @@
" <td>(GD1685421)</td>\n",
" <td>0.000000</td>\n",
" <td>1</td>\n",
- " <td>0</td>\n",
+ " <td>0.0</td>\n",
" <td>0</td>\n",
" <td>0.0</td>\n",
" <td>0.0</td>\n",
@@ -330,7 +330,7 @@
" <td>(GD2032795)</td>\n",
" <td>0.000000</td>\n",
" <td>1</td>\n",
- " <td>0</td>\n",
+ " <td>0.0</td>\n",
" <td>0</td>\n",
" <td>0.0</td>\n",
" <td>0.0</td>\n",
@@ -346,12 +346,12 @@
" <td>(GD1626932, GD3274230)</td>\n",
" <td>0.000000</td>\n",
" <td>2</td>\n",
- " <td>0</td>\n",
+ " <td>0.0</td>\n",
" <td>0</td>\n",
" <td>0.0</td>\n",
" <td>0.0</td>\n",
" <td>[A-PCLI, P-PPL]</td>\n",
- " <td>P-PPL</td>\n",
+ " <td>A-PCLI</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
@@ -362,7 +362,7 @@
" <td>(GD639917, GD3789919, GD1316637, GD2055944)</td>\n",
" <td>0.166667</td>\n",
" <td>4</td>\n",
- " <td>0</td>\n",
+ " <td>0.0</td>\n",
" <td>1</td>\n",
" <td>0.0</td>\n",
" <td>0.0</td>\n",
@@ -378,7 +378,7 @@
" <td>(GD639917, GD3995806, GD3789919, GD1316637, GD...</td>\n",
" <td>0.200000</td>\n",
" <td>5</td>\n",
- " <td>0</td>\n",
+ " <td>0.0</td>\n",
" <td>2</td>\n",
" <td>0.0</td>\n",
" <td>0.0</td>\n",
@@ -394,7 +394,7 @@
" <td>(GD639917, GD3789919, GD2055944)</td>\n",
" <td>0.000000</td>\n",
" <td>3</td>\n",
- " <td>0</td>\n",
+ " <td>0.0</td>\n",
" <td>0</td>\n",
" <td>0.0</td>\n",
" <td>0.0</td>\n",
@@ -410,7 +410,7 @@
" <td>(GD5526704, GD976842, GD1316637, GD2055944)</td>\n",
" <td>0.333333</td>\n",
" <td>4</td>\n",
- " <td>2</td>\n",
+ " <td>1.0</td>\n",
" <td>0</td>\n",
" <td>0.0</td>\n",
" <td>0.0</td>\n",
@@ -426,12 +426,12 @@
" <td>(GD2908705, GD1404948, GD9642903, GD3995806, G...</td>\n",
" <td>0.285714</td>\n",
" <td>7</td>\n",
- " <td>4</td>\n",
+ " <td>2.0</td>\n",
" <td>2</td>\n",
" <td>1.0</td>\n",
" <td>1.0</td>\n",
" <td>[A-ADM1, P-PPL, P-PPL, P-PPL, A-ADM1, A-ADM1, ...</td>\n",
- " <td>A-ADM1</td>\n",
+ " <td>P-PPL</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
@@ -472,29 +472,29 @@
"8 (GD2908705, GD1404948, GD9642903, GD3995806, G... 0.285714 7 \n",
"\n",
" NB_ED_ADJ NB_ED_INC R_ADJ_INC R_INC_ADJ \\\n",
- "0 0 0 0.0 0.0 \n",
- "1 0 0 0.0 0.0 \n",
- "2 0 0 0.0 0.0 \n",
- "3 0 0 0.0 0.0 \n",
- "4 0 1 0.0 0.0 \n",
- "5 0 2 0.0 0.0 \n",
- "6 0 0 0.0 0.0 \n",
- "7 2 0 0.0 0.0 \n",
- "8 4 2 1.0 1.0 \n",
+ "0 0.0 0 0.0 0.0 \n",
+ "1 0.0 0 0.0 0.0 \n",
+ "2 0.0 0 0.0 0.0 \n",
+ "3 0.0 0 0.0 0.0 \n",
+ "4 0.0 1 0.0 0.0 \n",
+ "5 0.0 2 0.0 0.0 \n",
+ "6 0.0 0 0.0 0.0 \n",
+ "7 1.0 0 0.0 0.0 \n",
+ "8 2.0 2 1.0 1.0 \n",
"\n",
" CLASS MEAN_LVL \n",
"0 [P-PPLA, P-PPL, P-PPLA, A-ADM1, P-PPLA] P-PPLA \n",
"1 [P-PPL] P-PPL \n",
"2 [A-PCLI] A-PCLI \n",
- "3 [A-PCLI, P-PPL] P-PPL \n",
+ "3 [A-PCLI, P-PPL] A-PCLI \n",
"4 [A-PCLI, A-ADM1, P-PPLA, P-PPLC, A-PCLI, A-PCLI] A-PCLI \n",
"5 [A-PCLI, P-PPL, A-ADM1, P-PPLA, P-PPLC, A-PCLI... A-PCLI \n",
"6 [A-PCLI, A-ADM1, P-PPLA, P-PPLC, A-PCLI] A-PCLI \n",
"7 [A-PCLI, A-PCLI, A-PCLI, A-PCLI] A-PCLI \n",
- "8 [A-ADM1, P-PPL, P-PPL, P-PPL, A-ADM1, A-ADM1, ... A-ADM1 "
+ "8 [A-ADM1, P-PPL, P-PPL, P-PPL, A-ADM1, A-ADM1, ... P-PPL "
]
},
- "execution_count": 10,
+ "execution_count": 26,
"metadata": {},
"output_type": "execute_result"
}
@@ -514,29 +514,29 @@
},
{
"cell_type": "code",
- "execution_count": 11,
+ "execution_count": 27,
"metadata": {
"ExecuteTime": {
- "end_time": "2018-07-16T13:19:13.367470Z",
- "start_time": "2018-07-16T13:19:13.158508Z"
+ "end_time": "2018-07-16T13:58:27.641675Z",
+ "start_time": "2018-07-16T13:58:27.457994Z"
}
},
"outputs": [
{
"data": {
"text/plain": [
- "<matplotlib.axes._subplots.AxesSubplot at 0x10a1db828>"
+ "<matplotlib.axes._subplots.AxesSubplot at 0x11409d5c0>"
]
},
- "execution_count": 11,
+ "execution_count": 27,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
- "image/png": "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\n",
+ "image/png": "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\n",
"text/plain": [
- "<matplotlib.figure.Figure at 0x104a11ac8>"
+ "<matplotlib.figure.Figure at 0x114971860>"
]
},
"metadata": {},
@@ -556,29 +556,29 @@
},
{
"cell_type": "code",
- "execution_count": 12,
+ "execution_count": 28,
"metadata": {
"ExecuteTime": {
- "end_time": "2018-07-16T13:19:13.528700Z",
- "start_time": "2018-07-16T13:19:13.369629Z"
+ "end_time": "2018-07-16T13:58:28.848371Z",
+ "start_time": "2018-07-16T13:58:28.698404Z"
}
},
"outputs": [
{
"data": {
"text/plain": [
- "<matplotlib.axes._subplots.AxesSubplot at 0x10a2c9ba8>"
+ "<matplotlib.axes._subplots.AxesSubplot at 0x114c629b0>"
]
},
- "execution_count": 12,
+ "execution_count": 28,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
- "image/png": "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\n",
+ "image/png": "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\n",
"text/plain": [
- "<matplotlib.figure.Figure at 0x10a2d4fd0>"
+ "<matplotlib.figure.Figure at 0x114c6bcc0>"
]
},
"metadata": {},
@@ -598,11 +598,11 @@
},
{
"cell_type": "code",
- "execution_count": 13,
+ "execution_count": 31,
"metadata": {
"ExecuteTime": {
- "end_time": "2018-07-16T13:19:13.545917Z",
- "start_time": "2018-07-16T13:19:13.530312Z"
+ "end_time": "2018-07-16T13:58:53.263760Z",
+ "start_time": "2018-07-16T13:58:53.250263Z"
}
},
"outputs": [
@@ -655,7 +655,7 @@
" <td>0.014706</td>\n",
" <td>0.293545</td>\n",
" <td>6.455882</td>\n",
- " <td>4.117647</td>\n",
+ " <td>2.058824</td>\n",
" <td>2.705882</td>\n",
" <td>0.559194</td>\n",
" <td>1.049048</td>\n",
@@ -666,7 +666,7 @@
" <td>0.030303</td>\n",
" <td>0.379501</td>\n",
" <td>4.636364</td>\n",
- " <td>2.969697</td>\n",
+ " <td>1.484848</td>\n",
" <td>1.303030</td>\n",
" <td>0.398990</td>\n",
" <td>0.240657</td>\n",
@@ -678,8 +678,8 @@
"text/plain": [
" ID_TEXT MIXED DENSITY NB_NODE NB_ED_ADJ NB_ED_INC \\\n",
"IS_BILAN \n",
- "BILAN 51.588235 0.014706 0.293545 6.455882 4.117647 2.705882 \n",
- "EPIDEMIE 46.727273 0.030303 0.379501 4.636364 2.969697 1.303030 \n",
+ "BILAN 51.588235 0.014706 0.293545 6.455882 2.058824 2.705882 \n",
+ "EPIDEMIE 46.727273 0.030303 0.379501 4.636364 1.484848 1.303030 \n",
"\n",
" R_ADJ_INC R_INC_ADJ \n",
"IS_BILAN \n",
@@ -687,7 +687,7 @@
"EPIDEMIE 0.398990 0.240657 "
]
},
- "execution_count": 13,
+ "execution_count": 31,
"metadata": {},
"output_type": "execute_result"
}
@@ -730,101 +730,6 @@
"\n",
"Est-ce qu'un ratio élevé INC/ADJ traduit une information concernant la dispertion d'une maladie ? \n"
]
- },
- {
- "cell_type": "code",
- "execution_count": 29,
- "metadata": {
- "ExecuteTime": {
- "end_time": "2018-07-16T13:32:14.236322Z",
- "start_time": "2018-07-16T13:32:14.233834Z"
- }
- },
- "outputs": [],
- "source": [
- "from ipywidgets import interact, interactive, fixed, interact_manual\n",
- "import ipywidgets as widgets"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 15,
- "metadata": {
- "ExecuteTime": {
- "end_time": "2018-07-16T13:19:13.645466Z",
- "start_time": "2018-07-16T13:19:13.554626Z"
- }
- },
- "outputs": [
- {
- "data": {
- "application/vnd.jupyter.widget-view+json": {
- "model_id": "479f1072f38a4a5991a18cb3f381a035",
- "version_major": 2,
- "version_minor": 0
- },
- "text/plain": [
- "interactive(children=(IntSlider(value=0, description='x'), Output()), _dom_classes=('widget-interact',))"
- ]
- },
- "metadata": {},
- "output_type": "display_data"
- }
- ],
- "source": [
- "from nxpd import draw\n",
- "\n",
- "def f(x):\n",
- " global data_bilan\n",
- " return draw(data_bilan[data_bilan[\"IS_BILAN\"]==\"BILAN\"].iloc[x][\"GRAPH_C\"],show=\"ipynb\")\n",
- "interact(f, x=widgets.IntSlider(min=0,max=100,step=1));"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 16,
- "metadata": {
- "ExecuteTime": {
- "end_time": "2018-07-16T13:19:33.768901Z",
- "start_time": "2018-07-16T13:19:33.762369Z"
- }
- },
- "outputs": [],
- "source": [
- "dd=data_bilan.groupby(\"IS_BILAN\").mean()"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 28,
- "metadata": {
- "ExecuteTime": {
- "end_time": "2018-07-16T13:27:13.438513Z",
- "start_time": "2018-07-16T13:27:13.433856Z"
- }
- },
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "|:---------|---------:|---------:|\n",
- "| BILAN | 0.559194 | 1.04905 |\n",
- "| EPIDEMIE | 0.39899 | 0.240657 |\n"
- ]
- }
- ],
- "source": [
- "from tabulate import tabulate\n",
- "print(tabulate(dd[[\"R_ADJ_INC\", \"R_INC_ADJ\"]],tablefmt=\"pipe\"))"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {},
- "outputs": [],
- "source": []
}
],
"metadata": {