From 2f9183c77eb5e609e5418b40538e09274d80011b Mon Sep 17 00:00:00 2001 From: Erwan O <106835222+ErwanR123@users.noreply.github.com> Date: Wed, 4 Jun 2025 10:52:23 +0200 Subject: [PATCH] Add files via upload --- logistic_regression.ipynb | 928 +++++++++++++++++++++----------------- 1 file changed, 523 insertions(+), 405 deletions(-) diff --git a/logistic_regression.ipynb b/logistic_regression.ipynb index ba95d04..c76ab2c 100644 --- a/logistic_regression.ipynb +++ b/logistic_regression.ipynb @@ -896,50 +896,96 @@ }, { "cell_type": "code", - "execution_count": null, - "id": "a2e1af5c", + "execution_count": 35, + "id": "d58417ad", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ - "Nombre de prédictions positives : 12\n", - "Nombre réel de cas positifs dans y_test : 11\n" + "Évaluation du modèle simple (régression logistique)\n", + "Accuracy : 0.792\n", + "F1-score : 0.783\n", + "\n", + "Classification report :\n", + " precision recall f1-score support\n", + "\n", + " 0 0.83 0.77 0.80 13\n", + " 1 0.75 0.82 0.78 11\n", + "\n", + " accuracy 0.79 24\n", + " macro avg 0.79 0.79 0.79 24\n", + "weighted avg 0.80 0.79 0.79 24\n", + "\n" ] } ], "source": [ - "# Entraînement initial du modèle de régression logistique (sans régularisation explicite)\n", - "\n", "from sklearn.linear_model import LogisticRegression\n", - "# Initialisation du modèle\n", - "logreg_simple = LogisticRegression(random_state=42)\n", + "from sklearn.metrics import accuracy_score, f1_score, classification_report\n", "\n", - "# Entraînement sur les données standardisées\n", + "# Initialisation et entraînement du modèle sans régularisation explicite\n", + "logreg_simple = LogisticRegression(random_state=42)\n", "logreg_simple.fit(X_train_scaled, y_train)\n", "\n", - "# Prédictions sur le test\n", - "y_pred = logreg_simple.predict(X_test_scaled)\n", + "# Prédictions sur le jeu de test\n", + "y_pred_simple = logreg_simple.predict(X_test_scaled)\n", "\n", - "# Évaluation des performances (jsp si c'est utile de faire ça ici)\n", - "print(\"Nombre de prédictions positives :\", y_pred.sum())\n", - "print(\"Nombre réel de cas positifs dans y_test :\", y_test.sum())\n", - "\n" + "# Évaluation des performances de base\n", + "print(\"Évaluation du modèle simple (régression logistique)\")\n", + "print(\"Accuracy :\", round(accuracy_score(y_test, y_pred_simple), 3))\n", + "print(\"F1-score :\", round(f1_score(y_test, y_pred_simple), 3))\n", + "print(\"\\nClassification report :\\n\", classification_report(y_test, y_pred_simple))\n" ] }, { "cell_type": "markdown", - "id": "6562cf4a", + "id": "76d3da3a", "metadata": {}, "source": [ - "# Étape 9 — Validation croisée (cross-validation)" + "Nous commençons par entraîner un modèle de régression logistique simple, sans régularisation explicite, sur les données prétraitées (après transformation log et standardisation).\n", + "\n", + "### Objectif\n", + "\n", + "Prédire la probabilité qu’une patiente soit atteinte d’un cancer à partir des neuf variables explicatives. \n", + "La régression logistique est un modèle linéaire interprétable, particulièrement adapté pour une première approche supervisée.\n", + "\n", + "### Résultats sur le jeu de test\n", + "\n", + "Le modèle a été évalué sur l’ensemble de test après entraînement sur les données standardisées. Les résultats sont les suivants :\n", + "\n", + "- **Accuracy** : 0.792\n", + "- **F1-score** : 0.783\n", + "\n", + "### Détails du rapport de classification :\n", + "\n", + "| Classe | Précision | Rappel | F1-score | Support |\n", + "|--------|-----------|--------|----------|---------|\n", + "| 0 (sain) | 0.83 | 0.77 | 0.80 | 13 |\n", + "| 1 (malade) | 0.75 | 0.82 | 0.78 | 11 |\n", + "\n", + "- **Macro avg F1** : 0.79\n", + "- **Weighted avg F1** : 0.79\n", + "\n", + "Ces résultats montrent un bon équilibre entre précision et rappel, ce qui est important dans un contexte médical. \n", + "Cependant, ce modèle simple ne prend pas en compte d’éventuelles interactions entre variables ni d’optimisation des hyperparamètres, ce qui sera abordé dans les étapes suivantes.\n", + "\n", + "---\n" + ] + }, + { + "cell_type": "markdown", + "id": "4b7c1cdf", + "metadata": {}, + "source": [ + "# Étape 9 — Validation croisée sur les données d'entraînement\n" ] }, { "cell_type": "code", "execution_count": null, - "id": "7afddd86", + "id": "d8804f23", "metadata": {}, "outputs": [ { @@ -953,7 +999,7 @@ }, { "data": { - "image/png": "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", + "image/png": "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", "text/plain": [ "
" ] @@ -965,262 +1011,173 @@ "source": [ "from sklearn.model_selection import cross_val_score\n", "\n", - "# Initialisation du modèle\n", - "logreg_cv = LogisticRegression(random_state=42)\n", - "\n", - "# Validation croisée avec scoring F1\n", - "f1_scores = cross_val_score(logreg_cv , X_train_scaled, y_train, cv=10, scoring='f1')\n", - "acc_scores = cross_val_score(logreg_cv , X_train_scaled, y_train, cv=10, scoring='accuracy')\n", + "# Validation croisée (10 folds) sur le modèle simple\n", + "f1_scores = cross_val_score(logreg_simple, X_train_scaled, y_train, cv=10, scoring='f1')\n", + "acc_scores = cross_val_score(logreg_simple, X_train_scaled, y_train, cv=10, scoring='accuracy')\n", "\n", "# Résumé\n", "print(\"Validation croisée (10 folds)\")\n", - "print(f\"F1-score moyen : {f1_scores.mean():.3f} ± {f1_scores.std():.3f}\") \n", + "print(f\"F1-score moyen : {f1_scores.mean():.3f} ± {f1_scores.std():.3f}\")\n", "print(f\"Accuracy moyen : {acc_scores.mean():.3f} ± {acc_scores.std():.3f}\")\n", "\n", - "\n", - "# Visualisation des scores de validation croisée\n", + "# Visualisation des scores par fold\n", "folds = range(1, len(f1_scores) + 1)\n", "plt.figure(figsize=(8, 5))\n", "plt.plot(folds, f1_scores, marker='o', label='F1-score')\n", "plt.plot(folds, acc_scores, marker='s', label='Accuracy')\n", - "\n", - "plt.title(\"Performance par fold (validation croisée)\")\n", + "plt.title(\"Scores par fold (validation croisée)\")\n", "plt.xlabel(\"Fold\")\n", "plt.ylabel(\"Score\")\n", "plt.ylim(0, 1)\n", - "plt.xticks(folds)\n", "plt.grid(True)\n", "plt.legend()\n", "plt.tight_layout()\n", - "plt.show()\n", - "\n" + "plt.show()\n" ] }, { "cell_type": "markdown", - "id": "f7c68db8", + "id": "8407908d", "metadata": {}, "source": [ - "Afin d’évaluer la robustesse du modèle de régression logistique, nous avons effectué une validation croisée à 10 folds. \n", - "Chaque fold permet d’entraîner le modèle sur 90 % des données d’entraînement, puis de l’évaluer sur les 10 % restants.\n", + "Afin d'évaluer la stabilité du modèle de régression logistique et sa capacité de généralisation, nous avons effectué une **validation croisée à 10 plis** (k-fold CV). \n", + "Deux métriques principales ont été suivies : **le F1-score** (adapté au contexte de déséquilibre de classes) et **l'accuracy** (exactitude globale).\n", "\n", - "Nous avons suivi l’évolution du **F1-score** et de l’**accuracy** à chaque fold, ce qui permet d’observer la stabilité du modèle.\n", + "### Résultats de la validation croisée\n", "\n", - "- Le **F1-score moyen** est de **0.696**, avec un écart-type de **0.187**\n", - "- L'**accuracy moyenne** est de **0.731**, avec un écart-type de **0.175**\n", + "- **F1-score moyen** : 0.696 ± 0.187\n", + "- **Accuracy moyenne** : 0.731 ± 0.175\n", "\n", - "Ces résultats montrent que le modèle présente des performances correctes mais **variables** selon la partition des données. \n", - "Cette variabilité est attendue dans le cas d’un **petit jeu de données**, et justifie la suite de la démarche par une **optimisation via régularisation**.\n" + "### Analyse\n", + "\n", + "La dispersion importante des scores (écarts types > 0.17) reflète une **sensibilité à la composition des folds**, probablement due à la **petite taille du jeu d'entraînement** (92 observations). \n", + "Cela justifie l'intérêt d’une future **régularisation ou sélection de variables**.\n", + "\n", + "### Visualisation\n", + "\n", + "La figure suivante présente les scores individuels obtenus sur chaque fold. On note une **variabilité significative d’un pli à l’autre**, notamment sur les F1-scores (certains folds tombant à 0.4–0.5, d'autres dépassant 0.8), ce qui confirme le besoin d’optimiser le modèle pour gagner en robustesse.\n", + "\n", + "---\n", + "\n", + "Cette étape permet de constituer un point de référence pour les performances du modèle simple avant d’envisager une **optimisation par recherche d’hyperparamètres**.\n" ] }, { "cell_type": "markdown", - "id": "96408f32", + "id": "5c8f1ec9", "metadata": {}, "source": [ - "# Étape 10 — Évaluation du modèle sur le test set" + "# Étape 10 — Optimisation des hyperparamètres (GridSearchCV)" ] }, { "cell_type": "code", - "execution_count": null, - "id": "d0267905", + "execution_count": 38, + "id": "ba6b4fd9", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ - "Accuracy : 0.792\n", - "F1-score : 0.783\n", - "\n", - "Classification Report :\n", - " precision recall f1-score support\n", - "\n", - " 0 0.83 0.77 0.80 13\n", - " 1 0.75 0.82 0.78 11\n", - "\n", - " accuracy 0.79 24\n", - " macro avg 0.79 0.79 0.79 24\n", - "weighted avg 0.80 0.79 0.79 24\n", - "\n" + "Best parameters : {'C': 100.0, 'penalty': 'l1', 'solver': 'liblinear'}\n", + "Best F1-score (cross-val) : 0.737\n" ] } ], "source": [ - "from sklearn.metrics import accuracy_score, f1_score, classification_report\n", + "from sklearn.model_selection import GridSearchCV\n", "\n", - "# Accuracy\n", - "accuracy = accuracy_score(y_test, y_pred)\n", - "print(\"Accuracy : \", round(accuracy, 3))\n", + "# Grille d’hyperparamètres\n", + "param_grid = {\n", + " 'penalty': ['l1', 'l2'],\n", + " 'C': np.logspace(-2, 4, 10), # de 0.01 à 10000 (échelle log)\n", + " 'solver': ['liblinear'] # nécessaire pour 'l1'\n", + "}\n", "\n", - "# F1-score (binaire par défaut)\n", - "f1 = f1_score(y_test, y_pred)\n", - "print(\"F1-score :\",round(f1, 3))\n", + "# Initialisation du modèle + recherche par validation croisée\n", + "logreg_grid = LogisticRegression(random_state=42, max_iter=1000)\n", + "grid_search = GridSearchCV(logreg_grid, param_grid, cv=10, scoring='f1')\n", + "grid_search.fit(X_train_scaled, y_train)\n", "\n", - "# Rapport complet\n", - "print(\"\\nClassification Report :\")\n", - "print(classification_report(y_test, y_pred))\n" + "# Meilleur modèle\n", + "best_logreg = grid_search.best_estimator_\n", + "\n", + "print(\"Best parameters :\", grid_search.best_params_)\n", + "print(\"Best F1-score (cross-val) :\", round(grid_search.best_score_, 3))\n" + ] + }, + { + "cell_type": "markdown", + "id": "090475a0", + "metadata": {}, + "source": [ + "Afin d'améliorer les performances du modèle, nous avons mis en œuvre une **recherche par grille (GridSearchCV)** pour identifier la meilleure combinaison d’hyperparamètres de la régression logistique. \n", + "L’objectif est de sélectionner automatiquement le niveau optimal de régularisation, afin d’équilibrer la capacité d’apprentissage du modèle et sa généralisation.\n", + "\n", + "---\n", + "\n", + "### Grille d’hyperparamètres testée\n", + "\n", + "Les combinaisons suivantes ont été évaluées dans le cadre d’une validation croisée à 10 plis :\n", + "\n", + "- **Type de régularisation (`penalty`)** : \n", + " `'l1'` (Lasso, régularisation favorisant la sparsité) et `'l2'` (Ridge, régularisation quadratique classique),\n", + "- **Coefficient de régularisation (`C`)** : \n", + " `[0.01, 0.1, 1, 10, 100]` \n", + " *(plus `C` est grand, plus la pénalisation est faible ; `C` est l’inverse du poids de régularisation)*,\n", + "- **Solveur (`solver`)** : \n", + " `'liblinear'`, choisi car c’est le seul compatible avec la régularisation `l1` dans `scikit-learn`.\n", + "\n", + "La métrique d’évaluation choisie pour guider la sélection est le **F1-score**, bien adaptée aux problèmes de classification binaire avec déséquilibre de classes.\n", + "\n", + "---\n", + "\n", + "### Résultats de la recherche\n", + "\n", + "- **Meilleurs hyperparamètres identifiés** : \n", + " `{'C': 100, 'penalty': 'l1', 'solver': 'liblinear'}`\n", + "\n", + "- **F1-score moyen en validation croisée (10 folds)** : \n", + " **0.737**\n", + "\n", + "---\n", + "\n", + "### Analyse et interprétation\n", + "\n", + "Le modèle optimal utilise :\n", + "- une **régularisation L1 faible** (`C = 100`), ce qui réduit l’effet de la pénalisation tout en permettant la mise à zéro automatique de certains coefficients (sparsité),\n", + "- une formulation **plus parcimonieuse**, limitant potentiellement le surapprentissage sur un petit jeu de données,\n", + "- une structure **interprétable**, les coefficients non nuls pouvant être analysés directement comme poids des variables.\n", + "\n", + "Le fait que la meilleure valeur de `C` se situe à l’extrémité supérieure de la grille initiale a motivé une extension du domaine de recherche, mais celle-ci a confirmé que `C = 100` restait optimal. Cela suggère que la régularisation, bien que contrôlée, **n’a pas d’impact significatif** sur les performances dans ce contexte.\n", + "\n", + "---\n", + "\n", + "### Impact attendu\n", + "\n", + "Le modèle régularisé ainsi obtenu sera évalué dans l’étape suivante (étape 11) sur le jeu de test. \n", + "L’objectif est de vérifier si cette optimisation permet **un gain significatif de performance** par rapport au modèle de base (non régularisé) de l’étape 8.\n" + ] + }, + { + "cell_type": "markdown", + "id": "57f03133", + "metadata": {}, + "source": [ + "# Étape 11 — Évaluation du modèle optimisé" ] }, { "cell_type": "code", "execution_count": null, - "id": "dc106736", - "metadata": {}, - "outputs": [ - { - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "\n", - "Coefficients du modèle :\n", - " Feature Coefficient\n", - "1 BMI 0.815422\n", - "0 Age 0.286680\n", - "5 Insulin_log 0.048883\n", - "3 Leptin 0.026291\n", - "7 MCP.1_log -0.290218\n", - "4 Adiponectin -0.301358\n", - "6 HOMA_log -0.583908\n", - "8 Resistin_log -0.700961\n", - "2 Glucose -1.673655\n" - ] - }, - { - "name": "stderr", - "output_type": "stream", - "text": [ - "/var/folders/kz/yn6r06915yl8rm1zpm_51j800000gn/T/ipykernel_69987/2534653591.py:23: FutureWarning: \n", - "\n", - "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `y` variable to `hue` and set `legend=False` for the same effect.\n", - "\n", - " sns.barplot(x='Coefficient', y='Feature', data=coefficients, palette='viridis')\n" - ] - }, - { - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - }, - { - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "# Matrice de confusion\n", - "from sklearn.metrics import confusion_matrix\n", - "cm = confusion_matrix(y_test, y_pred)\n", - "plt.figure(figsize=(6, 5))\n", - "sns.heatmap(cm, annot=True, fmt='d', cmap='Blues', cbar=False,\n", - " xticklabels=['Négatif', 'Positif'], yticklabels=['Négatif', 'Positif'])\n", - "plt.title(\"Matrice de confusion\")\n", - "plt.xlabel(\"Prédictions\")\n", - "plt.ylabel(\"Réel\")\n", - "plt.tight_layout()\n", - "plt.show()\n", - "\n", - "\n", - "# Coefficients du modèle\n", - "coefficients = pd.DataFrame({\n", - " 'Feature': X_train_scaled.columns,\n", - " 'Coefficient': logreg_simple.coef_[0]\n", - "}).sort_values(by='Coefficient', ascending=False)\n", - "print(\"\\nCoefficients du modèle :\")\n", - "print(coefficients)\n", - "# Visualisation des coefficients\n", - "plt.figure(figsize=(10, 6))\n", - "sns.barplot(x='Coefficient', y='Feature', data=coefficients, palette='viridis')\n", - "plt.title(\"Coefficients du modèle de régression logistique\")\n", - "plt.xlabel(\"Coefficient\")\n", - "plt.ylabel(\"Feature\")\n", - "plt.tight_layout()\n", - "plt.show()\n", - "\n", - "#Courbe ROC et AUC\n", - "from sklearn.metrics import roc_curve, auc\n", - "# Calcul des probabilités de prédiction\n", - "y_proba = logreg_simple.predict_proba(X_test_scaled)[:, 1]\n", - "# Calcul de la courbe ROC\n", - "fpr, tpr, thresholds = roc_curve(y_test, y_proba)\n", - "# Calcul de l'AUC\n", - "roc_auc = auc(fpr, tpr) \n", - "# Affichage de la courbe ROC\n", - "plt.figure(figsize=(8, 6))\n", - "plt.plot(fpr, tpr, color='blue', label=f'AUC = {roc_auc:.2f}')\n", - "plt.plot([0, 1], [0, 1], color='red', linestyle='--') # Diagonale\n", - "plt.title('Courbe ROC')\n", - "plt.xlabel('Taux de faux positifs (FPR)')\n", - "plt.ylabel('Taux de vrais positifs (TPR)')\n", - "plt.xlim([0.0, 1.0])\n", - "plt.ylim([0.0, 1.05])\n", - "plt.grid()\n", - "plt.legend(loc='lower right')\n", - "plt.tight_layout()\n", - "plt.show()" - ] - }, - { - "cell_type": "markdown", - "id": "5e3cf2bb", - "metadata": {}, - "source": [ - "### Analyse finale du modèle de régression logistique\n", - "\n", - "L'évaluation complète du modèle sur le jeu de test montre des **résultats satisfaisants**, malgré la taille réduite du dataset :\n", - "\n", - "- **Matrice de confusion** : le modèle identifie correctement **9 cas positifs sur 11**, et limite les fausses alertes à **3 faux positifs**.\n", - "- **Courbe ROC** : l'AUC atteint **0.81**, ce qui traduit une **bonne capacité de discrimination** entre patients atteints et non atteints.\n", - "- **Analyse des coefficients** :\n", - " - Les variables les plus influentes sont `BMI` (positif) et `Glucose` (négatif).\n", - " - Cela donne une **lecture interprétable** du comportement du modèle.\n", - "- **Équilibre des performances** : la **précision** et le **rappel** sont proches sur les deux classes, ce qui reflète un bon compromis.\n", - "\n", - "Ces résultats montrent que la régression logistique, bien prétraitée (transformation + standardisation), permet d'obtenir un modèle à la fois **efficace** et **interprétable**, ce qui est essentiel dans un contexte biomédical.\n" - ] - }, - { - "cell_type": "markdown", - "id": "f05ce0b4", - "metadata": {}, - "source": [ - "# Étape 11: Régularisation" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "id": "ff150ac7", + "id": "bd749019", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ - "Best params : {'C': 100, 'penalty': 'l1', 'solver': 'liblinear'}\n", - "Best F1-score : 0.7371428571428572\n", "\n", "Évaluation sur le jeu de test :\n", "Accuracy : 0.792\n", @@ -1237,48 +1194,7 @@ "weighted avg 0.79 0.79 0.79 24\n", "\n" ] - } - ], - "source": [ - "from sklearn.model_selection import GridSearchCV\n", - "# Grille d’hyperparamètres à tester\n", - "param_grid = {\n", - " 'penalty': ['l1', 'l2'],\n", - " 'C': [0.01, 0.1, 1, 10, 100],\n", - " 'solver': ['liblinear'] # nécessaire pour l1\n", - "}\n", - "# Initialisation du modèle de régression logistique pour la recherche de grille\n", - "logreg_grid = LogisticRegression(random_state=42, max_iter=1000)\n", - "\n", - "# Recherche de grille avec validation croisée\n", - "grid_search = GridSearchCV(logreg_grid, param_grid, cv=10, scoring='f1')\n", - "grid_search.fit(X_train_scaled, y_train)\n", - "\n", - "# Meilleur modèle\n", - "best_logreg = grid_search.best_estimator_\n", - "\n", - "# Prédictions finales\n", - "y_pred_best = best_logreg.predict(X_test_scaled)\n", - "\n", - "# Meilleurs paramètres\n", - "print(\"Best params :\", grid_search.best_params_)\n", - "print(\"Best F1-score :\", grid_search.best_score_)\n", - "\n", - "\n", - "print(\"\\nÉvaluation sur le jeu de test :\")\n", - "print(\"Accuracy :\", round(accuracy_score(y_test, y_pred_best), 3))\n", - "print(\"F1-score :\", round(f1_score(y_test, y_pred_best), 3))\n", - "print(\"\\nClassification Report :\")\n", - "print(classification_report(y_test, y_pred_best))\n", - "\n" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "id": "5dc01116", - "metadata": {}, - "outputs": [ + }, { "data": { "image/png": "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", @@ -1289,48 +1205,9 @@ "metadata": {}, "output_type": "display_data" }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "\n", - "Coefficients du modèle :\n", - " Feature Coefficient\n", - "5 Insulin_log 2.661888\n", - "1 BMI 1.017807\n", - "0 Age 0.423608\n", - "3 Leptin -0.014790\n", - "4 Adiponectin -0.374322\n", - "7 MCP.1_log -0.434532\n", - "8 Resistin_log -0.844157\n", - "2 Glucose -1.678957\n", - "6 HOMA_log -3.778132\n" - ] - }, - { - "name": "stderr", - "output_type": "stream", - "text": [ - "/var/folders/kz/yn6r06915yl8rm1zpm_51j800000gn/T/ipykernel_69987/3329826649.py:22: FutureWarning: \n", - "\n", - "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `y` variable to `hue` and set `legend=False` for the same effect.\n", - "\n", - " sns.barplot(x='Coefficient', y='Feature', data=coefficients, palette='viridis')\n" - ] - }, { "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - }, - { - "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAxUAAAJOCAYAAADBIyqKAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjAsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvlHJYcgAAAAlwSFlzAAAPYQAAD2EBqD+naQAAetpJREFUeJzt3Xt8jvUfx/H3vfPGZhhzaM7kzBANhaI5pwNKMXIMkaUQOSWHypmSs1LpIKXIIZGcy6GczyFMzocNu23X74/rZ1rbavd2b9cOr+fjsUf39b2v+74+97rM3r4nm2EYhgAAAAAghVysLgAAAABA5kaoAAAAAJAqhAoAAAAAqUKoAAAAAJAqhAoAAAAAqUKoAAAAAJAqhAoAAAAAqUKoAAAAAJAqhAoAAAAAqUKoAAA4zfDhw2Wz2XThwgWrSwEApCNCBQBkUkePHlX37t1VokQJeXl5yc/PT3Xq1NHkyZN18+ZNq8tLU/Pnz5fNZov7cnNzU+HChdWxY0edPn060dcYhqGPPvpIDz/8sPz9/eXj46NKlSpp5MiRioyMTPJaS5YsUZMmTRQQECAPDw8VKlRIbdq00Y8//phWHw8AMh03qwsAADhu2bJlat26tTw9PdWhQwdVrFhR0dHR2rBhg1599VXt3btXM2fOtLrMNDdy5EgVL15ct27d0pYtWzR//nxt2LBBe/bskZeXV9x5MTExateunT7//HM99NBDGj58uHx8fPTzzz9rxIgR+uKLL/TDDz8oMDAw7jWGYeiFF17Q/PnzFRwcrPDwcBUoUEBnz57VkiVL9Oijj2rjxo2qXbu2FR8dADIUQgUAZDLHjx/XM888o6JFi+rHH39UwYIF457r1auXjhw5omXLlqVrTZGRkcqRI0e6XlOSmjRpoho1akiSunTpooCAAI0bN05Lly5VmzZt4s57++239fnnn6t///5655134tq7deumNm3aqFWrVurYsaO+//77uOfGjx+v+fPn6+WXX9aECRNks9ninhs8eLA++ugjubnx1ygASAx/AoBM5+2339aNGzc0Z86ceIHirlKlSqlv375xx3fu3NGbb76pkiVLytPTU8WKFdPrr7+u27dvx3udzWbT8OHDE7xfsWLF1LFjx7jju0OPfvrpJ/Xs2VP58+fXfffdF+81Fy5cUJs2beTn56e8efOqb9++unXrVoL3XrhwoapXry5vb2/lyZNHzzzzjE6dOuXgd+Sehx56SJI5NOyumzdv6p133lGZMmU0ZsyYBK9p0aKFwsLCtGLFCm3ZsiXuNWPGjFHZsmX17rvvxgsUd7Vv3141a9ZMca0AkJUQKgAgk/n2229VokSJZA+76dKli4YOHapq1app4sSJqlevnsaMGaNnnnkmVXX07NlT+/bt09ChQzVw4MB4z7Vp00a3bt3SmDFj1LRpU02ZMkXdunWLd85bb72lDh06qHTp0powYYJefvllrVmzRg8//LCuXLmSopr++OMPSVLu3Lnj2jZs2KDLly+rXbt2SfYsdOjQQZL03Xffxb3m0qVLateunVxdXVNUCwBkJ/TbAkAmcu3aNZ0+fVqPP/54ss7/7bfftGDBAnXp0kWzZs2SpLjehXfffVdr165VgwYNUlRLnjx5tGbNmkR/6S5evLi++eYbSeaQLD8/P7333nvq37+/KleurBMnTmjYsGEaNWqUXn/99bjXPfnkkwoODtZ7770Xrz0pV69e1YULF3Tr1i1t3bpVI0aMkKenp5o3bx53zr59+yRJVapUSfJ97j63f//+eP+tVKnSf9YAAKCnAgAylWvXrkmSfH19k3X+8uXLJUnh4eHx2l955RVJStXci65duyb5r/i9evWKd/zSSy/Fq+err75SbGys2rRpowsXLsR9FShQQKVLl9batWuTVUPDhg2VL18+BQUF6emnn1aOHDm0dOnSeMOxrl+/Lunfv2d3n7v7/XX0+wwA2R09FQCQifj5+Um694vyfzlx4oRcXFxUqlSpeO0FChSQv7+/Tpw4keJaihcvnuRzpUuXjndcsmRJubi4xA1POnz4sAzDSHDeXe7u7smqYfr06SpTpoyuXr2quXPnav369fL09Ix3zt1g8G/fs38GD0e/zwCQ3REqACAT8fPzU6FChbRnzx6HXpfYROPkiomJSbTd29s7xdePjY2VzWbT999/n2hvR86cOZP1vjVr1oxb/alVq1aqW7eu2rVrp4MHD8a9R7ly5SRJv//+u1q1apXo+/z++++SpPLly0uSypYtK0navXt3kq8BANzD8CcAyGSaN2+uo0ePavPmzf95btGiRRUbG6vDhw/Haz937pyuXLmiokWLxrXlzp07wQTp6OhonT171uEa/3m9I0eOKDY2VsWKFZNk9lwYhqHixYurYcOGCb4efPBBh6/p6uqqMWPG6MyZM5o2bVpce926deXv769PPvkkyYD04YcfSlLcXIy6desqd+7c+vTTT5N8DQDgHkIFAGQyr732mnLkyKEuXbro3LlzCZ4/evSoJk+eLElq2rSpJGnSpEnxzpkwYYIkqVmzZnFtJUuW1Pr16+OdN3PmzBT9Uj19+vR4x1OnTpVk7ishmROyXV1dNWLECBmGEe9cwzB08eJFh68pSfXr11fNmjU1adKkuCVsfXx81L9/fx08eFCDBw9O8Jply5Zp/vz5Cg0NjQszPj4+GjBggPbv368BAwYkqFEyl8Pdtm1biuoEgKyG4U8AkMmULFlSn3zyidq2baty5crF21F706ZN+uKLL+L2lahSpYrCwsI0c+ZMXblyRfXq1dO2bdu0YMECtWrVKt7KT126dFGPHj301FNPqVGjRvrtt9+0cuVKBQQEOFzj8ePH1bJlSzVu3FibN2/WwoUL1a5du7hVlkqWLKlRo0Zp0KBB+uOPP9SqVSv5+vrq+PHjWrJkibp166b+/fun6Pvz6quvqnXr1po/f7569OghSRo4cKB27typcePGafPmzXrqqafk7e2tDRs2aOHChSpXrpwWLFiQ4H327t2r8ePHa+3atXr66adVoEABRURE6Ouvv9a2bdu0adOmFNUIAFmOAQDIlA4dOmR07drVKFasmOHh4WH4+voaderUMaZOnWrcunUr7jy73W6MGDHCKF68uOHu7m4EBQUZgwYNineOYRhGTEyMMWDAACMgIMDw8fExQkNDjSNHjhhFixY1wsLC4s6bN2+eIcn45ZdfEtQ0bNgwQ5Kxb98+4+mnnzZ8fX2N3LlzG7179zZu3ryZ4PzFixcbdevWNXLkyGHkyJHDKFu2rNGrVy/j4MGD//rZ/62GmJgYo2TJkkbJkiWNO3fuxGufN2+eUadOHcPPz8/w8vIyKlSoYIwYMcK4ceNGktf68ssvjccee8zIkyeP4ebmZhQsWNBo27atsW7dun+tEQCyE5thJNKnCwAAAADJxJwKAAAAAKlCqAAAAACQKoQKAAAAAKlCqAAAAACQKoQKAAAAAKlCqAAAAACQKtlu87vY2FidOXNGvr6+stlsVpcDAAAApCvDMHT9+nUVKlRILi7O6WPIdqHizJkzCgoKsroMAAAAwFKnTp3Sfffd55T3ynahwtfXV5J0/Phx5cmTx+JqkFnY7XatWrVKjz32mNzd3a0uB5kA9wxSgvsGjuKeQUpcunRJxYsXj/u92BmyXai4O+TJ19dXfn5+FleDzMJut8vHx0d+fn780EaycM8gJbhv4CjuGaSE3W6XJKdOBWCiNgAAAIBUIVQAAAAASBVCBQAAAIBUIVQAAAAASBVCBQAAAIBUIVQAAAAASBVCBQAAAIBUIVQAAAAASBVCBQAAAIBUIVQAAAAASBVCBQAAAIBUIVQAAAAASBVCBQAAAIBUIVQAAAAASBVCBQAAAIBUIVQAAAAASBVCBQAAAIBUsTRUrF+/Xi1atFChQoVks9n09ddf/+dr1q1bp2rVqsnT01OlSpXS/Pnz07xOAAAAAEmzNFRERkaqSpUqmj59erLOP378uJo1a6YGDRpo165devnll9WlSxetXLkyjSsFAAAAkBQ3Ky/epEkTNWnSJNnnz5gxQ8WLF9f48eMlSeXKldOGDRs0ceJEhYaGplWZAAAAqWYYUlSUc9/Tbpdu3XJVZKTk7u7c90bWFRnp/Pe0NFQ4avPmzWrYsGG8ttDQUL388stJvub27du6fft23PG1a9ckSXa7XXa7PU3qRNZz917hnkFycc8gJbhvsi7DkOrXd9Xmzc4eJOIuqbmT3xNZmY8iVVVbnf6+mSpUREREKDAwMF5bYGCgrl27pps3b8rb2zvBa8aMGaMRI0YkaF+7dq18fHzSrFZkTatXr7a6BGQy3DNICe6brOfWLVdt3swv/7BWiDZpgcLkqz9V0MnvnalCRUoMGjRI4eHhccfXrl1TUFCQGjRooLx581pYGTITu92u1atXq1GjRnKnfxnJwD2DlOC+ybr+Ptzkzz/typHDOe9rt9v1448/6pFHHuGeQdJu3ZLXmBHynDZRtthYXQ4sIJ2LcOolMlWoKFCggM6dOxev7dy5c/Lz80u0l0KSPD095enpmaDd3d2dP3xwGPcNHMU9g5Tgvsl6/v6/09/f3YmhQvLyipG/P/cM/kXXntLs2ebjsDDFDh0qlSzp1Etkqn0qQkJCtGbNmnhtq1evVkhIiEUVAQAAABncoEFmiPj6a2n+fClXLqdfwtJQcePGDe3atUu7du2SZC4Zu2vXLp08eVKSOXSpQ4cOcef36NFDx44d02uvvaYDBw7ovffe0+eff65+/fpZUT4AAACQ8ezZI02ceO+4RAnp4EHp8cfT7JKWhopff/1VwcHBCg4OliSFh4crODhYQ4cOlSSdPXs2LmBIUvHixbVs2TKtXr1aVapU0fjx4zV79myWkwUAAABiYqRx46Tq1aXwcOnHH+895+qappe2dE5F/fr1ZRhGks8ntlt2/fr1tXPnzjSsCgAAAMhkDh2SOnaUNm82j5s3l8qXT7fLZ6o5FQAAAAD+JjZWmjpVqlrVDBR+ftK8edLSpVKBAulWRqZa/QkAAADA37RpIy1ebD5+9FFp7lypSJF0L4OeCgAAACCzevJJycdHmj5dWrXKkkAh0VMBAAAAZB5nzkgnTkh3t1R49lmpfn2pUCFLy6KnAgAAAMjoDEP65BOpYkWzd+LiRbPdZrM8UEiECgAAACBjO39eat1aeu456fJl6b77pKtXra4qHkIFAAAAkFF99ZVUoYI5GdvNTRo5Utq0ydzQLgNhTgUAAACQ0dy5I3XqJC1caB5XqiQtWCD9f9PojIaeCgAAACCjcXMzd8F2cZEGDZJ++SXDBgqJngoAAAAgY7h+Xbp1S8qXzzyeNEl68UWpVi1Ly0oOeioAAAAAq61daw5x6tTJXOlJkvz9M0WgkAgVAAAAgHWioqS+faVHHjH3n9i7V4qIsLoqhxEqAAAAACts2iRVrSpNmWIed+8u/f67VLCgpWWlBKECAAAASE+3b0sDB0oPPSQdPiwVLix9/700Y4bk62t1dSlCqAAAAADS05070pdfSrGxUvv20u7dUuPGVleVKqz+BAAAAKQ1u/3eErE5ckgffiidOyc98YTVlTkFPRUAAABAWtq7VwoJkaZNu9dWu3aWCRQSoQIAAABIGzEx0jvvSNWqSdu3S2+/be5DkQURKgAAAABnO3JEevhh6bXXpOhoqVkzads2ycvL6srSBKECAAAAcJbYWGn6dKlKFXPJWF9fac4c6dtvpUKFrK4uzTBRGwAAAHCWQ4ekl182V3h65BFp7lypaFGrq0pzhAoAAADAWcqWld56y1zh6cUXzdWesoHs8SkBAACAtHD2rLmK065d99pee03q1SvbBAqJngoAAADAcYYhLVpkhofLl6XTp6WtWyWbzerKLEGoAAAAABxx/rzUs6e5K7YkVa8uzZuXbQOFxPAnAAAAIPm+/lqqUMEMFG5u0ogR0ubNZls2Rk8FAAAAkBzLl9/bBbtCBenDD82N7UCoAAAAAJIlNFSqX1968EFp+HDJ09PqijIMQgUAAACQmOvXpXHjpNdfl3x8JFdXafVqc9gT4uE7AgAAAPzTunVSp07SH3+Y4WLyZLOdQJEoJmoDAAAAd0VFmTtiN2hgBoqiRaVWrSwuKuMjagEAAACStGWLFBYmHTpkHnftKo0fL/n6WltXJkCoAAAAAD76SOrYUYqNlQoVkmbPlpo0sbqqTINQAQAAADzyiOTnJzVvLk2ZIuXObXVFmQqhAgAAANmP3S59/73UsqV5XLiwtGeP+V84jFABAJBhmHMTYS27Xbp1y1WRkZK7u9XVwJkiI62uAPHs22fOnfj1V+nbb83eCYlAkQqECgDI5gxDqltX2rTJ6koguUtqbnURQNYVEyNNnCgNGSLdvm0OcbLbra4qSyBUAEA2FxVFoADSS5065h5qsMCRI+ZE7I0bzeOmTaVZs8xJ2Ug1QgUAIM65c1KOHFZXkX3Z7XatXLlSoaGhcmf8U5bk4yPZbFZXkQ0tWCD17Gn+K4qvr9lb8cIL/M9wIkIFACBOjhyECivZ7ZKXV4xy5GBOBeBUuXKZgaJBA2nePHNDOzgVO2oDAAAgazEM6fjxe8etWkkrV0o//ECgSCOECgAAAGQdZ89KLVpINWqYj+967DHJhV990wrfWQAAAGR+hiEtWiRVqCAtWybduCFt3mx1VdkGoQIAAACZ24ULUtu20rPPSpcvS8HB0vbt0pNPWl1ZtkGoAAAAQOb1zTdm78QXX0hubtLw4dLWrVLFilZXlq2w+hMAAAAyr1WrpL/+MoPFggVS9epWV5QtESoAAACQuURHSx4e5uNx46T77pP69ZO8vKytKxtj+BMAAAAyh+vXpR49pCZNpNhYsy1nTmnQIAKFxeipAAAAQMb3009Sp0739p/4+WepXj1ra0IceioAAACQcd28aQ5tatDADBRFi0pr1hAoMhh6KgAAAJAxbd0qhYVJBw+ax126SOPHS35+1taFBAgVAAAAyHhiY835EwcPSgULSnPmmHMpkCEx/AkAAAAZj4uLGSTat5f27CFQZHD0VAAAAMB6d+6Yy8O6u0uvvWa2VasmffihtXUhWQgVAAAAsNb+/ebciV9+MUPF009LJUpYXRUcwPAnAAAAWCMmRpowQQoONgOFv780b55UvLjVlcFB9FQAAAAg/R09au478fPP5nHjxtLs2VLhwtbWhRQhVAAAACB9RUZKtWpJFy+aO2JPmGAuF2uzWV0ZUohQAQAAgPSVI4c0aJD03XfmcKdixayuCKnEnAoAAACkLcOQ5s+Xtmy519avn7kzNoEiS6CnAgAAAGknIkLq1k369lupTBlp507Jx8fchwJZBv83AQAAkDY++0yqUMEMFB4e0gsvSJ6eVleFNEBPBQAAAJzrwgWpVy/p88/N4+BgcxO7ihWtrQtphlABAAAA5zlxwlzZ6dw5ydVVGjJEGjzY3NQOWRahAgAAAM5TpIhUtap06pTZO1G9utUVIR0QKgAAAJA6a9ZIDzwg+fmZe00sXGjuP+HlZXVlSCdM1AYAAEDK3Lghvfii1LChFB5+rz0ggECRzdBTAQAAAMetXy916iQdO2Yee3tLsbEsFZtN8X8dAAAAyXfzpvTKK1L9+magKFJE+uEHaepUAkU2Rk8FAAAAkmf/funJJ6UDB8zjzp2lCRPMuRTI1ggVAAAASJ6AAOniRalAAWn2bKlZM6srQgZBqAAAAEDS/vhDKlbMfJwvn7k7dqlSUt68VlaFDIaBbwAAAEjozh3prbekMmWkzz67116rFoECCRAqAAAAEN+BA1Lt2uZu2Ha7tGqV1RUhgyNUAAAAwBQbK02cKAUHS7/8IuXKZe6KPXu21ZUhg2NOBQAAAMzlYTt1MvefkKTQUDNM3HeftXUhU6CnAgAAANKRI2agyJlT+uAD6fvvCRRINnoqAAAAsqvoaMnDw3z82GPmBnbNmknFi1tbFzIdeioAAACyG8OQFiwwl4Y9ceJee+/eBAqkCD0VgIUMQ4qKsroKpAW7Xbp1y1WRkZK7u9XV/LvISKsrAJCuIiKk7t2lpUvN44kTpUmTLC0JmR+hArCIYUh160qbNlldCdKGu6TmVhcBAPF98YX04ovmrtju7tKIEdKrr1pdFbIAQgVgkagoAgUyljp1JB8fq6sAkCYuXjSHNi1aZB5XrWoOf6pc2dKykHUQKoAM4Nw5KUcOq6uAM9ntdq1cuVKhoaFyz+jjn/7Px0ey2ayuAkCamDbNDBSurtLrr5ub2t2doA04AaECyABy5CBUZDV2u+TlFaMcOTL+nAoA2cCAAdLvv0sDB0oPPGB1NciCWP0JAAAgq/nhB6l1a+nOHfPYy0tavJhAgTRjeaiYPn26ihUrJi8vL9WqVUvbtm371/MnTZqk+++/X97e3goKClK/fv1069atdKoWAAAgA4uMlHr1kho1kr78Unr/fasrQjZhaaj47LPPFB4ermHDhmnHjh2qUqWKQkND9ddffyV6/ieffKKBAwdq2LBh2r9/v+bMmaPPPvtMr7/+ejpXDgAAkMFs2CBVqSK995553KuX9MIL1taEbMPSUDFhwgR17dpVnTp1Uvny5TVjxgz5+Pho7ty5iZ6/adMm1alTR+3atVOxYsX02GOP6dlnn/3P3g0AAICsyiU6Wi4DBkgPPywdPSoFBUmrV5uTs5mwh3Ri2UTt6Ohobd++XYMGDYprc3FxUcOGDbV58+ZEX1O7dm0tXLhQ27ZtU82aNXXs2DEtX75c7du3T/I6t2/f1u3bt+OOr127JslcmcVutzvp0yCru3uvOPOeMd/KPe59uR2zlrS4Z5D1cd/AUXa7XVWnTZPr+vWSpNiOHRXzzjtSrlziLxYkJS1+xlgWKi5cuKCYmBgFBgbGaw8MDNSBAwcSfU27du104cIF1a1bV4Zh6M6dO+rRo8e/Dn8aM2aMRowYkaB97dq18mFBdjho9erVTnuvW7dcdXdztJUrV8rLK8Zp742Mw5n3DLIP7hs4wrd1a+U5cEC7u3bVuQcekDZutLokZHBRUVFOf89MtaTsunXrNHr0aL333nuqVauWjhw5or59++rNN9/UG2+8kehrBg0apPDw8Ljja9euKSgoSA0aNFDevHnTq3Rkcna7XatXr1ajRo2ctudAZOS9x6GhofRQZzFpcc8g6+O+QbL8/rtctmxRbLducfeM7fBhVff2troyZBIXL150+ntaFioCAgLk6uqqc+fOxWs/d+6cChQokOhr3njjDbVv315dunSRJFWqVEmRkZHq1q2bBg8eLBeXhFNEPD095enpmaDd3d2dH9hwmDPvm7+/jfm+TnlbZDD8rEFKcN8gUXfuSO+8Iw0bJsXEyLVGDalaNUmSu7c39wySLS3uFcsmant4eKh69epas2ZNXFtsbKzWrFmjkJCQRF8TFRWVIDi4urpKkgzDSLtiAQAArHTggFSnjrkbtt0utWghFStmdVVAHEuHP4WHhyssLEw1atRQzZo1NWnSJEVGRqpTp06SpA4dOqhw4cIaM2aMJKlFixaaMGGCgoOD44Y/vfHGG2rRokVcuAAAAMgyYmOlKVOkQYOkW7fMCdhTpkjt20s2G5OxkWFYGiratm2r8+fPa+jQoYqIiFDVqlW1YsWKuMnbJ0+ejNczMWTIENlsNg0ZMkSnT59Wvnz51KJFC7311ltWfQQAAIC0YRhSy5bSsmXmcaNG0pw55pKxQAZj+UTt3r17q3fv3ok+t27dunjHbm5uGjZsmIYNG5YOlQEAAFjIZjOHOa1bJ40fL3XrZrYBGZDloQIAAAD/9+ef0rlzUvXq5nG3blKzZtJ991lbF/AfLN1RGwAAADKHOn30kVSxovTUU9L/N+uVzUagQKZAqAAAALDSuXPSE09IHTpIV69K+fNLly9bXRXgEEIFAACAVb78UqpQQfrmG3MDo7fekjZtkooWtboywCHMqQAAAEhvt29LnTpJn35qHlepIi1YYP4XyIToqQAAAEhvHh5msHB1lYYMkbZtI1AgU6OnAgAAID1cvWpuZpc7tzkBe8YMacAAqWZNqysDUo2eCgAAgLS2Zo1UqZLUs+e9tnz5CBTIMggVAAAAaSUyUurdW2rYUDp1yhzmdPGi1VUBTkeoAAAASAsbN5rzJKZPN49ffFH67Tcpb15r6wLSAKECAADAmW7dkl59VXroIenoUXPzulWrpPfek3LmtLo6IE0QKgAAAJzp1i1zqVjDkDp2lHbvlho1sroqIE2x+hMAAEBq2e2Sm5u5qpO/v/Thh9KNG1LLllZXBqQLeioAAABSY/duqVYtaf78e22PPEKgQLZCqAAAAEiJmBhp3DipRg1p507pzTfNHgsgGyJUAAAAOOrQIaluXWngQCk6WmrRQtq0SXJ3t7oywBKECgAAgOSKjZWmTJGqVpW2bJH8/MxhT998IxUoYHV1gGWYqA0AAJBcv/8uvfyyubJTo0bSnDlSUJDVVQGWI1QAAAAkV9Wq0htvSAULSt27m6s9AWD4EwAAQJJOn5aefFI6ePBe24gRUo8eBArgb+ipAAAA+CfDkD7+WHrpJenKFenyZWntWqurAjIsQgUAAMDf/fWX2ROxZIl5/MAD0nvvWVsTkMEx/AkAAOCuxYulChXMQOHuLo0aZS4VW66c1ZUBGRo9FQAAAJIZKJ5+2nxcubK0YIE5MRvAfyJUAAAASFLLllLNmlLDhtLQoZKnp9UVAZkGoQIAAGRP165JEydKgwZJHh7mcKcNG9gVG0gBQgUyPMOQoqKsrcFul27dclVkpPP+romMdM77AABS4McfpU6dpJMnzR/yo0aZ7QQKIEUIFcjQDEOqW9ecI2ctd0nNrS4CAJBaUVHSwIHS1KnmcfHi0mOPWVsTkAUQKpChRUVlhECRturUkXx8rK4CALKBTZukjh2lw4fN4x49pHfekXLmtLQsICsgVCDTOHdOypHDmmvb7XatXLlSoaGhcndy17iPD5uyAkCamz1b6t5dio2VCheW5syRQkOtrgrIMggVyDRy5LAyVEheXjHKkYPhtgCQKTVoIHl7S089JU2eLPn7W10RkKUQKgAAQNZjt0s//CA1aWIelywp7dsnFSlibV1AFsWO2gAAIGvZs0d68EGpaVNp7dp77QQKIM043FNx+/Ztbd26VSdOnFBUVJTy5cun4OBgFS9ePC3qAwAASJ6YGOndd82N66KjpTx5pOvXra4KyBaSHSo2btyoyZMn69tvv5XdbleuXLnk7e2tS5cu6fbt2ypRooS6deumHj16yNfXNy1rBgAAiO/QIXNlp82bzePmzaWZM6WCBS0tC8gukjX8qWXLlmrbtq2KFSumVatW6fr167p48aL+/PNPRUVF6fDhwxoyZIjWrFmjMmXKaPXq1WldNwAAgGn2bKlqVTNQ+PpKc+dKS5cSKIB0lKyeimbNmmnx4sVJLqVZokQJlShRQmFhYdq3b5/Onj3r1CIBAACS5OUl3bwpPfqoGSiYOwGku2SFiu7duyf7DcuXL6/y5cunuCAAAIB/ZRjSn39KQUHm8XPPSblzmys9ubAGDWAFp/3JO3v2rHr37u2stwMAAEjozBmpWTOpZk3p4kWzzWYz2wgUgGUc+tO3d+9eTZs2TTNnztSVK1ckSRcuXFC/fv1UokQJrf37sm0AAADOYhjSJ59IFStK338vXb4sbdlidVUA/i/ZoWLp0qUKDg5Wnz591KNHD9WoUUNr165VuXLltH//fi1ZskR79+5Ny1oBAEB29Ndf0tNPm8OcLl+WatSQduwweycAZAjJDhWjRo1Sr169dO3aNU2YMEHHjh1Tnz59tHz5cq1YsUKNGzdOyzoBAEB29NVXZu/EV19Jbm7SyJHSpk0S8zeBDCXZoeLgwYPq1auXcubMqZdeekkuLi6aOHGiHnjggbSsDwAAZGdffy2dPy9VqiRt2ya98YaUxGqUAKyT7M3vrl+/Lj8/P0mSq6urvL29VaJEiTQrDAAAZFN2+73gMGWKdP/9Uv/+kqentXUBSFKyQ4UkrVy5Urly5ZIkxcbGas2aNdqzZ0+8c1q2bOm86gAAQPZx7Zr0yivSuXPSN9+Yqzr5+0uDB1tdGYD/4FCoCAsLi3f8z/0rbDabYmJiUl8VAADIXtaulTp1kk6cMI9/+cVcNhZAppDsORWxsbH/+UWgAAAADomKkvr2lR55xAwUxYqZAYNAAWQqDu8Sc/v2bUVGRqZFLQAAIDvZvFmqWtWcNyFJ3btLv/8u1a9vZVUAUiDZoeL8+fNq0qSJcubMKT8/Pz344IM6cuRIWtYGAACyqpgYqWNH6fBhqXBhc0O7GTMkX1+rKwOQAskOFQMGDNCuXbs0cuRIvfvuu7py5Yq6du2alrUBAICsytVVmjNHat9e2r1bYr8rIFNL9kTt1atXa/78+QoNDZUkNW/eXOXKldPt27flyRJvAADg39jt0pgxUkCA1LOn2Va3rvkFINNLdqg4c+aMqlSpEndcunRpeXp66uzZsypWrFha1AYAALKCvXulsDBp+3bJ21t6/HFzyBOALMOhidqurq4Jjg3DcGpBAAAgi4iJkd55R6pWzQwUuXObQ54KFbK6MgBOluyeCsMwVKZMGdlstri2GzduKDg4WC4u97LJpUuXnFshAADIfI4cMXsnNm0yj5s1k2bOJFAAWVSyQ8W8efPSsg4AAJBVXLki1aghXb1qruY0aZK5sd3f/mESQNaS7FBRvHhx1a5dW25uDm3CDQAAsht/f+nll6Wff5bmzpWKFrW6IgBpLNkJoUGDBjp79qzy58+flvUAAIDMxjCkefPM3onKlc22IUMkFxfzC0CWl+w/6UzIBgAACZw5IzVvLnXuLHXoIEVHm+1ubgQKIBtx6E+7jbGQAABAMnsnPvlEqlhRWr5c8vSUnn/e3NQOQLbj0ASJjh07/udGd1999VWqCgIAABnc+fPSiy9Kixebx9WrSx9+KJUvb21dACzjUKjw9fWVt7d3WtUCAAAyusOHpTp1zGDh5iYNHSoNHCi5u1tdGQALORQqpkyZwkRtAACysxIlpPvvlwIDzd6J4GCrKwKQASQ7VDCfAgCAbGrNGikkRPLxMedMfPmluWzsfwyJBpB9sPoTAABI3PXrUvfuUsOG0uDB99oDAwkUAOJJdk/F2rVrlSdPnrSsBQAAZBTr1pm7YP/xh3lsGOYXIxcAJCJZPRWLFi1SvXr1krWb9qlTp7Rx48ZUFwYAACwQFWXuht2ggRkoihaVfvxRmjSJQAEgSckKFe+//77KlSunt99+W/v370/w/NWrV7V8+XK1a9dO1apV08WLF51eKAAASGO7d5sTrydPNo+7dpV+/90MGADwL5I1/Omnn37S0qVLNXXqVA0aNEg5cuRQYGCgvLy8dPnyZUVERCggIEAdO3bUnj17FBgYmNZ1AwAAZ8udWzp3TipUSJo9W2rSxOqKAGQSyZ5T0bJlS7Vs2VIXLlzQhg0bdOLECd28eVMBAQEKDg5WcHCwXFwc2qAbAABY7dQpKSjIfHzffdK335q7ZOfObW1dADIVh/apkKSAgAC1atUqDUoBAADpxm6Xxo6V3nxT+vprqWlTs/2hhywtC0DmRNcCAADZzb595r4TQ4ea4eK776yuCEAmR6gAACC7iImR3n1XqlZN2r7d3MDu44+l6dOtrgxAJufw8CcAAJAJHTkidewo3V32vWlTadYsc1I2AKQSPRUAAGQHu3ebgcLX11zZ6bvvCBQAnCbVoSImJka7du3S5cuXnVEPAABwFrv93uMnnpDeftvcd6JzZzayA+BUDoeKl19+WXPmzJFkBop69eqpWrVqCgoK0rp165xdHwAAcJRhSHPnSvffL0VE3Gt/9VWpWDHLygKQdTkcKr788ktVqVJFkvTtt9/q+PHjOnDggPr166fBgwc7vUAAAOCAs2elFi3M3ojjx+/tjg0AacjhUHHhwgUVKFBAkrR8+XK1bt1aZcqU0QsvvKDdu3c7vUAAAJBMixZJFSpIy5ZJHh7SuHHSqFFWVwUgG3A4VAQGBmrfvn2KiYnRihUr1KhRI0lSVFSUXF1dnV4gAAD4DxcuSG3aSM8+K12+fG/J2Ndek/i7GUA6cHhJ2U6dOqlNmzYqWLCgbDabGjZsKEnaunWrypYt6/QCAQDAfxg3TvriC8nNTRoyRHr9dcnd3eqqAGQjDoeK4cOHq1KlSjp58qRat24tT09PSZKrq6sGDhzo9AIBAMB/GDZMOnjQ/G/16lZXAyAbSlaoyJMnjw4dOqSAgAC98MILmjx5snx9feOdExYWliYFAgCAf1i1SvroI2nBAsnFRcqZU1q61OqqAGRjyZpTER0drWvXrkmSFixYoFu3bqVpUQAAIBE3bkgvviiFhkoLF0rz51tdEQBISmZPRUhIiFq1aqXq1avLMAz16dNH3t7eiZ47d+5cpxYIAAAkrV8vdexoLhMrSS+9JLVta2lJAHBXskLFwoULNXHiRB09elQ2m01Xr16ltwIAgPRw86Y0eLA0aZK5qV3RoubGdo88YnVlABAnWaEiMDBQY8eOlSQVL15cH330kfLmzeuUAqZPn6533nlHERERqlKliqZOnaqaNWsmef6VK1c0ePBgffXVV7p06ZKKFi2qSZMmqWnTpk6pBwCADOW556QlS8zHXbtK774r+flZWxMA/IPDqz8dv9vt6gSfffaZwsPDNWPGDNWqVUuTJk1SaGioDh48qPz58yc4Pzo6Wo0aNVL+/Pn15ZdfqnDhwjpx4oT8/f2dVhMAABnK4MHmnhMzZkhNmlhdDQAkKlmhYsqUKerWrZu8vLw0ZcqUfz23T58+yb74hAkT1LVrV3Xq1EmSNGPGDC1btkxz585NdHnauXPn6tKlS9q0aZPc/7/+drFixZJ9PQAAMjq/48dl+/BDqXNns6F6denIEfadAJChJStUTJw4Uc8995y8vLw0ceLEJM+z2WzJDhXR0dHavn27Bg0aFNfm4uKihg0bavPmzYm+ZunSpQoJCVGvXr30zTffKF++fGrXrp0GDBiQ5G7et2/f1u3bt+OO765iZbfbZbfbk1UrrGP+L3L//2O7rPpfdvde4Z5BcnHPwGF37sgYN0713npLNptN9uBgqVKle89zLyER/KxBSqTF/ZKsUPH3IU/OGv504cIFxcTEKDAwMF57YGCgDhw4kOhrjh07ph9//FHPPfecli9friNHjqhnz56y2+0aNmxYoq8ZM2aMRowYkaB97dq18vHxSf0HQZq6dctVUnNJ0sqVK+XlFWNpPatXr7b0+sh8uGeQHDlPnVK1KVOU+/BhSdKZBx/Ub7t3K/rUKYsrQ2bBzxo4Iioqyunv6fCcipEjR6p///4JfiG/efOm3nnnHQ0dOtRpxf1TbGys8ufPr5kzZ8rV1VXVq1fX6dOn9c477yQZKgYNGqTw8PC442vXrikoKEgNGjRw2mRzpJ3IyHuPQ0NDlSOHNXXY7XatXr1ajRo1iht6B/wb7hkkS0yMXKZOlcsbb8h2+7YMf3/t6NRJ5d58Uw09PKyuDpkAP2uQEhcvXnT6ezocKkaMGKEePXokCBVRUVEaMWJEskNFQECAXF1dde7cuXjt586dU4ECBRJ9TcGCBeXu7h5vqFO5cuUUERGh6OhoeSTyA9jT01Oenp4J2t3d3fnDlwn8/X+R+f/Mulru1cB9g+TjnkGSDENq1ky6+y/MjRvrzvvv68/fflNlDw/uGziEnzVwRFrcK8naUfvvDMOQzWZL0P7bb78pT548yX4fDw8PVa9eXWvWrIlri42N1Zo1axQSEpLoa+rUqaMjR44oNjY2ru3QoUMqWLBgooECAIAMy2aTHntMyplTmjlTWr5cKlzY6qoAIEWS3VORO3du2Ww22Ww2lSlTJl6wiImJ0Y0bN9SjRw+HLh4eHq6wsDDVqFFDNWvW1KRJkxQZGRm3GlSHDh1UuHBhjRkzRpL04osvatq0aerbt69eeuklHT58WKNHj3ZoxSkAACxz6pR05cq9Cdj9+pm7YgcFWVoWAKRWskPFpEmTZBiGXnjhBY0YMUK5cuWKe87Dw0PFihVLsochKW3bttX58+c1dOhQRUREqGrVqlqxYkXc5O2TJ0/KxeVeZ0pQUJBWrlypfv36qXLlyipcuLD69u2rAQMGOHRdAADSlWFICxZIfftKBQtKO3dK3t6SqyuBAkCWkOxQERYWJsncUbt27dpOG4vVu3dv9e7dO9Hn1q1bl6AtJCREW7Zsccq1AQBIcxERUrdu0rffmsfly0uXLjHUCUCWkqxQce3aNfn5+UmSgoODdfPmTd28eTPRc++eBwBAtvfZZ1LPnmaI8PCQRo6U+vc3eygAIAtJVqjInTu3zp49q/z588vf3z/Ridp3J3DHxFi7jwAAAJaLipI6dZI+/9w8Dg42hz/9fTM7AMhCkhUqfvzxx7iVndauXZumBQEAkOl5e0uXL5s9EkOGSIMHy/I1sQEgDSUrVNSrVy/RxwAA4P+uXDFDhK+vuVzs3LnSuXNS9epWVwYAac7hfSpWrFihDRs2xB1Pnz5dVatWVbt27XT58mWnFgcAQKawapU5tOmVV+613XcfgQJAtuFwqHj11Vd17do1SdLu3bsVHh6upk2b6vjx4woPD3d6gQAAZFg3bkgvviiFhkp//imtXSv9/+9IAMhOkr2k7F3Hjx9X+fLlJUmLFy9WixYtNHr0aO3YsUNNmzZ1eoEAAGRI69ebk7GPHTOPe/WSxo2TcuSwti4AsIDDPRUeHh6KioqSJP3www967LHHJEl58uSJ68EAACDLunlTCg+X6tc3A0WRItIPP0jTphEoAGRbDvdU1K1bV+Hh4apTp462bdumzz77TJJ06NAh3XfffU4vEACADOX6demjj8xdsjt3liZMkNijCUA253BPxbRp0+Tm5qYvv/xS77//vgr/f0fQ77//Xo0bN3Z6gQAAWO7OnXuP8+eX5s+XvvtOmj2bQAEASkFPRZEiRfTdd98laJ84caJTCgIAIEP57TcpLMzca6J1a7OtWTNrawKADMbhUCFJMTEx+vrrr7V//35JUoUKFdSyZUu5uro6tTgAACxz54709tvS8OGS3S4NHSo9+aS5FwUAIB6HQ8WRI0fUtGlTnT59Wvfff78kacyYMQoKCtKyZctUsmRJpxcJAEC6OnDA7J3Yts08btVKmjGDQAEASXB4TkWfPn1UsmRJnTp1Sjt27NCOHTt08uRJFS9eXH369EmLGgEASB+xsdLEiVJwsBkocuWSPvxQ+uorKTDQ6uoAIMNyuKfip59+0pYtW5QnT564trx582rs2LGqU6eOU4sDACBdbd1qLhcrmRvazZ5t7owNAPhXDocKT09PXb9+PUH7jRs35OHh4ZSiAACwREiI9MorUpkyUteuks1mdUUAkCk4PPypefPm6tatm7Zu3SrDMGQYhrZs2aIePXqoZcuWaVEjAABp49Qp6amnpBMn7rW9+67UrRuBAgAc4HComDJlikqWLKmQkBB5eXnJy8tLderUUalSpTR58uS0qBEAAOcyDGnBAqlSJXO+RK9eVlcEAJmaw8Of/P399c033+jw4cPav3+/bDabypUrp1KlSqVFfXCAYUhRUVZX4VyRkVZXACDLiYiQuneXli41jx98UBo/3tqaACCTS9E+FZJUunTpuCBho4vYcoYh1a0rbdpkdSUAkIF98YX04ovSxYuSu7s0cqTUv7/kluK/DgEASsHwJ0maM2eOKlasGDf8qWLFipo9e7aza4MDoqKydqCoU0fy8bG6CgCZ2sKFUps2ZqCoWlXavl0aOJBAAQBO4PBP0qFDh2rChAl66aWXFBISIknavHmz+vXrp5MnT2rkyJFOLxKOOXdOypHD6iqcy8eHOZMAUunpp80dslu1koYMkVixEACcxuFQ8f7772vWrFl69tln49patmypypUr66WXXiJUZAA5cmS9UAEADrt6VZo2zeyNcHWVvLykX38lTABAGnA4VNjtdtWoUSNBe/Xq1XXnzh2nFAUAQKr88IP0wgvmkrFubtKAAWY7gQIA0oTDcyrat2+v999/P0H7zJkz9dxzzzmlKAAAUiQy0lwetlEjM1CULGlOygIApKkUzU6bM2eOVq1apQcffFCStHXrVp08eVIdOnRQeHh43HkTJkxwTpUAAPyXDRukjh2lo0fN4169pHHjGA8KAOnA4VCxZ88eVatWTZJ09P8/uAMCAhQQEKA9e/bEnccyswCAdDN9uvTSS+b62kFB0ty5UsOGVlcFANmGw6Fi7dq1aVEHAAApV7++OV+iXTtp4kQpVy6rKwKAbIXFuQEAmU90tLR+/b3eiAoVpAMHpGLFLC0LALKrFG1+BwCAZX7/XapZUwoNlbZuvddOoAAAyxAqAACZw5070pgxUo0a0m+/SblzS5cuWV0VAEAMfwIAZAYHDkhhYdK2beZxy5bSzJlSYKC1dQEAJNFTAQDI6N5/XwoONgNFrlzSggXS118TKAAgA3E4VCxYsEDLli2LO37ttdfk7++v2rVr68SJE04tDgAASdKtW+aGdrt3Sx06SCxbDgAZisOhYvTo0fL29pYkbd68WdOnT9fbb7+tgIAA9evXz+kFAgCyGcOQzpy5d9yjh7RkibRypbkHBQAgw3F4TsWpU6dUqlQpSdLXX3+tp556St26dVOdOnVUv359Z9cHAMhO/vxT6tJFOnjQXOXJ19fslWjVyurKAAD/wuGeipw5c+rixYuSpFWrVqlRo0aSJC8vL928edO51QEAsgfDkD78UKpY0eyRiIiIv1wsACBDc7inolGjRurSpYuCg4N16NAhNW3aVJK0d+9eFWONcACAo86dk7p3l775xjyuWdOcjF22rLV1AQCSzeGeiunTpyskJETnz5/X4sWLlTdvXknS9u3b9eyzzzq9QABAFvbll+Zu2N98I7m7S2+9JW3cSKAAgEzG4Z4Kf39/TZs2LUH7iBEjnFIQACCbMAzp44+lixelKlXM3okqVayuCgCQAskKFb///rsqVqwoFxcX/f777/96buXKlZ1SGAAgi7pzR3JzMydgf/CBVL269NprkoeH1ZUBAFIoWaGiatWqioiIUP78+VW1alXZbDYZhhH3/N1jm82mmJiYNCsWAJCJXb0qhYdLN29Kn3xituXPLw0ZYm1dAIBUS1aoOH78uPLlyxf3GAAAh6xZI3XqJJ06ZfZQDBwo0bMNAFlGskJF0aJFE30MAMC/ioyUBgyQpk83j0uUkObPJ1AAQBbj8ETtu/bt26eTJ08qOjo6XnvLli1TXRQAIAvYuFEKC5OOHjWPX3xRevttKWdOa+sCADidw6Hi2LFjeuKJJ7R79+54cytsNpskMacCACBFR0vt2kknT0r33SfNnSv9f7NUAEDW4/A+FX379lXx4sX1119/ycfHR3v37tX69etVo0YNrVu3Lg1KBABkOh4e0syZZk/F7t0ECgDI4hzuqdi8ebN+/PFHBQQEyMXFRS4uLqpbt67GjBmjPn36aOfOnWlRJwAgI7PbzY3rSpSQOnQw20JDzS8AQJbncKiIiYmRr6+vJCkgIEBnzpzR/fffr6JFi+rgwYNOLxAAkMHt3m32SOzcKfn5SU2bSgEBVlcFAEhHDoeKihUr6rffflPx4sVVq1Ytvf322/Lw8NDMmTNVokSJtKgRAJARxcRI774rDR1qzqHIk0d67z0CBQBkQw6HiiFDhigyMlKSNHLkSDVv3lwPPfSQ8ubNq88++8zpBQIAMqBDh8zeiS1bzOMWLcw5FAUKWFsXAMASDoeK0L+Njy1VqpQOHDigS5cuKXfu3HErQAEAsrC//pKqVTP3oPDzk6ZMMedR8HcAAGRbDq3+ZLfb5ebmpj179sRrz5MnD4ECALKL/Pml7t3NFZ327DF7LPg7AACyNYd6Ktzd3VWkSBH2ogCA7MQwpNmzpXr1pDJlzLaxYyU3N8IEAEBSCvapGDx4sF5//XVdunQpLeoBAGQkp0+bqzl16yZ17GhOzpYkd3cCBQAgjsNzKqZNm6YjR46oUKFCKlq0qHLkyBHv+R07djitOACARQxD+vhj6aWXpCtXJE9P6emnra4KAJBBORwqWrVqlQZlAAAyjL/+knr0kJYsMY8feEBasEAqV87augAAGZbDoWLYsGFpUQcAICPYu1eqX1+6cMEc4jRsmDRggDl/AgCAJDg8p6JLly5at25dGpQCALBcmTJSUJBUubK0bZs0eDCBAgDwnxwOFefPn1fjxo0VFBSkV199Vb/99lta1AUASC8//mjuiC2ZvRPffmsGiqpVLS0LAJB5OBwqvvnmG509e1ZvvPGGfvnlF1WrVk0VKlTQ6NGj9ccff6RBiQCANHHtmtSli/Too9Kbb95rL1zYnJgNAEAyORwqJCl37tzq1q2b1q1bpxMnTqhjx4766KOPVKpUKWfXBwBICz/+KFWqJM2ZYy4Ne+uW1RUBADKxVA2Utdvt+vXXX7V161b98ccfCgwMdFZdAIC0EBkpDRwoTZtmHhcvLs2fLz38sKVlAQAytxT1VKxdu1Zdu3ZVYGCgOnbsKD8/P3333Xf6888/nV0fAMBZdu4050ncDRQ9eki//06gAACkmsM9FYULF9alS5fUuHFjzZw5Uy1atJAnY28BIOPz9ZXOnDHnTMyZI4WGWl0RACCLcDhUDB8+XK1bt5a/v38alAMAcKozZ6RChczHpUpJS5dK1atL/AwHADiRw8OfunbtSqAAgIzObpeGD5eKFZP+vrfQo48SKAAATpeiORUAgAxszx6pVi1pxAgzXCxdanVFAIAsjlABAFlFTIw0bpw5vGnnTilPHmnRImnCBKsrAwBkcalaUhYAkEEcOiR17Cht3mweN28uzZwpFSxoaVkAgOyBngoAyAq2bjUDhZ+fNG+eOeSJQAEASCcpChUfffSR6tSpo0KFCunEiROSpEmTJumbb75xanEAgH9x5869x88/L735prR7t9ljYbNZVhYAIPtxOFS8//77Cg8PV9OmTXXlyhXFxMRIkvz9/TVp0iRn1wcA+CfDkGbPlipWlC5dMttsNmnIEKlIEWtrAwBkSw6HiqlTp2rWrFkaPHiwXF1d49pr1Kih3bt3O7U4AMA/nD4tNWsmde0qHTx4b3dsAAAs5HCoOH78uIKDgxO0e3p6KjIy0ilFAQD+wTCkjz82eye+/17y9JTeeUcaPNjqygAAcDxUFC9eXLt27UrQvmLFCpUrV84ZNQEA/u78eal1a3PexJUrUo0a0o4dUv/+0t96jAEAsIrDS8qGh4erV69eunXrlgzD0LZt2/Tpp59qzJgxmj17dlrUCADZ2/Dh0uLFkpubNHSoNHCg5O5udVUAAMRxOFR06dJF3t7eGjJkiKKiotSuXTsVKlRIkydP1jPPPJMWNQJA9jZqlHTsmDR6tJTI8FMAAKyWos3vnnvuOT333HOKiorSjRs3lD9/fmfXBQDZ14oV0pIl0owZ5qpOuXOb8ygAAMigUrWjto+Pj3x8fJxVCwBkb9evS6+8Is2aZR4/8ojUtq21NQEAkAzJChXBwcGyJXMjpR07dqSqIADIltaulV54QfrjD7N3om9fqUULq6sCACBZkhUqWrVqFff41q1beu+991S+fHmFhIRIkrZs2aK9e/eqZ8+eaVIkAGRZUVHSoEHSlCnmcfHi0rx5Ur161tYFAIADkhUqhg0bFve4S5cu6tOnj958880E55w6dcq51QFAVvfUU+YcCknq3t3ce8LX19qaAABwkMP7VHzxxRfq0KFDgvbnn39eixcvdkpRAJBtDBokBQWZwWLGDAIFACBTcjhUeHt7a+PGjQnaN27cKC8vL6cUBQBZ1o4d0mef3Tt++GHp8GEpNNS6mgAASCWHQ8XLL7+sF198UX369NHChQu1cOFCvfTSS+rVq5f69euXoiKmT5+uYsWKycvLS7Vq1dK2bduS9bpFixbJZrPFm/MBABmS3S6NGCHVqmVOyD5y5N5znp7W1QUAgBM4vKTswIEDVaJECU2ePFkLFy6UJJUrV07z5s1TmzZtHC7gs88+U3h4uGbMmKFatWpp0qRJCg0N1cGDB/91/4s//vhD/fv310MPPeTwNQEgXe3dK3XoYPZSSNLjj0u5cllbEwAATuRwT4UktWnTRhs3btSlS5d06dIlbdy4MUWBQpImTJigrl27qlOnTipfvrxmzJghHx8fzZ07N8nXxMTE6LnnntOIESNUokSJFF0XANJcTIxcxo+XqlUzA0Xu3NInn0hffCHly2d1dQAAOE2KQoWzREdHa/v27WrYsGFcm4uLixo2bKjNmzcn+bqRI0cqf/786ty5c3qUCQCOi4lRnaFD5TpokBQdLTVrJu3ZIz37rLkPBQAAWUiqdtROrQsXLigmJkaBgYHx2gMDA3XgwIFEX7NhwwbNmTNHu3btStY1bt++rdu3b8cdX7t2TZJkt9tlt9tTVngGZH4U9/8/tisLfbQM4e69kpXuGaQte2ysLlSsqLwnTypm/HgZYWFmmOAewr/gZw0cxT2DlEiL+8XSUOGo69evq3379po1a5YCAgKS9ZoxY8ZoxIgRCdrXrl0rHx8fZ5domVu3XCU1lyStXLlSXl4x1haURa1evdrqEpCBef/1l1ztdt0oXFiSZGvdWicbNtTNfPmk77+3uDpkJvysgaO4Z+CIqKgop7+nzTAMw+nvmkzR0dHy8fHRl19+GW8Fp7CwMF25ckXffPNNvPN37dql4OBgubq6xrXFxsZKModNHTx4UCVLloz3msR6KoKCgnT27FnlzZs3DT6VNSIjpdy5zZ6Ky5ftypHD4oKyGLvdrtWrV6tRo0Zyd3e3uhxkNIYh24IFcn3lFalECd3ZuFF2m417Bg7jZw0cxT2DlLh48aIKFiyoq1evys/Pzynv6XBPxa1bt5Lcj+Ls2bMqWLBgst/Lw8ND1atX15o1a+JCRWxsrNasWaPevXsnOL9s2bLavXt3vLYhQ4bo+vXrmjx5soKCghK8xtPTU56JLNfo7u6epf7w/f2jmJ/Nulqysqx238AJzpyRunaVli83j3PkkPuNG1KePJK4Z5Ay3DdwFPcMHJEW94rDE7WrVauW6HyGxYsXq3Llyg4XEB4erlmzZmnBggXav3+/XnzxRUVGRqpTp06SpA4dOmjQoEGSJC8vL1WsWDHel7+/v3x9fVWxYkV5eHg4fH0ASBHDMFdyqljRDBQeHtI770jr10v/mCcGAEBW53BPRf369fXggw9qxIgRGjBggCIjI9WrVy99/vnneuuttxwuoG3btjp//ryGDh2qiIgIVa1aVStWrIibvH3y5Em5uFi6SBUAxHf9utSpk7R4sXlcvbq0YIFUoYK1dQEAYBGHQ8V7772nZs2aqUuXLvruu+909uxZ5cyZU9u2bVPFihVTVETv3r0THe4kSevWrfvX186fPz9F1wSAFPPxkSIiJDc36Y03pEGDxJhDAEB2lqLVn5o0aaInn3xS77//vtzc3PTtt9+mOFAAQKZw+bLk5SV5e0uurmbPxLVrUnCw1ZUBAGA5h8cVHT16VCEhIfruu++0cuVKvfbaa2rZsqVee+011kgGkDWtWGHOnRg8+F5byZIECgAA/s/hUFG1alUVL15cv/32mxo1aqRRo0Zp7dq1+uqrr1SzZs20qBEArHH9utS9u9SkibnK0/LlUhqs7Q0AQGbncKh47733tGjRIvn7+8e11a5dWzt37lS1atWcWRsAWGfdOqlyZWnmTPO4Tx9pxw5zPgUAAIjH4TkV7du3T7Td19dXc+bMSXVBAGCpqCjp9delyZPN46JFpXnzpAYNrK0LAIAMzOFQ8eGHHyb5nM1mSzJ0AECmcPGiGSIkc1O78eMlX19rawIAIINzOFT07ds33rHdbldUVJQ8PDzk4+NDqACQ+cTEmCs6SVJQkDR7tpQzpzmXAgAA/CeH51Rcvnw53teNGzd08OBB1a1bV59++mla1AgAaWfnTnMVp++/v9fWujWBAgAABzhlq+rSpUtr7NixCXoxACDDstulN9+UataUdu8251EYhtVVAQCQKaVo87tE38jNTWfOnHHW2wFA2tm3TwoLk3791Tx+8klpxgzJZrO2LgAAMimHQ8XSpUvjHRuGobNnz2ratGmqU6eO0woDAKeLiZEmTpSGDJFu35Zy55amTZOefZZAAQBAKjgcKlq1ahXv2GazKV++fHrkkUc0fvx4Z9UFAM63bp306qvm46ZNpVmzpEKFLC0JAICswOFQERsbmxZ1AEDae/RRqUcPqUYN6YUX6J0AAMBJnDJRGwAypJMnpTZtpIiIe23vvy917kygAADAiVI0UfvPP//U0qVLdfLkSUVHR8d7bsKECU4pDABSzDDMDez69ZOuXZNcXKRFi6yuCgCALMvhULFmzRq1bNlSJUqU0IEDB1SxYkX98ccfMgxD1apVS4saASD5zp41d8Jetsw8DgmRRo60tiYAALI4h4c/DRo0SP3799fu3bvl5eWlxYsX69SpU6pXr55at26dFjUCwH8zDLM3okIFM1B4eEjjxkk//yyVKWN1dQAAZGkOh4r9+/erQ4cOksy9KW7evKmcOXNq5MiRGjdunNMLBIBkmT3bXBr28mWpWjVp+3bptdckV1erKwMAIMtzOFTkyJEjbh5FwYIFdfTo0bjnLly44LzKAMARzzwjlS4tDR8ubdkiVaxodUUAAGQbyQ4VI0eOVGRkpB588EFt2LBBktS0aVO98soreuutt/TCCy/owQcfTLNCASCeK1ekt9+W7i5z7esr7d4tDRsmubtbWhoAANlNsidqjxgxQj169NCECRN048aNuLYbN27os88+U+nSpVn5CUD6WLXK3Gfi9GkpZ06pZ0+z3dPT2roAAMimkh0qDMOQJJUoUSKuLUeOHJoxY4bzqwKAxNy4Ye6IfffnTqlSUtWqlpYEAAAcnFNhY7MoAFb56SepcuV7geKll6Rdu6TatS0tCwAAOLhPRZkyZf4zWFy6dClVBQFAAhMmSP37m8vGFi0qzZ0rPfKI1VUBAID/cyhUjBgxQrly5UqrWgAgcQ8/bC4N27GjNH685OdndUUAAOBvHAoVzzzzjPLnz59WtQCA6fZtaetWM0xIUo0a0oEDUsmS1tYFAAASlew5FcynAJAudu2SHnhAatTIXCL2LgIFAAAZVrJDxd3VnwAgTdy5I40aZQaK3bvNIU4REVZXBQAAkiHZw59i724wBQDOtm+fFBYm/fqrefzEE+YqTwy3BAAgU3BoSVkAcLopU6Rq1cxA4e8vLVwoLV5MoAAAIBNxaKI2ADjdrVvmxOzGjaXZs6XCha2uCAAAOIhQASB9GYb0119SYKB5/Mor5iTsJ5+UWBACAIBMieFPANLPqVNSaKhUv75086bZ5uoqPfUUgQIAgEyMUAEg7RmGNH++VLGitHq1dOKE9MsvVlcFAACchFABIG1FREiPPy516iRduyY9+KC5F8Xdje0AAECmR6gAkHY+/1yqUEH69lvJw0MaO1basEEqU8bqygAAgBMxURtA2jAMczWnS5ek4GDpww/N4U8AACDLIVQAcK6YGHPytc0mzZljzqUYOFByd7e6MgAAkEYY/gTAOa5ckTp2lHr2vNcWFCS98QaBAgCALI5QASD1Vq2SKlWSFiwwhzwdPmx1RQAAIB0RKgCk3I0b0osvmntP/PmnVKqUtH69VLq01ZUBAIB0RKgAkDLr10tVqkgzZpjHvXubS8XWqWNpWQAAIP0xURuA427elFq3lv76SypSRJo7V3r0UaurAgAAFiFUAHCct7c0fbq0YoU0YYLk52d1RQAAwEIMfwLw36KjzVWcFi++1/b00+akbAIFAADZHj0VAP7db79JYWHmfwMCpIYNpVy5rK4KAABkIPRUAEjcnTvS6NHSAw/cCxTvv0+gAAAACdBTASChAwfM3olt28zjVq3MVZ4CAy0tCwAAZEyECgDx/fmnVK2aucJTrlzS1KnS889LNpvVlQEAgAyKUAEgvvvuk9q3l06cMCdi33ef1RUBAIAMjlABZHeGIc2aJTVubO45IUlTpkgeHvROAACAZGGiNpCdnTplhonu3aXOnc2AIUmengQKAACQbIQKIDsyDGnBAqliRWnVKsnLS2re/F6oAAAAcADDn4DsJiLC7JlYutQ8rlXLDBj3329tXQAAINMiVADZyc6dUqNG0sWLkru7NHKk1L+/5MaPAgAAkHL8JgFkJ2XLSvnySUFBZu9E5cpWVwQAALIAQgWQ1a1bJz30kOTqKnl7SytWSAULmqs7AQAAOAETtYGs6upV6YUXpAYNpPHj77UXLUqgAAAATkVPBZAV/fCDGShOnTKXhr161eqKAABAFkaoALKSGzekAQOk994zj0uWlObPl+rWtbQsAACQtREqgKzil1+kZ5+Vjh41j3v1ksaNk3LksLYuAACQ5REqgKzC29sc7hQUJM2dKzVsaHVFAAAgmyBUAJnZuXNSYKD5uGJF6euvpdq1pVy5LC0LAABkL6z+BGRG0dHSG2+YKzn98su99iZNCBQAACDdESqAzOb336WaNaVRo6Tbt6UlS6yuCAAAZHOECiCzuHNHGjNGqlFD+u03KW9e6fPPpdGjra4MAABkc8ypADKDAweksDBp2zbzuGVLaebMe/MpAAAALERPBZAZrF1rBopcuaQFC8wJ2QQKAACQQdBTAWRUMTGSq6v5uHt36fRp879BQdbWBQAA8A/0VAAZjWGYQ5uqVZOuXzfbXFzMidkECgAAkAERKoCM5M8/zWVhu3c3V3n64AOrKwIAAPhPhAogIzAM6aOPzA3sVq6UvLykiROl8HCrKwMAAPhPzKkArHbunNSjhzn5WjL3oFiwQCpb1tKyAAAAkoueCsBqAwaYgcLdXXrrLWnjRgIFAADIVOipAKw2bpw5l2L8eKlKFaurAQAAcBg9FUB6W7ZM6tfv3nFgoPTDDwQKAACQadFTAaSXa9fMMDF3rnn86KNS8+bW1gQAAOAEhAogPaxZI73wgnTypGSzmas6Pfqo1VUBAAA4BaECSEuRkeZE7OnTzeMSJaT586WHHrK0LAAAAGciVABpqVkz6aefzMcvvii9/baUM6e1NQEAADgZoQJISwMGSMeOSXPmSI0aWV0NAABAmiBUAM7066/m8rCtWpnHTZpIhw6ZO2QDAABkUSwpCzhDdLQ0dKj04INSWJh06tS95wgUAAAgi6OnAkit3bulDh2kXbvM4yZNJB8fS0sCAABIT/RUACl15440dqxUvboZKPLkkRYtMr/y5rW6OgAAgHSTIULF9OnTVaxYMXl5ealWrVratm1bkufOmjVLDz30kHLnzq3cuXOrYcOG/3o+kCbsdqlePWnQIPNxixbS3r1S27ZWVwYAAJDuLA8Vn332mcLDwzVs2DDt2LFDVapUUWhoqP76669Ez1+3bp2effZZrV27Vps3b1ZQUJAee+wxnT59Op0rR7bm7i7VqiX5+Zn7TnzzjVSggNVVAQAAWMLyUDFhwgR17dpVnTp1Uvny5TVjxgz5+Pho7ty5iZ7/8ccfq2fPnqpatarKli2r2bNnKzY2VmvWrEnnypHd+Jw7Jx0/fq/hrbekPXvMidk2m3WFAQAAWMzSUBEdHa3t27erYcOGcW0uLi5q2LChNm/enKz3iIqKkt1uV548edKqTGR3hiHb7Nlq0LevXDt2lGJizHZvbykoyNLSAAAAMgJLV3+6cOGCYmJiFBgYGK89MDBQBw4cSNZ7DBgwQIUKFYoXTP7u9u3bun37dtzxtWvXJEl2u112uz2FlWc85kdx//9ju7LQR7PW6dNy7d5dbqtWSZJiXFxkP3+eidj4T3d/vmSlnzNIe9w3cBT3DFIiLe6XTL2k7NixY7Vo0SKtW7dOXknsBTBmzBiNGDEiQfvatWvlk4WW/bx1y1VSc0nSypUr5eUVY21BmZ1h6L6fflKlWbPkEhmpGHd37X/+eR1t3lzautXq6pCJrF692uoSkAlx38BR3DNwRFRUlNPf09JQERAQIFdXV507dy5e+7lz51TgPya9vvvuuxo7dqx++OEHVa5cOcnzBg0apPDw8Ljja9euKSgoSA0aNFDeLPSvzZGR9x6HhoYqRw7rasn0rlyRa9eucvnmG0lSbI0aiv7gAx09dUqNGjWSu7u7xQUiM7Db7Vq9ejX3DBzCfQNHcc8gJS5evOj097Q0VHh4eKh69epas2aNWrVqJUlxk6579+6d5OvefvttvfXWW1q5cqVq1Kjxr9fw9PSUp6dngnZ3d/cs9Yfv7x/F/GzW1ZLp5colnTxpflOHDZPLgAFyMwzp1Kksd98g7XHPICW4b+Ao7hk4Ii3uFcuHP4WHhyssLEw1atRQzZo1NWnSJEVGRqpTp06SpA4dOqhw4cIaM2aMJGncuHEaOnSoPvnkExUrVkwRERGSpJw5cypnzpyWfQ5kcpcuSTlzSh4e5tfCheZElSpVzOcZqwoAAJAky0NF27Ztdf78eQ0dOlQRERGqWrWqVqxYETd5++TJk3JxubdI1fvvv6/o6Gg9/fTT8d5n2LBhGj58eHqWjqxi2TKpa1fz6+78m/Llra0JAAAgE7E8VEhS7969kxzutG7dunjHf/zxR9oXhOzh2jUpPFyaM8c8XrxYGjzY7KkAAABAsmWIUGGFyEgpiQWjMqW/T9RGMvz4o9Spkzl3wmaTXn7Z3MyOQAEAAOCwbBsqihZlMlO2FBUlDRwoTZ1qHhcvLs2fLz38sKVlAQAAZGaW7qgN56tTR8pC228435kz94Y79egh/f47gQIAACCVsm1Pxf79dgUFWV2F8/n4mKN58DexsdLdyf6lSkkzZkj580uhodbWBQAAkEVk21Dh4yM2iMsOtm83505Mm3avR6J9e2trAgAAyGIY/oSsyW6Xhg+XatWSdu8251EYhtVVAQAAZEnZtqcCWdiePVKHDtLOneZx69bSe+8xLgwAACCN0FOBrCMmRho3Tqpe3QwUefJIixZJn38uBQRYXR0AAECWRU8Fso5ly8xhTpLUvLk0a5ZUoIC1NQEAAGQDhApkHS1amJOwH3lECgtjuBMAAEA6YfgTMq8//pCefVa6fNk8ttmkDz+UOnYkUAAAAKQjeiqQ+RiGuYFdv37SjRvm+sB3N7QDAABAuiNUIHM5c0bq0kX6/nvzuE4dadAga2sCAADI5hj+hMzBMKRPPpEqVjQDhaen9O670k8/mbtkAwAAwDL0VCBzmDZN6tPHfFyjhrRggVS+vLU1AQAAQBI9FcgsnntOKlpUevNNadMmAgUAAEAGQk8FMqbLl82VnPr0MVdyypNH2r9f8va2ujIAAAD8A6ECGc/335uTsc+cMcNE+/ZmO4ECAAAgQ2L4EzKOa9ekrl2lpk3NQFGmjPkFAACADI1QgYxh7VqpcmVp9mzz+OWXpZ07pVq1LC0LAAAA/43hT7DemDHS66+bj4sXl+bNk+rVs7YmAAAAJBs9FbBe7drmZOzu3aXffiNQAAAAZDL0VCD93b4t7dghhYSYx/XqSQcOMH8CAAAgk6KnAulrxw6penWpYUPpyJF77QQKAACATItQgfRht0sjRpgTr/fulXLkkP780+qqAAAA4AQMf0La27tX6tDB7KWQpKeekt5/X8qXz9q6AAAA4BT0VCBtTZwoVatmBorcuaVPPpG++IJAAQAAkIXQU4G0demSFB0tNWsmzZwpFSpkdUUAAABwMkIFnCs2Vrp48V5PxBtvmJvaPf20uWwsAAAAshyGP8F5TpyQGjWSQkPNidmS5OEhtW5NoAAAAMjCCBVIPcOQ5s6VKlWSfvxROnjw3qRsAAAAZHmECqTOmTNS8+ZS587S9evm7ti7dplLxwIAACBbIFQgZQxD+vRTqWJFaflyc5jTO+9I69dLpUtbXR0AAADSERO1kTKGIU2bJl2+bO6QvWCBVKGC1VUBAADAAoQKOCY2VnJxMb/mz5c++0waMEByd7e6MgAAAFiE4U9InsuXpfbtpddeu9dWurQ0ZAiBAgAAIJujpwL/bcUKcyL2mTOSm5vUp49UpIjVVQEAACCDoKcCSbt+XereXWrSxAwUpUubE7EJFAAAAPgbQgUSt26duRP2zJnmcZ8+5lKxISFWVgUAAIAMiOFPSOjaNemJJ6QrV6SiRaV586QGDayuCgAAABkUoQIJ+flJEydKmzZJ48dLvr5WVwQAAIAMjOFPkG7fll5/3ZyQfVfHjubQJwIFAAAA/gM9Fdndzp1SWJi0e7f04YfSoUOSj4/VVQEAACAToaciu7LbpTfflGrWNANFvnzS1KkECgAAADiMnorsaN8+s3fi11/N46eekt5/3wwWAAAAgIMIFdnNkSNStWrmPIrcuaXp06VnnpFsNqsrAwAAQCZFqMhuSpWSnnxSunpVmjVLKlTI6ooAAACQyREqsrrYWDM8tGolBQaabXPmSF5e9E4AAADAKZionZWdPCk99pjUo4f5ZRhmu7c3gQIAAABOQ6jIigxDmjtXqlhRWrPGDBGPPmp1VQAAAMiiGP6U1Zw9K3XtKi1bZh6HhEgLFkilS1tbFwAAALIsQkVWsnWr1KSJdPmy5OFh7kPxyiuSq6vVlQEAACALI1RkJeXLS7lyScWLm70TFStaXREAAACyAUJFZvfzz1LduubEa19f6YcfpCJFJHd3qysDAABANsFE7czqyhVzV+yHH5ZmzLjXXrIkgQIAAADpip6KzGjVKumFF6TTp80einPnrK4IAAAA2RihIjO5fl169VXpgw/M41KlzLkTtWtbWxcAAACyNUJFZrF5s/Tcc9Lx4+bxSy9JY8ZIOXJYWxcAAMhwDMPQnTt3FBMTY3UpsICrq6vc3NxkS8fNjgkVmYWrq7lDdtGi5sZ2jzxidUUAACADio6O1tmzZxUVFWV1KbCQj4+PChYsKA8Pj3S5HqEiIzt/XsqXz3xcs6a0eLHUoIHk52dtXQAAIEOKjY3V8ePH5erqqkKFCsnDwyNd/7Ua1jMMQ9HR0Tp//ryOHz+u0qVLy8Ul7ddmIlRkRLdvSyNGSJMnS9u2SRUqmO2PP25tXQAAIEOLjo5WbGysgoKC5OPjY3U5sIi3t7fc3d114sQJRUdHy8vLK82vSajIaHbtkjp0kHbvNo8XL74XKgAAAJIhPf5lGhlbet8D3HEZxZ070qhR0gMPmIEiIED68ktp6FCrKwMAAAD+FT0VGcH+/eZGdr/8Yh4/8YS5oV3+/NbWBQAAACQDPRUZwbffmoHC319auNAc8kSgAAAA2dDmzZvl6uqqZs2aJXhu3bp1stlsunLlSoLnihUrpkmTJsVrW7t2rZo2baq8efPKx8dH5cuX1yuvvKLTp0+nUfXSrVu31KtXL+XNm1c5c+bUU089pXP/sVHxjRs31Lt3b913333y9vZW+fLlNWPGjHjnzJw5U/Xr15efn1+S3wMrESqsEht77/Err0j9+0t79ph7UbBKAwAAyKbmzJmjl156SevXr9eZM2dS/D4ffPCBGjZsqAIFCmjx4sXat2+fZsyYoatXr2r8+PFOrDi+fv366dtvv9UXX3yhn376SWfOnNGTTz75r68JDw/XihUrtHDhQu3fv18vv/yyevfuraVLl8adExUVpcaNG+v1119Ps9pTg+FP6c0wzKFN8+ZJ69dLXl7mHhTvvGN1ZQAAAJa6ceOGPvvsM/3666+KiIjQ/PnzU/RL9J9//qk+ffqoT58+mjhxYlx7sWLF9PDDD6fZv/JfvXpVc+bM0SeffKJH/r+n2Lx581SuXDlt2bJFDz74YKKv27Rpk8LCwlS/fn1JUrdu3fTBBx9o27ZtatmypSTp5ZdflmT21mRE9FSkp1OnpNBQqWdPc7jT3LlWVwQAALI4w5AiI635MgzHav38889VtmxZ3X///Xr++ec1d+5cGY6+iaQvvvhC0dHReu211xJ93t/fP8nXNmnSRDlz5kzyq8K/rMq5fft22e12NWzYMK6tbNmyKlKkiDZv3pzk62rXrq2lS5fq9OnTMgxDa9eu1aFDh/TYY4/994fNIOipSA+GIS1YIPXtK127Jnl7S2PHSj16WF0ZAADI4qKipJw5rbn2jRtSjhzJP3/OnDl6/vnnJUmNGzfW1atX9dNPP8X9C35yHT58WH5+fipYsKBDr5Ok2bNn6+bNm0k+7+7unuRzERER8vDwSBBaAgMDFRERkeTrpk6dqm7duum+++6Tm5ubXFxcNGvWLD388MMO128VQkVai4iQunUzJ2NL0oMPmgGjTBlr6wIAAMhADh48qG3btmnJkiWSJDc3N7Vt21Zz5sxxOFQYhpHincQLFy6cotelxtSpU7VlyxYtXbpURYsW1fr169WrVy8VKlQoXq9HRkaoSGt9+piBwsNDGjnSnJDt6mp1VQAAIJvw8TF7DKy6dnLNmTNHd+7cUaFCheLaDMOQp6enpk2bply5csnPz0+SOXfhn70BV65cUa5cuSRJZcqU0dWrV3X27FmHeyuaNGmin3/+OcnnixYtqr179yb6XIECBRQdHa0rV67Eq+/cuXMqUKBAoq+5efOmXn/9dS1ZsiRuxavKlStr165devfddwkV+L9335XOn5emTpUqVrS6GgAAkM3YbI4NQbLCnTt39OGHH2r8+PEJ5hG0atVKn376qXr06KHSpUvLxcVF27dvV9GiRePOOXbsmK5evaoy/x8J8vTTT2vgwIF6++23403Uvuufv/T/XWqGP1WvXl3u7u5as2aNnnrqKUlmD8zJkycVEhKS6GvsdrvsdnuCHbBdXV0V+/fVQjM4QoWzLV0qbd0qvfWWeVykiLR2rbU1AQAAZGDfffedLl++rM6dO8f1Ntz11FNPac6cOerRo4d8fX3VpUsXvfLKK3Jzc1OlSpV06tQpDRgwQA8++KBq164tSQoKCtLEiRPVu3dvXbt2TR06dFCxYsX0559/6sMPP1TOnDmTXFY2NcOfcuXKpc6dOys8PFx58uSRn5+fXnrpJYWEhMRb+als2bIaM2aMnnjiCfn5+alevXp69dVX5e3traJFi+qnn37Shx9+qAkTJsS9JiIiQhERETpy5Igkaffu3fL19VWRIkWUJ0+eFNfsLKz+5CxXr0qdOkmPPy6NHk2QAAAASKY5c+aoYcOGCQKFZIaKX3/9Vb///rskafLkyQoLC9OAAQNUoUIFdezYUZUrV9a3334bbx5Fz549tWrVKp0+fVpPPPGEypYtqy5dusjPz0/9+/dPs88yceJENW/eXE899ZQefvhhFShQQF999VW8cw4ePKirV6/GHS9atEgPPPCAnnvuOZUvX15jx47VW2+9pR5/W9RnxowZCg4OVteuXSVJDz/8sIKDg+PtZWElm5GSdboysWvXrilXrlw6ceKCihTJ65w3XbVK6txZ+vNPs4/x1VelESPMPSiQJdjtdi1fvlxNmzb9125P4C7uGaQE9w0c9c975tatWzp+/LiKFy8uL34Pydb+7V64ePGiAgICdPXq1bh5KqnF8KfUuHHDDBB3t1EvVUqaP1+qU8fSsgAAAID0RKhIKcOQHn1U2rbNPO7d29x7IqPPhAIAAACcjDkVKWWzmcvDFiki/fCDuboTgQIAAADZED0Vjti2Tbp8WQoNNY9bt5aaNXNsEWYAAAAgi6GnIjmio6UhQ6TataX27aW//rr3HIECAAAA2Rw9Ff/lt9+ksDDzv5LUsKHkxrcNAABkXNlscU8kIr3vAXoqknLnjrnfxAMPmIEiIED64gvpk0+kDLDBCAAAwD/dXYo4KirK4kpgtbv3QHotT80/uSfm5k2pQQNzZ2xJatXKXDY2MNDSsgAAAP6Nq6ur/P399df/h2r7+PjE2xAOWZ9hGIqKitJff/0lf39/ubq6pst1CRWJ8faWKleWDhwwV3V6/nlztScAAIAMrkCBApIUFyyQPfn7+8fdC+mBUHHXsWOSp6dUuLB5PH68NHSodN991tYFAADgAJvNpoIFCyp//vyy2+1WlwMLuLu7p1sPxV0ZIlRMnz5d77zzjiIiIlSlShVNnTpVNWvWTPL8L774Qm+88Yb++OMPlS5dWuPGjVPTpk1TdnHDkD74wNxzonZtaeVKs1fC19f8AgAAyIRcXV3T/RdLZF+WT9T+7LPPFB4ermHDhmnHjh2qUqWKQkNDk+yy27Rpk5599ll17txZO3fuVKtWrdSqVSvt2bPH8YufOiU1biy9+KIUGWkuHXvtWio/EQAAAJC9WB4qJkyYoK5du6pTp04qX768ZsyYIR8fH82dOzfR8ydPnqzGjRvr1VdfVbly5fTmm2+qWrVqmjZtmkPXdf/yU6lSJWnVKsnLS5o0SfrxRylXLid8KgAAACD7sDRUREdHa/v27WrYsGFcm4uLixo2bKjNmzcn+prNmzfHO1+SQkNDkzw/KTleeUm6elWqVUvatUvq21dysTxjAQAAAJmOpXMqLly4oJiYGAX+Y6nWwMBAHThwINHXREREJHp+REREouffvn1bt2/fjju+evWq+V9XV8UMGqTY3r3NzewuXkzNR0EWZ7fbFRUVpYsXL6bbes/I3LhnkBLcN3AU9wxS4tKlS5Kcu0FehpionZbGjBmjESNGJGgvEhMjjRplfgEAAADZzMWLF5XLSUP/LQ0VAQEBcnV11blz5+K1nzt3Lsl1dQsUKODQ+YMGDVJ4eHjc8ZUrV1S0aFGdPHnSad9EZH3Xrl1TUFCQTp06JT8/P6vLQSbAPYOU4L6Bo7hnkBJXr15VkSJFlCdPHqe9p6WhwsPDQ9WrV9eaNWvUqlUrSVJsbKzWrFmj3r17J/qakJAQrVmzRi+//HJc2+rVqxUSEpLo+Z6envL09EzQnitXLv7wwWF+fn7cN3AI9wxSgvsGjuKeQUq4OHE+seXDn8LDwxUWFqYaNWqoZs2amjRpkiIjI9WpUydJUocOHVS4cGGNGTNGktS3b1/Vq1dP48ePV7NmzbRo0SL9+uuvmjlzppUfAwAAAMi2LA8Vbdu21fnz5zV06FBFRESoatWqWrFiRdxk7JMnT8ZLUbVr19Ynn3yiIUOG6PXXX1fp0qX19ddfq2LFilZ9BAAAACBbszxUSFLv3r2THO60bt26BG2tW7dW69atU3QtT09PDRs2LNEhUUBSuG/gKO4ZpAT3DRzFPYOUSIv7xmY4cy0pAAAAANkOu70BAAAASBVCBQAAAIBUIVQAAAAASJUsGSqmT5+uYsWKycvLS7Vq1dK2bdv+9fwvvvhCZcuWlZeXlypVqqTly5enU6XISBy5b2bNmqWHHnpIuXPnVu7cudWwYcP/vM+Q9Tj6s+auRYsWyWazxe3Pg+zF0fvmypUr6tWrlwoWLChPT0+VKVOGv6eyGUfvmUmTJun++++Xt7e3goKC1K9fP926dSudqkVGsH79erVo0UKFChWSzWbT119//Z+vWbdunapVqyZPT0+VKlVK8+fPd+yiRhazaNEiw8PDw5g7d66xd+9eo2vXroa/v79x7ty5RM/fuHGj4erqarz99tvGvn37jCFDhhju7u7G7t2707lyWMnR+6Zdu3bG9OnTjZ07dxr79+83OnbsaOTKlcv4888/07lyWMXRe+au48ePG4ULFzYeeugh4/HHH0+fYpFhOHrf3L5926hRo4bRtGlTY8OGDcbx48eNdevWGbt27UrnymEVR++Zjz/+2PD09DQ+/vhj4/jx48bKlSuNggULGv369UvnymGl5cuXG4MHDza++uorQ5KxZMmSfz3/2LFjho+PjxEeHm7s27fPmDp1quHq6mqsWLEi2dfMcqGiZs2aRq9eveKOY2JijEKFChljxoxJ9Pw2bdoYzZo1i9dWq1Yto3v37mlaJzIWR++bf7pz547h6+trLFiwIK1KRAaTknvmzp07Ru3atY3Zs2cbYWFhhIpsyNH75v333zdKlChhREdHp1eJyGAcvWd69eplPPLII/HawsPDjTp16qRpnci4khMqXnvtNaNChQrx2tq2bWuEhoYm+zpZavhTdHS0tm/froYNG8a1ubi4qGHDhtq8eXOir9m8eXO88yUpNDQ0yfOR9aTkvvmnqKgo2e125cmTJ63KRAaS0ntm5MiRyp8/vzp37pweZSKDScl9s3TpUoWEhKhXr14KDAxUxYoVNXr0aMXExKRX2bBQSu6Z2rVra/v27XFDpI4dO6bly5eradOm6VIzMidn/D6cITa/c5YLFy4oJiYmbjfuuwIDA3XgwIFEXxMREZHo+REREWlWJzKWlNw3/zRgwAAVKlQowR9IZE0puWc2bNigOXPmaNeuXelQITKilNw3x44d048//qjnnntOy5cv15EjR9SzZ0/Z7XYNGzYsPcqGhVJyz7Rr104XLlxQ3bp1ZRiG7ty5ox49euj1119Pj5KRSSX1+/C1a9d08+ZNeXt7/+d7ZKmeCsAKY8eO1aJFi7RkyRJ5eXlZXQ4yoOvXr6t9+/aaNWuWAgICrC4HmUhsbKzy58+vmTNnqnr16mrbtq0GDx6sGTNmWF0aMqh169Zp9OjReu+997Rjxw599dVXWrZsmd58802rS0MWl6V6KgICAuTq6qpz587Faz937pwKFCiQ6GsKFCjg0PnIelJy39z17rvvauzYsfrhhx9UuXLltCwTGYij98zRo0f1xx9/qEWLFnFtsbGxkiQ3NzcdPHhQJUuWTNuiYbmU/KwpWLCg3N3d5erqGtdWrlw5RUREKDo6Wh4eHmlaM6yVknvmjTfeUPv27dWlSxdJUqVKlRQZGalu3bpp8ODBcnHh35ORUFK/D/v5+SWrl0LKYj0VHh4eql69utasWRPXFhsbqzVr1igkJCTR14SEhMQ7X5JWr16d5PnIelJy30jS22+/rTfffFMrVqxQjRo10qNUZBCO3jNly5bV7t27tWvXrrivli1bqkGDBtq1a5eCgoLSs3xYJCU/a+rUqaMjR47EhVBJOnTokAoWLEigyAZScs9ERUUlCA53Q6k5ZxdIyCm/Dzs+hzxjW7RokeHp6WnMnz/f2Ldvn9GtWzfD39/fiIiIMAzDMNq3b28MHDgw7vyNGzcabm5uxrvvvmvs37/fGDZsGEvKZkOO3jdjx441PDw8jC+//NI4e/Zs3Nf169et+ghIZ47eM//E6k/Zk6P3zcmTJw1fX1+jd+/exsGDB43vvvvOyJ8/vzFq1CirPgLSmaP3zLBhwwxfX1/j008/NY4dO2asWrXKKFmypNGmTRurPgIscP36dWPnzp3Gzp07DUnGhAkTjJ07dxonTpwwDMMwBg4caLRv3z7u/LtLyr766qvG/v37jenTp7OkrGEYxtSpU40iRYoYHh4eRs2aNY0tW7bEPVevXj0jLCws3vmff/65UaZMGcPDw8OoUKGCsWzZsnSuGBmBI/dN0aJFDUkJvoYNG5b+hcMyjv6s+TtCRfbl6H2zadMmo1atWoanp6dRokQJ46233jLu3LmTzlXDSo7cM3a73Rg+fLhRsmRJw8vLywgKCjJ69uxpXL58Of0Lh2XWrl2b6O8pd++VsLAwo169egleU7VqVcPDw8MoUaKEMW/ePIeuaTMM+sIAAAAApFyWmlMBAAAAIP0RKgAAAACkCqECAAAAQKoQKgAAAACkCqECAAAAQKoQKgAAAACkCqECAAAAQKoQKgAAAACkCqECADKR4cOHq2rVqql6D8Mw1K1bN+XJk0c2m027du1ySm0Z0bp162Sz2XTlypV/Pa9YsWKaNGlS3HFERIQaNWqkHDlyyN/fP8XXf/jhh/XJJ5+k+PUpER0drWLFiunXX39N1+sCyN4IFQAgyWaz/evX8OHDrS7RaVasWKH58+fru+++09mzZ1WxYkWrS0oztWvX1tmzZ5UrVy5J0vz58xMNCb/88ou6desWdzxx4kSdPXtWu3bt0qFDh1J07aVLl+rcuXN65pln4tqKFSuW4N667777En0+R44cqlatmr744ou454cPHx73vKurq4KCgtStWzddunQp7hwPDw/1799fAwYMSFHdAJAShAoAkHT27Nm4r0mTJsnPzy9eW//+/a0u0WmOHj2qggULqnbt2ipQoIDc3NysLinNeHh4qECBArLZbP96Xr58+eTj4xN3fPToUVWvXl2lS5dW/vz5U3TtKVOmqFOnTnJxif9X7ciRI+PdWzt37kz0+Z07d+qBBx5Q27ZttWnTprjnK1SooLNnz+rkyZOaN2+eVqxYoRdffDHeezz33HPasGGD9u7dm6LaAcBRhAoAkFSgQIG4r1y5cslms8UdR0ZG6rnnnlNgYKBy5sypBx54QD/88EO819tsNn399dfx2vz9/TV//nxJ0ocffqicOXPq8OHDcc/37NlTZcuWVVRUVJJ1jR07VoGBgfL19VXnzp1169atBOfMnj1b5cqVk5eXl8qWLav33nsvyffr2LGjXnrpJZ08eVI2m03FihWTZPZe1K1bV/7+/sqbN6+aN2+uo0ePxr0usWFEu3btks1m0x9//CFJeuGFF1S5cmXdvn1bkjkMJzg4WB06dEiynvr166t3797q3bu3cuXKpYCAAL3xxhsyDCPunMuXL6tDhw7KnTu3fHx81KRJk3jfxxMnTqhFixbKnTu3cuTIoQoVKmj58uUJ6l63bp06deqkq1evJuiB+vvwp2LFimnx4sX68MMPZbPZ1LFjRxmGoeHDh6tIkSLy9PRUoUKF1KdPnyQ/1/nz5/Xjjz+qRYsWCZ7z9fWNd7/ly5cv0efLlCmj6dOny9vbW99++23c825ubipQoIAKFy6shg0bqnXr1lq9enW898idO7fq1KmjRYsWJVkjADgToQIA/sONGzfUtGlTrVmzRjt37lTjxo3VokULnTx5Mtnv0aFDBzVt2lTPPfec7ty5o2XLlmn27Nn6+OOP4/0L+d99/vnnGj58uEaPHq1ff/1VBQsWTBAYPv74Yw0dOlRvvfWW9u/fr9GjR+uNN97QggULEn3PyZMna+TIkbrvvvt09uxZ/fLLL5KkyMhIhYeH69dff9WaNWvk4uKiJ554QrGxscn+jFOmTFFkZKQGDhwoSRo8eLCuXLmiadOm/evrFixYIDc3N23btk2TJ0/WhAkTNHv27LjnO3bsqF9//VVLly7V5s2bZRiGmjZtKrvdLknq1auXbt++rfXr12v37t0aN26ccubMmeA6tWvXTtALlVgP1C+//KLGjRurTZs2Onv2rCZPnqzFixdr4sSJ+uCDD3T48GF9/fXXqlSpUpKfacOGDfLx8VG5cuWS9b1Lipubm9zd3RUdHZ3o83/88YdWrlwpDw+PBM/VrFlTP//8c6quDwDJlXX7vAHASapUqaIqVarEHb/55ptasmSJli5dqt69eyf7fT744ANVrlxZffr00VdffaXhw4erevXqSZ4/adIkde7cWZ07d5YkjRo1Sj/88EO83ophw4Zp/PjxevLJJyVJxYsX1759+/TBBx8oLCwswXvmypVLvr6+cnV1VYECBeLan3rqqXjnzZ07V/ny5dO+ffuSPeciZ86cWrhwoerVqydfX19NmjRJa9eulZ+f37++LigoSBMnTpTNZtP999+v3bt3a+LEieratasOHz6spUuXauPGjapdu7YkM0gFBQXp66+/VuvWrXXy5Ek99dRTcb/klyhRItHreHh4xOuFSkq+fPnk6ekpb2/vuPNOnjypAgUKqGHDhnJ3d1eRIkVUs2bNJN/jxIkTCgwMTDD0SZIGDBigIUOGxB2PHj060V6P6OhojR8/XlevXtUjjzwS1757927lzJlTMTExcffChAkTEry+UKFCOnHiRJI1AoAz0VMBAP/hxo0b6t+/v8qVKyd/f3/lzJlT+/fvd6inQjKHpMyZM0fvv/++SpYsGfcv+knZv3+/atWqFa8tJCQk7nFkZKSOHj2qzp07K2fOnHFfo0aNijd0KTkOHz6sZ599ViVKlJCfn1/csChHP2NISIj69++vN998U6+88orq1q37n6958MEH4815CAkJ0eHDhxUTE6P9+/fLzc0t3vchb968uv/++7V//35JUp8+fTRq1CjVqVNHw4YN0++//+5QzcnRunVr3bx5UyVKlFDXrl21ZMkS3blzJ8nzb968KS8vr0Sfe/XVV7Vr1664r38ODxswYIBy5swpHx8fjRs3TmPHjlWzZs3inr///vu1a9cu/fLLLxowYIBCQ0P10ksvJbiOt7f3vw6tAwBnIlQAwH/o37+/lixZotGjR+vnn3/Wrl27VKlSpXhDUmw2W7x5AJLihuf83fr16+Xq6qqzZ88qMjIyVXXduHFDkjRr1qx4v6Tu2bNHW7Zscei9WrRooUuXLmnWrFnaunWrtm7dKklxn/Huv7j//TMm9vliY2O1ceNGubq66siRIyn6XI7q0qWLjh07pvbt22v37t2qUaOGpk6d6tRrBAUF6eDBg3rvvffk7e2tnj176uGHH070eyBJAQEBunz5cpLPlSpVKu7rn6tR3Q0df/75py5fvpxgFScPDw+VKlVKFStW1NixY+Xq6qoRI0YkuM6lS5cSzNcAgLRCqACA/7Bx40Z17NhRTzzxhCpVqqQCBQrETU6+K1++fDp79mzc8eHDhxP8K/GmTZs0btw4ffvtt8qZM+d/Dp0qV65c3C/3d/09LAQGBqpQoUI6duxYvF9SS5UqpeLFiyf78128eFEHDx7UkCFD9Oijj6pcuXIJfiG++8vp3z9jYvtbvPPOOzpw4IB++uknrVixQvPmzfvP6yf2GUuXLi1XV1eVK1dOd+7ciXfO3XrLly8f1xYUFKQePXroq6++0iuvvKJZs2Ylei0PDw/FxMT8Z02J8fb2VosWLTRlyhStW7dOmzdv1u7duxM9Nzg4WBEREUkGi39zN3QkZ9UqSRoyZIjeffddnTlzJl77nj17FBwc7PD1ASAlCBUA8B9Kly6tr776Srt27dJvv/2mdu3aJZjA/Mgjj2jatGnauXOnfv31V/Xo0UPu7u5xz1+/fl3t27dXnz591KRJE3388cf67LPP9OWXXyZ53b59+2ru3LmaN2+eDh06pGHDhiVYInTEiBEaM2aMpkyZokOHDmn37t2aN29eomPsk5I7d27lzZtXM2fO1JEjR/Tjjz8qPDw83jmlSpVSUFCQhg8frsOHD2vZsmUaP358vHN27typoUOHavbs2apTp44mTJigvn376tixY/96/ZMnTyo8PFwHDx7Up59+qqlTp6pv376SzO/9448/rq5du2rDhg367bff9Pzzz6tw4cJ6/PHHJUkvv/yyVq5cqePHj2vHjh1au3ZtkhOkixUrphs3bmjNmjW6cOFCsocHzZ8/X3PmzNGePXt07NgxLVy4UN7e3ipatGii5wcHBysgIEAbN25M1vunRkhIiCpXrqzRo0fHa//555/12GOPpfn1AUAiVADAf5owYYJy586t2rVrq0WLFgoNDVW1atXinTN+/HgFBQXpoYceUrt27dS/f/94qzr17dtXOXLkiPvFr1KlSho9erS6d++u06dPJ3rdtm3b6o033tBrr72m6tWr68SJEwn2I+jSpYtmz56tefPmqVKlSqpXr57mz5/vUE+Fi4uLFi1apO3bt6tixYrq16+f3nnnnXjnuLu769NPP9WBAwdUuXJljRs3TqNGjYp7/tatW3r++efVsWPHuGVUu3XrpgYNGqh9+/b/2jvQoUMH3bx5UzVr1lSvXr3Ut2/feBvRzZs3T9WrV1fz5s0VEhIiwzC0fPnyuNAWExOjXr16qVy5cmrcuLHKlCmT5LK6tWvXVo8ePdS2bVvly5dPb7/9drK+R/7+/po1a5bq1KmjypUr64cfftC3336rvHnzJnq+q6urOnXqpI8//jhZ759a/fr10+zZs3Xq1ClJ0ubNm3X16lU9/fTT6XJ9ALAZ/xwEDABAOqlfv76qVq0at0dEVhIREaEKFSpox44dSfZopJW2bduqSpUqev3119P1ugCyL3oqAABIAwUKFNCcOXMcXkErtaKjo1WpUiX169cvXa8LIHtjnwoAANJIq1at0v2aHh4e8fbBAID0wPAnAAAAAKnC8CcAAAAAqUKoAAAAAJAqhAoAAAAAqUKoAAAAAJAqhAoAAAAAqUKoAAAAAJAqhAoAAAAAqUKoAAAAAJAqhAoAAAAAqfI/4/wRu6LkgDIAAAAASUVORK5CYII=", + "image/png": "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", "text/plain": [ "
" ] @@ -1340,6 +1217,18 @@ } ], "source": [ + "from sklearn.metrics import confusion_matrix, roc_curve, auc\n", + "\n", + "# Prédictions\n", + "y_pred_best = best_logreg.predict(X_test_scaled)\n", + "\n", + "# Évaluation globale\n", + "print(\"\\nÉvaluation sur le jeu de test :\")\n", + "print(\"Accuracy :\", round(accuracy_score(y_test, y_pred_best), 3))\n", + "print(\"F1-score :\", round(f1_score(y_test, y_pred_best), 3))\n", + "print(\"\\nClassification Report :\")\n", + "print(classification_report(y_test, y_pred_best))\n", + "\n", "# Matrice de confusion\n", "cm = confusion_matrix(y_test, y_pred_best)\n", "plt.figure(figsize=(6, 5))\n", @@ -1351,110 +1240,222 @@ "plt.tight_layout()\n", "plt.show()\n", "\n", + "# Courbe ROC et AUC\n", + "y_proba_best = best_logreg.predict_proba(X_test_scaled)[:, 1]\n", + "fpr, tpr, thresholds = roc_curve(y_test, y_proba_best)\n", + "roc_auc = auc(fpr, tpr)\n", "\n", - "# Coefficients du modèle\n", - "coefficients = pd.DataFrame({\n", - " 'Feature': X_train_scaled.columns,\n", - " 'Coefficient': best_logreg.coef_[0]\n", - "}).sort_values(by='Coefficient', ascending=False)\n", - "print(\"\\nCoefficients du modèle :\")\n", - "print(coefficients)\n", - "# Visualisation des coefficients\n", - "plt.figure(figsize=(10, 6))\n", - "sns.barplot(x='Coefficient', y='Feature', data=coefficients, palette='viridis')\n", - "plt.title(\"Coefficients du modèle de régression logistique\")\n", - "plt.xlabel(\"Coefficient\")\n", - "plt.ylabel(\"Feature\")\n", - "plt.tight_layout()\n", - "plt.show()\n", - "\n", - "#Courbe ROC et AUC\n", - "# Calcul des probabilités de prédiction\n", - "y_proba_grid = best_logreg.predict_proba(X_test_scaled)[:, 1]\n", - "# Calcul de la courbe ROC\n", - "fpr, tpr, thresholds = roc_curve(y_test, y_proba_grid)\n", - "# Calcul de l'AUC\n", - "roc_auc = auc(fpr, tpr) \n", - "# Affichage de la courbe ROC\n", "plt.figure(figsize=(8, 6))\n", - "plt.plot(fpr, tpr, color='blue', label=f'AUC = {roc_auc:.2f}')\n", - "plt.plot([0, 1], [0, 1], color='red', linestyle='--') # Diagonale\n", - "plt.title('Courbe ROC')\n", - "plt.xlabel('Taux de faux positifs (FPR)')\n", - "plt.ylabel('Taux de vrais positifs (TPR)')\n", - "plt.xlim([0.0, 1.0])\n", - "plt.ylim([0.0, 1.05])\n", - "plt.grid()\n", - "plt.legend(loc='lower right')\n", + "plt.plot(fpr, tpr, label=f'AUC = {roc_auc:.2f}', color='blue')\n", + "plt.plot([0, 1], [0, 1], linestyle='--', color='red')\n", + "plt.title(\"Courbe ROC - Modèle optimisé\")\n", + "plt.xlabel(\"Taux de faux positifs (FPR)\")\n", + "plt.ylabel(\"Taux de vrais positifs (TPR)\")\n", + "plt.grid(True)\n", + "plt.legend()\n", "plt.tight_layout()\n", - "plt.show()" + "plt.show()\n" ] }, { "cell_type": "markdown", - "id": "622c579f", + "id": "6bcbfdc8", "metadata": {}, "source": [ - "### Analyse finale du modèle régularisé (Logistic Regression + GridSearchCV)\n", - "\n", - "Après optimisation des hyperparamètres par validation croisée (GridSearchCV), le modèle final de régression logistique a été évalué sur le jeu de test. Voici les principaux résultats :\n", - "\n", - "#### Matrice de confusion :\n", - "- Le modèle détecte **8 vrais positifs** sur 11, et **11 vrais négatifs** sur 13.\n", - "- Cela montre un bon équilibre entre sensibilité et spécificité.\n", - "\n", - "#### Coefficients du modèle :\n", - "- Les variables ayant le plus d'influence sont :\n", - " - **Insulin_log** (positif) → facteur prédictif fort de la classe 1 (cancer),\n", - " - **BMI** (positif) et **Age** (positif),\n", - " - **Glucose**, **HOMA_log** et **Resistin_log** (coefficients négatifs) → indicateurs protecteurs selon le modèle.\n", - "- Ces coefficients sont **interprétables** grâce à la régression linéaire et confirment des liens biomédicaux plausibles.\n", - "\n", - "#### Courbe ROC :\n", - "- L'aire sous la courbe (**AUC**) est de **0.81**, indiquant une **bonne capacité de discrimination** du modèle.\n", - "- La courbe ROC montre que le modèle est capable de bien différencier les patients atteints ou non de cancer.\n", + "Après avoir sélectionné le meilleur modèle de régression logistique par validation croisée (cf. étape 10), nous l’avons évalué sur le **jeu de test**, resté totalement indépendant durant l’entraînement.\n", "\n", "---\n", "\n", - "Ces résultats sont encourageants, d'autant plus que le dataset est **petit**. \n", - "Le modèle a été rigoureusement entraîné avec :\n", - "- Prétraitement des variables (log + standardisation),\n", - "- Validation croisée,\n", - "- Optimisation par régularisation (`l1`/`l2` via `GridSearchCV`),\n", - "- Évaluation finale sur un jeu de test.\n", + "### Résultats globaux\n", "\n", - "**Conclusion** : La régression logistique régularisée est un **modèle performant, stable et interprétable**, bien adapté à cette tâche de détection binaire.\n" + "- **Accuracy** : 0.792 \n", + "- **F1-score** : 0.762\n", + "\n", + "**Rapport de classification :**\n", + "\n", + "| Classe | Précision | Rappel | F1-score | Support |\n", + "|--------|-----------|--------|----------|---------|\n", + "| 0 (sain) | 0.79 | 0.85 | 0.81 | 13 |\n", + "| 1 (malade) | 0.80 | 0.73 | 0.76 | 11 |\n", + "| **Moyenne (macro)** | 0.79 | 0.79 | 0.79 | 24 |\n", + "\n", + "Le **compromis entre précision et rappel est bien équilibré**, et les performances globales sont comparables (légèrement inférieures en F1) à celles obtenues avec le modèle simple (étape 8), malgré une régularisation L1 imposée.\n", + "\n", + "---\n", + "\n", + "### Matrice de confusion\n", + "\n", + "La matrice ci-dessous permet de visualiser les performances de prédiction classe par classe :\n", + "\n", + "| | Prédit : 0 | Prédit : 1 |\n", + "|---------|------------|------------|\n", + "| **Réel : 0** | 11 | 2 |\n", + "| **Réel : 1** | 3 | 8 |\n", + "\n", + "- **11 vrais négatifs** et **8 vrais positifs** correctement identifiés\n", + "- **2 faux positifs** et **3 faux négatifs**\n", + "\n", + "Le modèle présente une **bonne capacité à identifier les deux classes**, même si quelques cas malades sont encore manqués (ce qui est toujours critique dans un contexte médical).\n", + "\n", + "---\n", + "\n", + "### Courbe ROC et AUC\n", + "\n", + "La courbe ROC du modèle sur le jeu de test confirme une **bonne capacité discriminante**, avec une **aire sous la courbe (AUC) de 0.81**.\n", + "\n", + "Cela signifie qu’en moyenne, le modèle classe un individu malade avec une probabilité plus élevée qu’un individu sain dans **81 % des cas**.\n", + "\n", + "La courbe ROC est bien au-dessus de la diagonale (modèle aléatoire), ce qui valide la qualité du modèle.\n", + "\n", + "---\n", + "\n", + "### Conclusion\n", + "\n", + "Le modèle optimisé offre un **compromis robuste entre performance et parcimonie**, avec :\n", + "- une régularisation L1 qui limite le surajustement,\n", + "- une stabilité démontrée par la validation croisée,\n", + "- des performances solides sur le test.\n", + "\n", + "Cette évaluation confirme que la régression logistique constitue un **modèle fiable et interprétable** sur ce jeu de données, bien qu’il soit possible d’envisager des modèles non linéaires pour améliorer encore les résultats.\n" ] }, { "cell_type": "markdown", - "id": "378032aa", + "id": "918960cc", "metadata": {}, - "source": [] + "source": [ + "# Étape 12 — Comparaison des variantes de régression logistique\n" + ] }, { "cell_type": "code", "execution_count": null, - "id": "ffe2bff1", + "id": "e9c2a95f", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ - "\n", - "Comparaison des modèles :\n", - " Accuracy F1-score AUC\n", - "Modèle simple 0.792 0.783 0.811\n", - "Modèle CV 0.792 0.783 0.811\n", - "Modèle régularisé 0.792 0.762 0.811\n" + " Modèle Accuracy F1-score AUC\n", + "0 LogReg simple 0.792 0.783 0.811\n", + "1 LogReg optimisé (L1, C=100) 0.792 0.762 0.811\n" ] }, { "data": { - "image/png": "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", + "image/png": "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", "text/plain": [ - "
" + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAAArIAAAHqCAYAAAD4TK2HAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjAsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvlHJYcgAAAAlwSFlzAAAPYQAAD2EBqD+naQAAX3JJREFUeJzt3XlcVGX///H3gDKACIqyGu5rLmh6S26piSGZS5tm3rncpi1aKZpFpbiVaaW2WLZKu0uZVhpqKJmmlluWmSuKG6ikIqigcH5/+GO+juwIjMdez8djHjXXuc45nzPnMLw9XHONxTAMQwAAAIDJODm6AAAAAKA4CLIAAAAwJYIsAAAATIkgCwAAAFMiyAIAAMCUCLIAAAAwJYIsAAAATIkgCwAAAFMiyAIAAMCUCLIAcINISEjQhAkT9Mcffzi6lDxdvHhRL7/8sr777jtHlwLgBkCQxXVnwoQJslgsZbKvTp06qVOnTrbncXFxslgs+uqrr8pk/4MGDVLNmjXLZF+FkX38cXFxji6lVERHR8tisWjTpk0lts1OnTqpSZMmBfY7cOCALBaLoqOjbW25Xes1a9bUoEGDilzHxYsX1adPH23fvl2NGzcu8vpl5dlnn9UHH3ygW2+9tVD9s8/ZgQMHbG1X/9xeD6ZPn66GDRsqKyurSOsV93z/mxXl/Jf2e+zFixcVFBSkt99+u9T2gfwRZFGqsn8JZT9cXV0VGBiosLAwvfHGGzp79myJ7Ofo0aOaMGGCtm3bViLbK0nXc224/v3111+aMGGCXZDLzdixY+Xs7KzPP/9cTk7X51v7kiVL9NlnnykmJkY+Pj4ltl1H/4ylpKRo2rRpeuaZZ67b1/5G5sjzX758eUVEROjFF1/UhQsXynz/kMo5ugD8O0yaNEm1atXSxYsXlZiYqLi4OI0cOVIzZszQt99+q2bNmtn6vvDCC3r22WeLtP2jR49q4sSJqlmzppo3b17o9VasWFGk/RRHfrW9//77Rb6DA3OqUaOGzp8/r/Lly+fbb9euXXZh6K+//tLEiRPVqVOnPO8snT59WpUrV9a3334rNze3kiy7RB04cEA//PCD6tate03bufrntrg//yXlo48+0qVLl9SvX78y3/e/UVHOf1m8xw4ePFjPPvusvvjiC/3vf/8r1X0hJ4IsykR4eLhatWplex4ZGalVq1bprrvuUs+ePbVz507bL+By5cqpXLnSvTTPnTsnd3d3ubi4lOp+ClJQqEH+srKylJGRIVdXV0eXUqDsv0gUxGq1FnnblSpV0vjx44tTVqm5dOmSsrKy7H7GnnrqqRLZtqN/bq82d+5c9ezZ0xTX4Y2gKOe/LN5jK1WqpDvuuEPR0dEEWQfgbyBwmNtvv13jxo3TwYMH9dlnn9nacxs3uHLlSrVv316VKlWSh4eHGjRooOeee07S5XGd//nPfyRd/pdx9jCG7LGI2WMYN2/erNtuu03u7u62dfMaa5WZmannnntO/v7+qlChgnr27KlDhw7Z9clrbNuV2yyottzGb6WlpWn06NEKCgqS1WpVgwYN9Oqrr8owDLt+FotFI0aM0OLFi9WkSRNZrVY1btxYMTExub/gVzl8+LB69+6tChUqyNfXV6NGjVJ6enqufTdu3Khu3brJy8tL7u7u6tixo9atW2fX5+zZsxo5cqRq1qwpq9UqX19fde3aVVu2bMm3juzz/ffff6tPnz7y9PRUlSpV9NRTT+X4U132MX/++edq3LixrFar7Xi3bt2q8PBweXp6ysPDQ126dNGGDRty3ee5c+f0yCOPqEqVKvL09NSAAQN06tQpuz5LlixR9+7dFRgYKKvVqjp16mjy5MnKzMzMdZubN29W27Zt5ebmplq1amnOnDl2y3MbI5ubK6+r6Oho3X///ZKkzp07266fK8cw//DDD+rQoYMqVKigihUrqnv37tqxY4fdNhMTEzV48GDddNNNslqtCggIUK9evQocrjBo0CB5eHho//79CgsLU4UKFRQYGKhJkybZXY/Zx/bqq69q1qxZqlOnjqxWq/766y9J0t9//6377rtP3t7ecnV1VatWrfTtt9/m2N+OHTt0++23y83NTTfddJOmTJmS6920ovyMSaV7/cbHx2v79u0KDQ3NsSwrK0uvv/66mjZtKldXV/n4+Khbt275jtH+559/NGbMGDVt2lQeHh7y9PRUeHi4fv/99xx933zzTTVu3Fju7u6qXLmyWrVqpS+++KLIx1SY1yc32WPq58+fX+D7pSQtXLhQLVu2lJubm6pWrar//ve/OnLkiF2fwlyrxX2PvXjxory9vTV48OActaWkpMjV1VVjxoyxtaWnpysqKkp169aV1WpVUFCQxo4dm+v7ZNeuXbV27Vr9888/Bb5uKFnckYVDPfTQQ3ruuee0YsUKDR06NNc+O3bs0F133aVmzZpp0qRJslqt2rt3r+2NtlGjRpo0aZLGjx+vYcOGqUOHDpKktm3b2raRnJys8PBwPfDAA/rvf/8rPz+/fOt68cUXZbFY9Mwzz+j48eOaNWuWQkNDtW3btiL96bYwtV3JMAz17NlTq1ev1pAhQ9S8eXMtX75cTz/9tI4cOaKZM2fa9V+7dq0WLVqkxx9/XBUrVtQbb7yhe++9VwkJCapSpUqedZ0/f15dunRRQkKCnnzySQUGBurTTz/VqlWrcvRdtWqVwsPD1bJlS0VFRcnJyUlz587V7bffrp9//lmtW7eWJD366KP66quvNGLECN18881KTk7W2rVrtXPnTt1yyy0FvlZ9+vRRzZo1NXXqVG3YsEFvvPGGTp06pU8++SRHPQsWLNCIESNUtWpV1axZUzt27FCHDh3k6empsWPHqnz58nr33XfVqVMn/fTTTwoJCbHbxogRI1SpUiVNmDBBu3bt0jvvvKODBw/afjFLl0Okh4eHIiIi5OHhoVWrVmn8+PFKSUnRK6+8Yre9U6dO6c4771SfPn3Ur18/LViwQI899phcXFyu6Q7NbbfdpieffFJvvPGGnnvuOTVq1EiSbP/99NNPNXDgQIWFhWnatGk6d+6c3nnnHbVv315bt261/QK/9957tWPHDj3xxBOqWbOmjh8/rpUrVyohIaHAD8JkZmaqW7duuvXWWzV9+nTFxMQoKipKly5d0qRJk+z6zp07VxcuXNCwYcNktVrl7e2tHTt2qF27dqpWrZqeffZZVahQQQsWLFDv3r319ddf6+6775Z0OcB07txZly5dsvV77733Cvx5K+hnrLSv319++UWScu0zZMgQRUdHKzw8XA8//LAuXbqkn3/+WRs2bLD7C9WV9u/fr8WLF+v+++9XrVq1lJSUpHfffVcdO3bUX3/9pcDAQEmX/2T+5JNP6r777rP9o2/79u3auHGjHnzwwUIfU2Ffn/wU5v0yOjpagwcP1n/+8x9NnTpVSUlJev3117Vu3Tpt3bpVlSpVklT0a7Uo77Hly5fX3XffrUWLFundd9+1u7O7ePFipaen64EHHpB0+R8hPXv21Nq1azVs2DA1atRIf/zxh2bOnKndu3dr8eLFdttu2bKlDMPQL7/8orvuuqvA1wwlyABK0dy5cw1Jxm+//ZZnHy8vL6NFixa251FRUcaVl+bMmTMNScaJEyfy3MZvv/1mSDLmzp2bY1nHjh0NScacOXNyXdaxY0fb89WrVxuSjGrVqhkpKSm29gULFhiSjNdff93WVqNGDWPgwIEFbjO/2gYOHGjUqFHD9nzx4sWGJGPKlCl2/e677z7DYrEYe/futbVJMlxcXOzafv/9d0OS8eabb+bY15VmzZplSDIWLFhga0tLSzPq1q1rSDJWr15tGIZhZGVlGfXq1TPCwsKMrKwsW99z584ZtWrVMrp27Wpr8/LyMoYPH57vfnOTfb579uxp1/74448bkozff//d7pidnJyMHTt22PXt3bu34eLiYuzbt8/WdvToUaNixYrGbbfdZmvLvh5btmxpZGRk2NqnT59uSDKWLFlid4xXe+SRRwx3d3fjwoULtrbs6+u1116ztaWnpxvNmzc3fH19bfuJj4/PcR1cfa0bRs7rauHChXbnJNvZs2eNSpUqGUOHDrVrT0xMNLy8vGztp06dMiQZr7zySo7jKcjAgQMNScYTTzxha8vKyjK6d+9uuLi42H4ms4/N09PTOH78uN02unTpYjRt2tTuNcvKyjLatm1r1KtXz9Y2cuRIQ5KxceNGW9vx48cNLy8vQ5IRHx9vay/sz1hZXL8vvPCCIck4e/asXfuqVasMScaTTz6ZY50ra7n6fF+4cMHIzMy06x8fH29YrVZj0qRJtrZevXoZjRs3zre2go6pKK9Pbgr7fpmRkWH4+voaTZo0Mc6fP2/r9/333xuSjPHjxxuGUfhr9VreY5cvX25IMr777ju7fnfeeadRu3Zt2/NPP/3UcHJyMn7++We7fnPmzDEkGevWrbNrP3r0qCHJmDZtWr61o+QxtAAO5+Hhke/sBdn/Ul+yZEmxB+1brdZc/5yUlwEDBqhixYq25/fdd58CAgK0bNmyYu2/sJYtWyZnZ2c9+eSTdu2jR4+WYRj64Ycf7NpDQ0NVp04d2/NmzZrJ09NT+/fvL3A/AQEBuu+++2xt7u7uGjZsmF2/bdu2ac+ePXrwwQeVnJyskydP6uTJk0pLS1OXLl20Zs0a2zmpVKmSNm7cqKNHjxbr2IcPH273/IknnrDVeqWOHTvq5ptvtj3PzMzUihUr1Lt3b9WuXdvWHhAQoAcffFBr165VSkqK3TaGDRtmN3buscceU7ly5ez2deWdwLNnz+rkyZPq0KGDzp07p7///ttue+XKldMjjzxie+7i4qJHHnlEx48f1+bNmwv9GhTFypUrdfr0afXr1892Xk6ePClnZ2eFhIRo9erVtuNwcXFRXFxcjuEThTVixAjb/2cP78jIyNCPP/5o1+/ee++1m43gn3/+0apVq9SnTx/ba3jy5EklJycrLCxMe/bssf1pedmyZbr11lvt7gD6+Piof//+xapZKpvrNzk5WeXKlZOHh4dd+9dffy2LxaKoqKgc6+Q3vaDVarV92C8zM1PJycm24VRXDgmoVKmSDh8+rN9++y3PbRV0TEV5ffJT0Pvlpk2bdPz4cT3++ON244i7d++uhg0baunSpZJK5lotyO23366qVatq/vz5trZTp05p5cqV6tu3r61t4cKFatSokRo2bGj383X77bdLku3nK1vlypUlSSdPniyVupE3giwcLjU11e5N8Gp9+/ZVu3bt9PDDD8vPz08PPPCAFixYUKRQW61atSJ9QKBevXp2zy0Wi+rWrVvgmMJrdfDgQQUGBuZ4PbL/lHzw4EG79urVq+fYRuXKlQv8JXDw4EHVrVs3xy/UBg0a2D3fs2ePJGngwIHy8fGxe3zwwQdKT0/XmTNnJF2eR/PPP/9UUFCQWrdurQkTJhQYqK909Wtep04dOTk55XjNa9WqZff8xIkTOnfuXI7apcuvW1ZWVo7xelfvy8PDQwEBAXb72rFjh+6++255eXnJ09NTPj4++u9//ytJtmPOFhgYqAoVKti11a9fX5JK7ZrJPje33357jnOzYsUKHT9+XNLlYDRt2jT98MMP8vPz02233abp06crMTGxUPtxcnKy+weClPexXX1u9u7dK8MwNG7cuBw1Zge87DoPHjyY47xIOa/JoijL6/dq+/btU2BgoLy9vYu0XlZWlmbOnKl69erJarWqatWq8vHx0fbt2+2uu2eeeUYeHh5q3bq16tWrp+HDh+cY11rQMRXl9clPQe+X2e9buZ3Lhg0b2pZf67VaGOXKldO9996rJUuW2Ma6Llq0SBcvXrQLsnv27NGOHTtyvC7Z1372dZvN+P9jxstqDnT8H8bIwqEOHz6sM2fO5Dsdj5ubm9asWaPVq1dr6dKliomJ0fz583X77bdrxYoVcnZ2LnA/pTElUV5vWJmZmYWqqSTktR/jqg+GFVf2PxZeeeWVPKc1yr4T1adPH3Xo0EHffPONVqxYoVdeeUXTpk3TokWLFB4eXuR95/X6lsX0UqdPn1bHjh3l6empSZMmqU6dOnJ1ddWWLVv0zDPPXBdTpmXX8Omnn8rf3z/H8itn/hg5cqR69OihxYsXa/ny5Ro3bpymTp2qVatWqUWLFiVW09XnJrvGMWPGKCwsLNd1rnUqrvyUxfVbpUoVXbp0SWfPns33H+SF9dJLL2ncuHH63//+p8mTJ8vb21tOTk4aOXKk3XXXqFEj7dq1S99//71iYmL09ddf6+2339b48eM1ceLEQh1TUV6fslIW1+oDDzygd999Vz/88IN69+6tBQsWqGHDhgoODrb1ycrKUtOmTTVjxoxctxEUFGT3PPvmQdWqVUukRhQeQRYO9emnn0pSnr/ksjk5OalLly7q0qWLZsyYoZdeeknPP/+8Vq9erdDQ0BL/V3D2nYpshmFo7969dvPdVq5cWadPn86x7sGDB+3uYBWltho1aujHH3/M8Usx+0/ZNWrUKPS2CtrPn3/+KcMw7OrbtWuXXb/sYQuenp65fir7agEBAXr88cf1+OOP6/jx47rlllv04osvFirI7tmzx+6O3t69e5WVlVXgh5F8fHzk7u6eo3bp8uvm5OSU45fOnj171LlzZ9vz1NRUHTt2THfeeaeky5+ETk5O1qJFi3TbbbfZ+sXHx+daw9GjR5WWlmZ3V3b37t2SdM3fKpTX9ZN9bnx9fQt1burUqaPRo0dr9OjR2rNnj5o3b67XXnvNbsaQ3GRlZWn//v22O1FS4Y8t++egfPnyBdZYo0aNHD93Us5rMjcFvUalef02bNhQ0uVr48r3hzp16mj58uX6559/inRX9quvvlLnzp314Ycf2rWfPn06R0iqUKGC+vbtq759+yojI0P33HOPXnzxRUVGRtr+hJ/fMRX19clLQe+X2e9bu3btsv1pPtuuXbtyvK8V9Vot6vv/bbfdpoCAAM2fP1/t27fXqlWr9Pzzz+eo4ffff1eXLl0Ktf3s94bsv56h7DC0AA6zatUqTZ48WbVq1cp3HFxu05lk3z3I/tNQdoDILVgWxyeffGI3bverr77SsWPH7H6h1alTRxs2bFBGRoat7fvvv8/xZ+yi1HbnnXcqMzNTb731ll37zJkzZbFYinVnM6/9HD161O6reM+dO6f33nvPrl/Lli1Vp04dvfrqq0pNTc2xnRMnTki6fBf66j9B+vr6KjAwMM8pva42e/Zsu+dvvvmmJBV4zM7Ozrrjjju0ZMkSuz91JyUl6YsvvlD79u3l6elpt857772nixcv2p6/8847unTpkm1f2Xe6r7yznZGRkefXUF66dEnvvvuuXd93331XPj4+atmyZb71FySv6ycsLEyenp566aWX7I4lW/a5OXfuXI5pzOrUqaOKFSsW+txceT0ahqG33npL5cuXV5cuXfJdz9fXV506ddK7776rY8eO5VmjdPma3LBhg3799Ve75Z9//nmB9eX1GpXF9dumTRtJyjGl1r333ivDMGx3R6+U319MnJ2dcyxfuHBhjmmqkpOT7Z67uLjo5ptvlmEYunjxYqGOqbCvT0EKer9s1aqVfH19NWfOHLvX84cfftDOnTvVvXt3ScW/Vov6/u/k5KT77rtP3333nT799FNdunTJbliBdPlu9pEjR/T+++/nWP/8+fNKS0uza9u8ebMsFovtekDZ4Y4sysQPP/ygv//+W5cuXVJSUpJWrVqllStXqkaNGvr222/znUh80qRJWrNmjbp3764aNWro+PHjevvtt3XTTTepffv2ki6/2VWqVElz5sxRxYoVVaFCBYWEhOQYs1dY3t7eat++vQYPHqykpCTNmjVLdevWtZsi7OGHH9ZXX32lbt26qU+fPtq3b58+++wzuw9fFbW2Hj16qHPnznr++ed14MABBQcHa8WKFVqyZIlGjhyZY9vFNXToUL311lsaMGCANm/erICAAH366adyd3e36+fk5KQPPvhA4eHhaty4sQYPHqxq1arpyJEjWr16tTw9PfXdd9/p7Nmzuummm3TfffcpODhYHh4e+vHHH/Xbb7/ptddeK1RN8fHx6tmzp7p166b169frs88+04MPPmj35768TJkyxTbX8OOPP65y5crp3XffVXp6uqZPn56jf0ZGhrp06aI+ffpo165devvtt9W+fXv17NlT0uWpeypXrqyBAwfqySeflMVi0aeffppnAAkMDNS0adN04MAB1a9fX/Pnz9e2bdv03nvvXfOE7M2bN5ezs7OmTZumM2fOyGq16vbbb5evr6/eeecdPfTQQ7rlllv0wAMPyMfHRwkJCVq6dKnatWunt956S7t377Yd680336xy5crpm2++UVJSkm2qofy4uroqJiZGAwcOVEhIiH744QctXbpUzz33XKG+Znb27Nlq3769mjZtqqFDh6p27dpKSkrS+vXrdfjwYdv8qGPHjtWnn36qbt266amnnrJNv1WjRg1t3749333k9zNW2tdv7dq11aRJE/344492U6117txZDz30kN544w3t2bNH3bp1U1ZWln7++Wd17tzZ7gN0V7rrrrs0adIkDR48WG3bttUff/yhzz//PMc45TvuuEP+/v5q166d/Pz8tHPnTr311lvq3r27KlasqNOnTxd4TIX9+S5IQe+X5cuX17Rp0zR48GB17NhR/fr1s02/VbNmTY0aNUqSin2tFuf9v2/fvnrzzTcVFRWlpk2b5riT+tBDD2nBggV69NFHtXr1arVr106ZmZn6+++/tWDBAi1fvtxuCrWVK1eqXbt2+U57iFLiiKkS8O+RPd1R9sPFxcXw9/c3unbtarz++ut2U7Zku3pKotjYWKNXr15GYGCg4eLiYgQGBhr9+vUzdu/ebbfekiVLjJtvvtkoV66c3VQsHTt2zHOamrym3/ryyy+NyMhIw9fX13BzczO6d+9uHDx4MMf6r732mlGtWjXDarUa7dq1MzZt2pRjm/nVdvXUMIZxeVqlUaNGGYGBgUb58uWNevXqGa+88ord9DiGcXkqqtym1slrWrCrHTx40OjZs6fh7u5uVK1a1XjqqaeMmJiYXKd62rp1q3HPPfcYVapUMaxWq1GjRg2jT58+RmxsrGEYl6ebevrpp43g4GCjYsWKRoUKFYzg4GDj7bffLrCO7PP9119/Gffdd59RsWJFo3LlysaIESPspurJ75gNwzC2bNlihIWFGR4eHoa7u7vRuXNn45dffrHrk309/vTTT8awYcOMypUrGx4eHkb//v2N5ORku77r1q0zbr31VsPNzc0IDAw0xo4da5u658rXJ/v62rRpk9GmTRvD1dXVqFGjhvHWW2/Zba+4028ZhmG8//77Ru3atQ1nZ+cc+1+9erURFhZmeHl5Ga6urkadOnWMQYMGGZs2bTIMwzBOnjxpDB8+3GjYsKFRoUIFw8vLywgJCbGbei0vAwcONCpUqGDs27fPuOOOOwx3d3fDz8/PiIqKspsiKvvY8po2ad++fcaAAQMMf39/o3z58ka1atWMu+66y/jqq6/s+m3fvt3o2LGj4erqalSrVs2YPHmy8eGHHxY4/ZZh5P0zZhile/0ahmHMmDHD8PDwyDFl26VLl4xXXnnFaNiwoeHi4mL4+PgY4eHhxubNm219cpt+a/To0UZAQIDh5uZmtGvXzli/fn2OY3733XeN2267zXZMderUMZ5++mnjzJkzRT6mgl6fvBT1/XL+/PlGixYtDKvVanh7exv9+/c3Dh8+bFte2Gv1Wt9jDePy1GNBQUG5TneYLSMjw5g2bZrRuHFjw2q1GpUrVzZatmxpTJw40fY6G4ZhnD592nBxcTE++OCDfF8vlA6LYZTQp0IAoBgmTJigiRMn6sSJE3xQ4jozaNAgffXVV7n+2Rn/58yZM6pdu7amT5+uIUOGOLqcMhMXF6fOnTtr4cKFdlP5/dvMmjVL06dP1759+8rkw6iwxxhZAACugZeXl8aOHatXXnnlupjRAmXn4sWLmjFjhl544QVCrIMwRhYAgGv0zDPP6JlnnnF0GShj5cuXV0JCgqPL+FfjjiwAAABMiTGyAAAAMCXuyAIAAMCUCLIAAAAwJT7slYusrCwdPXpUFStWLPGvPgUAAEDeDMPQ2bNnFRgYKCen/O+5EmRzcfTo0RzfzQ4AAICyc+jQId1000359iHI5qJixYqSLr+AV39HOwAAAEpPSkqKgoKCbHksPwTZXGQPJ/D09CTIAgAAOEBhhnfyYS8AAACYEkEWAAAApkSQBQAAgCkxRhYAAJhaZmamLl686OgyUEjly5eXs7NziWyLIAsAAEzJMAwlJibq9OnTji4FRVSpUiX5+/tf83z9BFkAAGBK2SHW19dX7u7ufImRCRiGoXPnzun48eOSpICAgGvaHkEWAACYTmZmpi3EVqlSxdHloAjc3NwkScePH5evr+81DTPgw14AAMB0ssfEuru7O7gSFEf2ebvWsc0EWQAAYFoMJzCnkjpvBFkAAIDriMVi0eLFi0t9P506ddLIkSNLfT+liSALAABQhk6cOKHHHntM1atXl9Vqlb+/v8LCwrRu3TpJ0rFjxxQeHu7gKs2BD3sBAIAbxsyVu8t0f6O61i/yOvfee68yMjL08ccfq3bt2kpKSlJsbKySk5MlSf7+/iVd5g2LO7IAAABl5PTp0/r55581bdo0de7cWTVq1FDr1q0VGRmpnj17SrIfWnDgwAFZLBYtWLBAHTp0kJubm/7zn/9o9+7d+u2339SqVSt5eHgoPDxcJ06csO1n0KBB6t27tyZOnCgfHx95enrq0UcfVUZGRp61paena8yYMapWrZoqVKigkJAQxcXFlebLcc0IsgAAAGXEw8NDHh4eWrx4sdLT0wu9XlRUlF544QVt2bJF5cqV04MPPqixY8fq9ddf188//6y9e/dq/PjxduvExsZq586diouL05dffqlFixZp4sSJee5jxIgRWr9+vebNm6ft27fr/vvvV7du3bRnz55iH29pI8gCAACUkXLlyik6Oloff/yxKlWqpHbt2um5557T9u3b811vzJgxCgsLU6NGjfTUU09p8+bNGjdunNq1a6cWLVpoyJAhWr16td06Li4u+uijj9S4cWN1795dkyZN0htvvKGsrKwc209ISNDcuXO1cOFCdejQQXXq1NGYMWPUvn17zZ07t0Rfg5JEkAUAAChD9957r44ePapvv/1W3bp1U1xcnG655RZFR0fnuU6zZs1s/+/n5ydJatq0qV1b9rdlZQsODrabZ7dNmzZKTU3VoUOHcmz/jz/+UGZmpurXr2+7a+zh4aGffvpJ+/btK+6hljo+7HW9WD3V0RWUns6Rjq6gdHHuAABF5Orqqq5du6pr164aN26cHn74YUVFRWnQoEG59i9fvrzt/7PnYL26Lbc7rYWVmpoqZ2dnbd68Occ3bXl4eBR7u6WNIAsAAOBgN998c4nPHfv777/r/Pnztq+E3bBhgzw8PBQUFJSjb4sWLZSZmanjx4+rQ4cOJVpHaWJoAQAAQBlJTk7W7bffrs8++0zbt29XfHy8Fi5cqOnTp6tXr14luq+MjAwNGTJEf/31l5YtW6aoqCiNGDFCTk4541/9+vXVv39/DRgwQIsWLVJ8fLx+/fVXTZ06VUuXLi3RukoSd2QBAADKiIeHh0JCQjRz5kzt27dPFy9eVFBQkIYOHarnnnuuRPfVpUsX1atXT7fddpvS09PVr18/TZgwIc/+c+fO1ZQpUzR69GgdOXJEVatW1a233qq77rqrROsqSRbDMAxHF3G9SUlJkZeXl86cOSNPT8+y2SnjLM2LcwcAZe7ChQuKj49XrVq15Orq6uhyrjuDBg3S6dOny+Srbosjv/NXlBzG0AIAAACYEkEWAAAApsQYWQAAgBtMfnPS3ki4IwsAAABTIsgCAADAlAiyAAAAMCWCLAAAAEyJIAsAAABTIsgCAADAlAiyAAAAKDEHDhzQlClTlJqaWur7IsgCAADcQCZMmKDmzZtf0zYOHDggi8Wibdu2FWm99PR03X///apatao8PDyuqYbC4AsRAADAjWP11LLdX+fIYq126NAhRUVFKSYmRidPnlRAQIB69+6t8ePHq0qVKoXejsVi0TfffKPevXvb2saMGaMnnniiWHVlCwoK0rFjx1S1atUirTdq1CjdcccdevTRR69p/4VFkAUAAChD+/fvV5s2bVS/fn19+eWXqlWrlnbs2KGnn35aP/zwgzZs2CBvb+9ib9/Dw+Oa74Y6OzvL39+/yOu9/fbb17TfomJoAQAAQBkaPny4XFxctGLFCnXs2FHVq1dXeHi4fvzxRx05ckTPP/+8JKlmzZqaPHmy+vXrpwoVKqhatWqaPXu2bTs1a9aUJN19992yWCy251cPLRg0aJB69+6tl156SX5+fqpUqZImTZqkS5cu6emnn5a3t7duuukmzZ0717bO1UMLTp06pf79+8vHx0dubm6qV6+eXf9Dhw6pT58+qlSpkry9vdWrVy8dOHCgVF6/Kzk0yE6dOlX/+c9/VLFiRfn6+qp3797atWtXgestXLhQDRs2lKurq5o2baply5bZLTcMQ+PHj1dAQIDc3NwUGhqqPXv2lNZhAAAAFMo///yj5cuX6/HHH5ebm5vdMn9/f/Xv31/z58+XYRiSpFdeeUXBwcHaunWrnn32WT311FNauXKlJOm3336TJM2dO1fHjh2zPc/NqlWrdPToUa1Zs0YzZsxQVFSU7rrrLlWuXFkbN27Uo48+qkceeUSHDx/Odf1x48bpr7/+0g8//KCdO3fqnXfesQ07uHjxosLCwlSxYkX9/PPPWrdunTw8PNStWzdlZGRc82uWH4cG2Z9++knDhw/Xhg0btHLlSl28eFF33HGH0tLS8lznl19+Ub9+/TRkyBBt3bpVvXv3Vu/evfXnn3/a+kyfPl1vvPGG5syZo40bN6pChQoKCwvThQsXyuKwAAAAcrVnzx4ZhqFGjRrlurxRo0Y6deqUTpw4IUlq166dnn32WdWvX19PPPGE7rvvPs2cOVOS5OPjI0mqVKmS/P39bc9z4+3trTfeeEMNGjTQ//73PzVo0EDnzp3Tc889p3r16ikyMlIuLi5au3ZtrusnJCSoRYsWatWqlWrWrKnQ0FD16NFDkjR//nxlZWXpgw8+UNOmTdWoUSPNnTtXCQkJiouLK+5LVSgODbIxMTEaNGiQGjdurODgYEVHRyshIUGbN2/Oc53XX39d3bp109NPP61GjRpp8uTJuuWWW/TWW29Junw3dtasWXrhhRfUq1cvNWvWTJ988omOHj2qxYsXl9GRAQAA5C37jmtB2rRpk+P5zp07i7y/xo0by8np/2Kfn5+fmjZtanvu7OysKlWq6Pjx47mu/9hjj2nevHlq3ry5xo4dq19++cW27Pfff9fevXtVsWJF2/hcb29vXbhwQfv27StyrUVxXY2RPXPmjCTlO8B5/fr1Cg0NtWsLCwvT+vXrJUnx8fFKTEy06+Pl5aWQkBBbn6ulp6crJSXF7gEAAFDS6tatK4vFkmcY3blzpypXrpzv3dXiKF++vN1zi8WSa1tWVlau64eHh+vgwYMaNWqUjh49qi5dumjMmDGSpNTUVLVs2VLbtm2ze+zevVsPPvhgiR7H1a6bIJuVlaWRI0eqXbt2atKkSZ79EhMT5efnZ9fm5+enxMRE2/Lstrz6XG3q1Kny8vKyPYKCgq7lUAAAAHJVpUoVde3aVW+//bbOnz9vtywxMVGff/65+vbtK4vFIknasGGDXZ8NGzbYDUsoX768MjMzS79wXR7KMHDgQH322WeaNWuW3nvvPUnSLbfcoj179sjX11d169a1e3h5eZVqTddNkB0+fLj+/PNPzZs3r8z3HRkZqTNnztgehw4dKvMaAADAv8Nbb72l9PR0hYWFac2aNTp06JBiYmLUtWtXVatWTS+++KKt77p16zR9+nTt3r1bs2fP1sKFC/XUU0/ZltesWVOxsbFKTEzUqVOnSq3m8ePHa8mSJdq7d6927Nih77//3hao+/fvr6pVq6pXr176+eefFR8fr7i4OD355JN5fnispFwXQXbEiBH6/vvvtXr1at1000359vX391dSUpJdW1JSkm2us+z/5tfnalarVZ6ennYPAACA0lCvXj1t2rRJtWvXVp8+fVSnTh0NGzZMnTt31vr16+2GWI4ePVqbNm1SixYtNGXKFM2YMUNhYWG25a+99ppWrlypoKAgtWjRotRqdnFxUWRkpJo1a6bbbrtNzs7OtpuP7u7uWrNmjapXr6577rlHjRo10pAhQ3ThwoVSz1QWo7CjjUuBYRh64okn9M033yguLk716tUrcJ2+ffvq3Llz+u6772xtbdu2VbNmzTRnzhwZhqHAwECNGTNGo0ePliSlpKTI19dX0dHReuCBBwrcR0pKiry8vHTmzJmyC7Vl/U0kZamY33piGpw7AChzFy5cUHx8vGrVqiVXV1dHl1MqatasqZEjR2rkyJGOLqXE5Xf+ipLDHPrNXsOHD9cXX3yhJUuWqGLFirYxrF5eXra51QYMGKBq1app6tTLYeGpp55Sx44d9dprr6l79+6aN2+eNm3aZBunYbFYNHLkSE2ZMkX16tVTrVq1NG7cOAUGBtp9fRsAAADMzaFB9p133pEkderUya597ty5GjRokKTL85ZdOV1E27Zt9cUXX+iFF16wzX22ePFiuw+IjR07VmlpaRo2bJhOnz6t9u3bKyYm5ob9FxsAAMC/kUODbGFGNeQ2ke7999+v+++/P891LBaLJk2apEmTJl1LeQAAAA5TFl/xanbXxYe9AAAAgKIiyAIAAMCUCLIAAMC08vomKlzfSuq8OXSMLAAAQHG4uLjIyclJR48elY+Pj1xcXGzfhoXrl2EYysjI0IkTJ+Tk5CQXF5dr2h5BFgAAmI6Tk5Nq1aqlY8eO6ejRo44uB0Xk7u6u6tWr281MVRwEWQAAYEouLi6qXr26Ll26pMzMTEeXg0JydnZWuXLlSuQOOkEWAACYlsViUfny5VW+fHlHlwIH4MNeAAAAMCWCLAAAAEyJIAsAAABTIsgCAADAlAiyAAAAMCWCLAAAAEyJIAsAAABTIsgCAADAlAiyAAAAMCWCLAAAAEyJIAsAAABTIsgCAADAlAiyAAAAMCWCLAAAAEyJIAsAAABTIsgCAADAlAiyAAAAMCWCLAAAAEyJIAsAAABTIsgCAADAlAiyAAAAMCWCLAAAAEyJIAsAAABTIsgCAADAlAiyAAAAMKVyji4A+Vu/P9nRJVyzDZd22z0f1bW+gypBUazfn5zj3Jkd1x4A3Fi4IwsAAABTcmiQXbNmjXr06KHAwEBZLBYtXrw43/6DBg2SxWLJ8WjcuLGtz4QJE3Isb9iwYSkfCQAAAMqaQ4NsWlqagoODNXv27EL1f/3113Xs2DHb49ChQ/L29tb9999v169x48Z2/dauXVsa5QMAAMCBHDpGNjw8XOHh4YXu7+XlJS8vL9vzxYsX69SpUxo8eLBdv3Llysnf37/E6gQAAMD1x9RjZD/88EOFhoaqRo0adu179uxRYGCgateurf79+yshISHf7aSnpyslJcXuAQAAgOubaYPs0aNH9cMPP+jhhx+2aw8JCVF0dLRiYmL0zjvvKD4+Xh06dNDZs2fz3NbUqVNtd3u9vLwUFBRU2uUDAADgGpk2yH788ceqVKmSevfubdceHh6u+++/X82aNVNYWJiWLVum06dPa8GCBXluKzIyUmfOnLE9Dh06VMrVAwAA4FqZch5ZwzD00Ucf6aGHHpKLi0u+fStVqqT69etr7969efaxWq2yWq0lXSYAAABKkSnvyP7000/au3evhgwZUmDf1NRU7du3TwEBAWVQGQAAAMqKQ4Nsamqqtm3bpm3btkmS4uPjtW3bNtuHsyIjIzVgwIAc63344YcKCQlRkyZNciwbM2aMfvrpJx04cEC//PKL7r77bjk7O6tfv36leiwAAAAoWw4dWrBp0yZ17tzZ9jwiIkKSNHDgQEVHR+vYsWM5Zhw4c+aMvv76a73++uu5bvPw4cPq16+fkpOT5ePjo/bt22vDhg3y8fEpvQMBAABAmXNokO3UqZMMw8hzeXR0dI42Ly8vnTt3Ls915s2bVxKlAQAA4DpnyjGyAAAAAEEWAAAApmTK6bcAAACuO6unOrqC0tM50tEV5Io7sgAAADAlgiwAAABMiSALAAAAUyLIAgAAwJQIsgAAADAlgiwAAABMiSALAAAAUyLIAgAAwJQIsgAAADAlgiwAAABMiSALAAAAUyLIAgAAwJQIsgAAADAlgiwAAABMiSALAAAAUyLIAgAAwJQIsgAAADAlgiwAAABMiSALAAAAUyLIAgAAwJQIsgAAADAlgiwAAABMiSALAAAAUyLIAgAAwJQIsgAAADAlgiwAAABMiSALAAAAUyLIAgAAwJQIsgAAADAlgiwAAABMiSALAAAAU3JokF2zZo169OihwMBAWSwWLV68ON/+cXFxslgsOR6JiYl2/WbPnq2aNWvK1dVVISEh+vXXX0vxKAAAAOAIDg2yaWlpCg4O1uzZs4u03q5du3Ts2DHbw9fX17Zs/vz5ioiIUFRUlLZs2aLg4GCFhYXp+PHjJV0+AAAAHKicI3ceHh6u8PDwIq/n6+urSpUq5bpsxowZGjp0qAYPHixJmjNnjpYuXaqPPvpIzz777LWUCwAAgOuIKcfINm/eXAEBAeratavWrVtna8/IyNDmzZsVGhpqa3NyclJoaKjWr1/viFIBAABQSkwVZAMCAjRnzhx9/fXX+vrrrxUUFKROnTppy5YtkqSTJ08qMzNTfn5+duv5+fnlGEd7pfT0dKWkpNg9AAAAcH1z6NCComrQoIEaNGhge962bVvt27dPM2fO1Kefflrs7U6dOlUTJ04siRIBAABQRkx1RzY3rVu31t69eyVJVatWlbOzs5KSkuz6JCUlyd/fP89tREZG6syZM7bHoUOHSrVmAAAAXDvTB9lt27YpICBAkuTi4qKWLVsqNjbWtjwrK0uxsbFq06ZNntuwWq3y9PS0ewAAAOD65tChBampqba7qZIUHx+vbdu2ydvbW9WrV1dkZKSOHDmiTz75RJI0a9Ys1apVS40bN9aFCxf0wQcfaNWqVVqxYoVtGxERERo4cKBatWql1q1ba9asWUpLS7PNYgAAAIAbg0OD7KZNm9S5c2fb84iICEnSwIEDFR0drWPHjikhIcG2PCMjQ6NHj9aRI0fk7u6uZs2a6ccff7TbRt++fXXixAmNHz9eiYmJat68uWJiYnJ8AAwAAADm5tAg26lTJxmGkefy6Ohou+djx47V2LFjC9zuiBEjNGLEiGstDwAAANcx04+RBQAAwL8TQRYAAACmRJAFAACAKRFkAQAAYEoEWQAAAJgSQRYAAACmRJAFAACAKRFkAQAAYEoEWQAAAJgSQRYAAACmRJAFAACAKRFkAQAAYEoEWQAAAJgSQRYAAACmRJAFAACAKRFkAQAAYEoEWQAAAJgSQRYAAACmRJAFAACAKRFkAQAAYEoEWQAAAJgSQRYAAACmVM7RBQAAim/myt2OLqHEjepa39EloAA34nVXEm5NSHZ0CXba1K7i6BJKHXdkAQAAYEoEWQAAAJgSQRYAAACmRJAFAACAKRFkAQAAYEoEWQAAAJgSQRYAAACmRJAFAACAKRFkAQAAYEoEWQAAAJgSQRYAAACm5NAgu2bNGvXo0UOBgYGyWCxavHhxvv0XLVqkrl27ysfHR56enmrTpo2WL19u12fChAmyWCx2j4YNG5biUQAAAMARHBpk09LSFBwcrNmzZxeq/5o1a9S1a1ctW7ZMmzdvVufOndWjRw9t3brVrl/jxo117Ngx22Pt2rWlUT4AAAAcqJwjdx4eHq7w8PBC9581a5bd85deeklLlizRd999pxYtWtjay5UrJ39//5IqEwAAANchU4+RzcrK0tmzZ+Xt7W3XvmfPHgUGBqp27drq37+/EhISHFQhAAAASotD78heq1dffVWpqanq06ePrS0kJETR0dFq0KCBjh07pokTJ6pDhw76888/VbFixVy3k56ervT0dNvzlJSUUq8dAAAA1+aa7sju3btXy5cv1/nz5yVJhmGUSFGF8cUXX2jixIlasGCBfH19be3h4eG6//771axZM4WFhWnZsmU6ffq0FixYkOe2pk6dKi8vL9sjKCioLA4BAAAA16BQQTYrK8vueXJyskJDQ1W/fn3deeedOnbsmCRpyJAhGj16dMlXeZV58+bp4Ycf1oIFCxQaGppv30qVKql+/frau3dvnn0iIyN15swZ2+PQoUMlXTIAAABKWKGC7IwZM7Rs2TLb81GjRqlcuXJKSEiQu7u7rb1v376KiYkp+Sqv8OWXX2rw4MH68ssv1b179wL7p6amat++fQoICMizj9Vqlaenp90DAAAA17dCjZHt2rWr7r33Xh07dkxDhgzRihUrtHz5ct100012/erVq6eDBw8Weuepqal2d0rj4+O1bds2eXt7q3r16oqMjNSRI0f0ySefSLo8nGDgwIF6/fXXFRISosTEREmSm5ubvLy8JEljxoxRjx49VKNGDR09elRRUVFydnZWv379Cl0XAAAArn+FuiMbHBysX3/91faFBWlpaXZ3YrP9888/slqthd75pk2b1KJFC9vUWREREWrRooXGjx8vSTp27JjdjAPvvfeeLl26pOHDhysgIMD2eOqpp2x9Dh8+rH79+qlBgwbq06ePqlSpog0bNsjHx6fQdQEAAOD6V+hZC7y9vfXdd99Jkjp06KBPPvlEkydPliRZLBZlZWVp+vTp6ty5c6F33qlTp3w/IBYdHW33PC4ursBtzps3r9D7BwAAgHkVa/qt6dOnq0uXLtq0aZMyMjI0duxY7dixQ//884/WrVtX0jUCAAAAORRr+q0mTZpo9+7dat++vXr16qW0tDTdc8892rp1q+rUqVPSNQIAAAA5FOuObEJCgoKCgvT888/nuqx69erXXBgAAACQn2Ldka1Vq5ZOnDiRoz05OVm1atW65qIAAACAghQryBqGIYvFkqM9NTVVrq6u11wUAAAAUJAiDS2IiIiQdHmWgnHjxtlNwZWZmamNGzeqefPmJVogAAAAkJsiBdmtW7dKunxH9o8//pCLi4ttmYuLi4KDgzVmzJiSrRAAAADIRZGC7OrVqyVJgwcP1uuvv85XuQIAAMBhijVrwdy5c0u6DgAAAKBICh1k77nnHkVHR8vT01P33HNPvn0XLVp0zYUBAAAA+Sl0kPXy8rLNVODl5VVqBQEAAACFUegge+VwAoYWAAAAwNGKNY/s+fPnde7cOdvzgwcPatasWVqxYkWJFQYAAADkp1hBtlevXvrkk08kSadPn1br1q312muvqVevXnrnnXdKtEAAAAAgN8UKslu2bFGHDh0kSV999ZX8/f118OBBffLJJ3rjjTdKtEAAAAAgN8UKsufOnVPFihUlSStWrNA999wjJycn3XrrrTp48GCJFggAAADkplhBtm7dulq8eLEOHTqk5cuX64477pAkHT9+nC9JAAAAQJkoVpAdP368xowZo5o1ayokJERt2rSRdPnubIsWLUq0QAAAACA3xfpmr/vuu0/t27fXsWPHFBwcbGvv0qWL7r777hIrDgAAAMhLsYKsJPn7+8vf39+urXXr1tdcEAAAAFAYxQqyaWlpevnllxUbG6vjx48rKyvLbvn+/ftLpDgAAAAgL8UKsg8//LB++uknPfTQQwoICLB9dS0AAABQVooVZH/44QctXbpU7dq1K+l6AAAAgEIp1qwFlStXlre3d0nXAgAAABRasYLs5MmTNX78eJ07d66k6wEAAAAKpVhDC1577TXt27dPfn5+qlmzpsqXL2+3fMuWLSVSHAAAAJCXYgXZ3r17l3AZAAAAQNEUK8hGRUWVdB0AAABAkRRrjKwknT59Wh988IEiIyP1zz//SLo8pODIkSMlVhwAAACQl2Ldkd2+fbtCQ0Pl5eWlAwcOaOjQofL29taiRYuUkJCgTz75pKTrBAAAAOwU645sRESEBg0apD179sjV1dXWfuedd2rNmjUlVhwAAACQl2IF2d9++02PPPJIjvZq1aopMTHxmosCAAAAClKsIGu1WpWSkpKjfffu3fLx8bnmogAAAICCFCvI9uzZU5MmTdLFixclSRaLRQkJCXrmmWd07733lmiBAAAAQG6KFWRfe+01paamysfHR+fPn1fHjh1Vt25dVaxYUS+++GKht7NmzRr16NFDgYGBslgsWrx4cYHrxMXF6ZZbbpHValXdunUVHR2do8/s2bNVs2ZNubq6KiQkRL/++msRjg4AAABmUKwg6+XlpZUrV2rp0qV64403NGLECC1btkw//fSTKlSoUOjtpKWlKTg4WLNnzy5U//j4eHXv3l2dO3fWtm3bNHLkSD388MNavny5rc/8+fMVERGhqKgobdmyRcHBwQoLC9Px48eLfJwAAAC4fhV5+q2srCxFR0dr0aJFOnDggCwWi2rVqiV/f38ZhiGLxVLobYWHhys8PLzQ/efMmaNatWrptddekyQ1atRIa9eu1cyZMxUWFiZJmjFjhoYOHarBgwfb1lm6dKk++ugjPfvss0U4UgAAAFzPinRH1jAM9ezZUw8//LCOHDmipk2bqnHjxjp48KAGDRqku+++u7TqlCStX79eoaGhdm1hYWFav369JCkjI0ObN2+26+Pk5KTQ0FBbHwAAANwYinRHNjo6WmvWrFFsbKw6d+5st2zVqlXq3bu3PvnkEw0YMKBEi8yWmJgoPz8/uzY/Pz+lpKTo/PnzOnXqlDIzM3Pt8/fff+e53fT0dKWnp9ue5zYjAwAAAK4vRboj++WXX+q5557LEWIl6fbbb9ezzz6rzz//vMSKKytTp06Vl5eX7REUFOTokgAAAFCAIgXZ7du3q1u3bnkuDw8P1++//37NReXF399fSUlJdm1JSUny9PSUm5ubqlatKmdn51z7+Pv757ndyMhInTlzxvY4dOhQqdQPAACAklOkIPvPP//k+LP9lfz8/HTq1KlrLiovbdq0UWxsrF3bypUr1aZNG0mSi4uLWrZsadcnKytLsbGxtj65sVqt8vT0tHsAAADg+lakIJuZmaly5fIeVuvs7KxLly4Venupqanatm2btm3bJuny9Frbtm1TQkKCpMt3Sq8cb/voo49q//79Gjt2rP7++2+9/fbbWrBggUaNGmXrExERoffff18ff/yxdu7cqccee0xpaWm2WQwAAABwYyjSh70Mw9CgQYNktVpzXX7lB6YKY9OmTXbjbSMiIiRJAwcOVHR0tI4dO2YLtZJUq1YtLV26VKNGjdLrr7+um266SR988IFt6i1J6tu3r06cOKHx48crMTFRzZs3V0xMTL53kgEAAGA+RQqyAwcOLLBPUWYs6NSpkwzDyHN5bt/a1alTJ23dujXf7Y4YMUIjRowodB0AAAAwnyIF2blz55ZWHQAAAECRFOsragEAAABHI8gCAADAlAiyAAAAMCWCLAAAAEyJIAsAAABTIsgCAADAlAiyAAAAMCWCLAAAAEyJIAsAAABTIsgCAADAlAiyAAAAMCWCLAAAAEypnKMLwL/PzJW7HV1Cibo1IVltaldxdBnADeNGe48Y1bW+o0sAbljckQUAAIApEWQBAABgSgRZAAAAmBJBFgAAAKZEkAUAAIApEWQBAABgSgRZAAAAmBLzyAL491o91dEVXLNbE5IdXUKZ21B9mKNLQDHdmvCeo0vADYY7sgAAADAlgiwAAABMiSALAAAAUyLIAgAAwJQIsgAAADAlgiwAAABMiSALAAAAUyLIAgAAwJQIsgAAADAlgiwAAABMiSALAAAAUyLIAgAAwJSuiyA7e/Zs1axZU66urgoJCdGvv/6aZ99OnTrJYrHkeHTv3t3WZ9CgQTmWd+vWrSwOBQAAAGWknKMLmD9/viIiIjRnzhyFhIRo1qxZCgsL065du+Tr65uj/6JFi5SRkWF7npycrODgYN1///12/bp166a5c+fanlut1tI7CAAAAJQ5h9+RnTFjhoYOHarBgwfr5ptv1pw5c+Tu7q6PPvoo1/7e3t7y9/e3PVauXCl3d/ccQdZqtdr1q1y5clkcDgAAAMqIQ4NsRkaGNm/erNDQUFubk5OTQkNDtX79+kJt48MPP9QDDzygChUq2LXHxcXJ19dXDRo00GOPPabk5OQSrR0AAACO5dChBSdPnlRmZqb8/Pzs2v38/PT3338XuP6vv/6qP//8Ux9++KFde7du3XTPPfeoVq1a2rdvn5577jmFh4dr/fr1cnZ2zrGd9PR0paen256npKQU84gAAABQVhw+RvZafPjhh2ratKlat25t1/7AAw/Y/r9p06Zq1qyZ6tSpo7i4OHXp0iXHdqZOnaqJEyeWer0AAAAoOQ4dWlC1alU5OzsrKSnJrj0pKUn+/v75rpuWlqZ58+ZpyJAhBe6ndu3aqlq1qvbu3Zvr8sjISJ05c8b2OHToUOEPAgAAAA7h0CDr4uKili1bKjY21taWlZWl2NhYtWnTJt91Fy5cqPT0dP33v/8tcD+HDx9WcnKyAgICcl1utVrl6elp9wAAAMD1zeGzFkREROj999/Xxx9/rJ07d+qxxx5TWlqaBg8eLEkaMGCAIiMjc6z34Ycfqnfv3qpSpYpde2pqqp5++mlt2LBBBw4cUGxsrHr16qW6desqLCysTI4JAAAApc/hY2T79u2rEydOaPz48UpMTFTz5s0VExNj+wBYQkKCnJzs8/auXbu0du1arVixIsf2nJ2dtX37dn388cc6ffq0AgMDdccdd2jy5MnMJQsAAHADcXiQlaQRI0ZoxIgRuS6Li4vL0dagQQMZhpFrfzc3Ny1fvrwkywMAAMB1yOFDCwAAAIDiIMgCAADAlAiyAAAAMCWCLAAAAEyJIAsAAABTIsgCAADAlAiyAAAAMCWCLAAAAEyJIAsAAABTIsgCAADAlAiyAAAAMCWCLAAAAEyJIAsAAABTIsgCAADAlAiyAAAAMCWCLAAAAEyJIAsAAABTIsgCAADAlAiyAAAAMCWCLAAAAEyJIAsAAABTIsgCAADAlAiyAAAAMCWCLAAAAEyJIAsAAABTIsgCAADAlAiyAAAAMCWCLAAAAEyJIAsAAABTIsgCAADAlAiyAAAAMCWCLAAAAEyJIAsAAABTIsgCAADAlK6LIDt79mzVrFlTrq6uCgkJ0a+//ppn3+joaFksFruHq6urXR/DMDR+/HgFBATIzc1NoaGh2rNnT2kfBgAAAMqQw4Ps/PnzFRERoaioKG3ZskXBwcEKCwvT8ePH81zH09NTx44dsz0OHjxot3z69Ol64403NGfOHG3cuFEVKlRQWFiYLly4UNqHAwAAgDLi8CA7Y8YMDR06VIMHD9bNN9+sOXPmyN3dXR999FGe61gsFvn7+9sefn5+tmWGYWjWrFl64YUX1KtXLzVr1kyffPKJjh49qsWLF5fBEQEAAKAsODTIZmRkaPPmzQoNDbW1OTk5KTQ0VOvXr89zvdTUVNWoUUNBQUHq1auXduzYYVsWHx+vxMREu216eXkpJCQk320CAADAXBwaZE+ePKnMzEy7O6qS5Ofnp8TExFzXadCggT766CMtWbJEn332mbKystS2bVsdPnxYkmzrFWWb6enpSklJsXsAAADg+ubwoQVF1aZNGw0YMEDNmzdXx44dtWjRIvn4+Ojdd98t9janTp0qLy8v2yMoKKgEKwYAAEBpcGiQrVq1qpydnZWUlGTXnpSUJH9//0Jto3z58mrRooX27t0rSbb1irLNyMhInTlzxvY4dOhQUQ8FAAAAZcyhQdbFxUUtW7ZUbGysrS0rK0uxsbFq06ZNobaRmZmpP/74QwEBAZKkWrVqyd/f326bKSkp2rhxY57btFqt8vT0tHsAAADg+lbO0QVERERo4MCBatWqlVq3bq1Zs2YpLS1NgwcPliQNGDBA1apV09SpUyVJkyZN0q233qq6devq9OnTeuWVV3Tw4EE9/PDDki7PaDBy5EhNmTJF9erVU61atTRu3DgFBgaqd+/ejjpMAAAAlDCHB9m+ffvqxIkTGj9+vBITE9W8eXPFxMTYPqyVkJAgJ6f/u3F86tQpDR06VImJiapcubJatmypX375RTfffLOtz9ixY5WWlqZhw4bp9OnTat++vWJiYnJ8cQIAAADMy+FBVpJGjBihESNG5LosLi7O7vnMmTM1c+bMfLdnsVg0adIkTZo0qaRKBAAAwHXGdLMWAAAAABJBFgAAACZFkAUAAIApEWQBAABgStfFh70As1u/P9nRJaAQZq7cbff81gTOG0rf1dcdgJLDHVkAAACYEkEWAAAApsTQApS6WxPec3QJKCbOHa5HXJcAsnFHFgAAAKZEkAUAAIApEWQBAABgSgRZAAAAmBJBFgAAAKZEkAUAAIApEWQBAABgSgRZAAAAmBJBFgAAAKZEkAUAAIApEWQBAABgSgRZAAAAmBJBFgAAAKZEkAUAAIApEWQBAABgSgRZAAAAmBJBFgAAAKZEkAUAAIApEWQBAABgSgRZAAAAmBJBFgAAAKZEkAUAAIApEWQBAABgSgRZAAAAmBJBFgAAAKZEkAUAAIApEWQBAABgStdFkJ09e7Zq1qwpV1dXhYSE6Ndff82z7/vvv68OHTqocuXKqly5skJDQ3P0HzRokCwWi92jW7dupX0YAAAAKEMOD7Lz589XRESEoqKitGXLFgUHByssLEzHjx/PtX9cXJz69eun1atXa/369QoKCtIdd9yhI0eO2PXr1q2bjh07Znt8+eWXZXE4AAAAKCMOD7IzZszQ0KFDNXjwYN18882aM2eO3N3d9dFHH+Xa//PPP9fjjz+u5s2bq2HDhvrggw+UlZWl2NhYu35Wq1X+/v62R+XKlcvicAAAAFBGHBpkMzIytHnzZoWGhtranJycFBoaqvXr1xdqG+fOndPFixfl7e1t1x4XFydfX181aNBAjz32mJKTk/PcRnp6ulJSUuweAAAAuL45NMiePHlSmZmZ8vPzs2v38/NTYmJiobbxzDPPKDAw0C4Md+vWTZ988oliY2M1bdo0/fTTTwoPD1dmZmau25g6daq8vLxsj6CgoOIfFAAAAMpEOUcXcC1efvllzZs3T3FxcXJ1dbW1P/DAA7b/b9q0qZo1a6Y6deooLi5OXbp0ybGdyMhIRURE2J6npKQQZgEAAK5zDr0jW7VqVTk7OyspKcmuPSkpSf7+/vmu++qrr+rll1/WihUr1KxZs3z71q5dW1WrVtXevXtzXW61WuXp6Wn3AAAAwPXNoUHWxcVFLVu2tPugVvYHt9q0aZPnetOnT9fkyZMVExOjVq1aFbifw4cPKzk5WQEBASVSNwAAABzP4bMWRERE6P3339fHH3+snTt36rHHHlNaWpoGDx4sSRowYIAiIyNt/adNm6Zx48bpo48+Us2aNZWYmKjExESlpqZKklJTU/X0009rw4YNOnDggGJjY9WrVy/VrVtXYWFhDjlGAAAAlDyHj5Ht27evTpw4ofHjxysxMVHNmzdXTEyM7QNgCQkJcnL6v7z9zjvvKCMjQ/fdd5/ddqKiojRhwgQ5Oztr+/bt+vjjj3X69GkFBgbqjjvu0OTJk2W1Wsv02AAAAFB6LIZhGI4u4nqTkpIiLy8vnTlzpuzGy66emmvz+v15TxsGAACQlza1q5TcxjpHFtynhBQlhzl8aAEAAABQHARZAAAAmBJBFgAAAKZEkAUAAIApEWQBAABgSgRZAAAAmBJBFgAAAKZEkAUAAIApEWQBAABgSgRZAAAAmBJBFgAAAKZEkAUAAIApEWQBAABgSgRZAAAAmBJBFgAAAKZEkAUAAIApEWQBAABgSgRZAAAAmBJBFgAAAKZEkAUAAIApEWQBAABgSgRZAAAAmBJBFgAAAKZEkAUAAIApEWQBAABgSgRZAAAAmBJBFgAAAKZEkAUAAIApEWQBAABgSgRZAAAAmBJBFgAAAKZEkAUAAIApEWQBAABgSgRZAAAAmNJ1EWRnz56tmjVrytXVVSEhIfr111/z7b9w4UI1bNhQrq6uatq0qZYtW2a33DAMjR8/XgEBAXJzc1NoaKj27NlTmocAAACAMubwIDt//nxFREQoKipKW7ZsUXBwsMLCwnT8+PFc+//yyy/q16+fhgwZoq1bt6p3797q3bu3/vzzT1uf6dOn64033tCcOXO0ceNGVahQQWFhYbpw4UJZHRYAAABKmcUwDMORBYSEhOg///mP3nrrLUlSVlaWgoKC9MQTT+jZZ5/N0b9v375KS0vT999/b2u79dZb1bx5c82ZM0eGYSgwMFCjR4/WmDFjJElnzpyRn5+foqOj9cADDxRYU0pKiry8vHTmzBl5enqW0JEWYPXUXJvX708um/0DAIAbSpvaVUpuY50jS25bBShKDitXRjXlKiMjQ5s3b1Zk5P+9OE5OTgoNDdX69etzXWf9+vWKiIiwawsLC9PixYslSfHx8UpMTFRoaKhtuZeXl0JCQrR+/fpcg2x6errS09Ntz8+cOSPp8gtZZtJyv1ucdj4913YAAID8pOSRLYq3sbLLRNn5qzD3Wh0aZE+ePKnMzEz5+fnZtfv5+envv//OdZ3ExMRc+ycmJtqWZ7fl1edqU6dO1cSJE3O0BwUFFe5AAAAAbmiTynyPZ8+elZeXV759HBpkrxeRkZF2d3mzsrL0zz//qEqVKrJYLKW675SUFAUFBenQoUNlN4wBJYpzeGPgPJof59D8OIc3hms9j4Zh6OzZswoMDCywr0ODbNWqVeXs7KykpCS79qSkJPn7++e6jr+/f779s/+blJSkgIAAuz7NmzfPdZtWq1VWq9WurVKlSkU5lGvm6enJD63JcQ5vDJxH8+Mcmh/n8MZwLeexoDux2Rw6a4GLi4tatmyp2NhYW1tWVpZiY2PVpk2bXNdp06aNXX9JWrlypa1/rVq15O/vb9cnJSVFGzduzHObAAAAMB+HDy2IiIjQwIED1apVK7Vu3VqzZs1SWlqaBg8eLEkaMGCAqlWrpqlTL3+q/6mnnlLHjh312muvqXv37po3b542bdqk9957T5JksVg0cuRITZkyRfXq1VOtWrU0btw4BQYGqnfv3o46TAAAAJQwhwfZvn376sSJExo/frwSExPVvHlzxcTE2D6slZCQICen/7tx3LZtW33xxRd64YUX9Nxzz6levXpavHixmjRpYuszduxYpaWladiwYTp9+rTat2+vmJgYubq6lvnxFcRqtSoqKirH0AaYB+fwxsB5ND/OoflxDm8MZXkeHT6PLAAAAFAcDv9mLwAAAKA4CLIAAAAwJYIsAAAATIkgCwAAAFMiyJaB2bNnq2bNmnJ1dVVISIh+/fXXfPsvXLhQDRs2lKurq5o2baply5aVUaXIS1HO4fvvv68OHTqocuXKqly5skJDQws85ygbRf1ZzDZv3jxZLBam8LsOFPUcnj59WsOHD1dAQICsVqvq16/Pe6qDFfUczpo1Sw0aNJCbm5uCgoI0atQoXbhwoYyqxdXWrFmjHj16KDAwUBaLRYsXLy5wnbi4ON1yyy2yWq2qW7euoqOjS64gA6Vq3rx5houLi/HRRx8ZO3bsMIYOHWpUqlTJSEpKyrX/unXrDGdnZ2P69OnGX3/9ZbzwwgtG+fLljT/++KOMK0e2op7DBx980Jg9e7axdetWY+fOncagQYMMLy8v4/Dhw2VcOa5U1POYLT4+3qhWrZrRoUMHo1evXmVTLHJV1HOYnp5utGrVyrjzzjuNtWvXGvHx8UZcXJyxbdu2Mq4c2Yp6Dj///HPDarUan3/+uREfH28sX77cCAgIMEaNGlXGlSPbsmXLjOeff95YtGiRIcn45ptv8u2/f/9+w93d3YiIiDD++usv48033zScnZ2NmJiYEqmHIFvKWrdubQwfPtz2PDMz0wgMDDSmTp2aa/8+ffoY3bt3t2sLCQkxHnnkkVKtE3kr6jm82qVLl4yKFSsaH3/8cWmViEIoznm8dOmS0bZtW+ODDz4wBg4cSJB1sKKew3feeceoXbu2kZGRUVYlogBFPYfDhw83br/9dru2iIgIo127dqVaJwqnMEF27NixRuPGje3a+vbta4SFhZVIDQwtKEUZGRnavHmzQkNDbW1OTk4KDQ3V+vXrc11n/fr1dv0lKSwsLM/+KF3FOYdXO3funC5evChvb+/SKhMFKO55nDRpknx9fTVkyJCyKBP5KM45/Pbbb9WmTRsNHz5cfn5+atKkiV566SVlZmaWVdm4QnHOYdu2bbV582bb8IP9+/dr2bJluvPOO8ukZly70s41Dv9mrxvZyZMnlZmZafuWsmx+fn76+++/c10nMTEx1/6JiYmlVifyVpxzeLVnnnlGgYGBOX6QUXaKcx7Xrl2rDz/8UNu2bSuDClGQ4pzD/fv3a9WqVerfv7+WLVumvXv36vHHH9fFixcVFRVVFmXjCsU5hw8++KBOnjyp9u3byzAMXbp0SY8++qiee+65sigZJSCvXJOSkqLz58/Lzc3tmrbPHVmgFL388suaN2+evvnmm+vyK5KRu7Nnz+qhhx7S+++/r6pVqzq6HBRTVlaWfH199d5776lly5bq27evnn/+ec2ZM8fRpaGQ4uLi9NJLL+ntt9/Wli1btGjRIi1dulSTJ092dGm4TnBHthRVrVpVzs7OSkpKsmtPSkqSv79/ruv4+/sXqT9KV3HOYbZXX31VL7/8sn788Uc1a9asNMtEAYp6Hvft26cDBw6oR48etrasrCxJUrly5bRr1y7VqVOndIuGneL8LAYEBKh8+fJydna2tTVq1EiJiYnKyMiQi4tLqdYMe8U5h+PGjdNDDz2khx9+WJLUtGlTpaWladiwYXr++efl5MT9uOtdXrnG09Pzmu/GStyRLVUuLi5q2bKlYmNjbW1ZWVmKjY1VmzZtcl2nTZs2dv0laeXKlXn2R+kqzjmUpOnTp2vy5MmKiYlRq1atyqJU5KOo57Fhw4b6448/tG3bNtujZ8+e6ty5s7Zt26agoKCyLB8q3s9iu3bttHfvXts/QiRp9+7dCggIIMQ6QHHO4blz53KE1ex/mFz+rBGud6Wea0rkI2PI07x58wyr1WpER0cbf/31lzFs2DCjUqVKRmJiomEYhvHQQw8Zzz77rK3/unXrjHLlyhmvvvqqsXPnTiMqKorptxysqOfw5ZdfNlxcXIyvvvrKOHbsmO1x9uxZRx0CjKKfx6sxa4HjFfUcJiQkGBUrVjRGjBhh7Nq1y/j+++8NX19fY8qUKY46hH+9op7DqKgoo2LFisaXX35p7N+/31ixYoVRp04do0+fPo46hH+9s2fPGlu3bjW2bt1qSDJmzJhhbN261Th48KBhGIbx7LPPGg899JCtf/b0W08//bSxc+dOY/bs2Uy/ZTZvvvmmUb16dcPFxcVo3bq1sWHDBtuyjh07GgMHDrTrv2DBAqN+/fqGi4uL0bhxY2Pp0qVlXDGuVpRzWKNGDUNSjkdUVFTZFw47Rf1ZvBJB9vpQ1HP4yy+/GCEhIYbVajVq165tvPjii8alS5fKuGpcqSjn8OLFi8aECROMOnXqGK6urkZQUJDx+OOPG6dOnSr7wmEYhmGsXr06199x2edt4MCBRseOHXOs07x5c8PFxcWoXbu2MXfu3BKrx2IY3JsHAACA+TBGFgAAAKZEkAUAAIApEWQBAABgSgRZAAAAmBJBFgAAAKZEkAUAAIApEWQBAABgSgRZAPgXOXDggKZMmaLU1NRS31dGRoZeeukl7dy5s9T3BeDfiSALAEU0aNAg9e7d+5q2ceDAAVksFm3bti3PPnFxcbJYLDp9+rQkKTo6WpUqVbItnzBhgpo3b17ofaanp+v+++9X1apV5eHhUbzCi2D06NH6448/1LBhw3z7XetxAfj3IsgCuGENGjRIFotFFotFLi4uqlu3riZNmqRLly45urRCadu2rY4dOyYvL69cl48ZM0axsbG25wUF7FGjRumOO+7Qo48+WtKl5rBgwQLt2LFDH3/8sSwWS5HWLepxAfj3KufoAgCgNHXr1k1z585Venq6li1bpuHDh6t8+fKKjIzM0TcjI0MuLi4OqDJ3Li4u8vf3z3O5h4dHke6svv322yVRVq4yMzNlsVjk5HT5/kifPn3Up0+fYm2rqMcF4N+LO7IAbmhWq1X+/v6qUaOGHnvsMYWGhurbb7+V9H93+l588UUFBgaqQYMGkqQ//vhDt99+u9zc3FSlShUNGzYs1zGlEydOlI+Pjzw9PfXoo48qIyPDtiwmJkbt27dXpUqVVKVKFd11113at29fjm38/fffatu2rVxdXdWkSRP99NNPtmVXDy242pV/gp8wYYI+/vhjLVmyxHYXOi4uTpJ06NAh9enTR5UqVZK3t7d69eqlAwcO5PmaZe936dKlatasmVxdXXXrrbfqzz//tPXJHg7w7bff6uabb5bValVCQoLS09M1ZswYVatWTRUqVFBISIitjivXrV69utzd3XX33XcrOTm5TI4LwI2HIAvgX8XNzc0ucMbGxmrXrl1auXKlvv/+e6WlpSksLEyVK1fWb7/9poULF+rHH3/UiBEj7LYTGxurnTt3Ki4uTl9++aUWLVqkiRMn2panpaUpIiJCmzZtUmxsrJycnHT33XcrKyvLbjtPP/20Ro8era1bt6pNmzbq0aNHjmBXGGPGjFGfPn3UrVs3HTt2TMeOHVPbtm118eJFhYWFqWLFivr555+1bt06eXh4qFu3bnavQ26efvppvfbaa/rtt9/k4+OjHj166OLFi7bl586d07Rp0/TBBx9ox44d8vX11YgRI7R+/XrNmzdP27dv1/33369u3bppz549kqSNGzdqyJAhGjFihLZt26bOnTtrypQpZXpcAG4gBgDcoAYOHGj06tXLMAzDyMrKMlauXGlYrVZjzJgxtuV+fn5Genq6bZ333nvPqFy5spGammprW7p0qeHk5GQkJiba1vP29jbS0tJsfd555x3Dw8PDyMzMzLWWEydOGJKMP/74wzAMw4iPjzckGS+//LKtz8WLF42bbrrJmDZtmmEYhrF69WpDknHq1CnDMAxj7ty5hpeXl61/VFSUERwcnOvxZvv000+NBg0aGFlZWba29PR0w83NzVi+fHmutWbvd968eba25ORkw83NzZg/f76tFknGtm3bbH0OHjxoODs7G0eOHLHbXpcuXYzIyEjDMAyjX79+xp133mm3vG/fvmVyXABuPIyRBXBD+/777+Xh4aGLFy8qKytLDz74oCZMmGBb3rRpU7txsTt37lRwcLAqVKhga2vXrp2ysrK0a9cu+fn5SZKCg4Pl7u5u69OmTRulpqbq0KFDqlGjhvbs2aPx48dr48aNOnnypO1ObEJCgpo0aWK3XrZy5cqpVatWJTpd1e+//669e/eqYsWKdu0XLlzIdajDla6szdvbWw0aNLCrzcXFRc2aNbM9/+OPP5SZman69evbbSc9PV1VqlSRdPn1vfvuu3PsJyYmpsyOC8CNgyAL4IbWuXNnvfPOO3JxcVFgYKDKlbN/27sysJakHj16qEaNGnr//fcVGBiorKwsNWnSpMz/7J2amqqWLVvq888/z7HMx8fnmrbt5uZmNyNBamqqnJ2dtXnzZjk7O9v1LekPb5XmcQEwD4IsgBtahQoVVLdu3UL3b9SokaKjo5WWlmYLuevWrZOTk5Ptw2DS5TuC58+fl5ubmyRpw4YN8vDwUFBQkJKTk7Vr1y69//776tChgyRp7dq1ue5vw4YNuu222yRJly5d0ubNm3OMxy0sFxcXZWZm2rXdcsstmj9/vnx9feXp6Vmk7W3YsEHVq1eXJJ06dUq7d+9Wo0aN8uzfokULZWZm6vjx47bjvlqjRo20cePGHPvJT0kfF4AbBx/2AoAr9O/fX66urho4cKD+/PNPrV69Wk888YQeeugh27AC6fJUXUOGDNFff/2lZcuWKSoqSiNGjJCTk5MqV66sKlWq6L333tPevXu1atUqRURE5Lq/2bNn65tvvtHff/+t4cOH69SpU/rf//5XrNpr1qyp7du3a9euXTp58qQuXryo/v37q2rVqurVq5d+/vlnxcfHKy4uTk8++aQOHz6c7/YmTZqk2NhY/fnnnxo0aJCqVq2a73yu9evXV//+/TVgwAAtWrRI8fHx+vXXXzV16lQtXbpUkvTkk08qJiZGr776qvbs2aO33nqrwGEFJX1cAG4cBFkAuIK7u7uWL1+uf/75R//5z3903333qUuXLnrrrbfs+nXp0kX16tXTbbfdpr59+6pnz562sbdOTk6aN2+eNm/erCZNmmjUqFF65ZVXct3fyy+/rJdfflnBwcFau3atvv32W1WtWrVYtQ8dOlQNGjRQq1at5OPjo3Xr1snd3V1r1qxR9erVdc8996hRo0YaMmSILly4UOCdzJdffllPPfWUWrZsqcTERH333XcFzrM7d+5cDRgwQKNHj1aDBg3Uu3dv/fbbb7Y7u7feeqvef/99vf766woODtaKFSv0wgsvlOlxAbhxWAzDMBxdBADg+hEXF6fOnTvr1KlTdl8dCwDXG+7IAgAAwJQIsgAAADAlhhYAAADAlLgjCwAAAFMiyAIAAMCUCLIAAAAwJYIsAAAATIkgCwAAAFMiyAIAAMCUCLIAAAAwJYIsAAAATIkgCwAAAFP6f1DKpgapBppwAAAAAElFTkSuQmCC", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
y_testPred_simplePred_optimisé
66011
5100
80010
16110
103011
13100
\n", + "
" + ], + "text/plain": [ + " y_test Pred_simple Pred_optimisé\n", + "66 0 1 1\n", + "5 1 0 0\n", + "80 0 1 0\n", + "16 1 1 0\n", + "103 0 1 1\n", + "13 1 0 0" ] }, "metadata": {}, @@ -1462,52 +1463,169 @@ } ], "source": [ - "from sklearn.metrics import accuracy_score, f1_score, roc_auc_score\n", + "from sklearn.metrics import (roc_auc_score)\n", "\n", - "# Liste des modèles à comparer\n", - "model_names = [\"Modèle simple\", \"Modèle CV\", \"Modèle régularisé\"]\n", - "models = [logreg_simple, logreg_cv, best_logreg]\n", + "# 1. Calcul des probabilités \n", + "y_proba_simple = logreg_simple.predict_proba(X_test_scaled)[:, 1]\n", + "y_proba_best = best_logreg.predict_proba(X_test_scaled)[:, 1]\n", "\n", - "# Dictionnaires pour stocker les résultats\n", - "results = {\n", - " \"Accuracy\": [],\n", - " \"F1-score\": [],\n", - " \"AUC\": []\n", - "}\n", + "# 2. Récapitulatif des scores\n", + "results = pd.DataFrame({\n", + " 'Modèle': ['LogReg simple', 'LogReg optimisé (L1, C=100)'],\n", + " 'Accuracy': [\n", + " accuracy_score(y_test, y_pred_simple),\n", + " accuracy_score(y_test, y_pred_best)\n", + " ],\n", + " 'F1-score': [\n", + " f1_score(y_test, y_pred_simple),\n", + " f1_score(y_test, y_pred_best)\n", + " ],\n", + " 'AUC': [\n", + " roc_auc_score(y_test, y_proba_simple),\n", + " roc_auc_score(y_test, y_proba_best)\n", + " ]\n", + "})\n", + "print(results.round(3))\n", "\n", - "# Évaluation de chaque modèle\n", - "logreg_cv.fit(X_train_scaled, y_train)\n", + "# 3. Courbes ROC superposées\n", + "fpr_simple, tpr_simple, _ = roc_curve(y_test, y_proba_simple)\n", + "fpr_best, tpr_best, _ = roc_curve(y_test, y_proba_best)\n", "\n", - "for model in models:\n", - " y_pred = model.predict(X_test_scaled)\n", - " y_proba = model.predict_proba(X_test_scaled)[:, 1]\n", - " \n", - " results[\"Accuracy\"].append(accuracy_score(y_test, y_pred))\n", - " results[\"F1-score\"].append(f1_score(y_test, y_pred))\n", - " results[\"AUC\"].append(roc_auc_score(y_test, y_proba))\n", - "\n", - "# Affichage sous forme de DataFrame\n", - "df_results = pd.DataFrame(results, index=model_names)\n", - "print(\"\\nComparaison des modèles :\")\n", - "print(df_results.round(3))\n", - "\n", - "# Visualisation\n", - "df_results.plot(kind='bar', figsize=(10, 6))\n", - "plt.title(\"Comparaison des performances sur le jeu de test\")\n", - "plt.ylabel(\"Score\")\n", - "plt.ylim(0, 1.05)\n", - "plt.grid(axis='y')\n", - "plt.xticks(rotation=0)\n", + "plt.figure(figsize=(8, 6))\n", + "plt.plot(fpr_simple, tpr_simple, '--', label=f'Simple (AUC={roc_auc_score(y_test, y_proba_simple):.2f})')\n", + "plt.plot(fpr_best, tpr_best, '-', label=f'Optimisé (AUC={roc_auc_score(y_test, y_proba_best):.2f})')\n", + "plt.plot([0, 1], [0, 1], linestyle=':', color='grey')\n", + "plt.title(\"Comparaison des courbes ROC\")\n", + "plt.xlabel(\"Taux de faux positifs (FPR)\")\n", + "plt.ylabel(\"Taux de vrais positifs (TPR)\")\n", + "plt.legend()\n", + "plt.grid(True)\n", "plt.tight_layout()\n", - "plt.show()\n" + "plt.show()\n", + "\n", + "# 4. Comparaison des coefficients\n", + "coef_df = pd.DataFrame({\n", + " 'Feature': X_train_scaled.columns,\n", + " 'LogReg simple': logreg_simple.coef_[0],\n", + " 'LogReg optimisé (L1)': best_logreg.coef_[0]\n", + "}).set_index('Feature')\n", + "\n", + "plt.figure(figsize=(10, 6))\n", + "coef_df.T.plot(kind='barh', colormap='viridis')\n", + "plt.axvline(x=0, color='black', linestyle='--')\n", + "plt.title(\"Comparaison des coefficients entre les deux modèles\")\n", + "plt.xlabel(\"Coefficient\")\n", + "plt.ylabel(\"Modèle / Feature\")\n", + "plt.grid(True)\n", + "plt.tight_layout()\n", + "plt.show()\n", + "\n", + "# 5. Distribution des probabilités prédites (classe 1)\n", + "plt.figure(figsize=(7, 5))\n", + "plt.hist(y_proba_simple, bins=10, alpha=0.5, label=\"Simple\", density=True)\n", + "plt.hist(y_proba_best, bins=10, alpha=0.5, label=\"Optimisé\", density=True)\n", + "plt.title(\"Distribution des probabilités prédites (classe positive)\")\n", + "plt.xlabel(\"Probabilité prédite\")\n", + "plt.ylabel(\"Densité\")\n", + "plt.legend()\n", + "plt.tight_layout()\n", + "plt.show()\n", + "\n", + "# 6. Analyse des erreurs par individu\n", + "erreurs = pd.DataFrame({\n", + " 'y_test': y_test.values,\n", + " 'Pred_simple': y_pred_simple,\n", + " 'Pred_optimisé': y_pred_best\n", + "}, index=y_test.index)\n", + "\n", + "# Afficher uniquement les cas mal classés par au moins un des deux modèles\n", + "diff = erreurs.query(\"Pred_simple != y_test or Pred_optimisé != y_test\")\n", + "display(diff)\n" ] }, { "cell_type": "markdown", - "id": "33c1f1b0", + "id": "a30c65a7", "metadata": {}, "source": [ - " Les résultats très similaires entre les versions du modèle suggèrent que la régression logistique capture efficacement la structure du jeu de données, sans sur-apprentissage apparent. La régularisation n’a pas apporté de gain de performance notable, ce qui est cohérent avec la petite taille du dataset et la simplicité du modèle. " + "Nous avons comparé deux versions du modèle de régression logistique :\n", + "- un **modèle simple**, sans régularisation explicite (paramètres par défaut),\n", + "- un **modèle optimisé**, avec régularisation L1 (`penalty='l1'`, `C=100`), déterminé via validation croisée.\n", + "\n", + "---\n", + "\n", + "### Performances globales\n", + "\n", + "| Modèle | Accuracy | F1-score | AUC |\n", + "|-----------------------------|----------|----------|-------|\n", + "| LogReg simple | 0.792 | 0.783 | 0.811 |\n", + "| LogReg optimisé (L1, C=100) | 0.792 | 0.762 | 0.811 |\n", + "\n", + "Les deux modèles atteignent une **précision identique (0.792)** et une **capacité discriminante équivalente (AUC = 0.811)**. \n", + "Le **modèle simple obtient un F1-score légèrement supérieur**, indiquant un meilleur équilibre entre rappel et précision sur la classe positive.\n", + "\n", + "---\n", + "\n", + "### Courbes ROC\n", + "\n", + "Les courbes ROC confirment que les deux modèles sont très proches en termes de performance globale. \n", + "Ils possèdent la même AUC (0.81), et suivent des trajectoires similaires sur l’ensemble de test.\n", + "\n", + "---\n", + "\n", + "### Comparaison des coefficients\n", + "\n", + "Le graphe ci-dessous compare les poids attribués à chaque variable par les deux modèles.\n", + "\n", + "- Le modèle optimisé par L1 **modifie fortement certains coefficients**, en particulier ceux de `Insulin_log` et `HOMA_log`, ce qui indique une régularisation efficace.\n", + "- D'autres variables comme `Glucose` ou `BMI` conservent un rôle dominant dans les deux modèles.\n", + "\n", + "Cette comparaison renforce l’idée que le **modèle régularisé est plus parcimonieux**, tout en conservant l’essentiel du signal prédictif.\n", + "\n", + "---\n", + "\n", + "### Distribution des probabilités `predict_proba`\n", + "\n", + "Nous avons visualisé les distributions des probabilités prédites pour la classe positive par les deux modèles.\n", + "\n", + "- Les deux distributions couvrent des plages similaires, mais le modèle régularisé produit **des prédictions légèrement plus extrêmes** (plus proches de 0 ou 1),\n", + "- Cela peut indiquer une **plus grande confiance dans certaines décisions**, ou une **meilleure séparation des classes**.\n", + "\n", + "---\n", + "\n", + "### Comparaison des erreurs\n", + "\n", + "Nous avons identifié les cas où les prédictions diffèrent entre les deux modèles. \n", + "Seulement quelques individus sont classés différemment, ce qui explique les scores très proches. \n", + "Exemple :\n", + "\n", + "| id | Vérité | Simple | Optimisé |\n", + "|----|--------|--------|----------|\n", + "| 80 | 0 | 1 | 0 |\n", + "| 5 | 1 | 0 | 0 |\n", + "\n", + "Cela confirme que la **structure des erreurs est différente**, même si les performances agrégées sont proches.\n", + "\n", + "---\n", + "\n", + "### Analyse finale\n", + "\n", + "Les deux modèles sont proches en termes de performance, mais présentent :\n", + "- des **différences structurelles dans les coefficients**,\n", + "- des **légères différences de comportement** sur les individus en bord de décision,\n", + "- et une **forme de décision plus régulière** dans le cas du modèle L1.\n", + "\n", + "---\n", + "\n", + "### Conclusion\n", + "\n", + "| Modèle | Avantage principal |\n", + "|---------------------------|--------------------|\n", + "| **LogReg simple** | Légère supériorité en F1-score |\n", + "| **LogReg optimisé (L1)** | Parcimonie, coefficients régularisés, meilleure interprétabilité |\n", + "\n", + "En pratique, les deux modèles peuvent être retenus selon le compromis souhaité entre performance brute et simplicité structurelle. \n", + "Le modèle L1 est particulièrement pertinent dans un objectif de généralisation ou de sélection automatique de variables.\n" ] } ],