From a09ea2d841595864135a7199b5a6cb4c3ad8a666 Mon Sep 17 00:00:00 2001
From: Thomas PEROT <thomas.perot@inrae.fr>
Date: Mon, 19 Jun 2023 08:15:41 +0200
Subject: [PATCH] fichier jupiter d'un participant pour evaluation, exercice
 evalutation par les pairs

---
 .../evaluations/module3_exo3_exercice.ipynb   | 1341 +++++++++++++++++
 1 file changed, 1341 insertions(+)
 create mode 100644 module3/exo3/evaluations/module3_exo3_exercice.ipynb

diff --git a/module3/exo3/evaluations/module3_exo3_exercice.ipynb b/module3/exo3/evaluations/module3_exo3_exercice.ipynb
new file mode 100644
index 0000000..c120633
--- /dev/null
+++ b/module3/exo3/evaluations/module3_exo3_exercice.ipynb
@@ -0,0 +1,1341 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Analyse des dialogues dans l'Avare de Molière"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Auteur : Emile Pierret"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Nous proposons d'étudier la part de dialogue de chacun des personnages de l'Avare de Molière."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "%matplotlib inline\n",
+    "import numpy as np\n",
+    "import matplotlib.pylab as plt\n",
+    "import matplotlib.patches as mpatches\n",
+    "import copy"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Voici la liste des versions de package utilisées :\n",
+    "- Version numpy       : 1.15.2\n",
+    "- Version matplotlib  : 2.2.3\n",
+    "- Version copy        : 1.15.2\n",
+    "- Version python      : 3.6.4"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# 1. Etude des données"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Fichier du texte"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Le texte utilisé a été télécharger au lien suivant: http://dramacode.github.io/markdown/moliere_avare.txt, le 1er avril 2023 à 14h09. Il est disponible sur le git, au nom de \"moliere_avare.txt\""
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "fichier = open(\"moliere_avare.txt\", \"r\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "lignes = fichier.readlines()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Voici, ci-dessous, un extrait du texte :"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "---\n",
+      "\n",
+      "identifier: moliere_avare  \n",
+      "\n",
+      "creator: Molière.  \n",
+      "\n",
+      "date: 1668  \n",
+      "\n",
+      "title: L'Avare. Comédie  \n",
+      "\n",
+      "---\n",
+      "\n",
+      "\n",
+      "\n",
+      "\n",
+      "\n",
+      "L'AVARE,\n",
+      "\n",
+      "\n",
+      "\n",
+      "COMÉDIE.\n",
+      "\n",
+      "\n",
+      "\n",
+      "Par J.B.P. MOLIÈRE.\n",
+      "\n",
+      "\n",
+      "\n",
+      "À PARIS, Chez JEAN RIBOU, au Palais, vis à vis la Porte de l'Église de la Sainte Chapelle, à l'Image Saint-Louis. M. DC. LXIX. *AVEC PRIVILÈGE DU ROI*\n",
+      "\n",
+      "\n",
+      "\n",
+      "\n",
+      "\n",
+      "\n",
+      "\n",
+      "# ACTEURS.\n",
+      "\n",
+      " – Harpagon, Père de Cléante et d'Élise, et Amoureux de Mariane.\n",
+      "\n",
+      " – Cléante, Fils d'Harpagon, Amant de Mariane.\n",
+      "\n",
+      " – Élise, Fille d'Harpagon, Amante de Valère.\n",
+      "\n",
+      " – Valère, Fils d'Anselme, et Amant d'Élise.\n",
+      "\n",
+      " – Mariane, Amante de Cléante, et aimée d'Harpagon.\n",
+      "\n",
+      " – Anselme, Père de Valère et de Mariane.\n",
+      "\n",
+      " – Frosine, Femme d'Intrigue.\n",
+      "\n",
+      " – Maitre Simon, Courtier.\n",
+      "\n",
+      " – Maitre Jacques, Cuisinier et Cocher d'Harpagon.\n",
+      "\n",
+      " – La Flèche, Valet de Cléante.\n",
+      "\n",
+      " – Dame Claude, Servante d'Harpagon.\n",
+      "\n",
+      " – Brindavoine, laquais d'Harpagon.\n",
+      "\n",
+      " – La Merluche, laquais d'Harpagon.\n",
+      "\n",
+      " – Le commissaire, et son clerc.\n",
+      "\n",
+      "La Scène est à Paris.\n",
+      "\n",
+      "\n",
+      "\n",
+      "\n",
+      "\n",
+      "\n",
+      "\n",
+      "# L'Avare, *Comédie.*.\n",
+      "\n",
+      "\n",
+      "\n",
+      "\n",
+      "\n",
+      "## Acte Premier.\n",
+      "\n",
+      "\n",
+      "\n",
+      "\n",
+      "\n",
+      "### Scène Première.\n",
+      "\n",
+      "Valère, Élise\n",
+      "\n",
+      "\n",
+      "\n",
+      "\n",
+      "\n",
+      "    VALÈRE.\n",
+      "\n",
+      "Hé quoi, charmante Élise, vous devenez mélancolique, après les obligeantes assurances que vous avez eu la bonté de me donner de votre foi ?Je vous vois soupirer, hélas, au milieu de ma joie !Est-ce du regret, dites-moi, de m'avoir fait heureux ? et vous repentez-vous de cet engagement où mes feux ont pu vous contraindre ?\n",
+      "\n",
+      "\n",
+      "\n"
+     ]
+    }
+   ],
+   "source": [
+    "for ligne in lignes[:50] :\n",
+    "    print(ligne)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Liste des personnages"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "On peut lire dans le tete ci-dessus la liste des personnages. On peut également voir que chacune de leur réplique est précédée de leur nom en majuscule. Pour faciliter le travail par la suite, on commence par faire la liste des personnages au format majuscule."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "metadata": {
+    "scrolled": false
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      " – Harpagon, Père de Cléante et d'Élise, et Amoureux de Mariane.\n",
+      "\n",
+      " – Cléante, Fils d'Harpagon, Amant de Mariane.\n",
+      "\n",
+      " – Élise, Fille d'Harpagon, Amante de Valère.\n",
+      "\n",
+      " – Valère, Fils d'Anselme, et Amant d'Élise.\n",
+      "\n",
+      " – Mariane, Amante de Cléante, et aimée d'Harpagon.\n",
+      "\n",
+      " – Anselme, Père de Valère et de Mariane.\n",
+      "\n",
+      " – Frosine, Femme d'Intrigue.\n",
+      "\n",
+      " – Maitre Simon, Courtier.\n",
+      "\n",
+      " – Maitre Jacques, Cuisinier et Cocher d'Harpagon.\n",
+      "\n",
+      " – La Flèche, Valet de Cléante.\n",
+      "\n",
+      " – Dame Claude, Servante d'Harpagon.\n",
+      "\n",
+      " – Brindavoine, laquais d'Harpagon.\n",
+      "\n",
+      " – La Merluche, laquais d'Harpagon.\n",
+      "\n",
+      " – Le commissaire, et son clerc.\n",
+      "\n",
+      "['HARPAGON', 'CLÉANTE', 'ÉLISE', 'VALÈRE', 'MARIANE', 'ANSELME', 'FROSINE', 'MAITRE SIMON', 'MAITRE JACQUES', 'LA FLÈCHE', 'DAME CLAUDE', 'BRINDAVOINE', 'LA MERLUCHE', 'LE COMMISSAIRE']\n"
+     ]
+    }
+   ],
+   "source": [
+    "persos = []\n",
+    "for k in range(19,33) :\n",
+    "    l = lignes[k]\n",
+    "    print(l)\n",
+    "    i = 3\n",
+    "    perso = \"\"\n",
+    "    while l[i] != \",\" :\n",
+    "        perso += l[i]\n",
+    "        i+=1\n",
+    "    persos.append(perso.upper())\n",
+    "print(persos)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Les répliques pour Maître Simon et Maître Jacques sont précédes de leur nom, avec leurs accent circonflexes, contrairement à ce qui était indiqué dans la liste des personnages. On modifie donc la liste ci-dessous."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "persos[7] = 'MAÎTRE SIMON'\n",
+    "persos[8] = 'MAÎTRE JACQUES'"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Séparation des chaînes de caractères en mots"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "L'objectif par la suite est de compter le nombre de mots prononcé par chaque personnage. On commence donc par écrire une fonction qui sépare une réplique en mots. Notre méthodologie est la suivante :\n",
+    "1. Extraire les lettres qui apparaissent dans le texte de la pièce (hors ponctuation)\n",
+    "2. Ecrire une fonction qui sépare une réplique donnée en la liste de ces mots, constitués de lettres."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Dans un premier temps, on fait la liste des caractères présents dans le texte puis on sélectionne manuellement ce que l'on considère comme une lettre."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "caractères = []\n",
+    "for k in range(len(lignes)) :\n",
+    "    ligne = lignes[k]\n",
+    "    for i in range(len(ligne)) :\n",
+    "        if not(ligne[i] in caractères) :\n",
+    "            caractères.append(ligne[i])"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Ci-dessous, la liste des caractères :"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "['-',\n",
+       " '\\n',\n",
+       " 'i',\n",
+       " 'd',\n",
+       " 'e',\n",
+       " 'n',\n",
+       " 't',\n",
+       " 'f',\n",
+       " 'r',\n",
+       " ':',\n",
+       " ' ',\n",
+       " 'm',\n",
+       " 'o',\n",
+       " 'l',\n",
+       " '_',\n",
+       " 'a',\n",
+       " 'v',\n",
+       " 'c',\n",
+       " 'M',\n",
+       " 'è',\n",
+       " '.',\n",
+       " '1',\n",
+       " '6',\n",
+       " '8',\n",
+       " 'L',\n",
+       " \"'\",\n",
+       " 'A',\n",
+       " 'C',\n",
+       " 'é',\n",
+       " 'V',\n",
+       " 'R',\n",
+       " 'E',\n",
+       " ',',\n",
+       " 'O',\n",
+       " 'É',\n",
+       " 'D',\n",
+       " 'I',\n",
+       " 'P',\n",
+       " 'J',\n",
+       " 'B',\n",
+       " 'È',\n",
+       " 'À',\n",
+       " 'S',\n",
+       " 'h',\n",
+       " 'z',\n",
+       " 'N',\n",
+       " 'U',\n",
+       " 'u',\n",
+       " 's',\n",
+       " 'à',\n",
+       " 'g',\n",
+       " 'p',\n",
+       " 'X',\n",
+       " '*',\n",
+       " 'G',\n",
+       " '#',\n",
+       " 'T',\n",
+       " '–',\n",
+       " 'H',\n",
+       " 'x',\n",
+       " 'F',\n",
+       " 'q',\n",
+       " 'b',\n",
+       " '?',\n",
+       " 'j',\n",
+       " '!',\n",
+       " 'ù',\n",
+       " 'y',\n",
+       " 'î',\n",
+       " 'ê',\n",
+       " '\\xa0',\n",
+       " ';',\n",
+       " 'œ',\n",
+       " 'ç',\n",
+       " 'ô',\n",
+       " 'â',\n",
+       " 'û',\n",
+       " 'ï',\n",
+       " 'Q',\n",
+       " '…',\n",
+       " 'Ê',\n",
+       " 'Ô',\n",
+       " 'Y',\n",
+       " 'Î',\n",
+       " '<',\n",
+       " '>']"
+      ]
+     },
+     "execution_count": 8,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "caractères"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Ci-dessous, la liste des caractères étant considérés comme des lettres et pouvant donc constituer un mot."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "lettres = ['i',\n",
+    " 'd',\n",
+    " 'e',\n",
+    " 'n',\n",
+    " 't',\n",
+    " 'f',\n",
+    " 'r',\n",
+    " 'm',\n",
+    " 'o',\n",
+    " 'l',\n",
+    " 'a',\n",
+    " 'v',\n",
+    " 'c',\n",
+    " 'M',\n",
+    " 'è',\n",
+    " 'L',\n",
+    " 'A',\n",
+    " 'C',\n",
+    " 'é',\n",
+    " 'V',\n",
+    " 'R',\n",
+    " 'E',\n",
+    " 'O',\n",
+    " 'É',\n",
+    " 'D',\n",
+    " 'I',\n",
+    " 'P',\n",
+    " 'J',\n",
+    " 'B',\n",
+    " 'È',\n",
+    " 'À',\n",
+    " 'S',\n",
+    " 'h',\n",
+    " 'z',\n",
+    " 'N',\n",
+    " 'U',\n",
+    " 'u',\n",
+    " 's',\n",
+    " 'à',\n",
+    " 'g',\n",
+    " 'p',\n",
+    " 'X',\n",
+    " 'G',\n",
+    " 'T',\n",
+    " 'H',\n",
+    " 'x',\n",
+    " 'F',\n",
+    " 'q',\n",
+    " 'b',\n",
+    " 'j',\n",
+    " 'ù',\n",
+    " 'y',\n",
+    " 'î',\n",
+    " 'ê',\n",
+    " 'œ',\n",
+    " 'ç',\n",
+    " 'ô',\n",
+    " 'â',\n",
+    " 'û',\n",
+    " 'ï',\n",
+    " 'Q',\n",
+    " 'Ê',\n",
+    " 'Ô',\n",
+    " 'Y',\n",
+    " 'Î']"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Voici ci-dessous la fonction de séparation des répliques en mots, suivi d'un exemple."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def split_string(s) :\n",
+    "    n = len(s)\n",
+    "    liste_mots = []\n",
+    "    k = 0\n",
+    "    while k < n :\n",
+    "        mot = \"\"\n",
+    "        while (k < n) and not (s[k] in lettres) :\n",
+    "            k +=1\n",
+    "        if k <n :\n",
+    "            mot += s[k]\n",
+    "            k+=1\n",
+    "            while (k < n) and (s[k] in lettres) :\n",
+    "                mot += s[k]\n",
+    "                k +=1\n",
+    "            k+=1\n",
+    "            liste_mots.append(mot)\n",
+    "    return liste_mots"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Hé quoi, charmante Élise, vous devenez mélancolique, après les obligeantes assurances que vous avez eu la bonté de me donner de votre foi ?Je vous vois soupirer, hélas, au milieu de ma joie !Est-ce du regret, dites-moi, de m'avoir fait heureux ? et vous repentez-vous de cet engagement où mes feux ont pu vous contraindre ?\n",
+      "\n"
+     ]
+    },
+    {
+     "data": {
+      "text/plain": [
+       "['Hé',\n",
+       " 'quoi',\n",
+       " 'charmante',\n",
+       " 'Élise',\n",
+       " 'vous',\n",
+       " 'devenez',\n",
+       " 'mélancolique',\n",
+       " 'après',\n",
+       " 'les',\n",
+       " 'obligeantes',\n",
+       " 'assurances',\n",
+       " 'que',\n",
+       " 'vous',\n",
+       " 'avez',\n",
+       " 'eu',\n",
+       " 'la',\n",
+       " 'bonté',\n",
+       " 'de',\n",
+       " 'me',\n",
+       " 'donner',\n",
+       " 'de',\n",
+       " 'votre',\n",
+       " 'foi',\n",
+       " 'Je',\n",
+       " 'vous',\n",
+       " 'vois',\n",
+       " 'soupirer',\n",
+       " 'hélas',\n",
+       " 'au',\n",
+       " 'milieu',\n",
+       " 'de',\n",
+       " 'ma',\n",
+       " 'joie',\n",
+       " 'Est',\n",
+       " 'ce',\n",
+       " 'du',\n",
+       " 'regret',\n",
+       " 'dites',\n",
+       " 'moi',\n",
+       " 'de',\n",
+       " 'm',\n",
+       " 'avoir',\n",
+       " 'fait',\n",
+       " 'heureux',\n",
+       " 'et',\n",
+       " 'vous',\n",
+       " 'repentez',\n",
+       " 'vous',\n",
+       " 'de',\n",
+       " 'cet',\n",
+       " 'engagement',\n",
+       " 'où',\n",
+       " 'mes',\n",
+       " 'feux',\n",
+       " 'ont',\n",
+       " 'pu',\n",
+       " 'vous',\n",
+       " 'contraindre']"
+      ]
+     },
+     "execution_count": 11,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "print(lignes[48])\n",
+    "split_string(lignes[48])"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Structure du texte"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Après une étude manuelle du texte, on constate que le pièce a la structure synthaxique suivante :\n",
+    "\n",
+    "PERSONNAGE \\n\n",
+    "\n",
+    "réplique \\n\n",
+    "\n",
+    "\\n\n",
+    "\n",
+    "PERSONNAGE \\n\n",
+    "\n",
+    "réplique \\n\n",
+    "\n",
+    "\\n\n",
+    "\n",
+    "\n",
+    "excepté pour les passages à un nouvel acte ou une nouvelle scène. On utiliser cette structure pour étudier le texte dans la suite. Cependant, en cas de présence de didascalie, cette structure n'est plus respectée. On repère donc, dans le code suivant, ces cas pathologiques dans un liste \"non_habituel\"."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 12,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "acte = 0\n",
+    "perso_prec = 0\n",
+    "scene = 0\n",
+    "k = 40\n",
+    "non_habituel = []\n",
+    "new_scene = False\n",
+    "while k < len(lignes) :\n",
+    "    l = lignes[k]\n",
+    "    est_perso = False\n",
+    "    for perso in persos :\n",
+    "        if perso in l :\n",
+    "            if not new_scene :\n",
+    "                if (k-perso_prec) != 3 :\n",
+    "                    non_habituel.append((acte,scene,k,perso_prec))\n",
+    "            else :\n",
+    "                new_scene = False\n",
+    "            perso_prec = k\n",
+    "            k+=2\n",
+    "    if l[:3] == '###' :\n",
+    "        scene +=1\n",
+    "        new_scene = True\n",
+    "        k+=4\n",
+    "    elif l[:2] == '##' :\n",
+    "        scene = 0\n",
+    "        acte +=1\n",
+    "        k+=3\n",
+    "    else :\n",
+    "        k+=1"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "On peut ensuite afficher ces cas non habituels."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 13,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "['    HARPAGON.\\n', 'Non, coquin, je ne dis pas cela.\\n', 'À part.\\n', '\\n', \"J'enrage.\\n\", 'Haut.\\n', '\\n', \"Je demande si malicieusement tu n'irais point faire courir le bruit que j'en ai.\\n\", '\\n']\n",
+      "['    HARPAGON.\\n', 'Tu fais le raisonneur ; je te baillerai de ce raisonnement-ci par les oreilles.\\n', 'Il lève la main pour lui donner un soufflet.\\n', '\\n', \"Sors d'ici, encore une fois.\\n\", '\\n']\n",
+      "['    HARPAGON.\\n', 'Il tâte le bas de ses chausses.\\n', '\\n', \"Ces grands hauts-de-chausses sont propres à devenir les recéleurs des choses qu'on dérobe ; et je voudrais qu'on en eût fait pendre quelqu'un.\\n\", '\\n']\n",
+      "['    HARPAGON.\\n', \"C'est ce que je veux faire.\\n\", 'Il fouille dans les poches de la Flèche.\\n', '\\n', '\\n']\n",
+      "['    HARPAGON.\\n', \"Certes, ce n'est pas une petite peine que de garder chez soi une grande somme d'argent ; et bienheureux qui a tout son fait bien placé, et ne conserve seulement que ce qu'il faut pour sa dépense. On n'est pas peu embarrassé à inventer dans toute une Maison une cache fidèle : car pour moi les Coffres forts me sont suspects, et je ne veux jamais m'y fier.Je les tiens justement une franche amorce à Voleurs, et c'est toujours la première chose que l'on va attaquer.Cependant je ne sais si j'aurai bien fait, d'avoir enterré dans mon Jardin dix mille écus qu'on me rendit hier. Dix mille écus en or chez soi, est une somme assez…\\n\", \"Ici le frère et la sœur paraissent s'entretenant bas.\\n\", '\\n', \"Ô Ciel ! je me serai trahi moi-même.La chaleur m'aura emporté ; et je crois que j'ai parlé haut en raisonnant tout seul. Qu'est-ce ?\\n\", '\\n']\n",
+      "['    ÉLISE.\\n', 'Elle fait une révérence.\\n', '\\n', \"Je ne veux point me marier, mon Père, s'il vous plaît.\\n\", '\\n']\n",
+      "['    HARPAGON.\\n', 'Il contrefait sa révérence.\\n', '\\n', \"Et moi, ma petite Fille ma mie, je veux que vous vous mariiez, s'il vous plaît.\\n\", '\\n']\n",
+      "['    HARPAGON.\\n', 'Il regarde vers le jardin.\\n', '\\n', \"Ouais. Il me semble que j'entends un Chien qui aboie. N'est-ce point qu'on en voudrait à mon argent ? Ne bougez, je reviens tout à l'heure.\\n\", '\\n']\n",
+      "['    VALÈRE.\\n', \"Enfin notre dernier recours, c'est que la fuite nous peut mettre à couvert de tout ; et si votre amour, belle Élise, est capable d'une fermeté…\\n\", 'Il aperçoit Harpagon.\\n', '\\n', \"Oui, il faut qu'une fille obéisse à son Père. Il ne faut point qu'elle regarde comme un Mari est fait ; et lorsque la grande raison de *sans dot* s'y rencontre, elle doit être prête à prendre tout ce qu'on lui donne.\\n\", '\\n']\n",
+      "['    FROSINE.\\n', \"J'aurais, Monsieur, une petite prière à vous faire.\\n\", 'Il prend un air sévère.\\n', '\\n', \"J'ai un Procès que je suis sur le point de perdre, faute d'un peu d'argent ; et vous pourriez facilement me procurer le gain de ce Procès, si vous aviez quelque bonté pour moi.\\n\", 'Il reprend un air gai.\\n', '\\n', \"Vous ne sauriez croire le plaisir qu'elle aura de vous voir. Ah ! que vous lui plairez ! et que votre fraise à l'antique fera sur son esprit un effet admirable ! Mais, surtout, elle sera charmée de votre haut-de-chausses, attaché au pourpoint avec des aiguillettes.C'est pour la rendre folle de vous ; et un Amant aiguilletté sera pour elle un ragoût merveilleux.\\n\", '\\n']\n",
+      "['    FROSINE.\\n', 'Il reprend son visage sévère.\\n', '\\n', \"En vérité, Monsieur, ce Procès m'est d'une conséquence tout à fait grande. Je suis ruinée, si je le perds ; et quelque petite assistance me rétablirait mes affaires.\\n\", 'Il reprend un air gai.\\n', '\\n', \"Je voudrais que vous eussiez vu le ravissement où elle était, à m'entendre parler de vous. La joie éclatait dans ses yeux, au récit de vos qualités ; et je l'ai mise enfin dans une impatience extrême, de voir ce mariage entièrement conclu.\\n\", '\\n']\n",
+      "['    FROSINE.\\n', 'Il reprend son air sérieux.\\n', '\\n', 'Je vous prie, Monsieur, de me donner le petit secours que je vous demande. Cela me remettra sur pied ; et je vous en serai éternellement obligée.\\n', '\\n']\n",
+      "['    HARPAGON.\\n', 'Allons.Venez çà tous, que je vous distribue mes ordres pour tantôt, et règle à chacun son emploi.Approchez, Dame Claude. Commençons par vous.\\n', 'Elle tient un Balai.\\n', '\\n', \"Bon, vous voilà les armes à la mainJe vous commets au soin de nettoyer partout ; et surtout, prenez garde de ne point frotter les meubles trop fort, de peur de les user.Outre cela, je vous constitue, pendant le souper, au gouvernement des bouteilles ; et s'il s'en écarte quelqu'une, et qu'il se casse quelque chose, je m'en prendrai à vous, et le rabattrai sur vos gages.\\n\", '\\n']\n",
+      "['    HARPAGON.\\n', 'Paix. Rangez cela adroitement du côté de la muraille, et présentez toujours le devant au monde.\\n', \"Harpagon met son chapeau au-devant de son pourpoint, pour montrer à Brindavoine comment il doit faire pour cacher la tache d'huile.\\n\", '\\n', \"Et vous, tenez toujours votre chapeau ainsi, lorsque vous servirez. Pour vous, ma Fille, vous aurez l'œil sur ce que l'on desservira, et prendrez garde qu'il ne s'en fasse aucun dégât. Cela sied bien aux Filles. Mais cependant préparez-vous à bien recevoir ma Maîtresse, qui vous doit venir visiter, et vous mener avec elle à la Foire. Entendez-vous ce que je vous dis ?\\n\", '\\n']\n",
+      "['    MAÎTRE JACQUES.\\n', \"Attendez donc, s'il vous plaît.\\n\", 'Il ôte sa Casaque de Cocher, et paraît vêtu en Cuisinier.\\n', '\\n', '\\n']\n",
+      "['    MAÎTRE JACQUES.\\n', \"Attendez. Ceci s'adresse au cocher.\\n\", 'Il remet sa casaque.\\n', '\\n', 'Vous dites…\\n', '\\n']\n",
+      "['    MAÎTRE JACQUES.\\n', \"Il file doux. Je veux faire le brave, et s'il est assez sot pour me craindre, le frotter quelque peu. Savez-vous bien, Monsieur le rieur, que je ne ris pas, moi\\xa0; et que si vous m'échauffez la tête, je vous ferai rire d'une autre sorte ?\\n\", 'Maître Jacques pousse Valère jusques au bout du théâtre, en le menaçant.\\n', '\\n', '\\n']\n",
+      "['    VALÈRE.\\n', 'Comment, un bâton ?\\n', \"Valère le fait reculer autant qu'il l'a fait.\\n\", '\\n', '\\n']\n",
+      "['    VALÈRE.\\n', 'Et moi, je ne prends point de goût à votre raillerie.\\n', 'Il lui donne des coups de bâton.\\n', '\\n', 'Apprenez que vous êtes un mauvais railleur.\\n', '\\n']\n",
+      "['    CLÉANTE.\\n', \"Il l'ôte du doigt de son Père, et le donne à Mariane.\\n\", '\\n', 'Il faut que vous le voyiez de près.\\n', '\\n']\n",
+      "['    CLÉANTE.\\n', 'Il se met au devant de Mariane, qui le veut rendre.\\n', '\\n', \"Nenni, Madame, il est en de trop belles mains. C'est un présent que mon Père vous a fait.\\n\", '\\n']\n",
+      "['    LA MERLUCHE.\\n', 'Il vient en courant, et fait tomber Harpagon.\\n', '\\n', 'Monsieur…\\n', '\\n']\n",
+      "['    MAÎTRE JACQUES.\\n', 'Vous avez raison. Laissez-moi lui parler, et demeurez là.\\n', \"Il vient trouver Cléante à l'autre bout du théâtre.\\n\", '\\n', '\\n']\n",
+      "['    MAÎTRE JACQUES.\\n', 'Vous avez raison, il se moque. Laissez-moi lui dire deux mots.\\n', 'Il revient à Harpagon.\\n', '\\n', \"Hé bien, votre Fils n'est pas si étrange que vous le dites, et il se met à la raison. Il dit qu'il sait le respect qu'il vous doit, qu'il ne s'est emporté que dans la première chaleur, et qu'il ne fera point refus de se soumettre à ce qu'il vous plaira, pourvu que vous vouliez le traiter mieux que vous ne faites, et lui donner quelque Personne en mariage, dont il ait lieu d'être content.\\n\", '\\n']\n",
+      "['    MAÎTRE JACQUES.\\n', 'Il va au fils.\\n', '\\n', \"Laissez-moi faire. Hé bien, votre Père n'est pas si déraisonnable que vous le faites\\xa0; et il m'a témoigné que ce sont vos emportements qui l'ont mis en colère\\xa0; qu'il n'en veut seulement qu'à votre manière d'agir, et qu'il sera fort disposé à vous accorder ce que vous souhaitez, pourvu que vous vouliez vous y prendre par la douceur, et lui rendre les déférences, les respects, et les soumissions qu'un Fils doit à son père.\\n\", '\\n']\n",
+      "['    HARPAGON.\\n', \"Tu m'a fait plaisir, maître Jacques, et cela mérite une récompense. Va, je m'en souviendrai, je t'assure.\\n\", \"Il tire son mouchoir de sa poche\\xa0; ce qui fait croire à maître Jacques qu'il va lui donner quelque chose.\\n\", '\\n', '\\n']\n"
+     ]
+    }
+   ],
+   "source": [
+    "for s in non_habituel :\n",
+    "    print(lignes[s[3]:s[2]])"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "On voit effectivement apparaître les didascalies telles que \"'Il lève la main pour lui donner un soufflet.\\n'\".\n",
+    "\n",
+    "Ensuite, après une étude manuelle des cas précédents, on en déduit un nombre de mots et de répliques à ajouter au calcul automatique des répliques, reposant sur la structure générale.\n",
+    "\n",
+    "PERSONNAGE \\n\n",
+    "\n",
+    "réplique \\n\n",
+    "\n",
+    "\\n\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 14,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "Ajouter_mots = {}\n",
+    "Ajouter_répliques = {}\n",
+    "\n",
+    "## Harpagon\n",
+    "Ajouter_mots['HARPAGON'] = {}\n",
+    "Ajouter_répliques['HARPAGON'] = {}\n",
+    "#Harpagon, acte 1\n",
+    "Ajouter_mots['HARPAGON'][1] = {}\n",
+    "Ajouter_répliques['HARPAGON'][1] = {}\n",
+    "#Harpagon, acte 1, scène 3\n",
+    "Ajouter_mots['HARPAGON'][1][3] = 0\n",
+    "Ajouter_répliques['HARPAGON'][1][3] = 0\n",
+    "for k in [199,202,205,211,214,252,270] :\n",
+    "    Ajouter_mots['HARPAGON'][1][3] += len(split_string(lignes[k]))\n",
+    "Ajouter_répliques['HARPAGON'][1][3] += 5\n",
+    "#Harpagon, acte 1, scène 4\n",
+    "Ajouter_mots['HARPAGON'][1][4] = 0\n",
+    "Ajouter_répliques['HARPAGON'][1][4] = 0\n",
+    "for k in [382,385,629] :\n",
+    "    Ajouter_mots['HARPAGON'][1][4] += len(split_string(lignes[k]))\n",
+    "Ajouter_répliques['HARPAGON'][1][4] +=2\n",
+    "#Harpagon, acte 1, scène 5\n",
+    "Ajouter_mots['HARPAGON'][1][5] = 0\n",
+    "Ajouter_répliques['HARPAGON'][1][5] = 0\n",
+    "for k in [783] :\n",
+    "    Ajouter_mots['HARPAGON'][1][5] += len(split_string(lignes[k]))\n",
+    "Ajouter_répliques['HARPAGON'][1][5] += 1\n",
+    "\n",
+    "#Harpagon, acte 3\n",
+    "Ajouter_mots['HARPAGON'][3] = {}\n",
+    "Ajouter_répliques['HARPAGON'][3] = {}\n",
+    "#Harpagon, acte 3, scène 1\n",
+    "Ajouter_mots['HARPAGON'][3][1] = 0\n",
+    "Ajouter_répliques['HARPAGON'][3][1] = 0\n",
+    "for k in [1340,1343,1367,1370] :\n",
+    "    Ajouter_mots['HARPAGON'][3][1] += len(split_string(lignes[k]))\n",
+    "Ajouter_répliques['HARPAGON'][3][1] += 2\n",
+    "\n",
+    "#Harpagon, acte 4\n",
+    "Ajouter_mots['HARPAGON'][4] = {}\n",
+    "Ajouter_répliques['HARPAGON'][4] = {}\n",
+    "#Harpagon, acte 4, scène 4\n",
+    "Ajouter_mots['HARPAGON'][4][4] = 0\n",
+    "Ajouter_répliques['HARPAGON'][4][4] = 0\n",
+    "for k in [2355] :\n",
+    "    Ajouter_mots['HARPAGON'][4][4] += len(split_string(lignes[k]))\n",
+    "Ajouter_répliques['HARPAGON'][4][4] += 1\n",
+    "\n",
+    "## Elise\n",
+    "Ajouter_mots['ÉLISE'] = {}\n",
+    "Ajouter_répliques['ÉLISE'] = {}\n",
+    "#Elise, acte 1\n",
+    "Ajouter_mots['ÉLISE'][1] = {}\n",
+    "Ajouter_répliques['ÉLISE'][1] = {}\n",
+    "#Elise, acte 1, scène 4\n",
+    "Ajouter_mots['ÉLISE'][1][4] = 0\n",
+    "Ajouter_répliques['ÉLISE'][1][4] = 0\n",
+    "for k in [624] :\n",
+    "    Ajouter_mots['ÉLISE'][1][4] += len(split_string(lignes[k]))\n",
+    "Ajouter_répliques['ÉLISE'][1][4] += 1\n",
+    "\n",
+    "##Valère\n",
+    "Ajouter_mots['VALÈRE'] = {}\n",
+    "Ajouter_répliques['VALÈRE'] = {}\n",
+    "#Valère, acte 1\n",
+    "Ajouter_mots['VALÈRE'][1] = {}\n",
+    "Ajouter_répliques['VALÈRE'][1] = {}\n",
+    "#Valère, acte 1, scène 5\n",
+    "Ajouter_mots['VALÈRE'][1][5] = 0\n",
+    "Ajouter_répliques['VALÈRE'][1][5] = 0\n",
+    "for k in [813,816] :\n",
+    "    Ajouter_mots['VALÈRE'][1][5] += len(split_string(lignes[k]))\n",
+    "Ajouter_répliques['VALÈRE'][1][5] += 1\n",
+    "\n",
+    "#Valère, acte 3\n",
+    "Ajouter_mots['VALÈRE'][3] = {}\n",
+    "Ajouter_répliques['VALÈRE'][3] = {}\n",
+    "#Valère, acte 3, scène 2\n",
+    "Ajouter_mots['VALÈRE'][3][2] = 0\n",
+    "Ajouter_répliques['VALÈRE'][3][2] = 0\n",
+    "for k in [1631,1663,1666] :\n",
+    "    Ajouter_mots['VALÈRE'][3][2] += len(split_string(lignes[k]))\n",
+    "Ajouter_répliques['VALÈRE'][3][2] += 2\n",
+    "\n",
+    "##Frosine\n",
+    "Ajouter_mots['FROSINE'] = {}\n",
+    "Ajouter_répliques['FROSINE'] = {}\n",
+    "#Frosine, acte 2\n",
+    "Ajouter_mots['FROSINE'][2] = {}\n",
+    "Ajouter_répliques['FROSINE'][2] = {}\n",
+    "#Frosine, acte 2, scène 5\n",
+    "Ajouter_mots['FROSINE'][2][5] = 0\n",
+    "Ajouter_répliques['FROSINE'][2][5] = 0\n",
+    "for k in [1278,1281,1284,1292,1295,1303] :\n",
+    "    Ajouter_mots['FROSINE'][2][5] += len(split_string(lignes[k]))\n",
+    "Ajouter_répliques['FROSINE'][2][5] += 4\n",
+    "\n",
+    "## Maître Jacques\n",
+    "Ajouter_mots['MAÎTRE JACQUES'] = {}\n",
+    "Ajouter_répliques['MAÎTRE JACQUES'] = {}\n",
+    "## Maître Jacques, acte 3\n",
+    "Ajouter_mots['MAÎTRE JACQUES'][3] = {}\n",
+    "Ajouter_répliques['MAÎTRE JACQUES'][3] = {}\n",
+    "## Maître Jacques, acte 3, scène 1\n",
+    "Ajouter_mots['MAÎTRE JACQUES'][3][1] = 0\n",
+    "Ajouter_répliques['MAÎTRE JACQUES'][3][1] = 0\n",
+    "for k in [1409,1525,1528] :\n",
+    "    Ajouter_mots['MAÎTRE JACQUES'][3][1] += len(split_string(lignes[k]))\n",
+    "Ajouter_répliques['MAÎTRE JACQUES'][3][1] += 2\n",
+    "## Maître Jacques, acte 3, scène 2\n",
+    "Ajouter_mots['MAÎTRE JACQUES'][3][2] = 0\n",
+    "Ajouter_répliques['MAÎTRE JACQUES'][3][2] = 0\n",
+    "for k in [1608] :\n",
+    "    Ajouter_mots['MAÎTRE JACQUES'][3][2] += len(split_string(lignes[k]))\n",
+    "Ajouter_répliques['MAÎTRE JACQUES'][3][2] += 1\n",
+    "\n",
+    "## Maître Jacques, acte 4\n",
+    "Ajouter_mots['MAÎTRE JACQUES'][4] = {}\n",
+    "Ajouter_répliques['MAÎTRE JACQUES'][4] = {}\n",
+    "## Maître Jacques, acte 4, scène 4\n",
+    "Ajouter_mots['MAÎTRE JACQUES'][4][4] = 0\n",
+    "Ajouter_répliques['MAÎTRE JACQUES'][4][4] = 0\n",
+    "for k in [2297,2317,2320,2328] :\n",
+    "    Ajouter_mots['MAÎTRE JACQUES'][4][4] += len(split_string(lignes[k]))\n",
+    "Ajouter_répliques['MAÎTRE JACQUES'][4][4] +=3\n",
+    "\n",
+    "## Cléante\n",
+    "Ajouter_mots['CLÉANTE'] = {}\n",
+    "Ajouter_répliques['CLÉANTE'] = {}\n",
+    "#Cléante, acte 3\n",
+    "Ajouter_mots['CLÉANTE'][3] = {}\n",
+    "Ajouter_répliques['CLÉANTE'][3] = {}\n",
+    "#Cléante, acte 3, scène 4\n",
+    "Ajouter_mots['CLÉANTE'][3][4] = 0\n",
+    "Ajouter_répliques['CLÉANTE'][3][4] = 0\n",
+    "for k in [1867] :\n",
+    "    Ajouter_mots['CLÉANTE'][3][4] += len(split_string(lignes[k]))\n",
+    "Ajouter_répliques['CLÉANTE'][3][4] +=1\n",
+    "#Cléante, acte 3, scène 7\n",
+    "Ajouter_mots['CLÉANTE'][3][7] = 0\n",
+    "Ajouter_répliques['CLÉANTE'][3][7] = 0\n",
+    "for k in [1875] :\n",
+    "    Ajouter_mots['CLÉANTE'][3][7] += len(split_string(lignes[k]))\n",
+    "Ajouter_répliques['CLÉANTE'][3][7] += 1\n",
+    "\n",
+    "##La merluche\n",
+    "Ajouter_mots['LA MERLUCHE'] = {}\n",
+    "Ajouter_répliques['LA MERLUCHE'] = {}\n",
+    "#La merluche, acte 3\n",
+    "Ajouter_mots['LA MERLUCHE'][3] = {}\n",
+    "Ajouter_répliques['LA MERLUCHE'][3] = {}\n",
+    "#La merluche, acte 3, scène 9\n",
+    "Ajouter_mots['LA MERLUCHE'][3][9] = 0\n",
+    "Ajouter_répliques['LA MERLUCHE'][3][9] = 0\n",
+    "for k in [1971] :\n",
+    "    Ajouter_mots['LA MERLUCHE'][3][9] += len(split_string(lignes[k]))\n",
+    "Ajouter_répliques['LA MERLUCHE'][3][9] +=1"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Dénombrement mots et répliques."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Il ne reste plus qu'à traiter automatiquement la structure repérée et ajouter ensuite les cas pathologiques pour compter les mots et les répliques."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 15,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "dico_persos = {}\n",
+    "for perso in persos :\n",
+    "    dico_persos[perso] = 0\n",
+    "\n",
+    "acte = 0\n",
+    "perso_prec = 0\n",
+    "scene = 0\n",
+    "k = 40\n",
+    "new_scene = False\n",
+    "\n",
+    "dico_acte_nombre_mots = {}\n",
+    "dico_acte_nombre_répliques = {}\n",
+    "while k < len(lignes) :\n",
+    "    l = lignes[k]\n",
+    "    for perso in persos :\n",
+    "        if perso in l :\n",
+    "            if not new_scene :\n",
+    "                if (k-perso_prec) == 3 :\n",
+    "                    dico_scene_nombre_mots[perso] += len(split_string(lignes[k+1]))\n",
+    "                    dico_scene_nombre_répliques[perso] +=1\n",
+    "            else :\n",
+    "                new_scene = False\n",
+    "                dico_scene_nombre_mots[perso] += len(split_string(lignes[k+1]))\n",
+    "                dico_scene_nombre_répliques[perso] +=1\n",
+    "            perso_prec = k\n",
+    "            k+=2\n",
+    "    if l[:3] == '###' :\n",
+    "        if scene > 0 :\n",
+    "            dico_acte_nombre_mots[acte][scene] = copy.deepcopy(dico_scene_nombre_mots)\n",
+    "            dico_acte_nombre_répliques[acte][scene] = copy.deepcopy(dico_scene_nombre_répliques)\n",
+    "        scene +=1\n",
+    "        new_scene = True\n",
+    "        k+=4\n",
+    "        dico_scene_nombre_mots = copy.deepcopy(dico_persos)\n",
+    "        dico_scene_nombre_répliques = copy.deepcopy(dico_persos)\n",
+    "    elif l[:2] == '##' :\n",
+    "        if acte > 0 :\n",
+    "            dico_acte_nombre_mots[acte][scene] = copy.deepcopy(dico_scene_nombre_mots)\n",
+    "            dico_acte_nombre_répliques[acte][scene] = copy.deepcopy(dico_scene_nombre_répliques)\n",
+    "        scene = 0\n",
+    "        acte +=1\n",
+    "        k+=3\n",
+    "        dico_acte_nombre_mots[acte] = {}\n",
+    "        dico_acte_nombre_répliques[acte] = {}\n",
+    "    else :\n",
+    "        k+=1\n",
+    "dico_acte_nombre_mots[acte][scene] = copy.deepcopy(dico_scene_nombre_mots)\n",
+    "dico_acte_nombre_répliques[acte][scene] = copy.deepcopy(dico_scene_nombre_répliques)\n",
+    "\n",
+    "for perso_a_ajouter in Ajouter_mots.keys() :\n",
+    "    for acte_a_ajouter in Ajouter_mots[perso_a_ajouter].keys() :\n",
+    "        for scene_a_ajouter in Ajouter_mots[perso_a_ajouter][acte_a_ajouter].keys() :\n",
+    "            dico_acte_nombre_mots[acte][scene][perso_a_ajouter] += Ajouter_mots[perso_a_ajouter][acte_a_ajouter][scene_a_ajouter]\n",
+    "            dico_acte_nombre_répliques[acte][scene][perso_a_ajouter] += Ajouter_répliques[perso_a_ajouter][acte_a_ajouter][scene_a_ajouter]"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "On obtient, pour le nombre de mots par exemple, une donné de la forme suivante:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 16,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "{1: {1: {'HARPAGON': 0, 'CLÉANTE': 0, 'ÉLISE': 491, 'VALÈRE': 630, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 0, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 2: {'HARPAGON': 0, 'CLÉANTE': 762, 'ÉLISE': 154, 'VALÈRE': 0, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 0, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 3: {'HARPAGON': 421, 'CLÉANTE': 0, 'ÉLISE': 0, 'VALÈRE': 0, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 0, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 201, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 4: {'HARPAGON': 1101, 'CLÉANTE': 213, 'ÉLISE': 148, 'VALÈRE': 0, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 0, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 5: {'HARPAGON': 243, 'CLÉANTE': 0, 'ÉLISE': 26, 'VALÈRE': 662, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 0, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}}, 2: {1: {'HARPAGON': 0, 'CLÉANTE': 379, 'ÉLISE': 0, 'VALÈRE': 0, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 0, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 903, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 2: {'HARPAGON': 171, 'CLÉANTE': 127, 'ÉLISE': 0, 'VALÈRE': 0, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 0, 'MAÎTRE SIMON': 197, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 12, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 3: {'HARPAGON': 21, 'CLÉANTE': 0, 'ÉLISE': 0, 'VALÈRE': 0, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 1, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 4: {'HARPAGON': 0, 'CLÉANTE': 0, 'ÉLISE': 0, 'VALÈRE': 0, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 130, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 302, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 5: {'HARPAGON': 520, 'CLÉANTE': 0, 'ÉLISE': 0, 'VALÈRE': 0, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 1291, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}}, 3: {1: {'HARPAGON': 593, 'CLÉANTE': 76, 'ÉLISE': 0, 'VALÈRE': 272, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 0, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 775, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 23, 'LA MERLUCHE': 26, 'LE COMMISSAIRE': 0}, 2: {'HARPAGON': 0, 'CLÉANTE': 0, 'ÉLISE': 0, 'VALÈRE': 83, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 0, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 134, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 3: {'HARPAGON': 0, 'CLÉANTE': 0, 'ÉLISE': 0, 'VALÈRE': 0, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 19, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 11, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 4: {'HARPAGON': 0, 'CLÉANTE': 0, 'ÉLISE': 0, 'VALÈRE': 0, 'MARIANE': 185, 'ANSELME': 0, 'FROSINE': 191, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 5: {'HARPAGON': 105, 'CLÉANTE': 0, 'ÉLISE': 0, 'VALÈRE': 0, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 26, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 6: {'HARPAGON': 70, 'CLÉANTE': 0, 'ÉLISE': 17, 'VALÈRE': 0, 'MARIANE': 35, 'ANSELME': 0, 'FROSINE': 9, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 7: {'HARPAGON': 177, 'CLÉANTE': 580, 'ÉLISE': 0, 'VALÈRE': 5, 'MARIANE': 224, 'ANSELME': 0, 'FROSINE': 41, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 8: {'HARPAGON': 23, 'CLÉANTE': 0, 'ÉLISE': 0, 'VALÈRE': 0, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 0, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 20, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 9: {'HARPAGON': 73, 'CLÉANTE': 40, 'ÉLISE': 0, 'VALÈRE': 7, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 0, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 28, 'LE COMMISSAIRE': 0}}, 4: {1: {'HARPAGON': 0, 'CLÉANTE': 245, 'ÉLISE': 58, 'VALÈRE': 0, 'MARIANE': 236, 'ANSELME': 0, 'FROSINE': 436, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 2: {'HARPAGON': 54, 'CLÉANTE': 14, 'ÉLISE': 3, 'VALÈRE': 0, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 0, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 3: {'HARPAGON': 393, 'CLÉANTE': 418, 'ÉLISE': 0, 'VALÈRE': 0, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 0, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 4: {'HARPAGON': 111, 'CLÉANTE': 98, 'ÉLISE': 0, 'VALÈRE': 0, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 0, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 132, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 5: {'HARPAGON': 129, 'CLÉANTE': 163, 'ÉLISE': 0, 'VALÈRE': 0, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 0, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 6: {'HARPAGON': 0, 'CLÉANTE': 17, 'ÉLISE': 0, 'VALÈRE': 0, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 0, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 47, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 7: {'HARPAGON': 0, 'CLÉANTE': 0, 'ÉLISE': 0, 'VALÈRE': 0, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 0, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}}, 5: {1: {'HARPAGON': 89, 'CLÉANTE': 0, 'ÉLISE': 0, 'VALÈRE': 0, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 0, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 0, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 109}, 2: {'HARPAGON': 182, 'CLÉANTE': 0, 'ÉLISE': 0, 'VALÈRE': 0, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 0, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 348, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 159}, 3: {'HARPAGON': 441, 'CLÉANTE': 0, 'ÉLISE': 0, 'VALÈRE': 641, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 0, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 11, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 4: {'HARPAGON': 124, 'CLÉANTE': 0, 'ÉLISE': 143, 'VALÈRE': 22, 'MARIANE': 0, 'ANSELME': 0, 'FROSINE': 4, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 7, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 5: {'HARPAGON': 258, 'CLÉANTE': 0, 'ÉLISE': 0, 'VALÈRE': 376, 'MARIANE': 192, 'ANSELME': 403, 'FROSINE': 0, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 7, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 0, 'LE COMMISSAIRE': 0}, 6: {'HARPAGON': 570, 'CLÉANTE': 157, 'ÉLISE': 12, 'VALÈRE': 93, 'MARIANE': 36, 'ANSELME': 114, 'FROSINE': 209, 'MAÎTRE SIMON': 0, 'MAÎTRE JACQUES': 275, 'LA FLÈCHE': 0, 'DAME CLAUDE': 0, 'BRINDAVOINE': 0, 'LA MERLUCHE': 1, 'LE COMMISSAIRE': 26}}}\n"
+     ]
+    }
+   ],
+   "source": [
+    "print(dico_acte_nombre_mots)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Précédemment, on a compter le nombre de mots et répliques, par acte et par scène pour chaque personnage. On fait le total pour chaque personnage dans la cellule ci-dessous."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 17,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "Nombre_mots_perso = copy.deepcopy(dico_persos)\n",
+    "Nombre_répliques_perso = copy.deepcopy(dico_persos)\n",
+    "\n",
+    "for perso in persos :\n",
+    "    for acte in dico_acte_nombre_mots.keys() :\n",
+    "        for scene in dico_acte_nombre_mots[acte].keys() :\n",
+    "            Nombre_mots_perso[perso] += dico_acte_nombre_mots[acte][scene][perso]\n",
+    "            Nombre_répliques_perso[perso] += dico_acte_nombre_répliques[acte][scene][perso]"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Affichage des résultats"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Nombre de mots et de répliques par personnages"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Ci-dessous, le nombre de mots par personnage."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 18,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "{'DAME CLAUDE': 0,\n",
+       " 'BRINDAVOINE': 43,\n",
+       " 'LA MERLUCHE': 55,\n",
+       " 'MAÎTRE SIMON': 197,\n",
+       " 'LE COMMISSAIRE': 294,\n",
+       " 'ANSELME': 517,\n",
+       " 'MARIANE': 908,\n",
+       " 'ÉLISE': 1052,\n",
+       " 'LA FLÈCHE': 1465,\n",
+       " 'MAÎTRE JACQUES': 1700,\n",
+       " 'FROSINE': 2357,\n",
+       " 'VALÈRE': 2791,\n",
+       " 'CLÉANTE': 3289,\n",
+       " 'HARPAGON': 5869}"
+      ]
+     },
+     "execution_count": 18,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "dict(sorted(Nombre_mots_perso.items(), key=lambda item:item[1]))"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Ci-dessous, le nombre de répliques par personnage."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 19,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "{'DAME CLAUDE': 0,\n",
+       " 'BRINDAVOINE': 3,\n",
+       " 'MAÎTRE SIMON': 5,\n",
+       " 'LA MERLUCHE': 6,\n",
+       " 'LE COMMISSAIRE': 17,\n",
+       " 'ANSELME': 20,\n",
+       " 'MARIANE': 30,\n",
+       " 'ÉLISE': 49,\n",
+       " 'LA FLÈCHE': 62,\n",
+       " 'FROSINE': 64,\n",
+       " 'MAÎTRE JACQUES': 87,\n",
+       " 'VALÈRE': 103,\n",
+       " 'CLÉANTE': 160,\n",
+       " 'HARPAGON': 354}"
+      ]
+     },
+     "execution_count": 19,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "dict(sorted(Nombre_répliques_perso.items(), key=lambda item:item[1]))"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "On remarque que Dame Claude n'adresse pas la parole et que le plus bavard est Harpagon."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Graphe des personnages"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Le graphe n'est pas exactement le même que celui proposé par [l'OBVIL](https://obvil.sorbonne-universite.fr/corpus/moliere/moliere_avare). Pour chaque scène,\n",
+    "- la longueur de la barre donne le nombre de mots total.\n",
+    "- la longueur de la barre d'un personnage donne le nombre de mots qu'il prononce dans la scène."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 20,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAm8AAAHiCAYAAAC+8CsTAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAIABJREFUeJzt3XmUVdWd9vHnqSpGAbVkEMqSAkFKBksm54ADy1cSLaMYk6CidhIgkcSoHQHRpIOYblvtRI2JccoLtE0TwYEmHTWaQF5jGhmkwFIwREEpZkGRWar2+0edsi+kBgqruOzL97PWXdx79j77/PbVhY9nn3OPQwgCAABAHLLSXQAAAAAOHOENAAAgIoQ3AACAiBDeAAAAIkJ4AwAAiAjhDQAAICKENwAAgIgQ3gDsw/Yc21tsNzvA/gW2g+2cz3HMR20vt11h+/qDHQcAjgSENwCfsV0g6QuSgqTiQ3joEknfkbToEB4TAKJEeAOQaoSk/5H0fyVdl9pgu4Xt+22vsv2x7Vdtt5D0p6TLR7a32T4r6f8Ptt9OzuK9aLtzTQcNITwcQnhF0q5GmRUAZBDCG4BUIyQ9lbz+j+0OKW33Seov6WxJuZJuk1QhaVDSfkwIoVUI4S+2vyzpdklXSGon6f9JmnZopgAAmY3wBkCSZPtcSZ0l/SaEsFDS3yQNT9qyJP2DpJtCCGUhhPIQwmshhN01DDdK0j+HEN4OIeyV9BNJp9V29g0AcGAIbwCqXCfppRDCpuTzf+h/l07bSmquykB3IDpLesD2R7Y/krRZkiXlNWC9AHBEOui7wwBkjuTataskZdtel2xuJukY20WSlqryerSTVHlzQapQzZAfSLo7hPBUI5UMAEcszrwBkKQvSyqX1FPSacnrFFVeqzYihFAh6UlJ/2a7k+1s22clPyeyUZXXvnVNGe8RSeNt95Ik20fb/kpNB7fd1HZzVZ6da2K7ebJUCwDYD385ApAql0d/HUJ4P4Swruol6eeSrk5+w+0fVXkGbr4ql0HvkZQVQtgh6W5Jf06WSc8MITybtP+n7a2S3pQ0tJbjvyRppypvhng0eT+olv4AcMRyCNWteAAAAOBwxJk3AACAiBDeAAAAIkJ4AwAAiAjhDQAAICKENwAAgIhk7I/0tm3bNhQUFKS7DAAADomFCxduCiG0a6Sx2+fk5Dwuqbc48dPYKiS9uXfv3m/2799/Q3UdMja8FRQUaMGCBekuAwCAQ8L2qsYaOycn5/Hjjz/+lHbt2m3JysriN8YaUUVFhTdu3Nhz3bp1j0sqrq4P6RkAANSld7t27bYS3BpfVlZWaNeu3ceqPMtZfZ9DWA8AAIhTFsHt0Em+6xozGuENAAAc9lq2bNk39fODDz543IgRI05M3dajR4+el156aZfUbcOGDSvIy8vrU1hY2LNHjx49n3/++dZVbaeffnqPgoKC3j169OjZr1+/wpKSkmZVbWvWrMnJycnpd++997ZNHe/jjz/Ouvrqq0/Mz8/vfcopp/Ts1avXKffff/9nfRYsWND8zDPPPLmgoKB3586de//gBz/oWFFR8VnNWVlZ/efNm9eiqn/37t17LV++vGl9vouMveYNAAA0jodfnlK069PdDZYhmjdptvfGISNKPs8YixYtah5C0Lx581pv3bo1q02bNhVVbZMmTVp9ww03bPmv//qv1mPGjOl82WWXvVnVNmXKlHcHDRq047777mt788035//hD39YkWw/tqioaPvTTz993A9+8INNVf2vvvrqgoKCgt0rV658Mzs7W2vWrMl5+OGH20rStm3bfPnll3d74IEH3r/iiiu2fvLJJ1lf+tKXTrrnnnvajR8/fqMkdejQYc/EiRM7/va3v333YOfKmTcAAFAvDRncGmq8yZMn51511VUfDho0aOu0adOOqa7PhRdeuG3Dhg1NampbtWrVZ2fenn766dz77rvvg3Xr1jV57733mkhSaWlps8WLFx/1wAMPlGVnZ0uSOnXqtPfuu+9eJ0mPPfbYcQMGDNh2xRVXbJWk1q1bV/zyl798/4EHHuiYcpyP33nnnRapZ/nqi/AGAAAOe7t3784qLCzsWfX653/+506p7c8//3zuiBEjtgwfPnzz9OnTc6sbY+bMmUcPGTLko+rannnmmaMLCwt3StKKFSuabNq0qcn555+/o7i4eMvkyZNzJWnx4sXNTznllB1VwW1/paWlzfv167cjdVuvXr1279ixI2vz5s1ZkpSVlaWbbrpp3Y9//OOO1Q5yAAhvAADgsNesWbOKZcuWvVX1Gj9+/Jqqtrlz57bMzc3de/LJJ+8pLi7eWlpa2nLjxo2fJaw77rjjhBNOOKHPqFGjuvzwhz9cmzruiBEjuhYWFvb8y1/+0uqBBx74QKo8i1dcXLxFkq699trNM2bMqDYMjh079vjCwsKe7du3P1WSQgi2XW39qdtHjRr14aJFi1otW7asXte6VSG8AQCAqE2dOjX33XffbZ6Xl9enc+fOfbZv3549derUY6vaJ02atHrVqlVLx40bV3b99dfvc0PDlClT3l22bNlbL7/88t+6dev2qSTNnDkzd/r06cfl5eX1ueKKK7otX768xdKlS5sVFRXtevvtt1uWl5dLku655551y5Yte2vbtm3ZktSrV6+dCxcubJk6/ltvvdW0ZcuWFccee+xn1+A1adJEY8aMWTdx4sTjD2a+hDcAABCt8vJyzZ49O/eNN94oLSsrW1pWVrZ02rRpK55++ul9zpZlZ2frjjvu2FBRUeGZM2e2qWm8kpKSZjt27MjesGHDkqrxxowZs27KlCm5vXv33n3qqaduv+mmm/L27t0rSdqxY4dDqPwVlZEjR344f/781s8991xrqfIGhhtvvPHE7373u+v2P86YMWM+fPXVV9ts3ry53tf7Ed4AAEC0fve737Xu0KHDni5dunxatW3o0KGfrFixovmqVav2uTkhKytLY8eOXXPffffVeMZr8uTJx33xi1/ckrrta1/72pZnnnkmV5L+/d//feXmzZtzOnfu3KdXr16nfOELXzj5zjvvXC1JrVq1Cs8888yKn/zkJ50KCgp69+zZs1e/fv22jx8//u8ec9W8efMwcuTIDQcT3j5Li5lmwIABgcdjAQCOFLYXhhAGNMbYJSUlK4uKij77uYzD8adCMk1JSUnboqKigura+J03AABQLwSt9GLZFAAAICKENwAAgIgQ3gAAACJCeAMAABlp0qRJ7X/yk5+0S3cdDY0bFgAAQEZq1apV+fe///0P011HQ+PMGwAAOOy9//77OZdccknX/Pz83ieddFKvwYMHd1uyZEmz7t2799q/77Bhwwry8vL6PPLIIx0KCwt79u3btzC1/cILLzzptNNO22fbLbfc0qlFixZ9y8rKPjux1bJly77r1q3Lrnqeatu2bYvat29/atXnXbt2OTs7u3/qM1dvv/32g3pqQn1k7Jm3D/ZU6ObVO9NdBjLM1rUPpLsEIG2eGDgu3SXgMPHzcTOKdm3f03C/83ZU071j/uXKGn9+pKKiQsXFxd2GDx/+4ezZs9+VpNdee63FmjVrmtS0z6RJk1bfcMMNW/bfvmnTpuzS0tKjWrZsWb5s2bKmhYWFe6rajjnmmL2TJk3q8Mtf/rKsatvxxx9fvmzZsrekyoDXqlWr8okTJ66vaq965mr9Z33wOPMGAADqpSGD24GMN3v27NY5OTnhtttu21i17eyzz97ZpUuXPbXtV52pU6ceO2TIkI8uv/zyzZMnT97nEVpf//rXP5w1a1bu+vXrs2va/3BAeAMAAIe1JUuWtCgqKtpRn33uuOOOE6qWMouLiz97GP3TTz+de80112y+7rrrNs+cOXOf8NaqVavyr3/965v+5V/+pcOBHmf37t1Zqcumjz322LH1qfNgZOyyKQAAOHJVt2z6wQcf5KxatarZRRddtC0rK0s5OTlh/vz5zQcOHLirqs+4ceM2FBUV9bzzzjv/7mHy1WHZFAAAYD99+vTZWVJS0vLzjjN58uTcrVu3Zufn5/fJy8vrU1ZW1mzq1Kn7nH1r27Zt+eWXX775vvvua/95j9dYCG8AAOCwdumll36yZ88e33///W2rts2dO7flihUrmtZnnBkzZuQ+++yzfy0rK1taVla2dN68eW8999xzufv3mzBhwvrJkye3Ky8vd0PU39AIbwAA4LCWlZWlWbNm/e2VV15pk5+f37tbt269fvSjH3U68cQTP33vvfeadejQ4dSq15NPPnmstO81b4WFhT2XLFnSbM2aNU0vuOCC7VXjFhYW7mnVqlX5H/7wh6NSj9exY8e9Q4cO3bJnz546w9v+17x95zvfyWv4b2BfDiE09jHSosOp/cLw//5zustAhuGnQnAk46dCDm+2F4YQBjTG2CUlJSuLioo2VX0+1D8VciQqKSlpW1RUVFBdGzcsAACAeiFopRfLpgAAABEhvAEAAESE8AYAABARwhsAAIhCeXm5zj333O5//etf6/UTIZmG8AYAAKKwbNmyZuPHj1/bvXv3ej/TNJMQ3gAAwGEvOzu7/7Bhw0669dZbTywsLOx5++23Hy9Jp59+eo8//elP+zx9Yfbs2a3PP//8blLlI7HOP//8bj169Oh50kkn9Ro8eHA3SVq+fHnT5s2b90v9jbaf//znxx36mdUfPxUCAADqZcKanUU7KhouQ7TM0t67O7Wo9edHDvYZomPHjs274IILtt55550bJGnevHktqtry8/N3H+rnkjYEzrwBAIB6acjg1hjjpVq3bl2T/Pz8z5ZZzzjjjJ2NdaxDhfAGAAAOe/s/huqxxx479kD2u/HGGzd897vfLTjjjDNOHjt27PErV65sUtX2wQcfNEsd84UXXmjVeDNoOCybAgCAw97BLpsOGzZs67nnnrv02WefPfqFF144un///j2XLl1aKrFsCgAAcFjq0KFD+ejRozc/99xz75166qnbX3rppSjOsNWE8AYAADLWrFmzWn/yySdZkrRly5asVatWNevSpUvUPzXCsikAADjsVV3zVvX5ggsu+PgXv/hFmSRdfvnl3XNycoIk9evXb9uNN964sarf/PnzW958880nZmdnhxCCr7322k2DBw/esXz58qZV17xV9b3mmms23XHHHRsO5bwOBuENAADUS8ss7W3onwqpq095efnC6ra//vrry6vbfskll3wiSXfdddf6u+66a/3+7T169Niza9euRfWt9XBAeAMAAPVS12+yoXFxzRsAAEBECG8AAAARIbwBAABEhPAGAACOKFOmTDnm29/+dl666zhYDX7Dgu2LJK0PIXAxIwAAOOyMGDHio23btmXt2rXLzZs3D+mup77qPPNme4LtUttLbC+2fUYdu7wq6Vbb3RqmRMn2INuLbO+1fWVDjQsAAA5/p59+eo+ZM2e2Sd02ceLE9tdcc82JkvTjH/+4fbNmzfp9+OGH2VXts2fPbn3++ef/XRY5/fTTexQUFPR+8MEHjz/ttNNOufjii7tK0i233NKpffv2pxYWFvY86aSTev3qV7/Krdpn2LBhBXl5eX2qnoHat2/fwsabbd1qPfNm+yxJl0jqF0LYbbutpKa17RNC2CFpRMOVKEl6X9L1kv6xgccFAAD19L1FDxRtL9/ZYKt3R2W32Ptgv5tqXLH7yle+8uG0adNyhw0btrVq28yZM3Pvueee1ZI0Y8aM43r37r39qaeeOuZ73/veh3Udb8qUKe8OGjRox/7bR48evX7ixInrly5d2uyss87qef31129p1qxZkKRJkyatvuGGG7Yc3AwbVl1n3jpK2hRC2C1JIYRNIYQ1kmR7oO3XbJfYft12a9vZtu+1Pd/2Utujkr7n2Z5je4btZbafsu2krb/tubYX2n7Rdsf9iwghrAwhLJFU0aCzBwAA9daQwe1Axrv22mu3vPLKK0fv3LnTkrR8+fKmGzZsaHLRRRdtKy0tbbZjx46siRMnlv3mN7/JrW2cA9WnT5/dzZs3r9i0aVN23b0PvbrC20uS8m2/Y/sXtgdLku2mkqZLuimEUCRpiKSdkr4haWsIYaCkgZJG2e6ajNVX0vcl9ZTUVdI5tptIekjSlSGE/pKelHR3g84QAABE7fjjjy8vKiraPnPmzKMlafLkybnFxcVbsrKyNHny5Nwrrrhi88UXX7ztvffea15WVlZnsBwxYkTXqiXQUaNGnbB/+6uvvtqyc+fOu/Ly8j578sMdd9xxQtU+xcXFXRp2hvVT6wRDCNts95f0BUnnS5pue5ykhZLWhhDmJ/22Sp/drNDF9oXJEE1VGdT2Sno9hLA66bdYUoGkjyT1lvT75ERctqS1BzsZ2yMljZSk1nn5BzsMAAA4zFx11VWbp0+ffuw111zz0TPPPJP7+OOPr5SkZ599NveZZ55ZkZ2draFDh26ZMmXKsePHj99Y21g1LZs+8sgjHaZMmdJu9erVTWfOnPnX1LaYlk0VQigPIcwJIfxI0hhJwyRZUnV3Z1jShBDCecmrdwjh5aRtd0q/clUGR0sqDSGclrz6hBAuOtjJhBAeDSEMCCEMaJHb9mCHAQAAh5mrr776oz//+c9tXn311Za7du3KOvfcc3fMmzevxapVq5pdfPHFJ+fl5fWZNWtW7owZMw566XT06NHrV65c+eYTTzzx7re+9a0uO3bscEPOoaHUGt5s97DdPWXTaZJWSVomqZPtgUm/1rZzJL0oaXSyHFq1/1G1HGK5pHbJjRGy3cR2r4OfDgAAyERHH310xZlnnvnJN7/5zYIrrrhisyRNmTIl99Zbb11TVla2tKysbOmGDRuWrFu3ruk777xT682Vdbnuuus+6tOnz/aHH374uIapvmHVdeatlaTJtt+yvUSV16v9Uwhhj6SvSnrIdomk30tqLulxSW9JWmT7TUmPqJal2WScKyXdk4yzWNLZ+/dLbo5YLekrkn5lu7Se8wQAAJH72te+tnn58uUtrr322s2S9Nxzz+VeddVVH6X2GTp06JbJkyfnStJf/vKXNh06dDi16vXyyy8fJe17zdvZZ599cnXH+qd/+qe1Dz/88PHl5eWS9r3mrbCwsOeuXbvSdlbOIUT323QHpMOp/cLw//5zustAhtm69oF0lwCkzRMDx6W7BNTC9sIQwoDGGLukpGRlUVHRpqrPh/qnQo5EJSUlbYuKigqqa2vwJywAAIDMRtBKL55tCgAAEBHCGwAAQEQIbwAAABEhvAEAAKTYtWuXzznnnO4rV65sku5aqkN4AwAASNG8efPw5JNPriopKWme7lqqQ3gDAACHvZYtW/atqa1Hjx49L7300hqfN3rLLbd0at++/alVv9H2ne98J0+STj/99B5/+tOfWqb2nT17duvWrVufdtlll3UbO3ZsfmFhYc/nnnuutSS9//77OZdccknX/Pz83ieddFKvwYMHd1uyZEmz5cuXN+3evXuv/Y/5wx/+sIMkDRs2rCAvL69P1fH79u1b+Hm+C34qBAAA1MvWn95WpJ3bGy5DtDhqb5ub//Wgfn5k0aJFzUMImjdvXuutW7dmtWnTpqK6fqNHj14/ceLE9Qcy5oABA7b98Y9/XJG6raKiQsXFxd2GDx/+4ezZs9+VpNdee63FmjVrmnTp0mVPXWM25LNROfMGAADqpyGD2+ccb/LkyblXXXXVh4MGDdo6bdq0YxqyrFSzZ89unZOTE2677bbPHnp/9tln77z44ou3NdYxa0J4AwAA0Xr++edzR4wYsWX48OGbp0+fXuND6R955JEOVcuWM2fObFPbmAsWLGiV+iis0tLSZkuWLGlRVFS0o6Z9Pvjgg2ap+0yZMqVdanvq47WKi4trXOI9ECybAgCAKM2dO7dlbm7u3pNPPnlP165d93z7298u2LhxY3a7du3K9+/7eZdNn3322Vr3yc/P371s2bK3qj7fcsstnVLbWTYFAABHvKlTp+a+++67zfPy8vp07ty5z/bt27OnTp16bGMcq0+fPjtLSkpa1t2z8RHeAABAdMrLyzV79uzcN954o7SsrGxpWVnZ0mnTpq14+umna1w6/TwuvfTST/bs2eP777+/bdW2uXPntvztb3/bqjGOVxvCGwAAOOzt2rUrq0OHDqdWve66664OHTp02NOlS5dPq/oMHTr0kxUrVjRftWrVAf+47uWXX969asyhQ4d2lf7+mrdf//rXx2ZlZWnWrFl/e+WVV9rk5+f37tatW68f/ehHnU488cRP6zqGtO81b4WFhT137drl+n8LlRxCONh9D2sdTu0Xhv/3n9NdBjLM1rUPpLsEIG2eGDgu3SWgFrYXhhAGNMbYJSUlK4uKijZVfT6cfiokU5WUlLQtKioqqK6NGxYAAEC9ELTSi2VTAACAiBDeAAAAIkJ4AwAAdamoqKg46AvsUT/Jd13tY74kwhsAAKjbmxs3bjyaANf4KioqvHHjxqMlvVlTH25YAAAAtdq7d+83161b9/i6det6ixM/ja1C0pt79+79Zk0dCG8AAKBW/fv33yCpON11oBLpGQAAICKENwAAgIgQ3gAAACJCeAMAAIgI4Q0AACAihDcAAICIZOxPheQ3zdJPT2iR7jKQaU4Yl+4KAABHOM68AQAARITwBgAAEBHCGwAAQEQIbwAAABEhvAEAAESE8AYAABARwhsAAEBECG8AAAARIbwBAABEhPAGAAAQEcIbAABARAhvAAAAESG8AQAARITwBgAAEBHCGwAAQERy0l1AY1n//mbd993/SHcZAGpQNvbydJdw0H56Qot0lwDgCMaZNwAAgIgQ3gAAACJCeAMAAIgI4Q0AACAihDcAAICIEN4AAAAiQngDAACICOENAAAgIoQ3AACAiBDeAAAAIkJ4AwAAiAjhDQAAICKENwAAgIgQ3gAAACJCeAMAAIgI4Q0AACAihDcAAICIEN4AAAAiQngDAACICOENAAAgIoQ3AACAiBDeAAAAIkJ4AwAAiAjhDQAAICKENwAAgIgQ3gAAACJCeAMAAIhIg4c32xfZLmrocQEAAHAA4c32BNultpfYXmz7jDp2eVXSrba7NUyJku1bbL+V1PCK7c4NNTYAAEBMcmprtH2WpEsk9Qsh7LbdVlLT2vYJIeyQNKLhSpQkvSFpQAhhh+1vS/pXSV9t4GMAAAAc9uo689ZR0qYQwm5JCiFsCiGskSTbA22/ZrvE9uu2W9vOtn2v7fm2l9oelfQ9z/Yc2zNsL7P9lG0nbf1tz7W90PaLtjvuX0QI4Y9JKJSk/5F0QkN9AQAAADGpK7y9JCnf9ju2f2F7sCTZbippuqSbQghFkoZI2inpG5K2hhAGShooaZTtrslYfSV9X1JPSV0lnWO7iaSHJF0ZQugv6UlJd9dR0zck/a6e8wQAAMgItS6bhhC22e4v6QuSzpc03fY4SQslrQ0hzE/6bZUqb1aQ1MX2hckQTVUZ1PZKej2EsDrpt1hSgaSPJPWW9PvkRFy2pLU11WP7GkkDJA2uoX2kpJGSdEyrtnVMHQAAID61hjdJCiGUS5ojaY7tpZKuk7RIUqimuyVNCCG8sM9G+zxJu1M2lSfHtqTSEMJZddVhe4ikCZIGVy3jVlPro5IelaT89l2rqw8AACBqtS6b2u5hu3vKptMkrZK0TFIn2wOTfq1t50h6UdLoZDm0av+jajnEckntkhsjZLuJ7V7V1NFX0q8kFYcQNhz49AAAADJLXWfeWkl6yPYxqlz6XCFpZAhhj+2vJm0tVHm92xBJj6tyOXRRckPCRklfrmnwZJwrJT1o++iknp9JKt2v671JLU8ny6vvhxCK6zVTAACADFDXNW8LJZ1dQ9t8SWdW03R78ko1J3lV7Tsm5f1iSYPqqGNIbe0AAABHCh6PBQAAEBHCGwAAQEQIbwAAABEhvAEAAESE8AYAABARwhsAAEBECG8AAAARIbwBAABEhPAGAAAQEcIbAABARAhvAAAAESG8AQAARITwBgAAEBHCGwAAQEQIbwAAABEhvAEAAESE8AYAABARwhsAAEBECG8AAAARIbwBAABEhPAGAAAQEcIbAABARAhvAAAAESG8AQAARITwBgAAEBHCGwAAQEQIbwAAABEhvAEAAEQkJ90FNJYOJ+bqHx8anu4yAAAAGhRn3gAAACJCeAMAAIgI4Q0AACAihDcAAICIEN4AAAAiQngDAACICOENAAAgIoQ3AACAiBDeAAAAIkJ4AwAAiAjhDQAAICKENwAAgIgQ3gAAACJCeAMAAIgI4Q0AACAiOekuoLGs/3iT7v/dY+kuA2n0rTcWp7uEQ67N7Q+nuwQAQCPjzBsAAEBECG8AAAARIbwBAABEhPAGAAAQEcIbAABARAhvAAAAESG8AQAARITwBgAAEBHCGwAAQEQIbwAAABEhvAEAAESE8AYAABARwhsAAEBECG8AAAARIbwBAABEhPAGAAAQEcIbAABARAhvAAAAESG8AQAARITwBgAAEBHCGwAAQEQIbwAAABEhvAEAAESE8AYAABARwhsAAEBECG8AAAARIbwBAABEhPAGAAAQkQYPb7Yvsl3U0OMCAADgAMKb7Qm2S20vsb3Y9hl17PKqpFttd2uYEiXbo20vTY7/qu2eDTU2AABATHJqa7R9lqRLJPULIey23VZS09r2CSHskDSi4UqUJP1HCOGRpKZiSf8m6eIGPgYAAMBhr64zbx0lbQoh7JakEMKmEMIaSbI90PZrtktsv267te1s2/fanp+cKRuV9D3P9hzbM2wvs/2UbSdt/W3Ptb3Q9ou2O+5fRAhha8rHoySFhpg8AABAbGo98ybpJUk/tP2OpJclTQ8hzLXdVNJ0SV8NIcy33UbSTknfkLQ1hDDQdnNJr9n+fTJWX0m9JK2R9GdJ59ieJ+khSZeFEDba/qqkuyX9w/6F2L5R0i2qPPN3QXXF2h4paaQkHds+94C/BAAAgFjUGt5CCNts95f0BUnnS5pue5ykhZLWhhDmJ/22SpU3K0jqYvvCZIimkrpK2ivp9RDC6qTfYkkFkj6S1FvS75MTcdmS1tZQy8OSHrY9XNIdkq6rps+jkh6VpPzuBZydAwAAGaeuM28KIZRLmiNpju2lqgxNi1T90qUlTQghvLDPRvs8SbtTNpUnx7ahM6c1AAAQNklEQVSk0hDCWfWo+T8l/bIe/QEAADJGrde82e5hu3vKptMkrZK0TFIn2wOTfq1t50h6UdJo201S9j+qlkMsl9QuuTFCtpvY7lVNHak1fEnSX+ueGgAAQOap68xbK0kP2T5GlUufKySNDCHsSa5Pe8h2C1Ve7zZE0uOqXA5dlNyQsFHSl2saPBnnSkkP2j46qednkkr36zrG9hBJn0raomqWTAEAAI4EdV3ztlDS2TW0zZd0ZjVNtyevVHOSV9W+Y1LeL5Y0qI46bqqtHQAA4EjB47EAAAAiQngDAACICOENAAAgIoQ3AACAiBDeAAAAIkJ4AwAAiAjhDQAAICKENwAAgIgQ3gAAACJCeAMAAIgI4Q0AACAihDcAAICIEN4AAAAiQngDAACICOENAAAgIoQ3AACAiBDeAAAAIkJ4AwAAiAjhDQAAICKENwAAgIgQ3gAAACJCeAMAAIgI4Q0AACAihDcAAICIEN4AAAAiQngDAACICOENAAAgIjnpLqCxdDi6rW4d+q10l4F0GpruAgAAaHiceQMAAIgI4Q0AACAihDcAAICIEN4AAAAiQngDAACICOENAAAgIoQ3AACAiBDeAAAAIkJ4AwAAiAjhDQAAICKENwAAgIgQ3gAAACJCeAMAAIgI4Q0AACAihDcAAICIEN4AAAAikpPuAhrL+o836f7fPZbuMgAgOuG/j0p3CWlXNvbydJcA1IgzbwAAABEhvAEAAESE8AYAABARwhsAAEBECG8AAAARIbwBAABEhPAGAAAQEcIbAABARAhvAAAAESG8AQAARITwBgAAEBHCGwAAQEQIbwAAABEhvAEAAESE8AYAABARwhsAAEBECG8AAAARIbwBAABEhPAGAAAQEcIbAABARAhvAAAAESG8AQAARITwBgAAEBHCGwAAQEQIbwAAABEhvAEAAESE8AYAABCRBg9vti+yXdTQ4wIAAOAAwpvtCbZLbS+xvdj2GXXs8qqkW213a5gS96nlStvB9oCGHhsAACAGObU12j5L0iWS+oUQdttuK6lpbfuEEHZIGtFwJX5WS2tJ35M0r6HHBgAAiEVdZ946StoUQtgtSSGETSGENZJke6Dt12yX2H7ddmvb2bbvtT3f9lLbo5K+59meY3uG7WW2n7LtpK2/7bm2F9p+0XbHGmq5S9K/StrVIDMHAACIUF3h7SVJ+bbfsf0L24MlyXZTSdMl3RRCKJI0RNJOSd+QtDWEMFDSQEmjbHdNxuor6fuSekrqKukc200kPSTpyhBCf0lPSrp7/yJs95WUH0KY/fmmCwAAELdal01DCNts95f0BUnnS5pue5ykhZLWhhDmJ/22SpU3K0jqYvvCZIimqgxqeyW9HkJYnfRbLKlA0keSekv6fXIiLlvS2tQabGdJ+qmk6+uajO2RkkZK0rHtc+vqDgAAEJ1aw5skhRDKJc2RNMf2UknXSVokKVTT3ZImhBBe2GejfZ6k3SmbypNjW1JpCOGsWkporcqANycJeMdLmmW7OISwYL9aH5X0qCTldy+orj4AAICo1bpsaruH7e4pm06TtErSMkmdbA9M+rW2nSPpRUmjk+XQqv2PquUQyyW1S26MkO0mtnuldgghfBxCaBtCKAghFEj6H0l/F9wAAACOBHWdeWsl6SHbx6hy6XOFpJEhhD22v5q0tVDl9W5DJD2uyuXQRckNCRslfbmmwZNxrpT0oO2jk3p+Jqn0800LAAAgM9V1zdtCSWfX0DZf0pnVNN2evFLNSV5V+45Jeb9Y0qADqray/3kH2hcAACDT8HgsAACAiBDeAAAAIkJ4AwAAiAjhDQAAICKENwAAgIgQ3gAAACJCeAMAAIgI4Q0AACAihDcAAICIEN4AAAAiQngDAACICOENAAAgIoQ3AACAiBDeAAAAIkJ4AwAAiAjhDQAAICKENwAAgIgQ3gAAACJCeAMAAIgI4Q0AACAihDcAAICIEN4AAAAiQngDAACICOENAAAgIoQ3AACAiBDeAAAAIkJ4AwAAiAjhDQAAICIOIaS7hkYxYMCAsGDBgnSXAQDAIWF7YQhhQLrrQOPjzBsAAEBECG8AAAARIbwBAABEhPAGAAAQEcIbAABARAhvAAAAESG8AQAARITwBgAAEBHCGwAAQEQIbwAAABEhvAEAAESE8AYAABARwhsAAEBECG8AAAARIbwBAABEJCfdBTSW9R9v0v2/eyzdZRz2Vve5ptb2rWsfOESVAPF6YuC4dJcA4AjCmTcAAICIEN4AAAAiQngDAACICOENAAAgIoQ3AACAiBDeAAAAIkJ4AwAAiAjhDQAAICKENwAAgIgQ3gAAACJCeAMAAIgI4Q0AACAihDcAAICIEN4AAAAiQngDAACICOENAAAgIoQ3AACAiBDeAAAAIkJ4AwAAiAjhDQAAICKENwAAgIgQ3gAAACJCeAMAAIgI4Q0AACAihDcAAICIEN4AAAAiQngDAACISIOHN9sX2S5q6HEBAABwAOHN9gTbpbaX2F5s+4w6dnlV0q22uzVMiZLt621vTI6/2PY3G2psAACAmOTU1mj7LEmXSOoXQthtu62kprXtE0LYIWlEw5X4mekhhDGNMC4AAEA06jrz1lHSphDCbkkKIWwKIayRJNsDbb9mu8T267Zb2862fa/t+baX2h6V9D3P9hzbM2wvs/2UbSdt/W3Ptb3Q9ou2OzbmhAEAAGJWV3h7SVK+7Xds/8L2YEmy3VTSdEk3hRCKJA2RtFPSNyRtDSEMlDRQ0ijbXZOx+kr6vqSekrpKOsd2E0kPSboyhNBf0pOS7q6hlmHJ0u0M2/kHO2EAAICY1bpsGkLYZru/pC9IOl/SdNvjJC2UtDaEMD/pt1WqvFlBUhfbFyZDNFVlUNsr6fUQwuqk32JJBZI+ktRb0u+TE3HZktZWU8p/SZqWLN2OljRZ0gX7d7I9UtJISTq2fe4BfgUAAADxqDW8SVIIoVzSHElzbC+VdJ2kRZJCNd0taUII4YV9NtrnSdqdsqk8ObYllYYQzqqjhg9TPj4m6Z4a+j0q6VFJyu9eUF19AAAAUat12dR2D9vdUzadJmmVpGWSOtkemPRrbTtH0ouSRifLoVX7H1XLIZZLapfcGCHbTWz3qqaO1OvgiiW9XffUAAAAMk9dZ95aSXrI9jGqXPpcIWlkCGGP7a8mbS1Ueb3bEEmPq3I5dFFyQ8JGSV+uafBknCslPWj76KSen0kq3a/r92wXJzVslnR9vWYJAACQIeq65m2hpLNraJsv6cxqmm5PXqnmJK+qfcekvF8saVAddYyXNL62PgAAAEcCHo8FAAAQEcIbAABARAhvAAAAESG8AQAARITwBgAAEBHCGwAAQEQIbwAAABEhvAEAAESE8AYAABARwhsAAEBECG8AAAARIbwBAABEhPAGAAAQEcIbAABARAhvAAAAESG8AQAARITwBgAAEBHCGwAAQEQIbwAAABEhvAEAAESE8AYAABARwhsAAEBECG8AAAARIbwBAABEhPAGAAAQEcIbAABARAhvAAAAESG8AQAARMQhhHTX0CgGDBgQFixYkO4yAAA4JGwvDCEMSHcdaHyceQMAAIgI4Q0AACAihDcAAICIEN4AAAAiQngDAACICOENAAAgIoQ3AACAiBDeAAAAIkJ4AwAAiAjhDQAAICKENwAAgIgQ3gAAACJCeAMAAIgI4Q0AACAihDcAAICIEN4AAAAiQngDAACICOENAAAgIoQ3AACAiDiEkO4aGoXtTyQtT3cdh1hbSZvSXcQhxpyPDMz5yHAkzllquHl3DiG0a4BxcJjLSXcBjWh5CGFAuos4lGwvYM6ZjzkfGZjzkeNInTcOHsumAAAAESG8AQAARCSTw9uj6S4gDZjzkYE5HxmY85HjSJ03DlLG3rAAAACQiTL5zBsAAEDGycjwZvti28ttr7A9Lt31NATb+bb/aPtt26W2b0q259r+ve2/Jn8em7LP+OQ7WG77/6Sv+s/HdrbtN2zPTj5n9JxtH2N7hu1lyT/vs46AOd+c/Hv9pu1ptptn4pxtP2l7g+03U7bVe562+9temrQ9aNuHei4HqoY535v8+73E9rO2j0lpy8g5p7T9o+1gu23KtujnjEMshJBRL0nZkv4mqaukppJKJPVMd10NMK+Okvol71tLekdST0n/Kmlcsn2cpHuS9z2TuTeT1CX5TrLTPY+DnPstkv5D0uzkc0bPWdJkSd9M3jeVdEwmz1lSnqT3JLVIPv9G0vWZOGdJgyT1k/RmyrZ6z1PS65LOkmRJv5M0NN1zq+ecL5KUk7y/50iYc7I9X9KLklZJaptJc+Z1aF+ZeObtdEkrQgjvhhD2SPpPSZeluabPLYSwNoSwKHn/iaS3VfkfvctU+R97JX9+OXl/maT/DCHsDiG8J2mFKr+bqNg+QdKXJD2esjlj52y7jSr/4n9CkkIIe0IIHymD55zIkdTCdo6klpLWKAPnHEL4k6TN+22u1zxtd5TUJoTwlxBCkDQlZZ/DTnVzDiG8FELYm3z8H0knJO8zds6Jn0q6TVLqxeYZMWccWpkY3vIkfZDyeXWyLWPYLpDUV9I8SR1CCGulyoAnqX3SLVO+h5+p8i+7ipRtmTznrpI2Svp1slT8uO2jlMFzDiGUSbpP0vuS1kr6OITwkjJ4zvup7zzzkvf7b4/VP6jyrJKUwXO2XSypLIRQsl9Txs4ZjScTw1t11wRkzC21tltJminp+yGErbV1rWZbVN+D7UskbQghLDzQXarZFtWcVXkGqp+kX4YQ+krarsqltJpEP+fkGq/LVLlk1EnSUbavqW2XarZFNecDVNM8M2b+tidI2ivpqapN1XSLfs62W0qaIOmH1TVXsy36OaNxZWJ4W63K6wqqnKDKJZjo2W6iyuD2VAjhmWTz+uT0upI/NyTbM+F7OEdSse2Vqlz+vsD2vyuz57xa0uoQwrzk8wxVhrlMnvMQSe+FEDaGED6V9Iyks5XZc05V33mu1v8uM6Zuj4rt6yRdIunqZFlQytw5n6TK/zkpSf4+O0HSItvHK3PnjEaUieFtvqTutrvYbirpa5Jmpbmmzy25y+gJSW+HEP4tpWmWpOuS99dJej5l+9dsN7PdRVJ3VV78Go0QwvgQwgkhhAJV/nP8QwjhGmX2nNdJ+sB2j2TThZLeUgbPWZXLpWfabpn8e36hKq/pzOQ5p6rXPJOl1U9sn5l8XyNS9omC7YsljZVUHELYkdKUkXMOISwNIbQPIRQkf5+tVuUNaOuUoXNGI0v3HRON8ZL0RVXejfk3SRPSXU8DzelcVZ4yXyJpcfL6oqTjJL0i6a/Jn7kp+0xIvoPlivwuJUnn6X/vNs3oOUs6TdKC5J/1c5KOPQLm/GNJyyS9KWmqKu+8y7g5S5qmyuv6PlXlf8C/cTDzlDQg+a7+JunnSn5w/XB81TDnFaq8zqvq77JHMn3O+7WvVHK3aabMmdehffGEBQAAgIhk4rIpAABAxiK8AQAARITwBgAAEBHCGwAAQEQIbwAAABEhvAEAAESE8AYAABARwhsAAEBE/j8assvVHtbFrQAAAABJRU5ErkJggg==\n",
+      "text/plain": [
+       "<Figure size 576x576 with 1 Axes>"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 576x576 with 1 Axes>"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 576x576 with 1 Axes>"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 576x576 with 1 Axes>"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 576x576 with 1 Axes>"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "np.random.seed(0)\n",
+    "colors = np.random.rand(14,3)\n",
+    "for acte in range(1,6) :\n",
+    "    présents = []\n",
+    "    plt.figure(figsize=(8,8))\n",
+    "    x = np.zeros((len(dico_acte_nombre_mots[acte]),len(persos)))\n",
+    "    labels = persos\n",
+    "    for scene in range(1,len(dico_acte_nombre_mots[acte])+1) :\n",
+    "        for l,perso in enumerate(persos) :\n",
+    "                x[scene-1,l] += dico_acte_nombre_mots[acte][scene][perso]\n",
+    "    plt.title('Acte ' + str(acte))\n",
+    "    #plt.bar(labels,x)\n",
+    "    width = 1/2  # épaisseur de chaque bâton\n",
+    "    # Création du diagramme en bâtons (bâtons côte à côte)\n",
+    "    for scene in range(1,len(dico_acte_nombre_mots[acte])+1) :\n",
+    "        bt = 0\n",
+    "        for l,perso in enumerate(persos) :\n",
+    "            plt.barh(-scene, x[scene-1,l],left = bt,color=colors[l])\n",
+    "            bt += x[scene-1,l]\n",
+    "            if  x[scene-1,l] > 0 :\n",
+    "                if not (perso in présents) :\n",
+    "                    présents.append(perso)\n",
+    "    plt.yticks(-np.arange(scene)-1, np.array([\"Scène \"+str(k) for k in range(1,scene+1)]))\n",
+    "    patches = []\n",
+    "    présents.sort()\n",
+    "    for l,perso in enumerate(persos) :\n",
+    "        if perso in présents :\n",
+    "            patches.append(mpatches.Patch(color=colors[l], label=perso))\n",
+    "    plt.legend(handles=patches,bbox_to_anchor=(1.05, 1),loc='upper left', borderaxespad=0.)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Petite étude critique"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "L'étude suivante donne un bon ordre de grandeur du nombre de mots par personnage. Pourtant, il resterait, en partie à :\n",
+    "- vérifier si chaque réplique est précéde du nom du personnage, de façon homogène\n",
+    "- vérifier s'il n'y a pas de défaut synthaxique dans le texte."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Par exemple, nous avons repéré et non traité l'erreur suivante:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 21,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      " mainJe vous \n"
+     ]
+    }
+   ],
+   "source": [
+    "print(lignes[1343][30:43])"
+   ]
+  }
+ ],
+ "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": "3.6.4"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}
-- 
2.18.1