{ "cells": [ { "cell_type": "markdown", "metadata": { "id": "ZLsPh6QLqvsd" }, "source": [ "Autor: Jarosław Żygierewicz\n", "# Zanim przystąpimy do ćwiczenia\n", "W tym ćwiczeniu skorzystamy z modułu sklearn (scikit-learn) przeznaczonego do uczenia maszynowego. Zawiera wiele zoptymalizowanych i przydatnych funkcji i algorytmów. Alternatywnymi frameworkami są np. PyTorch, Tensorflow, Keras, Caffe2, chociaż z obserwacji środowiska wynika, że dominują raczej PyTorch i Tensorflow.\n" ] }, { "cell_type": "code", "execution_count": 1, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "executionInfo": { "elapsed": 1725, "status": "ok", "timestamp": 1605186097417, "user": { "displayName": "Anna Dawid", "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgUQAwZ7wyayL4BbiM0n_EANCgBjSdZ9H14lgcCFEE=s64", "userId": "02862484648310443813" }, "user_tz": -60 }, "id": "ZEDtJYrxqvsg", "outputId": "8e6705dc-8c21-442a-88f0-ad4dad0919c0" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Zainstalowana wersja scikit-learn: 0.23.2.\n" ] } ], "source": [ "import sklearn\n", "print('Zainstalowana wersja scikit-learn: {}.'.format(sklearn.__version__))" ] }, { "cell_type": "markdown", "metadata": { "id": "7wFBQnvOqvsr" }, "source": [ "Na stronie przedmiotu znajduje się instrukcja konfiguracji środowiska w przypadku pracy lokalnej, a nie w colabie. Może (nie musi) być przydatna dla osób pracujących na własnych laptopach:\n", "\n", "https://brain.fuw.edu.pl/edu/index.php/Uczenie_maszynowe_i_sztuczne_sieci_neuronowe/konfiguracja" ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "executionInfo": { "elapsed": 2734, "status": "ok", "timestamp": 1605186098429, "user": { "displayName": "Anna Dawid", "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgUQAwZ7wyayL4BbiM0n_EANCgBjSdZ9H14lgcCFEE=s64", "userId": "02862484648310443813" }, "user_tz": -60 }, "id": "mmIFNq4Bqvst" }, "outputs": [], "source": [ "import matplotlib.pyplot as plt\n", "plt.rcParams[\"figure.figsize\"] = (20,10)\n", "\n", "import numpy as np\n", "from scipy import diag, interp\n", "from itertools import cycle\n", "\n", "from sklearn.model_selection import StratifiedKFold, train_test_split\n", "from sklearn import metrics\n", "from sklearn.linear_model import LogisticRegression" ] }, { "cell_type": "markdown", "metadata": { "id": "YibBq2d4qvs0" }, "source": [ "# Ćwiczenie: Walidacja krzyżowa\n", "Już na ostatnich ćwiczeniach przerobiliśmy walidację krzyżową. Teraz przyjrzymy się jej bliżej i sprawdźmy efekt wspomniany na wykładzie, czyli efekt częstości występowania.\n", "\n", "* W tym ćwiczeniu przyjrzymy się jak miary jakości klasyfikatora zależą od proporcji klas w zbiorze uczącym i od rozmiaru zbioru uczącego\n", "* Klasyfikatorem będzie nadal regresja logistyczna, ale tym razem zamiast korzystać z własnej implementacji, skorzystamy z gotowej wersji bibliotecznej z modułu [scikit-learn] (http://scikit-learn.org/stable/index.html)\n" ] }, { "cell_type": "markdown", "metadata": { "id": "7GiQUZ_Cqvs1" }, "source": [ "Dzisiaj znowu wygenerujemy sztucznie dane, które później będziemy klasyfikować, a to po to, aby mieć pełną kontrolę nad częstością występowania klas.\n", "\n", "Poniższa funkcja przyjmuje za argument liczbę przykładów, którą ma wygenerować.\n", "Wygenerowane przykłady będą pochodziły z dwóch dwuwymiarowych rozkładów normalnych = klas określonych przez średnią mu[klasa] i macierz kowariancji cov[klasa]. Możemy te cechy dowolnie zmieniać." ] }, { "cell_type": "code", "execution_count": 3, "metadata": { "executionInfo": { "elapsed": 2732, "status": "ok", "timestamp": 1605186098430, "user": { "displayName": "Anna Dawid", "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgUQAwZ7wyayL4BbiM0n_EANCgBjSdZ9H14lgcCFEE=s64", "userId": "02862484648310443813" }, "user_tz": -60 }, "id": "WFTUaX6wqvs3" }, "outputs": [], "source": [ "def gen(ile):\n", " mu = [(-1,0.5),(1.2,4)] #średnie klas\n", " cov = [diag([3,3]), diag([4,1.7])] #macierze kowariancji dla klas\n", " \n", " X = np.zeros((ile*len(mu), 2)) # miejsce na dane wejściowe\n", " Y = np.zeros((ile*len(mu), 1),dtype = int) # miejsce na dane wyjściowe\n", " # użyj modułu np.random aby wygenerować liczby w X z rozkładu normalnego wielowymiarowego\n", " # Y powinno zawierać liczbowy identyfikator klasy, tj. 0, 1, 2, 3 ..\n", " for klasa in range(len(mu)):\n", " # miejsce na twój kod\n", " X[klasa*ile:(klasa+1)*ile] = np.random.multivariate_normal(mu[klasa],cov[klasa],ile)\n", " Y[klasa*ile:(klasa+1)*ile] = klasa\n", " # ravel spłaszczy nam Y do 1D\n", " Y = Y.ravel()\n", " return (X,Y)" ] }, { "cell_type": "markdown", "metadata": { "id": "QYxjhhHsqvs9" }, "source": [ "Testujemy tę funkcję, generujemy 50 przykładów, pierwszych 5 wypisujemy, wszystkie rysujemy za pomocą funkcji `scatter`. Pamiętajmy, że za każdym wywołaniem funkcji będziemy dostawać różne dane (jako że są one generowane przez np.random)." ] }, { "cell_type": "code", "execution_count": 4, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 728 }, "executionInfo": { "elapsed": 2726, "status": "ok", "timestamp": 1605186098431, "user": { "displayName": "Anna Dawid", "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgUQAwZ7wyayL4BbiM0n_EANCgBjSdZ9H14lgcCFEE=s64", "userId": "02862484648310443813" }, "user_tz": -60 }, "id": "8Df_pwsnqvs-", "outputId": "93b68b5b-92a8-43ca-ad2a-0c603f0b980f" }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ ":3: DeprecationWarning: scipy.diag is deprecated and will be removed in SciPy 2.0.0, use numpy.diag instead\n", " cov = [diag([3,3]), diag([4,1.7])] #macierze kowariancji dla klas\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "X: [[-0.51744353 4.17626645]\n", " [-4.12508267 2.07449487]\n", " [ 2.11435246 0.70771154]\n", " [-1.70383565 1.51593513]\n", " [-0.6522316 -2.21405951]]\n", "Y: [0 0 0 0 0]\n" ] }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "# miejsce na Twój kod\n", "X,Y = gen(50)\n", "print('X: ', X[0:5,:])\n", "print('Y: ', Y[0:5])\n", "plt.scatter(X[:,0], X[:,1], c = Y, cmap=plt.cm.Set1, alpha = 0.5)\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": { "id": "JWmqK-c4qvtB" }, "source": [ "# Klasy równoliczne" ] }, { "cell_type": "markdown", "metadata": { "id": "gmeJZgcxqvtC" }, "source": [ "## Zaobserwujmy zmienność miar jakości klasyfikatora przy wybieraniu podzbiorów do uczenia i testowania ze zbioru uczącego.\n", "* do podziałów zbioru zastosujemy funkcję [train_test_split](http://scikit-learn.org/stable/modules/generated/sklearn.model_selection.train_test_split.html)\n", "* do obliczania miar jakości zastosujemy funkcje z modułu [sklearn.metrics](http://scikit-learn.org/stable/modules/model_evaluation.html#classification-metrics)" ] }, { "cell_type": "markdown", "metadata": { "id": "ieaSg8JlqvtD" }, "source": [ "Wykonaj podział zbioru uczącego tak, aby zestaw testowy stanowił 20% całego zbioru uczącego. Zilustruj za pomocą `scatter` punkty należące do części uczącej i do części testowej.\n", "\n", "W razie wątpliwości, sprawdź funkcję w dokumentacji. Warto to robić dla wszelkich problemów, funkcje i metody z bibliotek ML są zazwyczaj dobrze opisane.\n", "\n", "* Podział:" ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "executionInfo": { "elapsed": 2725, "status": "ok", "timestamp": 1605186098433, "user": { "displayName": "Anna Dawid", "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgUQAwZ7wyayL4BbiM0n_EANCgBjSdZ9H14lgcCFEE=s64", "userId": "02862484648310443813" }, "user_tz": -60 }, "id": "EHbXlRRlqvtE" }, "outputs": [], "source": [ "X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size = 0.2)" ] }, { "cell_type": "markdown", "metadata": { "id": "sqsemxJQqvtG" }, "source": [ "* Ilustracja:" ] }, { "cell_type": "code", "execution_count": 6, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 592 }, "executionInfo": { "elapsed": 2718, "status": "ok", "timestamp": 1605186098433, "user": { "displayName": "Anna Dawid", "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgUQAwZ7wyayL4BbiM0n_EANCgBjSdZ9H14lgcCFEE=s64", "userId": "02862484648310443813" }, "user_tz": -60 }, "id": "3drSx4ScqvtH", "outputId": "5ad63834-2916-472a-834c-399d6aab948b" }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "plt.figure(figsize=(6,6))\n", "plt.scatter(X_train[:,0], X_train[:,1], c = y_train, cmap=plt.cm.Set1, alpha = 0.5, s=50)\n", "plt.scatter(X_test[:,0], X_test[:,1], c = y_test, cmap=plt.cm.Set1, alpha = 0.5, marker = '*', s=50 )\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": { "id": "sTMtVm5-qvtK" }, "source": [ "Regresja logistyczna zaimplementowana jest w klasie [`LogisticRegression`](http://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html). Tworzymy instancję obiektu tej klasy. Jeśli nie ustawimy żadnego algorytmu optymalizacji, to domyślnie zostanie ustawiony solver 'lbfgs', ale jednocześnie generuje to irytujący warning, więc lepiej zadeklarować go z góry." ] }, { "cell_type": "code", "execution_count": 7, "metadata": { "executionInfo": { "elapsed": 2716, "status": "ok", "timestamp": 1605186098434, "user": { "displayName": "Anna Dawid", "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgUQAwZ7wyayL4BbiM0n_EANCgBjSdZ9H14lgcCFEE=s64", "userId": "02862484648310443813" }, "user_tz": -60 }, "id": "rTYUHUEKqvtL" }, "outputs": [], "source": [ "model = LogisticRegression(solver = 'lbfgs')" ] }, { "cell_type": "markdown", "metadata": { "id": "Nzoph_sVqvtO" }, "source": [ "Uczymy go na zbiorze uczącym (fitujemy):" ] }, { "cell_type": "code", "execution_count": 8, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "executionInfo": { "elapsed": 2711, "status": "ok", "timestamp": 1605186098435, "user": { "displayName": "Anna Dawid", "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgUQAwZ7wyayL4BbiM0n_EANCgBjSdZ9H14lgcCFEE=s64", "userId": "02862484648310443813" }, "user_tz": -60 }, "id": "v4kNQhQ1qvtP", "outputId": "f4c3babf-a964-47cf-8f93-455d364b918b" }, "outputs": [ { "data": { "text/plain": [ "LogisticRegression()" ] }, "execution_count": 8, "metadata": {}, "output_type": "execute_result" } ], "source": [ "model.fit(X_train, y_train)" ] }, { "cell_type": "markdown", "metadata": { "id": "9VV0FB_gqvtR" }, "source": [ "Wykonujemy predykcje dla zbioru testowego:" ] }, { "cell_type": "code", "execution_count": 9, "metadata": { "executionInfo": { "elapsed": 2710, "status": "ok", "timestamp": 1605186098436, "user": { "displayName": "Anna Dawid", "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgUQAwZ7wyayL4BbiM0n_EANCgBjSdZ9H14lgcCFEE=s64", "userId": "02862484648310443813" }, "user_tz": -60 }, "id": "AwAdNSmXqvtS" }, "outputs": [], "source": [ "y_pred = model.predict(X_test) " ] }, { "cell_type": "markdown", "metadata": { "id": "OKSvOoqXqvtU" }, "source": [ "Efekty można obejrzeć za pomocą macierzy pomyłek: " ] }, { "cell_type": "code", "execution_count": 10, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "executionInfo": { "elapsed": 2701, "status": "ok", "timestamp": 1605186098436, "user": { "displayName": "Anna Dawid", "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgUQAwZ7wyayL4BbiM0n_EANCgBjSdZ9H14lgcCFEE=s64", "userId": "02862484648310443813" }, "user_tz": -60 }, "id": "cFK3Iy5HqvtV", "outputId": "bc6bb236-c4e9-43e6-a0d1-cccf8acfc54f" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[[8 4]\n", " [2 6]]\n", "TN: 8 FP: 4 FN: 2 TP: 6\n" ] } ], "source": [ "# użyj funkcji sklearn.metrics.confusion_matrix i wypisz wartości TN, FP, FN, TP\n", "print(metrics.confusion_matrix(y_test, y_pred))\n", "tn, fp, fn, tp = metrics.confusion_matrix(y_test, y_pred).ravel()\n", "print('TN: ',tn,'FP: ', fp, 'FN: ', fn, 'TP: ', tp )" ] }, { "cell_type": "markdown", "metadata": { "id": "-PATG0geqvtX" }, "source": [ "W pętli powtórzymy proces podziału zbioru uczącego i dla każdego podziału obliczmy miary jakości:\n", "* precyzja pozytywna: (positive predictive value (PPV), precision). Odpowiada na pytanie: \"Jeśli wynik testu jest pozytywny, jakie jest prawdopodobieństwo, że osoba badana jest chora?\"\n", "\n", "$\\qquad$ $PPV = \\frac{TP}{P'}=\\frac{TP}{ TP + FP}$\n", "\n", "* czułość: Prawdopodobieństwo, że klasyfikacja będzie poprawna pod warunkiem, że przypadek jest pozytywny (ang. True Positive Rate, Recall). Jest to np. prawdopodobieństwo, że test wykonany dla osoby chorej wykaże, że jest ona chora.\n", "\n", "$\\qquad$ $TPR = \\frac{TP}{ P} = \\frac{TP} { TP+FN}$\n", "\n", "\n", "* dokładność ( accuracy (ACC)): Prawdopodobieństwo prawidłowej klasyfikacji.\n", "\n", "$\\qquad$ $ACC = \\frac{TP + TN}{P + N}$\n", "\n", "* F1-score: średnia harmoniczna z precyzji i czułości:\n", "\n", "$\\qquad$ $F_1= 2 \\frac{PPV \\cdot TPR}{PPV+TPR}= \\frac{2TP}{ 2TP+FP+FN}$\n", "Miara ta daje ocenę balansu między czułością a precyzją. Miara ta nie uwzględnia wyników prawdziwie negatywnych.\n", "\n", "* współczynnik korelacji Matthews ( Matthews correlation coefficient):\n", "\n", "$\\qquad$ $\n", "\\text{MCC} = \\frac{ TP \\cdot TN - FP \\cdot FN } {\\sqrt{ (TP + FP) ( TP + FN ) ( TN + FP ) ( TN + FN ) } }\n", "$\n", "\n", " * Ten współczynnik uwzględnia wyniki zarówno prawdziwie jaki i fałszywie pozytywne i negatywne i jest na ogół uważany jako zrównoważona miara, która może być stosowana nawet wtedy, gdy klasy są bardzo różnej liczebności. \n", " * MCC jest w istocie współczynnikiem korelacji pomiędzy obserwowanymi i przewidywanymi klasyfikacjami binarnymi; zwraca wartość od -1 do +1. \n", " * Współczynnik +1 odpowiada idealnej klasyfikacji, \n", " * 0 nie lepiej niż losowe przypisanie wyniku i \n", " * -1 oznacza całkowitą niezgodę między klasyfikacją i stanem faktycznym." ] }, { "cell_type": "code", "execution_count": 11, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "executionInfo": { "elapsed": 2875, "status": "ok", "timestamp": 1605186098616, "user": { "displayName": "Anna Dawid", "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgUQAwZ7wyayL4BbiM0n_EANCgBjSdZ9H14lgcCFEE=s64", "userId": "02862484648310443813" }, "user_tz": -60 }, "id": "T_DmAtvsqvtY", "outputId": "1cbe9783-bf70-4466-c0e3-f8b29ad91781" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "PPV = 0.900 REC = 0.900 ACC = 0.900 F1 = 0.900 MCC = 0.800 \n", "PPV = 0.800 REC = 0.889 ACC = 0.850 F1 = 0.842 MCC = 0.704 \n", "PPV = 0.667 REC = 1.000 ACC = 0.800 F1 = 0.800 MCC = 0.667 \n", "PPV = 0.778 REC = 0.778 ACC = 0.800 F1 = 0.778 MCC = 0.596 \n", "PPV = 0.889 REC = 0.800 ACC = 0.850 F1 = 0.842 MCC = 0.704 \n", "PPV = 0.750 REC = 0.900 ACC = 0.800 F1 = 0.818 MCC = 0.612 \n", "PPV = 0.800 REC = 0.800 ACC = 0.800 F1 = 0.800 MCC = 0.600 \n", "PPV = 0.917 REC = 0.917 ACC = 0.900 F1 = 0.917 MCC = 0.792 \n", "PPV = 0.636 REC = 1.000 ACC = 0.800 F1 = 0.778 MCC = 0.664 \n", "PPV = 0.750 REC = 0.818 ACC = 0.750 F1 = 0.783 MCC = 0.492 \n" ] } ], "source": [ "# stwórz instancję klasyfikatora\n", "model = LogisticRegression(solver = 'lbfgs') \n", "for i in range(10):\n", " # podziel zbiór z 20% do testowania\n", " X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size = 0.2) \n", " # naucz klasyfikator\n", " model.fit(X_train, y_train)\n", " # wykonaj predykcję dla zbioru testowego\n", " y_pred = model.predict(X_test) \n", " # oblicz miary jakości korzystając z modułu sklearn.metrics\n", " PPV = metrics.precision_score(y_test, y_pred)\n", " REC = metrics.recall_score(y_test, y_pred)\n", " ACC = metrics.accuracy_score(y_test, y_pred)\n", " F1 = metrics.f1_score(y_test, y_pred)\n", " MCC = metrics.matthews_corrcoef(y_test, y_pred)\n", " # wypisywanie\n", " print('PPV = {p:.3f} REC = {r:.3f} ACC = {a:.3f} F1 = {f:.3f} MCC = {m:.3f} '.format(a=ACC,f=F1,m=MCC,p=PPV,r=REC))" ] }, { "cell_type": "markdown", "metadata": { "id": "gorayUYfqvtc" }, "source": [ "Widzimy, że miary zmieniają się przy każdym losowaniu.\n", "\n", "Najczęściej stosuje się nie takie losowe podziały, ale systematyczny podział `k`-krotny (k-fold cross-validation). Procedura wygląda wówczas następująco:\n", "* Dzielimy zbiór uczący (X i y) na `k` równych części\n", "* Odkładamy 1-szą część jako dane testowe, \n", "* Na pozostałych `k-1` częściach uczymy klasyfikator\n", "* Obliczamy miary jakości na tej odłożonej części\n", "* Wybieramy 2-gą część jako dane testowe\n", "* Na pozostałych `k-1` częściach uczymy klasyfikator\n", "* Obliczamy miary jakości na tej odłożonej części\n", "* $\\vdots$\n", "\n", "W bibliotece `sklearn` mamy do tego wygodną funkcję `cross_val_score`:" ] }, { "cell_type": "code", "execution_count": 12, "metadata": { "executionInfo": { "elapsed": 2873, "status": "ok", "timestamp": 1605186098617, "user": { "displayName": "Anna Dawid", "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgUQAwZ7wyayL4BbiM0n_EANCgBjSdZ9H14lgcCFEE=s64", "userId": "02862484648310443813" }, "user_tz": -60 }, "id": "Om4ZI2NCqvtd" }, "outputs": [], "source": [ "from sklearn.model_selection import cross_val_score" ] }, { "cell_type": "markdown", "metadata": { "id": "b3RLAd8aqvtf" }, "source": [ "Zobaczmy jak działa:" ] }, { "cell_type": "code", "execution_count": 13, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "executionInfo": { "elapsed": 2867, "status": "ok", "timestamp": 1605186098617, "user": { "displayName": "Anna Dawid", "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgUQAwZ7wyayL4BbiM0n_EANCgBjSdZ9H14lgcCFEE=s64", "userId": "02862484648310443813" }, "user_tz": -60 }, "id": "8xn_Q1HEqvtg", "outputId": "6fc69a65-915f-4030-959c-bb42fd0ff1d9" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "PPV = 0.84 +/- 0.11\n", "REC = 0.88 +/- 0.13\n", "ACC = 0.85 +/- 0.09\n", "F1 = 0.85 +/- 0.10\n" ] } ], "source": [ "# użyj funkcji cross_val_score, żeby policzyć miary jakości w k-krotnej walidacji krzyżowej\n", "ppv = cross_val_score(model, X, Y, cv=10, scoring='precision')\n", "# zwróć uwagę na estetyczny sposób wypisywania wyniku\n", "print('PPV = {0:.2f} +/- {1:.2f}'.format(ppv.mean(), ppv.std()))\n", "rec = cross_val_score(model, X, Y, cv=10, scoring='recall')\n", "print('REC = {0:.2f} +/- {1:.2f}'.format(rec.mean(),rec.std()))\n", "acc = cross_val_score(model, X, Y, cv=10, scoring='accuracy')\n", "print('ACC = {0:.2f} +/- {1:.2f}'.format(acc.mean(),acc.std()))\n", "f1 = cross_val_score(model, X, Y, cv=10, scoring='f1')\n", "print('F1 = {0:.2f} +/- {1:.2f}'.format(f1.mean(),f1.std()))" ] }, { "cell_type": "markdown", "metadata": { "id": "9OAmVCSKqvti" }, "source": [ "Dla kompletu zbadajmy jeszcze krzywą ROC. Tym razem też posłużymy się funkcjami bibliotecznymi." ] }, { "cell_type": "code", "execution_count": 14, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 723 }, "executionInfo": { "elapsed": 3660, "status": "ok", "timestamp": 1605186099417, "user": { "displayName": "Anna Dawid", "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgUQAwZ7wyayL4BbiM0n_EANCgBjSdZ9H14lgcCFEE=s64", "userId": "02862484648310443813" }, "user_tz": -60 }, "id": "g99d62Wvqvtj", "outputId": "3b3a37b9-73fc-4d82-cb12-054c1e27abfa" }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "skf = StratifiedKFold(n_splits=6)\n", "model = LogisticRegression(solver = 'lbfgs')\n", "tprs = []\n", "aucs = []\n", "mean_fpr = np.linspace(0, 1, 100)\n", "\n", "# pętla w której dokonujemy walidacji krzyżowej i liczymy krzywe roc\n", "i = 0\n", "for train, test in skf.split(X, Y):\n", " # fitujemy regresję\n", " model.fit(X[train], Y[train])\n", " # obliczamy prawdopodobieństwa przynależności przykładów testowych \n", " # do klas wg. wyuczonego klasyfikatora \n", " # (zwraca on w danym wierszu prawdopodobieństaw dla każdej z możliwych klas)\n", " probas_ = model.predict_proba(X[test]) \n", " \n", " # Obliczamy punkty krzywej ROC (wykorzystaj odpowiednią funckję z modułu sklearn.metrics)\n", " fpr, tpr, thresholds = metrics.roc_curve(Y[test], probas_[:, 1]) # względem prawdopodobieństwa klasy 1 \n", " \n", " # dostaliśmy pary wartości fpr (oś x) i tpr (oś y), ale te punkty nie są równoodległe\n", " # chcemy narysować krzywą roc dla punktów z tablicy mean_fpr które są równoodległe\n", " # za pomocą funkcji np.interp interpoluj wartości krzywej w punktach mean_fpr i dodaj je do tablicy tprs\n", " tprs.append(np.interp(mean_fpr, fpr, tpr))\n", " \n", " # ten punkt ustalamy, żeby krzywa zaczynała się w punkcie (0.0, 0.0), chodzi o estetykę wykresu\n", " tprs[-1][0] = 0.0\n", " # obliczamy powierzchnię pod krzywą (wykorzystaj odpowiednią funckję z modułu sklearn.metrics)\n", " roc_auc = metrics.auc(fpr, tpr)\n", " aucs.append(roc_auc)\n", " # rysujemy krzywą \n", " plt.plot(fpr, tpr, lw=1, alpha=0.3,\n", " label='ROC dla podziału %d (AUC = %0.2f)' % (i, roc_auc))\n", " i += 1\n", " \n", " \n", " \n", "# narysuj prostą odpowiadającą klasyfikatorowi losowemu = rzutowi monetą (podpowiedź: chodzi o diagonalę) \n", "# niech prosta będzie rysowana przerywaną linią o grubości dwa i w kolorze czerwonym\n", "plt.plot([0, 1], [0, 1], linestyle='--', lw = 2, color='r',\n", " label = 'Losowa klasa', alpha = 0.8)\n", " \n", "########################################\n", "# poniżej podsumowanie: oliczanie średnich i standardowych odchyleń, cieniowanie przedziału ufności \n", "########################################\n", "# policz średnie tpr po wszystkich klasach\n", "mean_tpr = np.mean(tprs, axis=0)\n", "# znów estetyka\n", "mean_tpr[-1] = 1.0\n", "# policz średnia auc korzystając z tablic mean_fpr i mean_tpr\n", "mean_auc = metrics.auc(mean_fpr, mean_tpr)\n", "# policz standardowe odchylenie danych w tablicy aucs\n", "std_auc = np.std(aucs)\n", "\n", "# proszę narysować krzywą ROC dla danych w mean_fpr i mean_tpr. Krzywa ma mieć grubość 2, kolor niebieski i \n", "# następującą etykietę: r'Średni ROC (AUC = %0.2f $\\pm$ %0.2f)' % (mean_auc, std_auc)\n", "plt.plot(mean_fpr, mean_tpr, color='b',\n", " label=r'Średni ROC (AUC = %0.2f $\\pm$ %0.2f)' % (mean_auc, std_auc),\n", " lw=2, alpha=.8)\n", "\n", "# policz odchylenie dla tprs\n", "std_tpr = np.std(tprs, axis=0)\n", "# poniższy kod zwróci nam przedział ufności 68% wokół średniej ROC (1 sigma)\n", "tprs_upper = np.minimum(mean_tpr + std_tpr, 1)\n", "tprs_lower = np.maximum(mean_tpr - std_tpr, 0)\n", "# użyj funkcji fill_between żeby zacieniować obszar między tprs_upper i tprs_lower na szaro z parametrem alpha=0.2\n", "plt.fill_between(mean_fpr, tprs_lower, tprs_upper, color='grey', alpha=.2,\n", " label=r'$\\pm$ 1 std. dev.')\n", "\n", "# estetyka\n", "plt.xlim([-0.05, 1.05])\n", "plt.ylim([-0.05, 1.05])\n", "plt.xlabel('False Positive Rate')\n", "plt.ylabel('True Positive Rate')\n", "plt.title('ROC')\n", "plt.legend(loc=\"lower right\")\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": { "id": "oJ-Nox9uqvtl" }, "source": [ "Sprawdźmy jak miary jakości od rozmiaru zbioru uczącego:" ] }, { "cell_type": "code", "execution_count": 15, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 935 }, "executionInfo": { "elapsed": 8829, "status": "ok", "timestamp": 1605186104593, "user": { "displayName": "Anna Dawid", "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgUQAwZ7wyayL4BbiM0n_EANCgBjSdZ9H14lgcCFEE=s64", "userId": "02862484648310443813" }, "user_tz": -60 }, "id": "_T0JqaIIqvtn", "outputId": "965b8530-2549-4950-f799-b2e8b0e576e4" }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ ":3: DeprecationWarning: scipy.diag is deprecated and will be removed in SciPy 2.0.0, use numpy.diag instead\n", " cov = [diag([3,3]), diag([4,1.7])] #macierze kowariancji dla klas\n", ":3: DeprecationWarning: scipy.diag is deprecated and will be removed in SciPy 2.0.0, use numpy.diag instead\n", " cov = [diag([3,3]), diag([4,1.7])] #macierze kowariancji dla klas\n", ":3: DeprecationWarning: scipy.diag is deprecated and will be removed in SciPy 2.0.0, use numpy.diag instead\n", " cov = [diag([3,3]), diag([4,1.7])] #macierze kowariancji dla klas\n", ":3: DeprecationWarning: scipy.diag is deprecated and will be removed in SciPy 2.0.0, use numpy.diag instead\n", " cov = [diag([3,3]), diag([4,1.7])] #macierze kowariancji dla klas\n", ":3: DeprecationWarning: scipy.diag is deprecated and will be removed in SciPy 2.0.0, use numpy.diag instead\n", " cov = [diag([3,3]), diag([4,1.7])] #macierze kowariancji dla klas\n", ":3: DeprecationWarning: scipy.diag is deprecated and will be removed in SciPy 2.0.0, use numpy.diag instead\n", " cov = [diag([3,3]), diag([4,1.7])] #macierze kowariancji dla klas\n", ":3: DeprecationWarning: scipy.diag is deprecated and will be removed in SciPy 2.0.0, use numpy.diag instead\n", " cov = [diag([3,3]), diag([4,1.7])] #macierze kowariancji dla klas\n", ":3: DeprecationWarning: scipy.diag is deprecated and will be removed in SciPy 2.0.0, use numpy.diag instead\n", " cov = [diag([3,3]), diag([4,1.7])] #macierze kowariancji dla klas\n", ":3: DeprecationWarning: scipy.diag is deprecated and will be removed in SciPy 2.0.0, use numpy.diag instead\n", " cov = [diag([3,3]), diag([4,1.7])] #macierze kowariancji dla klas\n", ":3: DeprecationWarning: scipy.diag is deprecated and will be removed in SciPy 2.0.0, use numpy.diag instead\n", " cov = [diag([3,3]), diag([4,1.7])] #macierze kowariancji dla klas\n", ":41: MatplotlibDeprecationWarning: The 'nonposx' parameter of __init__() has been renamed 'nonpositive' since Matplotlib 3.3; support for the old name will be dropped two minor releases later.\n", " ax.set_xscale(\"log\", nonposx='clip')\n" ] }, { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAABIcAAAJCCAYAAABahKemAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/Il7ecAAAACXBIWXMAAAsTAAALEwEAmpwYAACDb0lEQVR4nOzdd5hU5d2H8fvMbC/UpXeQolhAEMUKVuw99m5sMRqNyRtjoia2WGJLjDG2qLFGjd0Ye0UQFBQFpPfet+/MnPeP2V0WWGCBhQXm/lzXXHPmnOec8xt2gZ3vPiUIwxBJkiRJkiSlpkhDFyBJkiRJkqSGYzgkSZIkSZKUwgyHJEmSJEmSUpjhkCRJkiRJUgozHJIkSZIkSUphaQ1dwOoKCgrCzp07N3QZkiRJkiRJ242RI0cuDMOwRW3HtrpwqHPnzowYMaKhy5AkSZIkSdpuBEEwbW3HHFYmSZIkSZKUwgyHJEmSJEmSUpjhkCRJkiRJUgrb6uYckiRJkiRJ2pwqKiqYOXMmpaWlDV1KvcvKyqJ9+/akp6fX+RzDIUmSJEmSlFJmzpxJfn4+nTt3JgiChi6n3oRhyKJFi5g5cyZdunSp83kOK5MkSZIkSSmltLSU5s2bb1fBEEAQBDRv3nyDe0QZDkmSJEmSpJSzvQVDVTbmfRkOSZIkSZIkrccpDw3llIeGNnQZm4XhkCRJkiRJ0hYWjUbp06cPO++8MyeffDLFxcVr3T948GDeeeedVc6/9957ufTSS+ulFsMhSZIkSZKkLSw7O5tRo0YxZswYMjIy+Pvf/77W/aeddhrPPffcKuc/99xznHbaafVSi+GQJEmSJElSA9pvv/2YOHHiWvefdNJJvPnmm5SXlwMwdepUZs+ezX777Vcv93cpe0mSJEmSlLL+8Pr3/DB7+Xrb/TAn2aYu8w7t1LYRNxzdu073j8VivP322wwZMmSt+5s1a8aAAQN4++23OfbYY3nuuef4yU9+Um+TattzSJIkSZIkaQsrKSmhT58+9O/fn44dO3LBBResc3/NoWX1OaQM7DkkSZIkSZJSWF17+FT1GHr+4oH1ct+quYXquv/YY4/lqquu4uuvv6a4uJh+/frVSx1gzyFJkiRJkqStXl5eHoMHD+b888+v115DYDgkSZIkSZK0TTjttNMYPXp0vYdDDiuTJEmSJEnawgoLCzdoP8Bxxx1HGIb1XovhkCRJkiRJ0nrU11xDWyOHlUmSJEmSJKUwwyFJkiRJkqQUZjgkSZIkSZKUwgyHJEmSJEmSUpjhkCRJkiRJ0vo8fmTysR0yHGpI2/E3liRJkiRJWrtoNEqfPn3YeeedOfroo1m6dCkAU6dOJTs7mz59+lQ/nnzySSC5zP3FF19Mt27d6NevH4MGDWLYsGGbXItL2UuSJEmSJG1h2dnZjBo1CoBzzjmHBx54gOuuuw6Abt26VR+r6cILL6RLly5MmDCBSCTClClT+OGHHza5FsMhSZIkSZKkBjRw4EC+/fbbdbaZNGkSw4YN4+mnnyYSSQ4E69KlC126dNnk+xsOSZIkSZKk1PX2b2Dud+tvN7cyvKnL9DCtd4HD/1Sn28fjcd5//30uuOCC6n2TJk2iT58+1a//8pe/sGTJEvr06UM0Gq3TdTeE4ZAkSZIkSdIWVlJSQp8+fZg1axY77rgjhxxySPWx2oaVvfbaa5utFsMhSZIkSZKUuurYw6e6x9B5b9bLbavmHCouLuawww7jgQce4Iorrlhr+969ezN69Gji8Xi99x5ytTJJkiRJkqQGkpOTw/3338+f//xnYrHYWtt169aN/v37c8MNNxCGIZBc2ezNNzc9rDIckiRJkiRJakB9+/Zl11135dlnnwVWzjlU9bj//vsBeOSRR5g3bx477LADO++8M+eeey4tW7bc5Ps7rEySJEmSJGkLKywsXOX166+/Xr1dUlJS6zmNGjXi4YcfrvdaDIckSZIkSZLWp57mGtoaOaxMkiRJkiQphRkOSZIkSZIkpTDDIUmSJEmSpBRmOCRJkiRJkpTCDIdU7ZSHhnLKQ0MbugxJkiRJkrY65/33PM7773kNXcZmYTgkSZIkSZLUAF555RWCIGDcuHHV+4YPH87+++9Pz5496du3LxdeeCHFxcUAvP322/Tv35+ddtqJvn378stf/rJe6jAckiRJkiRJagDPPvss++67L88++ywA8+bN4+STT+b2229n/PjxfPPNNwwZMoQVK1YwZswYLr/8cv71r3/xww8/MGLECHbYYYd6qcNwSJIkSZIkaQsrLCzks88+49FHH+W5554D4IEHHuCcc85h4MCB1e1OOukkWrVqxR133MF1111Hr169AIhGo1x66aX1UktavVxFkiRJkiRpG3T78NsZt3jcettVtanLvEO9mvXi/wb83zrbvPrqqwwZMoQePXrQvHlzRo4cyZgxYzjnnHNqbT9mzJh6G0a2OnsOSZIkSZIkbWHPPvssp556KgCnnnpq9dCyhrDenkNBEDwGHAXMD8Nw51qOB8B9wBFAMXBuGIZfVx6LA99VNp0ehuEx9VW4trDHj0w+n/dmw9YhSZIkSVI9Wl8PnypVPYYeH/L4Jt9z8eLFfPDBB3z33XcEQUA8HicIAs455xxGjhzJscceu8Y5vXv3ZuTIkey2226bfP/V1aXn0D+BIes4fjjQvfJxEfBgjWMlYRj2qXwYDEmSJEmSpJT34osvctZZZzFt2jSmTp3KjBkz6NKlCwcffDBPPPEEw4YNq2778ssvM2/ePH71q19x66238uOPPwKQSCT4+9//Xi/1rLfnUBiGnwRB0HkdTY4FngzDMAS+DIKgSRAEbcIwnFMvFUqSJEmSJG1Hnn32Wf7v/1btsXTiiSfy3HPP8dxzz3HNNdcwf/58IpEI+++/P0OGDKFVq1bce++9nHbaaRQXFxMEAUcddVS91FMfE1K3A2bUeD2zct8cICsIghFADPhTGIav1MP9JEmSJEmStlkffvjhGvuuuOKK6u1PP/201vOOOuqoeguEatrcq5V1CsNwVhAEXYEPgiD4LgzDSas3CoLgIpJD0ujYseNmLkmSJEmSJGnD1MdcQ1ur+litbBbQocbr9pX7CMOw6nky8BHQt7YLhGH4jzAM+4dh2L9Fixb1UJIkSZIkSZLqoj7CodeAs4OkvYBlYRjOCYKgaRAEmQBBEBQA+wA/1MP9JEmSJEmSVE/qspT9s8AgoCAIgpnADUA6QBiGfwfeIrmM/USSS9mfV3nqjsBDQRAkSIZQfwrD0HBIkiRJkiRpK1KX1cpOW8/xEPhZLfu/AHbZ+NIkSZIkSZK0udXHsDJJkiRJkqTt2rSzzmbaWWc3dBmbheGQJEmSJEnSFhaNRunTp0/1Y+rUqSxatIjBgweTl5fH5ZdfvsVq2dxL2UuSJEmSJGk12dnZjBo1apV9RUVF3HTTTYwZM4YxY8ZssVrsOSRJkiRJkrQVyM3NZd999yUrK2uL3teeQ5vJKQ8NBeD5iwc2cCWSJEmSJGlt5t56K2Vjx623Xem4ZJu6zDuUuWMvWv/2t+tsU1JSQp8+fQDo0qUL//nPf9Zf7GZiOCRJkiRJkrSF1TasrKEYDkmSJEmSpJS1vh4+Vap6DHV66snNWU6DcM4hSZIkSZKkFGbPIUmSJEmSpK1E586dWb58OeXl5bzyyiv873//Y6eddtqs9zQckiRJkiRJ2sIKCwtr3T916tQtWwiGQ5IkSZIkSeu1Pc41VMU5hyRJkiRJklKY4ZAkSZIkSUo5YRg2dAmbxca8L8MhSZIkSZKUUrKysli0aNF2FxCFYciiRYvIysraoPOcc0iSJEmSJKWU9u3bM3PmTBYsWNDQpdS7rKws2rdvv0HnGA5JkiRJkqSUkp6eTpcuXRq6jK2Gw8okSZIkSZJSmOGQJEmSJElSCjMckiRJkiRJSmGGQ5IkSZIkSSnMcEiSJEmSJCmFGQ5JkiRJkiSlMMMhSZIkSZKkFGY4pE123n/P47z/ntfQZUiSJEmSpI1gOCRJkiRJkpTCDIe02Uw762ymnXV2Q5chSZIkSZLWwXBIkiRJkiQphRkOSZIkSZIkpTDDIUmSJEmSpBRmOCRJkiRJkpTCDIckSZIkSZJSmOGQJEmSJElSCjMckiRJkiRJSmGGQ5IkSZIkSSnMcEiSJEmSJCmFGQ5JkiRJkiSlMMMhSZIkSZKkFGY4JEmSJEmSlMIMhyRJkiRJklKY4ZAkSZIkSVIKMxySJEmSJElKYYZDkiRJkiRJKcxwSJIkSZIkKYUZDkmSJEmSJKUwwyFJkiRJkqQUZjgkSZIkSZKUwgyHJEmSJEmSUpjhkCRJkiRJUgozHJIkSZIkSUphhkOSJEmSJEkpzHBIkiRJkiQphRkOSZIkSZIkpTDDIUmSJEmSpBRmOCRJkiRJkpTCDIckSZIkSZJSmOGQJEmSJElSCjMckiRJkiRJSmGGQ5IkSZIkSSnMcEiSJEmSJCmFGQ5JkiRJkiSlMMMhSZIkSZKkFGY4JEmSJEmSlMIMhyRJkiRJklKY4ZAkSZIkSVIKMxySJEmSJElKYYZDkiRJkiRJKWy94VAQBI8FQTA/CIIxazkeBEFwfxAEE4Mg+DYIgt1rHDsnCIIJlY9z6rNwSZIkSZIkbbq69Bz6JzBkHccPB7pXPi4CHgQIgqAZcAOwJzAAuCEIgqabUqwkSZIkSZLq13rDoTAMPwEWr6PJscCTYdKXQJMgCNoAhwHvhmG4OAzDJcC7rDtkkiRJkiRJ0hZWH3MOtQNm1Hg9s3Lf2vavIQiCi4IgGBEEwYgFCxbUQ0mpY9pZZzPtrLMbugxJkiRJkrSN2iompA7D8B9hGPYPw7B/ixYtGrocSZIkSZKklFEf4dAsoEON1+0r961tvyRJkiRJkrYS9REOvQacXblq2V7AsjAM5wDvAIcGQdC0ciLqQyv3SZIkSZIkaSuRtr4GQRA8CwwCCoIgmElyBbJ0gDAM/w68BRwBTASKgfMqjy0OguAm4KvKS/0xDMN1TWwtSZIkSZKkLWy94VAYhqet53gI/Gwtxx4DHtu40iRJkiRJkrS5bRUTUkuSJEmSJKlhGA5JkiRJkiSlMMMhSZIkSZKkFGY4JEmSJEmSlMIMh7Zy5/33PM7773kNXYYkSZIkSdpOGQ5JkiRJkiSlMMMhSZIkSZKkFGY4JEmSJEmSlMIMhyRJkiRJklKY4ZAkSZIkSVIKMxySJEmSJElKYYZDkiRJkiRJKcxwSJK0dXj8yORDkiRJ0hZlOCRJkiRJkpTCDIckSZIkSZJSmOGQJEmSJElSCjMckiRJkiRJSmGGQ5IkSZIkSSnMcEiSJEmSJCmFGQ5JkiRJkiSlMMMhSZIkSZKkFGY4JEmSJEmSlMIMhyRJkiRJklKY4ZAkSZIkSVIKMxySJEmSJElKYYZDkrZtjx+ZfEiSJEmSNorhkCRJkiRJUgozHJIkSZIkSUphhkOSJEmSJEkpzHBIkiRJkiQphRkOSZIkSZIkpTDDIUmSJEmSpBRmOCRtKpdSlyRJkiRtwwyHJEmSJEmSUpjhkCRJkiRJUgozHJIkSZIkSUphhkOSJEmSJEkpzHBIkiRJkiQphRkOSZIkSZIkpTDDIUmSJEmSpBRmOCRtgFMeGsopDw1t6DIkSZIkSao3hkOSJEmSJEkpzHBIkiRJkiQphRkOSZIkSZIkpTDDIUmSJEmSpBRmOCRJkiRJkpTCDIckSZIkSZJSmOGQJEmSJElSCjMckiRJkiRJSmGGQ5IkSZIkSSnMcEiSJEmSJCmFGQ5JkiRJkiSlMMMhSZIkSZKkFGY4JEmSJEmSlMIMhyRJkiRJklKY4ZAkSZIkSVIKMxySJEmSJElKYYZDkiRJkiRJKcxwSJIkSZIkKYUZDkmSJEmSJKUwwyFJkiRJkqQUZjgkSZIkSZKUwgyHJG2Xpp11NtPOOruhy5AkSZKkrZ7hkCRJkiRJUgozHJIkSZIkSUphdQqHgiAYEgTB+CAIJgZB8JtajncKguD9IAi+DYLgoyAI2tc4Fg+CYFTl47X6LF6Sajrvv+dx3n/Pa+gyJG2Mx49MPiRJkrTFpa2vQRAEUeAB4BBgJvBVEASvhWH4Q41mdwFPhmH4RBAEBwK3AWdVHisJw7BP/ZYtSZIkSZKk+lCXnkMDgIlhGE4Ow7AceA44drU2OwEfVG5/WMtxbQPCMGzoEiRJkiRJ0hZWl3CoHTCjxuuZlftqGg2cULl9PJAfBEHzytdZQRCMCILgyyAIjtuUYrX5fD5xId/MWMr0xcUNXYokSZIkSdqC6mtC6muAA4Ig+AY4AJgFxCuPdQrDsD9wOnBvEATdVj85CIKLKgOkEQsWLKinklQXiUTIAx9O5KxHhxFPhMxZVsoXkxY2dFmSJEmSJGkLqUs4NAvoUON1+8p91cIwnB2G4QlhGPYFrqvct7TyeVbl82TgI6Dv6jcIw/AfYRj2D8Owf4sWLTbibWhjLC+t4OJ/jeTOd8Zz5K5t6dOhCVnpEX71729ZXlrR0OVJkiRJkqQtoC7h0FdA9yAIugRBkAGcCqyy6lgQBAVBEFRd61rgscr9TYMgyKxqA+wD1JzIWg1k3NzlHPOXz/hw3HxuOHon7j+1D+nRCF0L8pizrIQ/vu6XSZIkSZKkVLDecCgMwxhwOfAOMBZ4IQzD74Mg+GMQBMdUNhsEjA+C4EegFXBL5f4dgRFBEIwmOVH1n1Zb5UwN4JVvZnHcA59TXB7n2Yv24rx9uhAEAQD5WWlcNmgHXhw5k/99P7eBK5UkSZIkSZvbepeyBwjD8C3grdX2XV9j+0XgxVrO+wLYZRNrVD0pjyW45c0feGLoNAZ0acZfT+9Ly/ysNdpdcVB3Phg3n2tf/o7dOzWlIC+zAaqVlMqmnXU2AJ2eerKBK5EkSZK2f/U1IbW2cnOXlXLqP4byxNBpXLhvF56+cM9agyGAjLQI95zShxWlMa77z3cucS9JkiRJ0nbMcCgFfDFpIUf95VPGzV3BA6fvzu+O2on06Lq/9D1b5/PLQ3vwzvfz+M83s9bZVpIkSZIkbbsMh7ZjYRjy0MeTOPORYTTOTue1y/fhyF3b1Pn8C/fryh6dm3LDq98zuyJvM1YqSZIkSZIaiuHQdmpFaQWX/utrbnt7HEN2bs2rl+/LDi3zN+ga0UjAn0/uQzwM+dXcwSQcXSZJkiRJ0nbHcGg79OO8FRz71895d+w8fnfkjjxw+u7kZdZp7vE1dGyew++O3InPizvw1NKd67lSSZIkSZLU0AyHtjOvj57NcQ98zvLSGM9cuCcX7te1epn6jXXagA4Myp3GbQsGMmlBYT1VKkm1O++/5zFu8biGLkOSJElKGYZD24mKeII/vv4DP3/2G3Zq04g3r9iXPbs2r5drB0HAHa0/JCuIc/ULo4nFE/VyXUmSJEmS1PAMh7YDiyJZnP7wlzz2+RTO26czz160F60a1b5M/cZqmVbMTa0+YfSMpTz40aR6vbYkSZIkSWo4hkPbuNEZLbiw5aGMmbWc+07tww1H917vMvUb6+hGEzl6t7bc9/4ExsxatlnuIUlKbdPOOptpZ53d0GVIkiSlFMOhbVQYhjzy6WSuKhhEbqKCVy/fh2P7tNvs973p2N40y83g6hdGUVoR3+z3kySljvOCec43JUmS1AAMh7ZBhWUxLn/mG25+cyx7l87moQXv0qPVhi1Tv7Ga5GRwx0m78uO8Qu5+98ctck9JkiRJkrT5GA5tYybOX8FxD3zO22PmcO3hvbhp8efkhrEtWsOgni05fc+OPPzpZIZPWbxF7y1J2j59P2cZReVb9v8zSZIkJRkObUPe+m4Ox/71c5YUlfOvC/bk4gO6sWmL1G+8647YkQ5Nc/jlv0cRi6U1UBVKZac8NJRTHhra0GVIkiRJ0jbPcGgbkEgE3PLmD1z29Nf0aJ3PG1fsy947FDRoTbmZadz9k92YuaSEiZN2bdBaJEmSJEnSxrPLx1aurCyTH37Yi4+XTeHsgZ343ZE7kZG2dWR6/Ts346L9u/LQx9CiYHZDl7PVOO+/5wHw+JDHG7gSSZK2A48fmXw+782GrUOSpO3Y1pEyqFYjpi5mxMiDWb6iKfecsht/PHbnrSYYqnL1IT3IzV3GuPH9WFJU3tDlbLVcmllat1MeGsr3c5Y1dBmSJElSStq6kgYByWXqH/98Cqf+40ui0Rj9dv+Q4/u2b+iyapWZFmXHXsOpqMjkd6+MIQzDhi5JkiRJkiRtAIeVbWWKymJc+/J3vDZ6Ngfv2IrSZq+Qnl7R0GWtU37+Mrp0/p43v4tw6OhWHNunXUOXJEmSJEmS6sieQ1uRyQsKOf5vn/PGt7P51WE9+cdZ/bb6YKhKhw4/0rdjE65/9XvmLitt6HIkSZIkSVIdGQ5tJf47Zi7H/PVzFhaW8+T5e/KzwTsQiTTUQvUbLhIJufsnfSiPJfj1S986vEySJEmSpG2E4VADi4UBf3p7HJf8ayTdWuTy+s/3Zd/uDbtM/cbqUpDLb4/oxSc/LuDpYdMbuhxJkiRJklQHhkMNaGEsm7NmHs3fP57E6Xt25IVLBtKuSXZDl7VJztyrE/t1L+CWN8cyM5rX0OVIkiRJkqT1MBxqIF9PX8JR007m65LW3HnSrtx6/C5kpkUbuqwNVhYvY3bhbEpjyXmGgiDgjpN2JS0acFvTPYmz7QyNkyRJkiQpFRkObWFhGPLk0Kmc8tBQ0oM4L3d8iZP7d2josjZKIkxw3WfXMbtoNhOWTmBZ2TIA2jTO5qZjd2ZMZgHP5/Vs4ColSZIkSdK6GA5tQSXlca5+YTTXv/o9+3VvwRudXqR31qKGLmuj3TXiLt6Z+g4F2QWUx8u55uNriCViABzbpy0HlMzgsUY7M3bO8gauVJIkSZIkrY3h0BYydWERx//tc14ZNYurD+nBI2f3p3G0rKHL2mhPfP8ET/3wFGfueCad8jvRqVEnvpzzJXd+dSeQHF529dIR5CfKuer5UZTF4g1csSRJkiRJqo3h0Bbw7g/zOPqvnzF3eSn/PG8AVxzUfZtapn51b095m7tG3MWhnQ7lV3v8iiAIKMgu4OydzuaZcc/w4o8vAtAkUc41S0cwbu4K7ntvQgNXLUmSJEmSamM4tBmFYcid74zjp0+OoHPzXF6/fF8O6NGiocvaJMPnDOe3n/2Wfq36cet+txIJVn4LXd3vavZptw+3fHkLX839ikRxMXvOGM0p/drz948nMXLa4gasXJIkSZIk1cZwaDOpiCcYN3cFD3w4iVP36MC/LxlIh2Y5DV3WJvmRcq788Eo6N+rMfYPvIzOaucrxaCTKnfvfSYdGHbj5lSsp/nE85VOncuWSr2jTOJtfvjCa4vJYA1UvSZIkSZJqYzi0mUxfXMyKshi3n7gLfzpxV7LSt71l6muaQ4xLgwXkpOfw4MEP0jizca3t8jPyuX/P27n8meWUBTFo3Ihl99zNfZ0Kmba4mNveGreFK5ckSdJ27/Ejkw9J0kYxHNpMOjbLoXebRpyyR8eGLmWTLStbxqXBAopJ8ODBD9I6t/Va24aJBNGb/kq7hSF3HxfhvuPTyNxpJ/Lv/ANXdY3w1JfT+OTHBVuwekmSJEmStC6GQ5tJejRCbmZaQ5exycriZVzxwRVMp4L7whb0aNpjne0X3HMPhR98QOvrfsteFR35vM1y/ndpXyK5uRz+3J/pm5/g1y9+y7Liii30DiRJ0rbs+znL+H7OsoYuQ9L2wl5mUq0Mh7RWiTDBtZ9ey9fzv+bWsDkDyFpn+6X/eYVFDz9Ck9NOpdkZZ3DUzAKGzGzOA7OeY+p1pxFfuJBbvv4XS5cVccNrY7bQu5AkSduyPxQUMu2DxUw76+yGLkWSpO2W4ZBqFYYhd3x1B+9Oe5dr+l/DEHLX2b7d5BXMvf56cgbuRevf/haAgIBLxnegf6v+/GbBI8R+eynB99/ywJz/8so3s3jruzlb4q1IkiRJkqR1MBxSrZ74/gmeHvs0Z+10Fuf0PmedbRstKuW4x8aT3q4d7e+9lyA9vfpYWhhw96C7aZHTgiuC58m++FzaDv+Qny/4kuv+8x3zV5Ru7rciSZIkSdLGSZGhiIZDWsObk9/kzyP/zGGdD+Oa/tess228sIgTHxlPNB7S/sG/EW285ipmTbOa8pcD/0JRRRG/2eFrcoccyhFDX6b31NFc+9J3hGG4ud6KJEmSJElaD8MhraIwMpbfff479mi9B7fueyuRYO3fImE8zuxrrqH5vBJePbcHmV26rLVt96bduX3/2/lh8VgePDKNrJ135jcjn2HSsFG8MGLG5ngrkiRJkiSpDgyHVK00mMHM9Afp3Kgz9w6+l4xoxjrbz7/7bgo/+oj3TujMtJ5r9hha3aAOg7hy9yt5Y/b/+Phne5HZpDG3jXiC+/49jBmLi+vrbUiSJEmSpA1gOCQAZhfOZnrG/UTI5sGDH6RRRqNVG5StgFhZ9culL73M4kcfo+nppzNq39Z1vs/5O5/PUV2P4s/THmfO78+hSXkhv/78Mf7v2REkEg4vkyRJkiRpSzMcEsvKlnHpe5eSoJyO5VfQOrdG2FNeDG/9CuaOhjmjYNpQikeMYM6NN5K790Ba/fbaDbpXEATcuPeN7FKwC9fMe5Dg91fSc9FUBr7yDx77bHL9vjFJkiRJkrRehkMprixexhUfXMGMFTPoUH4ZWWG7lQdnfwP/OACG/wPyWkMkSvkDxzLz0ovJaN+edvfcQ5CWtsH3zIxmct/g+8jPyOdKniPnsgs5eMZIJt3/ID/OW1GP706SJEmSJK2P4VAKiyfiXPvptXwz/xtu3e9WcsOelQdi8PGd8MjBUFYIZ70CzXcg3nQXZn7RhrCskA5n9CCan7fR926R04L7B9/PotJFXL/DaNIPPYyzx7zJY396nIp4on7eoCRJkiRJWi/DoRQVhiG3f3U77057l1/t8SuGdB4CQKvYbHj8cPjwZtjxGLj0c+g2mDARMuvNxZQtTdD+/L3ImPg4PHd6ci6ijdS7oDc373MzXy/4hmdOyKVsh16c8r9HeOKJ/9XX25QkSVIKmPbMbKY9M7uhy5CkbZbhUIp6/PvHeXbcs5yz0zmctdNZEIYcWPw2dyy8DBaMhxMegZMfh5xmAMz/aDFFk0to/bvfkfuLJ+CIu2DCu/DooTQvLdroOoZ0GcJFu17Ev6e9xrjfHEosN58ef7mR0aMn1ddblSRJ0nZuHOWMo7yhy5CkbZbhUAp6Y/Ib3DPyHg7vfDhX978aChfAc6dz8bL7mJjeEy77AnY9ubr90hdfZPFXy2jarxFNTz0luXPAT+HMl2D5LG4Y8yHdly/c6Hp+1udnHNTxIG6b/HcKb7mERhUlzPr55RQXury9JEmSJEmbm+FQihk6eyi///z3DGg9gJv3vZnIj+/AgwNh4vs8kX8RNze7DRq3r25fNHw4c/7wR3I7Z9PqwOarXqzbYLjwAwrTMvj12E/hm39tVE2RIMKt+97KDk124LoFDzL/ysvoMn8qn/z0SsLQ5e21HrFSWDEXls9p6EokSVIDiYchcX9ulKSNtuFLTWmbNW7xOK766Cq6NO7CvXvfQsab18DXT0CrneHs13jrlWWrtC+fPp1ZV1xJRocOtDs6ThAJ1rxowQ7cvPMgLpswnN6v/gzmj4VD/giR6AbVlpOew18O/AunvXka9xe8zumHnUjfd17im1vvYffrrt6Ut63tVF5iOccVPg9FI4EQ7t2ZS5q25v1WXcEfDqVtSyJBTqKQM5ctplfHgKaZFfD2byCo+n+n8jmo8f9QzWO1tlv9nE1pV1sNdbnextbAFnhPW8N9qVO73mUlNG0UIyNzCcz7HtKyajwyk89Rf6SVJGlT+D9piphVOItL37uU/Ix8HtzpEvIfPxwWT4F9roTB1yV/uGJodfv4ihXMuPQyCEM6PPg3oh9cttZrF6dlcHevvXk0bAlD/woLJ8CJj2xwjW3z2nLv4Hs5/53z+eDw1lRM3IMBTz3MvN160eqoIzbmbWt7VFECw/7O/fPvJDsshtwWkNcauu7PLsMfYs9FM+Gh/cjLj1JU2H7915PUcBZPgdHPwehn6BKbTvsY0DxCNCiEUU8n21SHveGq29XHwo1vp23CDQC7AnwCD+5de6MgumpYlL5aeFTrc/Y6jldup6+rTY22NQM2SZK2QYZDKWBp6VIuefcSyuJlPNx4D1o9fQo0agfnvgGd912jfRiLMevqX1I+bRodH32UjE6d1nuPRBCBw++EFj3hrV/Do4eSltaOWCx3g2rt27Iv1+91Pdd/cT3Nfn48jf44n27X/pZGXTqR3bv3Bl1L25l4DEY/Ax/eBitmMy5zAM/mn8dduZUfIA+7hati0xm4cAbnFpXRvOV3NGn+Pfzv97DHhdB0/d/HkraAshXww6sw6hmY9jkQQNdBzFwR8psWEc56PYtezXrR6aknt1xNYT2FTRvcjrW029w1sBnfU23tqGO72q93w+sXcMRHFXQp2IGWV1wGsbLkkOI1nms+VjtWUQIlS5KvK0rWPL6pYWF0PQHS5gyrImmGU5KkTWY4tJ0rjZXy8w9+zuzCWfyjNIcdxj8Eu50Gh98OWY1rPWf+nXdS9OmntP7jH8jdc8CG3XCPC6H5DvDCObRuP4kFczfwfOD47sczYekEnvrhKSouPZtmd7/MxJ9eQq9XXiK9ZcsNvp62cWEI49+G9/8AC8ZBu/5w4sPc8c6aQxfLo2l83KoL5x72GHMvOpL8xpPJHfoAfPEX6Hk4DLgIug7yh2hpS0skYOqnyUBo7GtQUZz8v+Kg62HXU6Bxe5bdui/lkcKGqS+oZXiVthpjM7PZc1mU1jmtofdx9X+DMIR4RY2wqJbwaPXAqbaAaa1hVdnKYKq26yQqNq3+IFJLkLSeIKrWHlEbElbVOCfiFKba9kx7ZjZ8cPaW/UWEtlnTnpkNQKfzGriQzcxwaDsWT8T5zae/YfSCUdy1cBn9YhE4+Yl1/mC1+3cfs/i9J2l2ztk0/clPNu7GXQfBTz8g8ef9adX2c/j6Kdj9rA26xNX9rmby0sl8NPcZFh17Cle9+CxTL7mMbs/8i0hW1sbVpW3P9GHw3g0wfWjyg+RPnoQdj6n8ADd07ecFAWWlBZSVFpD7+1thxGMw8p8w/i0o6JEMiXY7FTLzt9Q7kVLTokkw+tnk0LFlMyCzMez6E+hzBrTfwzBGW4cggLSM5KMhJOLrCZrWFUTVMawqXbb2kGuTe01lrD9A2qiwavXQqpY20XT/HdEGOy+Yx6mU06uhC9HmU/Xvarx85SNWlvxFQLzyuebxNdpWbSfbxjsuYP7ibLb3cQiGQ9upMAy57fPf8/709/nNosUc2nJPOPYBaNRmred0njGOIz/4F7n77UfLX/1q0wpo3o05s/anRauvyH7t8mSPjw2YqDotksYdB9zBGW+ewaQeb3L3Xsdy7WcvMvu319Huz3cR+IPA9m3BeHj/jzDuDchrBUfdA33PSv4QuKEat0/2Ttj/1/D9f2D4Q/DWNfDeH6DPacmgqKB7/b8HKVWVLoPvX0n2EprxZbJXQ9fBcPCN0OvI5IdASStFopCRk3xsaWEIidh6QqYNCauqhvGVrhZMza+9V1W8fBPfQFAdFh24x3ISYQD37578eSGaDpH0ZHgVTU8Ov6vaXuVY2qrb0YzK1zW3qx4ZlddJX/PYOtvVuHck3d5W2j5UBzC1hC2rByyx8nUENeWrHq9TqFNzu5brhPF6faud20Fx2fYfnWz/7zBFPfrBNTw/83+ct7yIM/a9MTncax2BSvm0afzkjQdY3KQlO979Z4K0Tf/WCBPpzJ+zF51Oa1s5UfWPcOKjkNWoTuc3ymjEXw/6K6e9eRozDvyax5ccyvlvvcXCHbrR4rK1T5Ctbdjy2fDRbfDNvyA9Fwb/DgZeBhkbNndVrdKzkmFQn9Ng5ggY/g8Y8Xjyuetg2PNi6H7oBq+0J4nkD4hTPq4cNvZ68kNfQY9kILTrKdCobUNXKKk2QbAy0GiI3rSJROUHx/UETHUIqxa89i+CICSnf5/KD4sVySF7VdtVYVQ8lnxOVKy2XfUoZ7NPWB9E1xIc1RZg1RJmrSvAqmuYtc4Aq2Ydqx/z56Qtqup7tM4By9qCmvUFLHUNdWocCxP1+16rvsfTMiq/3zJrbFc+0jKTnyWjmcnvx7TMVY9F0yuPZdR+7sa0jWbw7mG7ALBz/b7jrY7h0PamdDmvv3Ye95X8yBHxDH5xyr+h5bo7TcaXL0+uTEbAM8deyf759fnDQQSOuKNyoupfwaOHwGnPQbMudTq7U6NO3HXAXVz23mV8dkRjuhX154D7/0Jm1640GjKkHutUgypZCp/fB18+mPwN5oCLYf9rILdg89yvff/k49CbYeQTMOJRePZUaNIpGaT2PRNymm2ee0vbk4UTkoHQ6OdgxezkXHZ9zkg+2u3ucA9J6xaJQCS7XnoUfnf3mwC0f/CxTb4WifjKoCgRq7FdI0RaPVBKVH2Ir1h1e41rrK1d1TUrA6ya2+XFaw+zVq+jnntMrCmoW4BVa5i1gQFWre3WF6RlrP1Y5eTtReUx4jUXIkjENzxgWW3YUe0BSy29YuoU6tQ4tjkCmFqDkdXClvTGdQtqqq9TW1BTs22N61S3rbFtj7qtguHQ9mTaF3zx+sVcnxeyZ1Yrbj7hdSLr6aJcvTLZ9Ok8f/wvWdJkM034vMcFlRNVnw0PHwinPFXrSmm12bvt3vxqj1/xp+F/4u+HDKLrmzsQ/OZa0tu1J3uX7T2/3c7FymD4w/DpXcnJOnc5GQZfV+fwcJPltYQDfgX7/iI5hG34w/Du7+HDW2HXk5NDzlrvsmVqkbYVJUvh+5eTodDMr5LDxnY4GIbcCj0OT/bSk6RtWSSafGyL/54lEisDo1oDrNpCqroGXRW1BFOrh1mrtYuVQ3lR3Xptberk7HURSefJMEZ0L4hGXoc/NKXee4pFVw9RMqi1h0p6k1UDlloDlLUELLW2rS3wWe2e/tJG62A4tF1IwHs3Mnb4A1zVtjVd8ztwz9EvkF6Hsevzbr+Dos8+o83NNzFtUbvNW2bXA+CnH8Azp8CTx8KRd0O/c+p06um9TmfCkgm8xEtcu89xPPr+Mmb+7Gd0/vcLpLdqtXnrVv1LJOC7F+CDW2DZ9OSwrkP+AG12a5h6ounQ+/jkY+53yZDo2xfg6yeh494w4Kew49Fs1JxH0vYgEYdJH8Kop2Hcm8nfaLbYEQ65KTnBdH7r+rtXGBKuvtS7JKluIhGIZCbDgm1N1RxY6wuzVgmpNjzMevebR9n5xzjNctrQ+Njjaw9Y6jTsqLa2TpK+PYqnyM8lhkPbuPT05RS0+pqZX77FZR270Di7GQ8e/gT5GesfGrbkuedY8tRTNDv3XJqcdBI8tI7Vn+pL825w4Xvw4nnw+hXJiaoPvXm945eDIOC6Pa9j6vKpfB2+yf8Vnci9H77MzJ9dTqenniSS7QSn24QwhInvw3s3wrzvoPWucMx90O3Ahq5spda7wDH3J+dKGfV0Mih68TzIbwP9z4d+5yZ7HEmpYP44GP1MMixdMQeymyZD/d1Og7Z96+0H4ERxMUwv5+ivK2g9vZySySMYu9tuBBkZRLKyiGRmEWRmEsnMJKh8rLKdlUmQUYftzOTr6u1VrpNFJDMD0tJc9ECSGkLNObDYfBO0PzX5Fc6dUk7vNr1pfODvNtt9pG2N4dC2KpGAYX+nTfuPWRykc2n3PpQnynj0kIdombP+D65FX37J3JtuJveA/Wn5q2u2QME1ZDeB0/8N/7sOvvxbcqLqkx5LzlWxDunRdO4ZdA8/ef1UZvd+i8cip3D+m/9k9rW/pd3dfyZwnOrWbdZIePcGmPppcm6fEx+F3idsveOLc5rB3j+HvS6DCe8mVzn78Bb4+I5kD6M9L4Z2/fztkLY/xYthzEvJJehnjUxOnNr9UDj8dugxpN5+G10xZw6L33+Xef97g8g3PxCtiLNTJozoHjC7WUB6rIKMWAXp8SIyKiAzHiErFiGzNCAzHpARC8iIQXoM0mIhaRUJ0mIJ0so3cb6NSIQgK4tIRkYyNMrKJFIZMK1zuzJ4qnW7tkArM5MgKysZgFXtjzrRqyRJahiGQ9uiZbPglUthyscsKW3Fpe2bMLt8GQ8f+jBdm3Rd7+nlU6cy88pfkNGlM+3+/OeG+WE0mpb8oNGiV3JZ8UcOgdOfg2brrr9pVlMeOPgvnPbGmby5w4fsfewZ9HrlXyzs1o0WP798CxWvDbJoEnxwU3IZ+ZzmcPgd0O+8ZPfcbUEkCj2HJB8LJyR7Eo16Jjksrm3f5OTZvY/fNuclkKrEYzDp/eT39vi3kl3yW/aGQ29JDhurh95yYSJB4ehRTHn735R9/Bl50xYCsKgJfN0nytxO6XzTMWTIiAw6N2pL3sXnUxovpTSWfCyp2o6XUhIrqd5fFi9Lvq46XlFCrLyUeGkp6fGQjApIjydDpIwKyFh9X+WjajsrHiE7ESc7Xk5mPE5WvGRlGLUc0isgPRZWB1LRijjRigSR8hiR2CYGU+npyVAqK4sgMyMZPm3IdmUPqPVu1wylKrftLaWtTZhIQCKRfI7HCeOJ5BDXmvuq2sQTpMWSwz7KZ84iiEYgGiWofFDbs9/zkrQKw6FtzXcvwptXQzxG/Kh7+L8PH2Vco+X8ef+72b3V7us9Pb5sGTMuuZQgEqHDgw8SzcvbAkWvQ//zkkPNqiaq/slT0GW/dZ7So2kP7jrgdq748Ep+23YsLx52JAsfeIDMbl1pdMQRW6hwrVfhfPj4dhj5z+TY7P1/neyJk9WooSvbeAXdk6vvHfT75OpMw/8Br1yS7AXX79zksLPG7Ru6Sqnu5v2wcthY4bxkgNv/fOhzenLY5yZ+eKpYsYIf33uRhe++Te5X48hdUUE0gBntA2Yc2Zqs/feld79DubRVP6bceSjTKGSXWVF6lTSlU/fjN+neYRhSniinNLYyTFojSKo8VhYvS27HSyr3lbK0qm1lILV625UhVYyQEAgIwihptQRO6THIiENGRUh6LBlM5cSj5CTSyU6kkZOIkhWPkp2IkhmLkBkPyIwFZMRjpMcKSY+FpK8ISVtSI4yqiCcDqfIKKI8RxDctmAqqezVlVA/j2+TtrKzkML7MysCr5nZmJpGMDEhP3yY+pIdhWDnR72rBxCoBRXzNNvEEhCuPrdKmMuzY4DbxBGEiDolw5bHq58pjG9smHicMEyuP1WxT8z2v1qbWP5f1tKl+v7XUQnzDV2jqUPk86eCD6/Y1jQQQiVYGSckwKfk6CtEIQTSt8rkqVEqrDpuCaJQgLY0gUvkcjUA0LdmLPa3qOVp5vPK6aTWvX9t9alynKtyKRNd+nbRosqdjWlryubbza9wruV17GwM0SWA4tO0oWZrsYfPdv6H9HoTH/Z3bJj7Hly2Xc/G49hxy7iHrvUQYizHrqqsonzWLTo8/RkaHDus9Z4vosj9c+H5yKfGnjoMj/5z8oL0OgzsO5uJdfsZD3/2VS3ZoxxMLdmf2tb8lvUMHsndxdakGVbYCvvgrfPEXiJUmv5YH/B/kb0cTh2fmJyep3uNCmPxRsjfRZ/fAZ/dCryOTq5x13tchZ9o6FS9O/qJh1NMwZ1Ryad/uhyUDoe6HblKvvjAMmTT+Sya/+QLh5yNoM34h6XHIzYSJPfMpHziA9gcfw0E9DqBx5rqHEm+qIAjIjGaSGc3crPcKw5CKRMU6g6Sq0Kn6EV/1eeE6wquabRNrXdI4IJKIrtETKj2WDKHywgxywgxyE2mVYVQa2fEoWYkoWfEIWfFIsmdUvKp3VdVQvThpFYVEVywjsrgyjKqIEZRVQEUFlJUnH5syUWcksu75pDIzOW1uBS0WhZQVTWTmz6+o0aOkKnSIE643vKhLm9oCjuQ5JOp5OektLIxEKsOQILkdJMOR6kdQtQ2JIKg+ngggEYFEJCAkJBEJSASVz4TJYwHEg5B4AGHlczw9QTyARBBWHguJU2M7SCSvXfWIrHwOg2QNiSCo8bpGm7Wek1xzKhJCNFHjuXI7kqi5P6zcHyeSiK/avrLtqu1r7I+t+jpaeTwaBpX7gzXOW/W6YfX1I4mwcn9IsBXOd5v8nqkKzqpCtJWvq0KkZKC0MiBLBlbJACtYJVhLI5JW9TotebwBArTOMxJkloXEV6ygaPjw5DdOGAJhclGERGKNfdWLJYRh5eJmycB4lX21tq/cHyZq2VfjnqtcJ1HLvtraU/lvU233q+XaiZX713ntqnbh6u1Zc3/NP6t1XTsMk//m1naNurYNw+QvY9Zon6i9vur2lfsTiTX3radtVX3t5oYsb+A+FVuC4dA2YMdl8+HBvWHF3OQy3/tezSPfP87z45/npKktOXpmizpdZ95tf6Loi6G0ueVmcvr338xVb6DqiarPh9evTE6CeujNyeFna/Gzvhfx9ZxxfMVb/PXwn3HZvHnMvKxyBbPW9bhyjuomVg5fP5HsLVS0AHY6Fg68Hgp2aOjKNp8ggG6Dk48l02DEo8kVzsa+Bi13SgZIu54CGbkNXalSXbwCJr6XDITG/ze5YkvrXWDIn2CXkyG3YKMvPWPZdL775GWWf/gBzUdOof28GB2ABc3TmHRQDxoNPpBdDvwJ/Ru3qb/3sxUJgoCMaAYZ0YwtGkLV2gsqXkJZbPWeTau2XR4rZf4avaBKVgm31h5C1VZYhGgCcuJp5IeZ5JFJXiKd3DCD3DCjOojKSaSRFU+GUZmJCFmxgIx4sEqQlR5LEC1Pzh0VKY8TrSgnWFFCo8Jkj6tYaTHFkydWBxwENcKNmoFHtCrwgDCSlgw7qkKPVQKFyqBjjfAhXBlsRCBOojIASe6LBQkSVB1Lvo6TIB6ExIKwcjtBrOpYkKCCZJsYCWJBggriyW0SxIhTUTMsqSX0qNpXa1Cy2vbqbdb8RcWqX99oEE0+IhHSgjTSImlEI8l9aZHK10GUaCS68njV60gaacFq7StfV7Vb5bnyGtFIlMxg1dc1263znjVqrHns8wuOAWDfx14nHsZJhAniYZx4YuV2zX3rahNLxGo9JxEmqKhqE1a2qdxf2754uOZ9ar6uahsP4yTiccJEjEQsXtnrrPI5ESeMxQnjMYLKMLOqBxnxleEn8QRBIgGx5HMYj9celNUWXK0zQEsQSSQ2LEArX+11GFbfJ1rz3BCiiWAdQdrK86tqrA/nVT6XLRjH9LPPqZ+LbmPCyjCVAEKC5HPlvxcr96/2uuqfksp9a7QPVl4bglVeh0FQ45q13LvG/ur21fdY/bzKewJhZOU5q997jXsFNdsGhEFY+/0jEEZX/XOq6AhlmQHrHt+y7TMc2ppVlHLKtG8ZMmciNN8BLnwX2vXj1Ymvcv8393Nk1yM5+71ZdbrU4meeYcnTT9Ps/PNpcuKJm7nwjZTVGE57Ht79/cqJqk9+fK0TVQdBwN8Ou41DnpvM/yoeZvDV19Hj97cw87Kf0elfTxHJ2XyrHKiGRAJ++A+8fxMsmQKd9oXTnoP2W1kAubk17QSH/BEGXZvslTH8IXjjKnj3Ruh7JuxxQTIElbakuWNWzpFVtAByCpI92/qclgyHNsKC4gV8NeVTZn3wJulDR9NzXBFdiiEegfk7NGfeEXvQ9YifsOMuA+v5zaS2LRlCxRKxVXo71T6crvZeUFUhVFmsjBXxEhZU95gqXqVtSaxkA0KoqrkR48CMzfTOV4oEkZWBRV2CjtVerxGuVF4jvbbgYx3hyupBybruufq16lp/NIgSCbbShSE20I/Lk8/dmvh/bZVaQ6i6BFUb0qYqAEusGYYlEskwbfVrrCt8W2ubRJx4PFYZjsUgFieeSD5XBWmJeGzlcM54LNkzsDJgC+Irh2UuL1pIdilEo1EieflQIyAIIyvDA6rC1RqhBJXPybBj7cFJENQIRyKVCUew8pqrBxdVxxJBQFAZblQ/V/Xmq1FnEAlq1Fl7++TrlSFLGFC9gE/AytC4artqGGHNY1Wb1W1qnldb+3UcW33fRt9nldNWu2ZQy/uqw3m11lfjWnM/+R/9Jm3bPUjrwnBoazX3O3j5IobMn8j7rbpy0AWfQkYOn8/6nBu/uJE92+zJTXvfxOx/XLDeSxV98QXzbrmVvEGDaPnLq7dA8ZsgmgZDboMWPeHNX8IjByeDhrV8qM5Ky+JfRz3I0S//hOvm3cerN11P6TXXMvs319Lu3ntcwWxzm/wxvHcDzP4m2VPm9H9D90NSezhVejbsflYyEJoxLDkv0fCHkoFn90OSH8y7HbT1rtKmbV/RwuQQ5FFPJ/8viaQnJ1TvcwbscHDlEsF1t6xsGSPmjuDbMe9T/vHndPxuAb2nh3SJQ1l2GsX9ehM95HC6H3oCOzdtupnelLaUIAhIj6aTHk2nUcbmmyOuZghV1dtpleF3NYKkv3z4O/qPitMuvyMtzz2v9nBlLWFLeiS9TsFLzQBlewlLpEgQWfn97GKI1X7yjz6c+EI5vdv0odNTTzZ0OdoGvHXzTg1dwhZhOLS1ScRh6F/hg5shuyl399yb75q25qCMHL5f9D1XfXQV3Zp0495B95Jehx/wy6ZMYeYvriKza1fa3nXntrNMbr9zoVk3eOEseOQg+MmTybmJatGpSVt+v8cd/GHEzzh38VP8+5pfsujOu1j417/S4oortmzdqWLud8ll6Se9D43aw3EPJodPRbaR768tIQig417Jx/I5yYm5Rz4OT5+UXJVvj59C3zPW2jNO2iCxcpjwv2QvoQnvQCIGbfrA4XfCzidCbvM6X6q4opiv53/N8FlfMmv4RzQfOYXdJyYYsiB5vKRtU9JO3pf2Q44nr19/gvQNC5tqWlFawZSFRUxeUMRXZfsya0kZHzSvYGpWM7r+MI/8rDQaZaXTKDuN/Kx08jPTiERSOHzejtQMoVjPNFf/ev969hsXo3ebFnTq+ZMtU6AkSSnGcGhrsnQ6/OcSmPY57Hg0HHUf3316DQAzVszgsvcuo2lmU/528N/Iy1j/jFjxZcuYecmlBGlptN8aVibbUF32g59+AM+cCk8dD0fclVzdrBYn77I3H0y6nM+W3cvvWnfhxhNPZOHfHiSjS1caH33UFi58O7ZkGnx4S3Jlo6zGyXmh9vipy7ivT6M2MPha2O+XyfmIhj0E71ybDIF3OyX5Z9gqNX4joXoUhjD328phY/+G4kWQ2xL2uhR2O73O31Pl8XJGLxjN8LnDGTVlKMFX39J3Qoy9J4U0Lk52h0/suiPNzz2CxgceRGaXLhtUZkU8wfTFxUxZUMTkhYVMWVjEpAVFTFlYxIIVZTVa7gdl8HzbypdPjljjWkEAeZnJwCg/K41G2enJ8Kh6OxkiNcquarNyu1F28pz0qL1CJEmSVlencCgIgiHAfSQ7JD4ShuGfVjveCXgMaAEsBs4Mw3Bm5bFzgN9VNr05DMMn6qn27UcYJpfFfvvXye3jHoTdTqsemlORqODS9y4lHsZ58JAHaZnTcv2XrKhg5i9+QcXs2XT85+NktG+3ud/F5tGsa3KupRcvgDd+AQvGwaG31DpR9b1HncvgR8YxdOF/efP4qzh0Wn/mXHcdGR07kL3bblu+9u1J0SL49C746hEIIrDPlbDvVZDdpKErW6swFtv6hrelZcAuJyUfs7+B4Y/AN0/DiMeg837JIWc9j1jnROwShfOTAe3oZ2HeGIhmJL9v+pyeHLK4nu+fWCLG2EVjGTZ3GMPnDGfGjyPZeXwp/SeGXDkdovGQRH4O+YMPoPGBB5G3775EG6+7h1sYhixYUcbkyl5AkxckQ6DJC4uYvriYeGLlLKLNcjPoWpDLoB4t6Noijy4FuXRrkUvRI0dxc8EKTnsznzbNdiL/5j+xvLSCFaUVLC+Jsby0guWlMZaXVCS3S2KsKK1g1tISxpYk260oi6134azs9Gh1T6RGNQKm/Fq3k+0a1wiYMtMiLu8sbQaJREhxRZzishiFZTGKy+MUVT2Xxyguq3yuub/G8Zk9LyUE/vn3oVROB0MkCKqfYdXXyWleql6vPBZUHltX24CgcnR4sMq5a95n5XlBZdvVr1mXtmvUHlSdu2rbWmuvdf9aaq9R28pza95j1XuttW2k7nXVfA+133/Vr0PNe9XWVtLGW+8nkCAIosADwCHATOCrIAheC8PwhxrN7gKeDMPwiSAIDgRuA84KgqAZcAPQn+TcWSMrz11S329kmxWvgH+fAz+8Ch0HwvF/h6adVx4O40xcOpFYIsbDhz5M18Zd63TZubfeSvHQL2lz223k9Ou3mYrfQrIaw+nPw/9+D18+AAsnwEmPrRFMZKZF+cfR1/GTV2Zw//f30vu3d9DyirnM+NnldPn3C6S32T5Xytmsyovgywfh8/ugvDA5Z8mga6Hx1hs2tpxZxOzfXEvJqFEQhkw54URy9uhPdv/+5PTrR1qzZg1dYlLbvnDcA8lJrL95Er56NDmMslF72ON82P3cDRoKpO1crAx+/C+MejY5fCyMQ7t+yR6VO58IOWv/vk6ECSYuncjwOcMZNncYI2d/RZtphfSbmOD0yem0mlcOQFrnTjQ65yDyBw8iu2/f5NK/qykqi1WHPtUBUGUvoMKyWHW7zLQIXQpy2bFNPkfs0pquBXl0aZFL14JcmuTUPobo+yBOJEiQFy+hXbyITu03fMhlIhFSWB5jRVWIVJIMlJIBU83tqrCpgsVF5UxdWFQdPMUS606X0qPBKj2Rqoe9Za7ZS2mV7cqwKTfDoXHa9sXiCYrK4xSXxygqW+25PBnwFFWGN6sEO7UEPFXnllTE63z/tEhAbmYauRlRciqfgzBBlJBoJCARhiQSlSvNhckAO1G9SnVYuar3ymdY9XX1KuGr7wvD1fYDhKvcI1F5cmK1tsndq7ZV/Vs9NKJmoFd5rLjid4zpE5IZRMi49T3SIhHSogFpkWDV7WiEaCQgPRoQjURIjwSVr5P7a7arOjfZtsa+Na67cn/Na61+j1XPj9Ryr5X7q64TMSDTJqrLr6cHABPDMJwMEATBc8CxQM1waCegaqbjD4FXKrcPA94Nw3Bx5bnvAkOAZze58u1ByRJYNAFmh3DwjbD3FavM2RJLxJiybApFFUXcO+he+rbsW6fLLn76aZY++xzNL7yAJscft3lq39IiURhyK7TslVwB6pGDk4HRahNV79yuCRfveB0PTbiaK0feyHN33k7402uYcdnP6Pz0v1zBrK7iMRj1L/jwNiicm+yVcND10HLHhq6sVmEiZMk7b3PSfaPpOqWEJVnjibVpRg4ZRPLzWfLc8yx+IjnhYEa3buT060fOHsmwKL1t2/VcfTPLbZ7shTXw58kP/8Mfgvf/CB/dnvzQv+dFySBJqScMYc6olcPGSpZAXmvY+/LksLGWvdZyWsiMFTOqewYNnzuckmWL2HVKyP5Tczh/QgVZhXFIi5LTry955w0if9AgMjp3BpIf+qYtKWHywkXJXkALi6qHhM1bvnIYWBBA28bZdG2Ry4m7t6vuBdS1RS5tG2c3SAASiVQGN1nptGuSvcHnh2FIaUWiusfSsqoQqaQiGTjV6LFUFSatKK1g7vLS6jbr+4AbBJCfuf4eS41qGRJX1SbNoXHaAOWxxCqhzcb0yikui1eel7xOeazuq/ZkpEWSIU5GGrmZyee8zDSa52aQm5lGTkZ05XNGGjmZlc8191c+52WmkZORRkbamn8H3jr4IgCO+Nsv6uuPbrMLVwui1hkk1Ro4hTXOrXG99bRdNfBaR9vEyv1UhmE161o9ZFs1fFuzbUgytFt5rzXbEq76vtcI9FarM7Hae6itLeGqgd6b3z1NrwkJGue2JnvP7sQSIbFEIvkcTxBPhFTEw8rnBGUVCSoSceKJBLF4WN2uqs3Kcyu3K9s0hPQ1AqiVgVV1qFUzdKotgIpEiEaDyqCqlvNrBl61BWGrB1ur7VslFIvUDMdWv0dkjXs11C9XyoIN/5liW1SXcKgdq64bOhPYc7U2o4ETSA49Ox7ID4Kg+VrOXaPLQRAEFwEXAXTs2LGutW/7ShZDJA3O/y+0WXPY05yiORRWFNIxvyMHdTqoTpcs/Oxz5t16G3kHHkiLq66q74ob3u5nJ4eaPX8WPHxgcqLqrges0uTyQTvz/o8/Z0rFrVw++Q7+efsfWfzza5j9f/9Hu/vucwWzdQlDGPcGvPeHZHDZfgCc/E/otHUuSZ1ZVsyPoxezYuRycpZdTV5jePLACB/sFqc4axk5sQi92nZgx4tOYrfF+XSaXEjaD9NY/vbbLH3hBQDS27Ylu38/cvr3J6d/fzK6dGmY37pE02DHo5KP+eOSq5yNfg5GPwPt94ABF8NOxyaHpmn7tmJuctjYqGdgwViIZkKvI5M997oOqnXY2LyieQyfO5xhc4YxfO5w5hTNoeWSkAOm53LDlExaT4AgliDSOEre4MHkDRpE2e57MLU0ypcLC5n8QxGTPxnB5AWFTF9cTEV85Q+1jbPT6doil313aEHXyt4/XVrk0rl5Llnp29ck9EEQkJ0RJTsjSqtGGzeXWnksUR0e1eyltOr2qsPjZiwurt63okYPrLXJyYiuEh7VHjCtFjbVCJi2t6/b9iIMQ8piiVVCmfX1yllfz53i8tgqf5/XJzs9Wh3gVIUyeZlptMrPWhnarBbeVPfgqRH+1LyG83ytXdWQKYBoLcuBa/MYM+dNTvmgnN5t+tPppCs3yz3CsCo4Sj7i8ZCKRKI6cFo9gKoOmeIrz4nF1wyd4onkdWqGVFVt4okEFYmwlnvUvG7t9yguj61aU2WtqwdesfjKe8QbIACLBKzRYyoZMgWVgVakuudVdeAUWUtPr5q9sCrDqFV6cdW4x9hW+7LTsvFb/P1uafU1scU1wF+DIDgX+ASYBdS5b2gYhv8A/gHQv3//1Olk2aRz8leItQRDAB3yO7Bz851Ji9Tty1Q2eTKzrrqKzG7daHvHHdvOymQbqvO+yYmqnz0V/nUCHH4H7HFB9eFoJOD+kw7i8IdmMivyD67Pf4Wbf30NC/90Bwvuu5+WV/2i4Wrfmk0bCu9eDzOHQ0EPOOXp5AfSrbB76rKJ4zjwo7vY4/uxxMtDprWHWaf05ePOJeRk5/PoHr/m8zt/zZT8Ema1SfDy1Nd4OlYCTSBtnzS6Ht6ZvYrasMvsKG0nLiX++ecsf+11AKLNmlX3LMru14+sXr22/N+llr3gqLvh4BuSAcHwh+HlC+Gd3yZX8ut/fnKSa20/Kkrhx7eTX++J7yeHjbXfA466B3ofD9mrLhG/pHQJX839qjoQmrp8KkEiZPf5uZw3u4BePzQla8YCYDlhx84sPWIwk7r15Zv8DkxaUsrkrwpZ8enw6utlRCN0Lshhh5Z5HNq7dfU8QF0K8miWayC5ITLSIjTPy6R5XuZGnR9PhBSWxVbrrbTusGlhYTmTFxZVt1vfD+wZaZHq3kn5tfRUWlvYVDVXU25GNOWHLoRhSElFfI1gZp29cmrrnVO1v3L41YZ81qo5pKoqkGmSk0G7psnwZkN75eRkpBF1yKNUL4KgMnDYTj+OQbJnWTysJUCqsR1PJNYIwVYJoOKr97yqGXglVoZrq52/8ljtQVt1KFYj4CqNrXqPmjWsHqpVnVP9b3L7Q2lSsbxB/7y3hLqkDrOADjVet6/cVy0Mw9kkew4RBEEecGIYhkuDIJgFDFrt3I82od7tSx2W/a5rMBRfupQZl15KkJFBhwf/RjQvd1Or27o16wIXvAsvXQBvXp2cqPqw26p/o965IJffHnQkf/xwPl8EL/HYjl04++STWfTQQ2R260rjY45p4DewFZk/NtlT6Me3k0NWjr4P+py51U2KHIYhY//3PHMef4TWo2axdwSG7phN0/6ZHNS6Mc3Oe4av/5tczW7ngp3Jn10AQKffPUkiTDBjxQzGLR7H+MXjGbd4HP8tH8+TLedDS2BgyM4lLdl3YXN6TU/Q8ruvWfHuuwBE8vLI3r0vOf36k7NHf7J23plIxhb6sJzVOLny1ICLYdIHyd5En9wJn92dXNFwwMXQca+tMsBTHYQhzPoaRj0NY16C0qWQ3zY54Xuf06Gge3XTwvJCvp7/dXXPoHGLxwHQPJbNMYs6stv4HrQcPZ20wuUkIkVMa9+TL/fYn3cbd2dOXvLvAlOgbeOldGmRy3F92lUPAevWIo+2TbL9ULiViEYCGmen0zg7faPOrwotaguRapvQu2rf7KUl1dtl6xkyFAlYdV6ltYRIq/dYaly5nZ+VvkW/3+KVvxVf15Cp1UObwvX0yimuiK938vMqkYDK3jWrhjMt8jPplJGz6v469srJSos6d5WkBhWJBEQISHZG3T5TsERlOPXykXuTlqj7vGjbqrp8+vsK6B4EQReSodCpwOk1GwRBUAAsDsMwAVxLcuUygHeAW4MgqPqV56GVx1WPwkSCmVf+gtjsOXR84gnS2229kwXXq6xGcNpzyd4uQ/+anKj65Merf8N+5p4defeHIxi5dD7/Gvc03U/7Pf2mTWPOdb8jvUMHcvqm+Dwuy2bBR7cmeypk5CXnFNrzUsjYuuZlWr58AV8+cRcZL71Dq7ll5OXAmKN68WrbQ4jl9uGFjFvWe41IEKFTo050atSJwzofVr1/celixi8enwyMlozjvcXjebjTVOL7xmm2PEqfORkMmJtNt0nf0viTTwEIMjPJ3nXXlUPR+vQhkruZw9hIBLofnHwsnpycvPqbp+D7/0DrXZKrnO180lb3tdNaLJ8N3z6fnFx64XhIy0qGfbudlhw2FolSFi9j1Jxh1WHQmIVjiIdx0oIMdi3pyoXjd2OHHxbTafo0oonvWZaRw0etdmT4jjsyvkNv2rRrQdeCXE6vHALWtSCPzgU55GRsXaGv6l8QBJXDedJo3XjjhsaVxeI1hr7VbXjctEXF1WFTYR2GxuVlplWvBre+HktLi3fgu8Yhc7I7kjN8ep175dTXRMc5GWm0bZJeGcysdmwt4U3NgMcV7iRp2xSJBGREAjITFQ1dyhax3p8SwzCMBUFwOcmgJwo8Fobh90EQ/BEYEYbhayR7B90WBEFIcljZzyrPXRwEwU0kAyaAP1ZNTq36EYYhFdOnE1uwgLa3/4mc3VMs8IhE4bBboEVPeOPqyomqX4Dm3QiCgDtO3JVD7jmKnLxF3DTyNh753T00vnwuMy//OV1eeD51grSaSpbAZ/fAsIcgTCQDof2vWedqR1taGIaMGvshEx67n04fjKdDMcxtk8X0nx/DXmdfw575LXj5oaGbPDq/WVYzBrYdyMC2K+dUKo2VMmnpJMYtHse4xeP435IfGbd4HNHlUXrNDNlpRozdZn9H24dGEHnw74SRCOk79qTRgL3I6d+P7N13J61p03XcdVOL7pr8nh/82+S8NMP/Aa/9PBmS9j0rOcSyxoqH2kpUlMC4N5PLz0/6IPl3r8NeyZ56vY+nIiOH7xd+zxejH+bTmV8ybsm3xMIKAiJkV3Six8Td2HV8GXvMnEPHFWMAmNOkDSP3PILSfgNpskdf9mzZmNNa5NI8N8MPotokmWlRMvOiFGzK0LjKIXHL6jg8bu7yUibMX9lu1eFV5/FjVUe6l7+r3ls10fHqvXLqOtFxbSFQbRMdS5JS259O7wwkJ1fentXpV4hhGL4FvLXavutrbL8IvLiWcx9jZU8i1bPY/PnEFiyg+U9/SuNjj23ochrO7mdDs27w/JmVE1U/AV0H0bpxFjcftxtXvnAy7Xs/zFVfX8+//vwnSs67MrmC2TNPb/5eH1uLitJkkPDpn6F0Gex6SjJgaNqpoSurtqR0Ce+/8xDlz77MLt+uoHcC5vRpR975P+WAg08msgUmE89Ky6J3QW96F/Su3pcIE8xaMYtxS5KB0auLxzNlzg80mTCPHWeE7DhjLN2fGkv6448DUN6pNTn9+9Ni4P7k7jGA9Fat6r/QjFzof15yDqJpnye/tkMfgC/+Aj0PhwE/ha6DHXLWkMIQZn6V7J035mUoWwaN2hPuczVzuhzLRyuKGDprGOPG/5z55WNJBKUAxEvbkLmkH/0mprPv7GX0nT2enNIJJKJpVOy8G+mDzqbd4QezY+et5++uVFM0EtA4J53GOemrzEtQV2EYUlQerw6OfvHCGRzzXim9m3en+713OdGxpE1yw8I8OBA6/fbJhi5F24jO5dc0dAlbhP3Lt2GJoiIq5swh2qQJLZxkGTrvAxd9CM+cCk+dAEfcAXtcyDG7teV/33fl3Qmn06z7Q/xiyl08fNdtzL/sSmb96te0/+tftu8VzBLx5BCWD26B5TNhh4Ph4BuTw5G2AokwwbCZX/D1vx+k3Vuj6D0zQVlmhOVH7c3Ol/ya3l17NnSJRIIIHRp1oEOjDhzS6ZDq/UtKlzB+SXJY2gvzfqDou9E0+mEmPWfMo+frbxB76Q0AilrkUb5LdxoP2JOO+w0hr2uP+uvZEQTJSdo77wvLZsKIx2HkP2H8W9C8e3LIWZ/TIDO/fu6n9Vs2C759jvjXTxNdMolYJItxzQ7gpYL+vFdRxuJxY2HaG0TSigEIy1vQiAH0Ke3IoFnldJs4luyxQwnicaJNmpB32MHkDR5E7r77Es3La9j3Jm0BQRCQV7lKVpvGkJ81nU7F5XRq3Jo2jVNjOWFtuFRZalqSNhfDoW1YJDeXrB13JEhL277DjQ3RtDNc8D946UJ485cwfxzBkD9x03E7M/zexWQtOYdJ8b/xh7zXuOHa3zD/lltZcM89tPzlLxu68voXhjDhf/DejTD/B2jbF477G3Q9oKErA2B+8XzeGPU8C154loFfLOHA5ckQJXLlKexy5sVE87f+MKNpVlP2arMXe7XZC3oDB0JZvIxJSycxfsEPzBn9JfFvvqPJuNn0GPoNkQ++Yeaf/s7y/CiLerSEPjvRfM/96NbvQApyW2x6QY3bw0G/h/1/BT+8khw6+Pav4P0/JgOiARetMsmxNl1ZLM70RcVMnbuQYNybdJz+H3YoGkmEkHciPXgq+wi+y4oQZk0lkj4WMiAv0pxu+XszoKA/hxQ3pmDUDxR++BHlUz4AILP7DuSdfz55gweRvdtu2+/Kk5IkSdpqGA5t4yKZGzcfwHYtqxGc9iy8d0NyiM2iCTQ7+Z/cfuIunP/PMgYPOIcPZzzODjt34+RTT2HRw4+Q0a0bTY47rqErrz8zR8C7N8C0z6BpFzjpcdjpuOTExg0olojx+azPee+zJyh4fRj7f5cgqwJKdu1Gqwt/RtODDt3mPwhnRjPZqflO7NR8J+h1EpyS7B01c8VMJo/+hCXDPif4dhwtxs+n2cg58Oj7TMm6kQ86ZrJsx3ak992V1v32oWfL3nTM70i0DqsariE9C3Y7NfmYORKGP5TsTTT8H8mhZnteDN0PrdOKiUoOcZm7vJTJC4qYvLCIyQsKmbKwiMnzC2m1bBQnRD7hqOiXlEfLeCe7Jbe03oMJ2RWsYBEwhiYZTdmzzUD2arsnA3J70+SbKRR+9BGFn/6FxLJlLE5PJ3ePPWh6+unkDR5ERvv2Df2WJWmbkypzgkjS5mI4pO1TJAqH3gwtesHrv4BHDubA057n1D068MJXIYcPPpKHxzzCDqffSu+p05j7++vJ6NiRnN13b+jKN83CCcleImNfg9wWcMRdsPs5kLaFll5fi1mFs3j5x5cY987z7P3ZEk6fFJJIi5A+5BA6XnApWTvu2KD1bW6RIELHRh3puN+ZsN+Z1fsXTRnHtE/fpnj4cDqOmUTjVyfDq5MpS3uFEe0Cnu2UTtFOHcntszs7tOlNr2a96N60O9lpG9B1vn0/aP+P5N+HkU/AiMfg2VOhSSfY40Loe+ZWNRl5Q1peWsGUBUVMXljIlAVFTFpYxJQFRUxZWLTKakfdMpZwTu5Qro9/yLQmi/kyO5czG3VkMiUA5KUX0r91f/ZsvScD2gyg49J0ij76mMLH3qB45B8ojseJNm1K/uDB5A0eTO4++xDNS5G5zyRpM0mVOUG06f7Y/E4Anm/gOqStjeGQtm99z6ycqPoMeORAbjjuMT6bmMN3ow+iT+/ZXD/8D/zz+r+QdemNzLz853R+4QUy2m+DK5itmAsf35788J+WBYOuhYE/a9B5ZiriFXww4wNeHfMCme8N4/ARcQ5eCPGm+TS7/Cyan3oaaQUFDVbf1qB5l14079ILzk6+ji1axPKvhjHviw/ZYeTX9P50DsEnE4lFJjK5NbzVIWB8xyjFO3akU7ve9GzWk17NetGzWU8KstfzZ5nXEg74Fez7Cxj3Bgx/GN79PXx4C+xycrI30VYyD9XmVBFPMH1xMZMXFDFlYWGN3kBFLCwsq24XCaBDsxy6FuSyV9fmdG8asEvRhyya9SLfrJjEf7My+XNmJomgBVnRTPq27MPRbQawZ+s96dW4O+WjvqXw9Y8o/PBKpkydCkBm9+40v+CC5HCxXXfd5nvJSZIkafthOKTtX6eB8NMP4NnTyH7+Jzy1x/Uc+Gl39iy9mAXZN3HliN/xr3tuZ/k5lzHz0kvp9Oyz285v8UuXwxf3J1epipdD//PhgF8ng4AGMnnZZF7+8WU++fo/7PXlEs4fFZBbkiDaszstrzmfRkccQSSjYXsyba3Smjen2ZAjaDbkCADihYWUfPMNxV99RebwoXT/ehzBsBhhMJnZrWbwXds3eKJjwNgOAdGCguqgqOq5U36nNYelRdOh9/HJx9zvkiHRty/AN09Bx4HJeYl2PDrZbhsVhiELVpQxqUYvoMkLkz2Api8uJl5jjezmuRl0bZHLgb1a0LVFHl0KcunWIpeOzXIJiPHd988ybOy/+d+cidyZkUZFNCCtSWN2bbYjF3XYnwGtB7Bbi92IFpZQ+OlnFP7zCSZ/+imJ5csJ0tPJGTCApmeeSd6gQdtm8CxJkqSUYDik1FBjououw2/g3x1O4NThx/HHU/7AfT9cwdWT/8zf/nwHcy+9nNnXXEP7B/66df9WP1aeHB70yR1QvAh6nwAH/g6ad2uQckpiJbw77V1e+vElln8zkqNGhNw2LiQA8g86kOZnn012//71t0JXiojm5ZG3337k7bcfLYFEaSml331H8YgR5H01gvbffMOQr5NDmZa3LGRip9EMb/U5b7QPmd8EstKy6dG0Bz2b9aRn0570bNaTHk17kJOek7xB613gmPvhkD/AN/9KBkUvngf5bZJB4+7nQH6rBnv/61NYFmPqwiImVc0BVDkEbMrCIgrLYtXtstIjdG6ey05tGnHkLm3o2iKXLgW5dC3Io3HOyhAsnogzbsk4Pp38DMM+/C9fl8yhJIAgDNkxpylnthnIgB7Hs3vrfuSk51A2eQqFb33EnA//QvHXX0M8TrRZM/IPOii5utjeDheT6sO09G6UBVMbugxJkrZrhkNKHZn5cOoz8N6N9P/ifl7IncSv37qa3//kZn77xS/5U6O3+b/f/pZ5N93E/LvvptWvftXQFa8pkYDvX07OK7R0GnTZHw7+A7RrmLmSxi0ex0s/vsR/J7zBjmOWc9rX6XSaESfIy6Xp2SfT9MwznFy3HkWyssjZYw9y9tgDLoUwFqN07FiKR4wkb8QImo4cye5fJUORWLNGzO3RnHHtl/BJizd5sfELhEFAQECnRp1W9jBqmnwuGHg5wV6XwYR3kxNXf3gLfHxHsofRgIugfX9ogHAvFk8wY0nJakPAkmHQvOUrh4EFAbRrkk3XFnn069SUri2S4U+XFrm0aZRFJLJm7WEYMmnpJIbNGcbw2V/w1ZxhLI+XAtCtvILjMpqzZ+dD6L/7RTTOb0tYUUHxyK9Z8fRfmPPhh5RPmwZAZs+eNP/pheQPGkTWrru6eqQkSZK2OYZDSi2RKBx6E7ToRZ/Xr+Th8v/jmS/u4Od9f87939xP9927c9Tpp7H40cfI7NqNJiee0NAVrzTpg+QKZHO/hVa7wJkvQbeDtvgH9sLyQt6e+jYv/fgSU2eO4bDREe4ZFSV3SYL0ji1pdt3ZND7+eHtMbAFBWhrZu+xC9i670Py8cwkTCconTaJ45EiKvxpB1ogRtP9yHgcDQeNGlOzUmVndGjG6XQWfV3zHO1Pfqb5Ws6xm1WFRz30uoNe+l9Np7NukjXoWvnsB2vZNhkS9T0iuhraBwjAklgiJxUNiiUTl88rt0oo4Y2Lt+W5peyYWt+VvnXZmSXZbZv/+v8RqDANrmpNOl4Jc9uveonoIWJeCPDo1zyErff29/WYVzmLYnGHJQGjucBaWLASgXSzOwSUlDIg2YkCPE2jR7zxo0pH40qUUfvQpsz78M4WffbZyuNhee9H0nLPJP+AA0ts5XEySpG3F8xcPbOgSpK2S4ZBSU98ziDTvRqunTuXyyZcwts39HN7lcO7/+n66nvlndpg6lTk33khGp47k9O/fsLXOHgXv3QiTP4TGHeH4fyQnEN6CvRPCMOTbhd8mewlN/S/N5hZz6rf57PFNQKS8gpyB/Wh29tnkHXCAvSY2UiIRUpFIEE+EVMRDYvHK7URIPF7z2Mo28USyXVXIktyXR0Wv/Yj32JfYKQki8+eSPXY0uePHkP/jGLoN/ZZuwDEZWSzq3IvpnVoxvkMm37cu4fu5Mxg6ezghyd5HEdLJbduHrhWwb9E09nr7Slq//hs+zzyMtzIPZw4tKmtMVNayWvATTyTrr3ysz2jOgmKIBHHiGQvZIbaMow7pmxwC1iKPrgW5NM3dsPmqFpYsrA6Chs0ZxqzCWQAURLIYUFLKnssXMSCeRvsdj4U+ZxC2H0D51KkseuldCj/8iOJvvkkOFysoIP+Qg8kfPJjcgQOJ5Bp+SpIkafthOKTU1XEvMi75iPl/O5b+X1zEDoP+wIzmvfnN0Ot46oa/k3Hxdcz8+RV0/vcLDTM0avEU+OBmGPMiZDeFw25NLj2elrnFSlhWtozXJ73OSxNeYtKSCewxLYM/fdeYNmNWEGQU0+iYY2l21tlk9eyxxWqCyl4o8QSJEGYGecTCCLH5hRQWNiIMI4yasZSZGQXECZgxYeEaPVVWhiuJ6iAmeSysJYBJrAxiamyv3maVXjHV26uGN2vsq7pWIkG4/uxkE7SFgrZQcChNS5fTe9EUdl40hV3mTabfj9+zByGxSJTJzTsyodW+/NiuORM7pFGUt4DytJl8lzaL0U3hgaatCULoEPuUXmUf0J62LM/cn8KsweREm5GeFiUtEpAWDUiLRCq3IzX21XgdCYhGI6RHAqKRgL9/PIlzih5lcOsKft8hm9P++j29mvWi05BLNuidLitbxoi5Ixg2dxjD5wxn0rJJAOSn5zMguzVnJ5qw5+wf6FoRJ+g6CPa6nrDboRR9N5bCpz9kxUc3UDFtOgCZvXrR/KKfkj94MFk772zwKUmSpO2W4ZBSWlrzziTOf4dPHjqdAz/6Pffufganp+dx5Yjf8uR9t7Pk7ItrrGCWt2WKKloIn9wJXz0KkTTY75ewz5WQ1XiL3D4MQ76a+xUvTXiJ96a9R1BaxqlT2nDD8KZkzlxIWos8mv7iSpr85CekNWu2WWspKY9XTzBcc86ZKQuLWFZSAcC+VevA3/0xcCgAx438HFoclNz/6LCNunckIBlwRJPhRXo0knyOBESjAenVxyKkV7WJREiLRMhKXy0IqRGCrD0siay5r+p19bHKGqKrhiyrHlvZpup1zWtU7YsEEAQB8eXLKf76a0pGjCB/xEh6jPmII8fEIAjI7NWLnP79ye73U4p26siEYAHjFo9j/LyvGTt/NDPiS4HXIP4aTcmmZ9Od6VXQOzmfUdNedG7cmbRI3f6beXHkTPqVTaFDemOCILvOX6fiimK+nv81w+cMZ9jcYYxdNJaQkOy0bHZvuTvHNtuZAfMm02v8B0Rj30PzHWD/3xLrdDhFoyZS+M+PKPz0ThIrVhBkZJCz1540P/dc8g44gPS2bTfqe0eSJEnbj1QZimg4pJTXtV1rPjn0H/z9vzdyyddPc1+XPTk3soBfT7mH++6+i9kXXcKsX/6SDn/72+YtpLwouST95/dDRRH0PQsGXQuN2mze+1ZaWLKQVye+yssTXmb6iul0Ksnl+h+70OPTqbBiJlk770yzK39No8MOI6jHpejjiZBZS0qYVLnk+JSFK5cfn72sdJW2bRpn0aUgl6N2bcOnExYQDQIuTX+DNBKkDf4VD43+G0EkwdX9r2TxXXcRDUPa/e63KwOVmgHMamHMKm0iQa0TGG+Poo0akT9oEPmDBgGQKCmhZPS3FI8YQfHIESx98UWWPPUUAO07d6bHHv3J7nc4OYOup6xpOhO+eYxxY19mfNkCxpUM5Zl5IyknAUBGJIMdmu6wysTXPZr2IC9j44PW8ng5oxeMZvjc4QyfM5xvF3xLLIyRHklntxa7cWmfS9kzpz27TBtJ+rf/huUzIbMx4a6nUNZsMIVjF7Li7x9R8s1jkEgkh4sddujK4WI5OZv8ZypJkiRtawyHJODsvbtx5tjLmT6jA7dMe4Q/tmjL/yVGck+jzlzxu+uY94c/Mv/Ou6Dp/vV/83gFfP0kfHw7FM6DXkfBQTdAi80/VCueiPPF7C94acJLfDzjY2KJCo4p6ckNI3cif+j3EI4l/9BDaXb2WWT37bvRS9GHYciiovLK5cYLk71/KnsBTV9UTHk8Ud02PyuNri3y2LNr88q5ZpLLjncpyCUnY+U/Wac8NBSAn2SMS+7YrS3/mZOcT+bAXq2YVjYPgE5dNm/vpu1NJDub3L32JHevPQEIKyoo/eGHZFg0YiTL3/kfS//9IgBpbdrQql8/uvS/mhN7ZZAx93XiY19jajRgXIfdGd+6J+MSpXww/QNenvBy9T065HdYJTDq2awnIbWPq4slYoxdNLZ6mNg387+hNF5KJIjQu3lvzul9DgPaDKBvo25kj38bRjwDM4ZBECHsPJji1uewYnIphX/5jIoZ/wMgc6cdKbjkYvIGDyard2+Hi0mSJCnlGQ5JQCQScOfJuzHknmVEmuzATStuZWI8wsMTXqL7gN9w0Jlnsvif/6TvIfDNzvUUEIUhjH0tuSz9oonQcSCc8i/oMKB+rr8Ocwrn8J+J/+E/E//D3KK5tEhryq+XDqTvR7MIf/iBSKNGNDn3HJqdfvoGrcRUXB6rHgY2uboXUDIQWlEaq26XHg3o1DyXrgW5HLRjS7pWTjjcpSCX5rkZGx1Cqf4F6elk77Yb2bvtRvMLLiBMJCibMJHiEV9RMnIkxcOGsfyNNwCINm1K9m4n0LSglIO+H8aRE4cRFHQl7H8h83seyviSOYxbPI5xi8fx45IfeXfau9X3iWbmclPzUvoFCRaWRPi0e4x/t5/MD8/tz4qKFQB0b9qdk3qcxIDWA+jXuh+N0nJh8kfwxaMw7g2IlRLL7UFR7umsmBpS9PpIEoVjCTIzyd1rL5pfcD55gwaR3rp1Q/xRSpI2o1QZ9iFJm4vhkFSpXZNsbjimN9f8O0bvwU9w+eRfM7F4PncMv50uZz9A+ylTOOr9p1jcpBWwiT+ATP0M3r0eZo2EFr3gtOegx5DNuix9RaKCj2d8zIsTXuSLWV8AcGCjPfjjnN1o/tZXxOd/SHqXLjS74XoaH3vsWofXxOIJZi4pWSX4qQqE5qw2DKxt4yy6tMjluD7tkr1/WuTSrSCPtk2ySIvaW2NbFEQiZPXskZyE/IwzCMOQiunTKR4xsnIo2kgKP5oORIlkdSa7NeR8eiv5bW5j30HHs/8+l8KuFwFQVFHEhCUTGLd4HA98/gmlkS94nkLKlq9gal9oU1zKYV2OZs/We9K/dX8KsguSRSz4ET65G0Y/T7h8NuVlTVhRvheF06Bk7ERIfES0RQGNDh9C3uDB5O61l8PFpG3cn07vzLDznmzoMiRJ2m4ZDkk1nLh7O/73/Vxu+HQB/S95mds+uZKzikZxzUdX8PQNj7PklEmc8voDlP/0QDI6dtzwG8z7Ht77A0x4B/LbwjF/hT6nQyRa/2+m0rTl03hpwku8OvFVFpcupmVOS65qeiL7fbGM2NsfEJaVkbXPPjS7+SZy992XIBIhDEMWrChbJfiZtCA5KfT0xcVUxFcOAWpUOQxsYNfmlUPA8qqHgWVnbL73pa1DEARkdOpERqdONDnxBAAq5s2nZOSI6qFoC74rhe8geO9dspq/Sc4Orcg5+ASyj76IPi370KdlH/7zcSeuX/QtPds04syCRhz3xEQGZu1Ep0tvSN6oZElykvbRzxJO+4rihVmsKOpO4dSmVMxfAvxI1k47UXDJJZXDxXZyuJgkSZJUR4ZDUg1BEHDrCbtw2D2f8IuXJ/CfS5/mL+9fy2mz3+SKD86j6zE/45zn/s6MSy+j83PPEs3Pr9N1o2nFNGk6Dh7cBzIbwcF/gD0vhvS6r8q0IcriZbw37T1emvASX839imgQZf+2+3H60l60eXMExUOfI5aVRe7Rx7Ds8BMYldcyORn0C6Or5wNaUbZyGFhGNELnghx2aJnHITu1pmuL5JCwLgW5NHMYmFaT3qol6UccQaMjjgAgvnQpxV9/Q/HQTyn+/H0WfTmfRUMfg5sfI6tjATn7DKbXipaQnSCNgOy0bJoWRyArARPehVHPEBv1FoUzAwoXtaBoZicSpRUEmYXkDhxI88sGkTd4EOmtWjXsG5ckSZK2UYZD0moK8jK57YRduOipkfz1w0lcPeRO7v68BT+d8CQFTe/hvcPP4Og3nmXW1b+kw4N/Y52xSPFi+Oxu2nV4P/l678th36shZ/NMkjxhyQRemvASr096neXly2mf154rel3CgK8SRP75Gmmz32NBo2Z8ue9JvNy2P5PL0+D1WUByIud2TbLp2iKX43dvVzkZdB5dC3Jp2ySbaIqs3qX6F23ShPwDB5N/4GDgehIrVlDy5sMUv/sSxRNmseT5Fzg1nvz+mtS8kOO6LaFb+iKa533Awus+pHBuLiULmkEIaS2b0eiYZBiUu9deRLI3T8AqaevRufyahi5BkqTtnuGQVItDe7fmxN3b88BHkzhwx1b03+fX/C4two3jnqB728dofclZzP3rc8y/805a1TK3bXoiDp/dA5/eA2XLKSpsz9IlvWh/6M31XmtxRTFvT3mb58e9yNglY4gGabRLH8BOS3uz6xs/0mfcI2RUlDKuaUde6X8G33XrR6dWjehbkMeJlSuBdW2RS+fmuWSlOwxMm18kP5/cU68m99SrYf44wqEPsezdF4gtgOIFWfT6poyKigxmjMsASA4X+8mB5A0eRNZOO9lTTZIkSapnhkPSWtxwzE4MnbSQq18YxZs/348T97yGh7/9kWcbD2XHhf9g4CH7s/iJJ8k4rICmfRolT0rE2W/+VI6b+QOUvwrdD4ODb2DRNbfXS02FZTGmLChi0oIVfDXnW75a9DZz40MJgzLiZS2pWHIEPScXcPzEr9hr9t+BgDl99mbesSfTcWB/7i3Io2luRr3UItWLlr0Ijr2HX8w+ktsWXk7HjFKeLWhN+xeKaJvWno6PPUZ6q5YNXaUkSZK0XTMcktaiUVY6d528G6c/Mozb/zuOG4/pTaPE2SyPL+aPBT/yaOn7FPTswdx3F5LRJI3c8W/Dezdy/oJxTMprSrPTX4LO+27wfSviCWYsLl5jKfgpC4uYX7SE9MajSG/yFdGsORCm0yzcg92yD+LQ+cvp+MnrpE/6kUjjxjS98EKannE6vV22W9uAkkgui6MFtGnZmP+1acmpad+T1qyZwZAkSZK0BRgOSeuw9w4FnLt3Z/75xVQO2akVAVHaVlxCWdN7uapjBs+UjSVjbgGzXplD58IzyOjchb9235ORzdry+DqCoTAMmb+ijMkLipi8sDA5GXRlEDR9cTHxxMrVwJrkpNGm1RyadhpGeWI48bCcro168pOe53Nk04FUvPwGSx77C/EFC8no1o1mf/gDjY852rlYJEmSJEl1Yjgkrcf/DenFJxMW8Kt/j6ZNkyzSIrncf+BfOOPNM7iidx8eKx7NvHebM+ObHen8u9cYOWzlxJkrSiuYsrCI4dkdmZGWz+Jnv2FKZRhUVB6vbpeZFqFLQS47tsnniF1a06Ugj4LG5fyw/APenvYqU5dPJS+Sx0k7HM8J3U+g64IIi598irlv3EZYXk7u/vvR7LZzyN1nb+djkSRJkiRtEMMhaT2yM6Lc/ZM+nPjgF5THE3RrkUeXxl2464C7uPT9S/ndTp35eUZLyl6dzrALfs6P++1JYWlj9hj5HgtWlCUv0mwgQRjSfsYSuhTk0b9TM7pWTwadR5tGWUQiAYkwwZezv+TFCf/gwx8+JJaI0bdlXy7c5UIObn8giS++YvHVdzBl2DCC7GyanHQiTc88k8yuXRv2D0mSJEmStM0yHJLqoE+HJvxs8A7c//4E0qPF3PTGD0xekEZW0XF8lP8y7zTvzqBd+nPVqH9zbHkmT+89iEE9WtC1RR5dCnLJ+vPNtI0V0uP2f9Z6/blFc3ll4iv8Z8J/mF00myaZTTi91+mc0P0EOqe1YtnLLzPnqROpmDGDtDZtaPmra2hy0klEGzfesn8QkiRJkqTtjuGQVEc/P3AHHv10MnOWlfL0sGl0KcijX/NjmFv4NRMKPqDvL28g950sjv7XU2T2nceVJ59Tfe602PI1rhdLxPhk5ie8NOElPpv1GYkwwV5t9uKqfldxYMcDYfY8ljz4Lya++BKJoiKyd9+dlr+8mvyDDyZI86+uJEmSJKl++AlTqqP0aIQd2zQilkjw8qX7EIkk5/apeDzOhWEmj47/E/v/9DGmfP06B780haIjvyR3r73WuM6M5TN4eeLLvDrxVRaULKBFdgsu2PkCju9+PO3z2lP81VfM+/PVFL7/AUSjNDr8cJqdfRbZu+yypd+yJEmSJCkFGA5JGyAaCYhGotXBEEA6AfeEBZyWFXLFx1fR8YxOnPeXMtKu/AWdn3uWzC5dqAgSfNFyGZ/870KGzRlGJIiwX7v9OLH7iezXfj8iFXGWv/kWU568grJx44g2bUrziy+i6Wmnu5S3pJTwx+Z3Mi39LmBGQ5ciSZKUcgyHpHrQjCj3H3g3Z719FmMrinjxwh5c/sB0Jl/8U/772wN4cb8xLM+I025FBZf3uZzjdjiOVrmtiC1YwJK/PsiS558nvmgRmd270+bmm2h01FFEsrIa+m1JkiRJklKA4ZBUT3o268lt+93GLz78BSMzZvDPM1py2t/G0+725+hzQFMOmVPAMfe+TCSIUPL998x+8h6WvfUWVFSQN2gQzc45m5y99nIpekkp7bmf9+bxIY83dBmSJEkpxXBIqkcHdTyIdnntmFU4i+/aNWOfS49glwfeom9RlPSO+RS+9z5LnniS4hEjCHJyaHrKKTQ78wwyOndu6NIlqcF1Lr+Gx4cMbOgyJEmSUo7hkFTPWue0pllWM54+4mmCIGBeaWsWP/oYscWLmTViBOnt2tHy//6PJieeQLRRo4YuV5IkSZKU4gyHpHoWBAGZ0czq4WEtr76aZa+8SlheTptbbyH/wAMJotEGrlKSJEmSpCTDIWkzC6JRMrt2BaDRIYc0cDWSJEmSJK0q0tAFSJIkSZIkqeEYDkmSJEmSJKUwwyFJkiRJkqQUZjgkSZIkSZKUwgyHJEmSJEmSUpjhkCRJkiRJUgpzKXtJkiRttZ6/eGBDlyBJ0nbPnkOSJEmSJEkpzHBIkiRJkiQphTmsTJIkNTiHDkmSJDUcew5JkiRJkiSlMMMhSZIkSZKkFGY4JEmSJEmSlMIMhyRJkiRJklKY4ZAkSZIkSVIKMxySJEmSJElKYYZDkiRJkiRJKcxwSJIkSZIkKYUZDkmSJEmSJKUwwyFJkiRJkqQUZjgkSZIkSZKUwgyHJEmSJEmSUpjhkCRJkiRJUgozHJIkSZIkSUphhkOSJEmSJEkpzHBIkiRJkiQphRkOSZIkSZIkpTDDIUmSJEmSpBRmOCRJkiRJkpTCDIckSZIkSZJSmOGQJEmSJElSCqtTOBQEwZAgCMYHQTAxCILf1HK8YxAEHwZB8E0QBN8GQXBE5f7OQRCUBEEwqvLx9/p+A5JS3HlvJh+SJEmSpI2Str4GQRBEgQeAQ4CZwFdBELwWhuEPNZr9DnghDMMHgyDYCXgL6Fx5bFIYhn3qtWpJkiRJkiTVi7r0HBoATAzDcHIYhuXAc8Cxq7UJgUaV242B2fVXoiRJkiRJkjaXuoRD7YAZNV7PrNxX043AmUEQzCTZa+jnNY51qRxu9nEQBPvVdoMgCC4KgmBEEAQjFixYUPfqJUmSJEmStEnqa0Lq04B/hmHYHjgCeCoIgggwB+gYhmFf4GrgmSAIGq1+chiG/wjDsH8Yhv1btGhRTyVJkiRJkiRpfeoSDs0COtR43b5yX00XAC8AhGE4FMgCCsIwLAvDcFHl/pHAJKDHphYtSZIkSZKk+rHeCamBr4DuQRB0IRkKnQqcvlqb6cBBwD+DINiRZDi0IAiCFsDiMAzjQRB0BboDk+utekmq4fEhj1dvd3rqyQasRJIkSZK2HesNh8IwjAVBcDnwDhAFHgvD8PsgCP4IjAjD8DXgl8DDQRBcRXJy6nPDMAyDINgf+GMQBBVAArgkDMPFm+3dSJIkSZIkaYPUpecQYRi+RXKi6Zr7rq+x/QOwTy3nvQS8tIk1SpIkSZIkaTOprwmpJUmSJEmStA0yHJIkSZIkSUphhkOSJEmSJEkpzHBIkiRJkiQphdVpQmqlhucvHtjQJUhKUc9fPBAeb9zQZUiSJEkpyZ5DkiRJkiRJKcxwSJIkSZIkKYUZDkmSJEmSJKUwwyFJkiRJkqQUZjgkSZIkSZKUwgyHJEmSJEmSUphL2UvaJj1/8cCGLkGSJEmStgv2HJIkSZIkSUphhkOSJEmSJEkpzHBIkiRJkiQphRkOSZIkSZIkpTDDIUmSJEmSpBRmOCRJkiRJkpTCDIckSZIkSZJSmOGQJEmSJElSCjMckiRJkiRJSmGGQ5IkSZIkSSnMcEiSJEmSJCmFpTV0Adr2PT7k8YYuQZIkSZIkbSR7DkmSJEmSJKUwwyFJkiRJkqQUZjgkSdqqPD7kcXo169XQZUiSJEkpw3BIkiRJkiQphRkOSZIkSZIkpTDDIUmSJEmSpBRmOCRJkiRJkpTC0hq6AG2/Oj31ZEOXIEmSJEmS1sOeQ5IkSZIkSSnMcEiSJEmSJCmFGQ5JkiRJkiSlMMMhSZIkSZKkFGY4JEmSJEmSlMIMhyRJkiRJklKY4ZAkSZIkSVIKMxySJEmSJElKYWkNXYAkSavr9NSTDV2CJEmSlDLsOSRJkiRJkpTCDIckSZIkSZJSmOGQJEmSJElSCjMckiRJkiRJSmGGQ5IkSZIkSSnMcEiSJEmSJCmFGQ5JkiRJkiSlMMMhSZIkSZKkFJbW0AVI25vHhzze0CVIkiRJklRn9hySJEmSJElKYYZDkiRJkiRJKcxhZdIW0OmpJxu6BEmSJEmSamXPIUmSJEmSpBRmz6HN5PmLBzZ0CZIkSZIkSetlzyFJkiRJkqQUZjgkSZIkSZKUwgyHJEmSJEmSUpjhkCRJkiRJUgpzQuqt3ONDHm/oEiRJkiRJ0nbMnkOSJEmSJEkpzHBIkiRJkiQphRkOSZIkSZIkpTDDIen/27ujUEnvs47jv4cNSUAwFLI3JjEuGjURwcISldwUNLotYooXuqleNJTGislF8cIIpZRCQa+UQrCmEBZ70W3ohaxkJSAavDDiLipitqQsEZuNF90ieiWGpI8XO4HpyaY7y84575zzfD6w7Mw777znmdkd/vDlPe8AAADAYOIQAAAAwGAbxaGqOlVVr1XV5ap65jqP/3BV/W1V/XNV/WtVfWTtsT9YPe+1qvrlbQ4PAAAAwK254VfZV9WxJM8meTTJlSQXqupcd19a2+0zSV7o7j+tqoeSnE/yI6vbp5P8VJIfSvLXVfXj3f3Otl8IAAAAADdvkzOHHk5yubtf7+63kpxN8tiefTrJD65u35XkP1e3H0tytrv/r7v/Pcnl1fEAAAAA2AGbxKF7kryxdv/Katu6zyX5raq6kmtnDT19E89NVT1ZVRer6uLVq1c3HB0AAACAW7WtC1I/nuRMd9+b5CNJvlJVGx+7u5/r7pPdffL48eNbGgkAAACAG7nhNYeSvJnkvrX79662rftEklNJ0t2vVNWdSe7e8LkAAAAALGSTs3suJHmgqk5U1e25doHpc3v2+VaSX0iSqnowyZ1Jrq72O11Vd1TViSQPJPnHbQ0PAAAAwK254ZlD3f12VT2V5KUkx5I8392vVtXnk1zs7nNJfi/Jl6vq07l2ceqPd3cnebWqXkhyKcnbSX7XN5UBAAAA7I5Nfq0s3X0+1y40vb7ts2u3LyV55H2e+4UkX7iFGQEAAADYJ9u6IDUAAAAAh5A4BAAAADCYOAQAAAAwmDgEAAAAMJg4BAAAADCYOAQAAAAw2EZfZc/uuv8rf770CAAAAMAh5swhAAAAgMHEIQAAAIDBxCEAAACAwcQhAAAAgMHEIQAAAIDBfFsZ3ISv/fbPLz0CAAAAbJU4xGaeeHHpCQAAAIB94NfKAAAAAAZz5hDcKmdVAQAAcIg5cwgAAABgMHEIAAAAYDBxCAAAAGAwcQgAAABgMHEIAAAAYDBxCAAAAGAwcQgAAABgMHEIAAAAYDBxCAAAAGAwcQgAAABgMHEIAAAAYDBxCAAAAGCw25YeAACSJE+8uPQEAAAwkjOHAAAAAAYThwAAAAAGE4cAAAAABhOHAAAAAAYThwAAAAAGE4cAAAAABhOHAAAAAAYThwAAAAAGE4cAAAAABhOHAAAAAAYThwAAAAAGE4cAAAAABhOHAAAAAAYThwAAAAAGE4cAAAAABhOHAAAAAAYThwAAAAAGE4cAAAAABhOHAAAAAAYThwAAAAAGE4cAAAAABhOHAAAAAAYThwAAAAAGE4cAAAAABhOHAAAAAAYThwAAAAAGE4cAAAAABhOHAAAAAAYThwAAAAAGE4cAAAAABhOHAAAAAAYThwAAAAAGE4cAAAAABhOHAAAAAAYThwAAAAAGE4cAAAAABhOHAAAAAAbbKA5V1amqeq2qLlfVM9d5/I+r6l9Wf75ZVf+99tg7a4+d2+LsAAAAANyi2260Q1UdS/JskkeTXElyoarOdfeld/fp7k+v7f90kg+uHeJ/u/tntjYxAAAAAFuzyZlDDye53N2vd/dbSc4meez77P94kq9uYzgAAAAA9tcmceieJG+s3b+y2vYeVXV/khNJ/mZt851VdbGq/qGqPvo+z3tytc/Fq1evbjY5AAAAALds2xekPp3k6939ztq2+7v7ZJKPJfmTqvrRvU/q7ue6+2R3nzx+/PiWRwIAAADg/WwSh95Mct/a/XtX267ndPb8Sll3v7n6+/UkL+d7r0cEAAAAwII2iUMXkjxQVSeq6vZcC0Dv+daxqvrJJB9I8sratg9U1R2r23cneSTJpb3PBQAAAGAZN/y2su5+u6qeSvJSkmNJnu/uV6vq80kudve7oeh0krPd3WtPfzDJn1XVd3MtRP3h+recAQAAALCsG8ahJOnu80nO79n22T33P3ed5/19kp++hfkAAAAA2EfbviA1AAAAAIeIOAQAAAAwmDgEAAAAMJg4BAAAADCYOAQAAAAwmDgEAAAAMJg4BAAAADCYOAQAAAAwmDgEAAAAMJg4BAAAADCYOAQAAAAwmDgEAAAAMJg4BAAAADCYOAQAAAAwmDgEAAAAMJg4BAAAADCYOAQAAAAw2G1LDzDaEy8uPQEAAAAwnDOHAAAAAAYThwAAAAAGE4cAAAAABhOHAAAAAAYThwAAAAAGE4cAAAAABhOHAAAAAAYThwAAAAAGE4cAAAAABhOHAAAAAAYThwAAAAAGE4cAAAAABhOHAAAAAAYThwAAAAAGE4cAAAAABhOHAAAAAAYThwAAAAAGE4cAAAAABhOHAAAAAAYThwAAAAAGE4cAAAAABhOHAAAAAAYThwAAAAAGE4cAAAAABqvuXnqG71FVV5P8x9JzbMldSf5n6SFuYFdmPOg59vvn7dfxt3ncu5N8Z0vH4vDZlc/+rjuK79NheE27MOMSM1ibrE3T7cJn/zA4iu/TYXhNuzCjtWmZ4x6lten+7j5+vQd2Lg4dJVX1XHc/ufQc38+uzHjQc+z3z9uv42/zuFV1sbtPbuNYHD678tnfdUfxfToMr2kXZlxiBmuTtWm6XfjsHwZH8X06DK9pF2a0Ni1z3Clrk18r219/ufQAG9iVGQ96jv3+eft1/F359+Lw839pM0fxfToMr2kXZlxiBmsT0/m/tJmj+D4dhte0CzNam5Y/7pHlzCEYakoBB+DwsDYBsGumrE3OHIK5nlt6AADYw9oEwK4ZsTY5cwgAAABgMGcOAQAAAAwmDgEAAAAMJg4BAAAADCYOAQAAAAwmDgFJkqr6aFV9uaq+VlW/tPQ8AFBVD1bVl6rq61X1O0vPAwBJUlU/UFUXq+pXlp5lW8QhOMKq6vmq+nZV/due7aeq6rWqulxVzyRJd/9Fd38yyaeS/MYS8wJw9N3k2vSN7v5Ukl9P8sgS8wJw9N3M2rTy+0leONgp95c4BEfbmSSn1jdU1bEkzyb5cJKHkjxeVQ+t7fKZ1eMAsB/O5CbWpqr61SQvJjl/sGMCMMiZbLg2VdWjSS4l+fZBD7mfxCE4wrr775L8157NDye53N2vd/dbSc4meayu+aMkf9Xd/3TQswIww82sTav9z3X3h5P85sFOCsAUN7k2fSjJzyX5WJJPVtWR6Cq3LT0AcODuSfLG2v0rSX42ydNJfjHJXVX1Y939pSWGA2Ck665NVfWhJL+W5I44cwiAg3Xdtam7n0qSqvp4ku9093cXmG3rxCEgSdLdX0zyxaXnAIB3dffLSV5eeAwAeI/uPrP0DNt0JE5/Am7Km0nuW7t/72obACzF2gTArhm1NolDMM+FJA9U1Ymquj3J6STnFp4JgNmsTQDsmlFrkzgER1hVfTXJK0l+oqquVNUnuvvtJE8leSnJN5K80N2vLjknAHNYmwDYNdampLp76RkAAAAAWIgzhwAAAAAGE4cAAAAABhOHAAAAAAYThwAAAAAGE4cAAAAABhOHAAAAAAYThwAAAAAGE4cAAAAABvt/VKD92OkLhLAAAAAASUVORK5CYII=\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "# przygotowujemy tablicę na wartości miar\n", "N = 10\n", "PPV_mean = np.zeros((N,1))\n", "PPV_std = np.zeros((N,1))\n", "REC_mean = np.zeros((N,1))\n", "REC_std = np.zeros((N,1))\n", "ACC_mean = np.zeros((N,1))\n", "ACC_std = np.zeros((N,1))\n", "F1_mean = np.zeros((N,1))\n", "F1_std = np.zeros((N,1))\n", "\n", "# ta tablica będzie nam mówić ile danych wygenerować w każdej iteracji poniższej pętli\n", "n = 30 + np.floor(np.logspace(1,4,N)).astype(int)\n", "\n", "for i in range(N):\n", " # wygeneruj dane używając wcześniej zdefiniowanej funkcji i tablicy n\n", " X,Y = gen(int(n[i]))\n", " lr = LogisticRegression(solver = 'lbfgs')\n", " # policz miary jakości, średnie i odchylenia\n", " ppv = cross_val_score(lr, X, Y, cv=10, scoring='precision')\n", " PPV_mean[i] = ppv.mean()\n", " PPV_std[i] = ppv.std()\n", " rec = cross_val_score(lr, X, Y, cv=10, scoring='recall')\n", " REC_mean[i] = rec.mean()\n", " REC_std[i] = rec.std()\n", " acc = cross_val_score(lr, X, Y, cv=10, scoring='accuracy')\n", " ACC_mean[i] = acc.mean()\n", " ACC_std[i] = acc.std()\n", " f1 = cross_val_score(lr, X, Y, cv=10, scoring='f1')\n", " F1_mean[i] = f1.mean()\n", " F1_std[i] = f1.std()\n", "\n", "# wykres ze słupkami błędów, na osi OX mamy liczebność zbioru, na osi OY wartość miary z niepewnością\n", "ax = plt.subplot(1,1,1)\n", "ppv_plot = plt.errorbar(n, PPV_mean, yerr = PPV_std.ravel())\n", "# uzupełnij wg powyższego przykładu\n", "rec_plot = plt.errorbar(n+2, REC_mean, yerr = REC_std.ravel())\n", "acc_plot = plt.errorbar(n+4, ACC_mean, yerr = ACC_std.ravel())\n", "f1_plot = plt.errorbar(n+6, F1_mean, yerr = F1_std.ravel())\n", "plt.legend(('PPV','REC','ACC','F1'))\n", "ax.set_xscale(\"log\", nonposx='clip')\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Widzimy, że im większa liczebność zbioru uczącego tym mniejszy błąd statystyczny używanych miar." ] }, { "cell_type": "markdown", "metadata": { "id": "Ffs87mqHqvtr" }, "source": [ "# Klasy niezrównoważone" ] }, { "cell_type": "markdown", "metadata": { "id": "ml0Tk0ngqvts" }, "source": [ "Wytwórzymy teraz dane, w których jedna z klas jest M-krotnie liczniejsza. " ] }, { "cell_type": "code", "execution_count": 16, "metadata": { "executionInfo": { "elapsed": 8825, "status": "ok", "timestamp": 1605186104594, "user": { "displayName": "Anna Dawid", "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgUQAwZ7wyayL4BbiM0n_EANCgBjSdZ9H14lgcCFEE=s64", "userId": "02862484648310443813" }, "user_tz": -60 }, "id": "PrYWH1_pqvtt" }, "outputs": [], "source": [ "def gen_rozne(ile, M):\n", " mu = [(-1,0.5),(1,4)]\n", " #mu = [(-1,0.5),(-1,0.5)]\n", " # macierz kowariancji\n", " cov = [diag([1.7,1.8]), diag([1.5,0.7])]\n", " X = np.zeros(((M+1)*ile, 2)) # miejsce na dane wejściowe\n", " Y = np.zeros(((M+1)*ile, 1),dtype = int) # miejsce na dane wyjściowe\n", " print(Y.shape)\n", " klasa = 0\n", " # wygeneruj dane w X z rozkładu normalnego wielowymiarowego z odp. parametrami dla klasy 0\n", " X[0:ile] = np.random.multivariate_normal(mu[klasa],cov[klasa],ile)\n", " Y[0:ile] = klasa\n", " klasa = 1 \n", " # wygeneruj dane w X z rozkładu normalnego wielowymiarowego z odp. parametrami dla klasy 1\n", " X[ile:ile+ile*M] = np.random.multivariate_normal(mu[klasa],cov[klasa],ile*M)\n", " Y[ile:ile+ile*M] = klasa\n", " Y = Y.ravel()\n", " print(np.sum(Y==0), np.sum(Y==1) )\n", " return (X,Y)" ] }, { "cell_type": "markdown", "metadata": { "id": "simmwRmjqvtw" }, "source": [ "Oglądamy dane:" ] }, { "cell_type": "code", "execution_count": 17, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 660 }, "executionInfo": { "elapsed": 9910, "status": "ok", "timestamp": 1605186105685, "user": { "displayName": "Anna Dawid", "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgUQAwZ7wyayL4BbiM0n_EANCgBjSdZ9H14lgcCFEE=s64", "userId": "02862484648310443813" }, "user_tz": -60 }, "id": "2DUzukmBqvtx", "outputId": "44339b67-0b37-42ff-e7f9-55ad8d42fa56" }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ ":5: DeprecationWarning: scipy.diag is deprecated and will be removed in SciPy 2.0.0, use numpy.diag instead\n", " cov = [diag([1.7,1.8]), diag([1.5,0.7])]\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "(3030, 1)\n", "30 3000\n" ] }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "X,Y = gen_rozne(30, 100)\n", "plt.scatter(X[:,0], X[:,1] ,c = Y, cmap=plt.cm.Set1, alpha = 0.5)\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": { "id": "ZL0aq9eBqvtz" }, "source": [ "Obliczamy miary jakości dla danych niezrównoważonych przy podziale 10-krotnym. Zwróćmy uwagę na różnicę w wartościach 4 pierwszych miar i miary MCC:" ] }, { "cell_type": "code", "execution_count": 18, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "executionInfo": { "elapsed": 9906, "status": "ok", "timestamp": 1605186105686, "user": { "displayName": "Anna Dawid", "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgUQAwZ7wyayL4BbiM0n_EANCgBjSdZ9H14lgcCFEE=s64", "userId": "02862484648310443813" }, "user_tz": -60 }, "id": "c5jm53Z3qvt0", "outputId": "f723a392-7da3-40e5-de79-8d01188a71d5" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "PPV = 1.00 +/- 0.00\n", "REC = 1.00 +/- 0.00\n", "ACC = 1.00 +/- 0.00\n", "F1 = 1.00 +/- 0.00\n", "-----\n", "30 2697\n", "25 2702\n", "26 2701\n", "27 2700\n", "25 2702\n", "28 2699\n", "23 2704\n", "27 2700\n", "27 2700\n", "26 2701\n", "MCC = 0.77 +/- 0.32\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "/Users/rafalmaselek/Library/Python/3.8/lib/python/site-packages/sklearn/metrics/_classification.py:846: RuntimeWarning: invalid value encountered in double_scalars\n", " mcc = cov_ytyp / np.sqrt(cov_ytyt * cov_ypyp)\n" ] } ], "source": [ "ppv = cross_val_score(model, X, Y, cv=10, scoring='precision')\n", "print('PPV = {0:.2f} +/- {1:.2f}'.format(ppv.mean(),ppv.std()))\n", "rec = cross_val_score(model, X, Y, cv=10, scoring='recall')\n", "print('REC = {0:.2f} +/- {1:.2f}'.format(rec.mean(),rec.std()))\n", "acc = cross_val_score(model, X, Y, cv=10, scoring='accuracy')\n", "print('ACC = {0:.2f} +/- {1:.2f}'.format(acc.mean(),acc.std()))\n", "f1 = cross_val_score(model, X, Y, cv=10, scoring='f1')\n", "print('F1 = {0:.2f} +/- {1:.2f}'.format(f1.mean(),f1.std()))\n", "print('-----')\n", "\n", "# MCC policzymy w pętli\n", "MCC=np.zeros((10,1))\n", "for i in range(10):\n", " X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size = 0.1) # test na 10% aby było podobnie jak dla podziału 10-krotnego\n", " # trenuj model\n", " model.fit(X_train, y_train) \n", " # predykcja dla zbioru testowego\n", " y_pred = model.predict(X_test) \n", " # policz MCC korzystając z odpowiedniej funkcji\n", " MCC[i] = metrics.matthews_corrcoef(y_test, y_pred)\n", " \n", " zeros = 0\n", " ones = 0\n", " # policz ile jest reprezentantów klasy 0 i klasy 1 w tablicy y_train i wpisz w powyższe zmienne\n", " # MIEJSCE NA TWÓJ KOD\n", " for j in range(len(y_train)):\n", " if y_train[j] == 0:\n", " zeros +=1\n", " else:\n", " ones += 1\n", " # wypisujemy liczebność klas w zbiorze uczącym, wykonaj kilka razy i zaobserwuj jak się zachowuje MCC a jak inne miary \n", " print(zeros, ones)\n", " \n", " \n", "print('MCC = {0:.2f} +/- {1:.2f}'.format(MCC.mean(),MCC.std())) \n", " " ] }, { "cell_type": "markdown", "metadata": { "id": "NC-4END_qvt4" }, "source": [ "Teraz spróbujemy zobaczyć czy da się to poprawić, jeśli w podziałach zadbać o zachowanie proporcji klas. Można to łatwo zrobić za pomocą funkcji `StratifiedKFold`, zwraca ona indeksy do zbioru treningowego i testowego:" ] }, { "cell_type": "code", "execution_count": 19, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "executionInfo": { "elapsed": 10613, "status": "ok", "timestamp": 1605186106397, "user": { "displayName": "Anna Dawid", "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgUQAwZ7wyayL4BbiM0n_EANCgBjSdZ9H14lgcCFEE=s64", "userId": "02862484648310443813" }, "user_tz": -60 }, "id": "lRgNmNCdqvt6", "outputId": "820c0535-fb02-4668-f4bf-528ebef6c561" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "22 2250 2272\n", "PPV = 0.999 REC = 0.999 ACC = 0.997 F1 = 0.999 MCC = 0.874 \n", "22 2250 2272\n", "PPV = 0.997 REC = 1.000 ACC = 0.997 F1 = 0.999 MCC = 0.865 \n", "23 2250 2273\n", "PPV = 0.997 REC = 1.000 ACC = 0.997 F1 = 0.999 MCC = 0.844 \n", "23 2250 2273\n", "PPV = 0.996 REC = 0.999 ACC = 0.995 F1 = 0.997 MCC = 0.674 \n" ] } ], "source": [ "skf = StratifiedKFold(n_splits = 4)\n", "for train, test in skf.split(X, Y):\n", " zeros = 0\n", " ones = 0\n", " # ponownie policz ile jest reprezentatnów poszczególnych klas w \n", " # zbiorze uczącym i wpisz wartości w powyższe zmienne\n", " # podpowiedź: wykorzystaj zmienną train aby dostać się do podzbioru danych uczących\n", " for i in range(len(Y[train])):\n", " if Y[train][i] == 0:\n", " zeros +=1\n", " else:\n", " ones += 1\n", "\n", " print(zeros, ones, len(Y[train]))\n", " # fituj model dla danego podzbioru uczącego\n", " model.fit(X[train,:],Y[train])\n", " # wykonaj predykcję dla danego podzbioru uczącego\n", " y_pred = model.predict(X[test,:]) \n", " # odczytaj etykiety klas podzbioru uczącego\n", " y_test = Y[test]\n", " # policz miary jakości\n", " PPV = metrics.precision_score(y_test, y_pred)\n", " REC = metrics.recall_score(y_test, y_pred)\n", " ACC = metrics.accuracy_score(y_test, y_pred)\n", " F1 = metrics.f1_score(y_test, y_pred)\n", " MCC = metrics.matthews_corrcoef(y_test, y_pred)\n", " \n", " print('PPV = {p:.3f} REC = {r:.3f} ACC = {a:.3f} F1 = {f:.3f} MCC = {m:.3f} '.format(a=ACC,f=F1,m=MCC,p=PPV,r=REC))" ] }, { "cell_type": "markdown", "metadata": { "id": "5uMyDC3Nqvt-" }, "source": [ "Zbadajmy jeszcze krzywą ROC:" ] }, { "cell_type": "code", "execution_count": 20, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 825 }, "executionInfo": { "elapsed": 10608, "status": "ok", "timestamp": 1605186106398, "user": { "displayName": "Anna Dawid", "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgUQAwZ7wyayL4BbiM0n_EANCgBjSdZ9H14lgcCFEE=s64", "userId": "02862484648310443813" }, "user_tz": -60 }, "id": "27l3I5V9qvt_", "outputId": "6979a864-9b27-40e9-fc8f-57dfe652b975" }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "skf = StratifiedKFold(n_splits=6)\n", "model = LogisticRegression(solver = 'lbfgs')\n", "tprs = []\n", "aucs = []\n", "mean_fpr = np.linspace(0, 1, 100)\n", "\n", "i = 0\n", "for train, test in skf.split(X, Y):\n", " # fitujemy regresję\n", " model.fit(X_train, y_train) \n", " # obliczamy prawdopodobieństwa przynależności przykładów testowych \n", " # do klas wg. wyuczonego klasyfikatora \n", " # (zwraca on w danym wierszu prawdopodobieństaw dla każdej z możliwych klas)\n", " probas_ = model.predict_proba(X[test]) \n", " \n", " # Obliczamy punkty krzywej ROC \n", " fpr, tpr, thresholds = metrics.roc_curve(Y[test], probas_[:, 1]) # względem prawdopodobieństwa klasy 1 \n", " # interpoluj wartości TPR dla mean_fpr, jak poprzednio \n", " tprs.append(np.interp(mean_fpr, fpr, tpr))\n", " # estetyka\n", " tprs[-1][0] = 0.0\n", " # policz powierzchnię pod krzywą\n", " roc_auc = metrics.auc(fpr, tpr)\n", " aucs.append(roc_auc)\n", " # rysujemy krzywą \n", " plt.plot(fpr, tpr, lw=1, alpha=0.3,\n", " label='ROC dla podziału %d (AUC = %0.2f)' % (i, roc_auc))\n", " i += 1\n", " \n", " \n", "plt.plot([0, 1], [0, 1], linestyle='--', lw=2, color='r',\n", " label='Losowa klasa', alpha=.8)\n", " \n", "########################################\n", "# poniżej podsumowanie: oliczanie średnich i standardowych odchyleń, cieniowanie przedziału ufności \n", "########################################\n", "# policz średnie tpr po wszystkich klasach\n", "mean_tpr = np.mean(tprs, axis=0)\n", "# znów estetyka\n", "mean_tpr[-1] = 1.0\n", "# policz średnia auc korzystając z tablic mean_fpr i mean_tpr\n", "mean_auc = metrics.auc(mean_fpr, mean_tpr)\n", "# policz standardowe odchylenie danych w tablicy aucs\n", "std_auc = np.std(aucs)\n", "\n", "# proszę narysować krzywą ROC dla danych w mean_fpr i mean_tpr. Krzywa ma mieć grubość 2, kolor niebieski i \n", "# następującą etykietę: r'Średni ROC (AUC = %0.2f $\\pm$ %0.2f)' % (mean_auc, std_auc)\n", "plt.plot(mean_fpr, mean_tpr, color='b',\n", " label=r'Średni ROC (AUC = %0.2f $\\pm$ %0.2f)' % (mean_auc, std_auc),\n", " lw=2, alpha=.8)\n", "\n", "# policz odchylenie dla tprs\n", "std_tpr = np.std(tprs, axis=0)\n", "# poniższy kod zwróci nam przedział ufności 68% wokół średniej ROC (1 sigma)\n", "tprs_upper = np.minimum(mean_tpr + std_tpr, 1)\n", "tprs_lower = np.maximum(mean_tpr - std_tpr, 0)\n", "# użyj funkcji fill_between żeby zacieniować obszar między tprs_upper i tprs_lower na szaro z parametrem alpha=0.2\n", "plt.fill_between(mean_fpr, tprs_lower, tprs_upper, color='grey', alpha=.2,\n", " label=r'$\\pm$ 1 std. dev.')\n", "\n", "# estetyka\n", "plt.xlim([-0.05, 1.05])\n", "plt.ylim([-0.05, 1.05])\n", "plt.xlabel('False Positive Rate')\n", "plt.ylabel('True Positive Rate')\n", "plt.title('ROC')\n", "plt.legend(loc=\"lower right\")\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": { "id": "NOKh4VeeqvuD" }, "source": [ "Powyższe obliczenia obliczenia proszę przeprowadzić dla klas, których rozkłady wyraźnie się różnią i dla takich które się pokrywają w znacznym stopniu. Trzeba podmienić średnie klas w funkcji generującej dane różnoliczne.\n", "\n", "# Jakie stąd płyną wnioski?\n" ] }, { "cell_type": "markdown", "metadata": { "collapsed": true, "id": "B-fCr9NPqvuE" }, "source": [ "Wyobraź sobie, że jesteś na egzaminie i w oparciu o dzisiejsze ćwiczenie dostajesz kilka pytań. Czy umiesz odpowiedzieć wyczerpująco i ze zrozumieniem na wszystkie?\n", "\n", "1) Jakie płyną konsekwencje z dużej różnicy częstości występowania dwóch klas?\n", "\n", "2) Które miary jakości są na ten efekt najmniej i najbardziej wrażliwe?\n", "\n", "3) Jak można przeciwdziałać efektowi częstości występowania?\n", "\n", "4) Czy efekt częstości występowania ma taki sam czy różny wpływ na miary jakości w przypadku, gdy klasy pochodzą z rozkładów, które się znacząco pokrywają lub są mocno rozseparowane?" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "colab": { "collapsed_sections": [], "name": "03M_Miary_jakości.ipynb", "provenance": [ { "file_id": "16srziWO2XRWYY8AVY9Px_05RQQdYY5jA", "timestamp": 1571847683761 } ] }, "kernelspec": { "display_name": "Python 3", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.8.6" } }, "nbformat": 4, "nbformat_minor": 2 }