46_Ejemplo_Clasificacion_Multiple_con_PCA.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 46 - Ejemplo de Clasificacion Multiple con PCA\n",
    "\n",
    "\n",
    "* En este Notebook vamos a resolver un problema de ***Clasificación Múltiple con Regresión Logística*** en el que vamos a ***aplicar al Dataset una reducción de la dimensionalidad usando el método del PCA (Análisis de Componentes Principales)***.\n",
    "\n",
    "\n",
    "* Por otro lado vamos a ***evaluar el rendimiento del modelo*** con los propios datos de entrenamiento aplicando las métricas del ***Accuracy, Precision, Recall y F1***, y viendo como clasifica el modelo con la ***matriz de confusión***.\n",
    "\n",
    "\n",
    "* En los siguientes enlaces esta la documentación de las clases a utilizar:\n",
    "\n",
    "    + Regresión Logística:<br>\n",
    "    https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html\n",
    "    + PCA:<br>\n",
    "    https://scikit-learn.org/stable/modules/generated/sklearn.decomposition.PCA.html\n",
    "    + Accuracy:<br>\n",
    "    https://scikit-learn.org/stable/modules/generated/sklearn.metrics.accuracy_score.html\n",
    "    + Precision:<br>\n",
    "    https://scikit-learn.org/stable/modules/generated/sklearn.metrics.precision_score.html\n",
    "    + Recall:<br>\n",
    "    https://scikit-learn.org/stable/modules/generated/sklearn.metrics.recall_score.html\n",
    "    + F1:<br>\n",
    "    https://scikit-learn.org/stable/modules/generated/sklearn.metrics.f1_score.html\n",
    "    + Classification Report:<br>\n",
    "    https://scikit-learn.org/stable/modules/generated/sklearn.metrics.classification_report.html\n",
    "    + Matriz de confusión:<br>\n",
    "    https://scikit-learn.org/stable/modules/generated/sklearn.metrics.confusion_matrix.html\n",
    "\n",
    "\n",
    "* El ejemplo con el que vamos a trabajar, consiste en ***clasificar tres tipos de flores, conociendo los tamaños (alto y ancho) del petalo y sépalo*** de la flor; es decir, que tendremos 4 variables.\n",
    "\n",
    "\n",
    "* Para resolver este problema vamos a realizar los siguientes pasos:\n",
    "<span></span><br>\n",
    "    1. [Carga de datos](#M1)\n",
    "<span></span><br>\n",
    "    2. [Cambio de estructura de datos a numpy y transformación de las etiquetas a números](#M2)\n",
    "<span></span><br>\n",
    "    3. [PCA: Estudio de la Varianza Acumulada](#M3)\n",
    "<span></span><br>\n",
    "    4. [PCA: Reducción de la dimensionalidad](#M4)\n",
    "<span></span><br>\n",
    "    5. [Creacción del modelo](#M5)\n",
    "<span></span><br>\n",
    "    6. [Evaluación del Modelo](#M6)\n",
    "<span></span><br>\n",
    "    7. [Visualización de Resultados](#M7)\n",
    "\n",
    "\n",
    "\n",
    "* En primer lugar vamos a importar las librería necesarias:\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import itertools\n",
    "\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.metrics import accuracy_score, classification_report, confusion_matrix\n",
    "from sklearn import preprocessing\n",
    "%matplotlib inline "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<hr>\n",
    "\n",
    "\n",
    "# <a name=\"M1\">1. Carga de datos</a>\n",
    "\n",
    "\n",
    "* El primer paso que vamos a realizar es el de cargar los datos que los tenemos en el siguiente fichero:\n",
    "    - ***DataSet***: \"./data/iris/iris.csv\"\n",
    "\n",
    "\n",
    "* Para ello ***leeremos el csv con pandas*** (pasandolo a un dataframe), teniendo las siguientes columnas:\n",
    "    - ***class***: Clase de la flor (Iris-versicolor, Iris-setosa)\n",
    "    - ***sepal_length***: El \"largo\" del Sepalo\n",
    "    - ***sepal_width***: El \"ancho\" del Sepalo\n",
    "    - ***petal_length***: El \"largo\" del Pétalo\n",
    "    - ***petal_width***: El \"ancho\" del Pétalo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>sepal_length</th>\n",
       "      <th>sepal_width</th>\n",
       "      <th>petal_length</th>\n",
       "      <th>petal_width</th>\n",
       "      <th>class</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>140</th>\n",
       "      <td>6.7</td>\n",
       "      <td>3.1</td>\n",
       "      <td>5.6</td>\n",
       "      <td>2.4</td>\n",
       "      <td>Iris-virginica</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>6.6</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.4</td>\n",
       "      <td>1.4</td>\n",
       "      <td>Iris-versicolor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>145</th>\n",
       "      <td>6.7</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.2</td>\n",
       "      <td>2.3</td>\n",
       "      <td>Iris-virginica</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>4.8</td>\n",
       "      <td>3.4</td>\n",
       "      <td>1.6</td>\n",
       "      <td>0.2</td>\n",
       "      <td>Iris-setosa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4.7</td>\n",
       "      <td>3.2</td>\n",
       "      <td>1.3</td>\n",
       "      <td>0.2</td>\n",
       "      <td>Iris-setosa</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     sepal_length  sepal_width  petal_length  petal_width            class\n",
       "140           6.7          3.1           5.6          2.4   Iris-virginica\n",
       "75            6.6          3.0           4.4          1.4  Iris-versicolor\n",
       "145           6.7          3.0           5.2          2.3   Iris-virginica\n",
       "11            4.8          3.4           1.6          0.2      Iris-setosa\n",
       "2             4.7          3.2           1.3          0.2      Iris-setosa"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv(\"./data/iris/iris.csv\")\n",
    "df.sample(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<hr>\n",
    "\n",
    "\n",
    "# <a name=\"M2\">2. Cambio de estructura de datos a numpy y transformación de las etiquetas a números</a>\n",
    "\n",
    "\n",
    "* Vamos a pasar los datos las flores a una nueva estructura de datos para que scikit-learn puedan trabajar con los datos:\n",
    "<span></span><br><br>\n",
    "    - Las ***columnas sepal_length, sepal_width, petal_length, petal_with*** representan las variables de entrada y las vamos a pasar a un array de array (***numpy array***)\n",
    "<span></span><br><br>\n",
    "    - La ***columna class*** que representa la clase (o variable de salida) y la vamos a pasar a un ***numpy array***\n",
    "    \n",
    "    \n",
    "* Por otro lado tenemos que transformar la clase de la flor a números, por lo que utilizaremos la clase ***LabelEncoder()*** para transformar las etiquetas de las clases a números:\n",
    "\n",
    "    - ***0***: ***Iris-Setosa***\n",
    "    - ***1***: ***Iris-Versicolor***\n",
    "    - ***2***: ***Iris-Virginica***\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df[['sepal_length', 'sepal_width', \n",
    "        'petal_length', 'petal_width']].values   # Obtengo el alto y ancho del sepalo\n",
    "y_labels = df['class'].values                    # Obtengo las etiquetas\n",
    "\n",
    "\n",
    "le = preprocessing.LabelEncoder()     # Instancio un objeto de la clase LabelEncoder()\n",
    "y = le.fit_transform(y_labels)        # Transformo las etiquetas a números"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<hr>\n",
    "\n",
    "\n",
    "# <a name=\"M3\">3. PCA: Estudio de la Varianza Acumulada</a>\n",
    "\n",
    "\n",
    "* Con el estudio de la Varianza Acumulada con los valores singulares, vamos a ver el ***compromiso entre la reducción de la dimensionalidad y la perdida de información***.\n",
    "\n",
    "\n",
    "* Tras realizar este análisis podemos ver que cogiendo 2 componentes, reducimos en 2 variables la dimensión de nuestro dataset y nos quedamos con un 98% de la información original del Dataset:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Valores Singulares: ['25.09', '6.01', '3.42', '1.88']\n",
      "Varianza Acumulada: ['0.92', '0.98', '0.99', '1.00']\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pca = PCA()\n",
    "pca.fit(X)\n",
    "\n",
    "varianza_acumulada = np.cumsum(pca.explained_variance_ratio_)\n",
    "print('Valores Singulares: {s}'.format(s=['{:.2f}'.format(i) for i in pca.singular_values_]))\n",
    "print('Varianza Acumulada: {v}'.format(v=['{:.2f}'.format(i) for i in varianza_acumulada]))\n",
    "\n",
    "# Dibujamos la varianza acumulada\n",
    "fig, ax = plt.subplots(figsize=(6, 4))\n",
    "ax.plot(np.insert(varianza_acumulada, 0, 0))\n",
    "ax.set_xlabel('Número de componentes')\n",
    "ax.set_ylabel('Varianza acumulada (%)')\n",
    "ax.set_title('Varianza Acumulada Vs Nº Componentes')\n",
    "ax.grid()\n",
    "ax.set_xticks(range(1, 5))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<hr>\n",
    "\n",
    "\n",
    "# <a name=\"M4\">4. PCA: Reducción de la dimensionalidad</a>\n",
    "\n",
    "\n",
    "* Vamos a obtener el Dataset reducido con 2 componentes y vamos a ver una muestra de como quedaría nuestro nuevo Dataset:\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>C1</th>\n",
       "      <th>C2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>-2.647637</td>\n",
       "      <td>0.819983</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>103</th>\n",
       "      <td>1.970815</td>\n",
       "      <td>-0.181126</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>131</th>\n",
       "      <td>3.232343</td>\n",
       "      <td>1.370524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>-0.174864</td>\n",
       "      <td>-0.251816</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91</th>\n",
       "      <td>0.890160</td>\n",
       "      <td>-0.033812</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           C1        C2\n",
       "32  -2.647637  0.819983\n",
       "103  1.970815 -0.181126\n",
       "131  3.232343  1.370524\n",
       "64  -0.174864 -0.251816\n",
       "91   0.890160 -0.033812"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pca = PCA(n_components=2)\n",
    "X_latente = pca.fit_transform(X)\n",
    "\n",
    "pd.DataFrame(X_latente, columns=['C1', 'C2']).sample(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<hr>\n",
    "\n",
    "\n",
    "# <a name=\"M5\">5. Creacción del modelo</a>\n",
    "\n",
    "\n",
    "* A continuación pasamos a crear un objeto de la clase \"***LogisticRegression()***\" para la resolución del problema de clasificación.\n",
    "\n",
    "\n",
    "* Una vez instanciado, procedemos a ajustar (entrenar) el modelo (o hipótesis) llamando al método ***fit(X,y)*** que recibe como parámetros el conjunto de datos de las variables de entrada 'X' y la variable de salida 'Y'.\n",
    "\n",
    "\n",
    "* En este caso las variables de entrada será nuestro Dataset reducido con 2 componentes:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LogisticRegression(max_iter=50, multi_class='ovr', solver='liblinear')"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Creamos un objeto de la clase LogisticRegression pasandole una serie de parámetros\n",
    "logreg = LogisticRegression(solver='liblinear', multi_class='ovr', max_iter=50)\n",
    "\n",
    "# Entrenamos el modelo\n",
    "logreg.fit(X_latente, y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Obtenemos las funciones de decisión:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Iris-setosa: Z = -1.86 + -2.22 · C1 + 1.27 · C2\n",
      "\n",
      "Iris-versicolor: Z = -0.81 + 0.24 · C1 + -1.59 · C2\n",
      "\n",
      "Iris-virginica: Z = -3.11 + 2.51 · C1 + -0.99 · C2\n"
     ]
    }
   ],
   "source": [
    "b0 = logreg.intercept_\n",
    "b = logreg.coef_\n",
    "\n",
    "print(\"\\nIris-setosa: Z = {b0:0.2f} + {b1:0.2f} · C1 + {b2:0.2f} · C2\"\n",
    "      .format(b0=b0[0], b1=b[0,0], b2=b[0,1]))\n",
    "print(\"\\nIris-versicolor: Z = {b0:0.2f} + {b1:0.2f} · C1 + {b2:0.2f} · C2\"\n",
    "      .format(b0=b0[1], b1=b[1,0], b2=b[1,1]))\n",
    "print(\"\\nIris-virginica: Z = {b0:0.2f} + {b1:0.2f} · C1 + {b2:0.2f} · C2\"\n",
    "      .format(b0=b0[2], b1=b[2,0], b2=b[2,1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<hr>\n",
    "\n",
    "\n",
    "# <a name=\"M6\">6. Evaluación del Modelo</a>\n",
    "\n",
    "\n",
    "* Vamos a continuación a ver los resultados de la evaluación de nuestro modelo con los mismos datos con los que los hemos entrenado:\n",
    "\n",
    "\n",
    "#### Acurracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy = 0.9133333333333333\n"
     ]
    }
   ],
   "source": [
    "accuracy = accuracy_score(y_true=y, y_pred=logreg.predict(X_latente))\n",
    "print(\"Accuracy = {acc}\".format(acc=accuracy))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Precision, Recall, F1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                 precision    recall  f1-score   support\n",
      "\n",
      "    Iris-setosa       1.00      1.00      1.00        50\n",
      "Iris-versicolor       0.97      0.76      0.85        50\n",
      " Iris-virginica       0.80      0.98      0.88        50\n",
      "\n",
      "       accuracy                           0.91       150\n",
      "      macro avg       0.93      0.91      0.91       150\n",
      "   weighted avg       0.93      0.91      0.91       150\n",
      "\n"
     ]
    }
   ],
   "source": [
    "classification = classification_report(y_true=y, y_pred=logreg.predict(X_latente), target_names =le.classes_)\n",
    "print(classification)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Matriz de Confusión"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAVMAAAEmCAYAAADfpHMGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3dd5hU5fn/8fdnd0GqBRVFLICxISpS7FHsCtYoNlSMJqhBjdGoxMQS/ZkvJhrFFsUGoihggiJGxKCI2AABAQs2UJoidmy0+/fH86yO6+zswszOmZm9X7nm2p1zzpy5ZyL3Pv2RmeGccy47ZUkH4JxzpcCTqXPO5YAnU+ecywFPps45lwOeTJ1zLgc8mTrnXA54MnWJkNRL0tgc3GeQpP+Xi5hySdJGkiZI+krS9Vnc51JJd6U5vqekSZLWyy5SlyueTN0PJM2VtEzSBlWOT5dkktrU4h5t4rUVma4zswfM7KDsIs6OgvMkzZL0taT5kkZI2iEHt+8DLAHWNrML1/QmZvY3M/tN6jFJmwF/Aw4zs8+yC9PliidTV9Uc4MTKJzGxNM7lG9SUaPNoAPB74DygBbA18AjQIwf33gJ43epgVoyZzTOzfcxsca7v7dacJ1NX1RDg1JTnvYH7Ui+Q1EPSNElfSpon6cqU0xPiz88lLZW0u6TTJD0v6QZJnwJXxmMT4/0ujtdWPpZLGpQuOEk7S5oaq8/DgEZVzh8WS9KfS3pB0o7V3GcroC9wopk9bWbfm9k3scTcP16zjqT7JH0s6X1Jf5FUFs+dJmmipOskfSZpjqRD47lB8Xur/FwHVG2OkNRN0vyU55dIWhA/12xJ+8fjV0q6P+W6IyS9Fj/feEnbpZybK+mPkmZI+kLSMEk/+X5c3fFk6qp6CVhb0naSyoHjgfurXPM1IeGuSyjFnS3pqHhu7/hzXTNrZmYvxue7Au8BLYFrUm9mZn+P1zYDtgM+BoZXDUxSQ0LJcQihJDkCOCblfCfgHuBMYH3gDmCUpLXSfM79gflmNinDd3EzsA7QDtgnfuZfp5zfFZgNbAD8HbhbkszsNOABoPJz/S/DeyBpG+AcoKuZNQcOBuamuW5r4EHgfGBD4L/AY/F7qXQccAjQFtgROC3Te7vc8WTq0qksnR4IvAksSD1pZuPNbKaZrTKzGYR/4PvUcM+FZnazma0ws2/TXSCpMSFZDjCz/6a5ZDegAXCjmS03s4eBySnnfwvcYWYvm9lKMxsMfB9fV9X6wKLqgk35Q/InM/vKzOYC1wOnpFz2vpndaWYrgcFAK2Cj6u6ZwUpgLaC9pAZmNtfM3k1z3fHA42b2lJktB64jNMHskXLNTWa20Mw+BR4DOq5BPG4NeDJ16QwBTiKUau6relLSrpKeidXfL4CzCKWzTObV4n3vBmab2bXVnN8EWFClHfL9lN+3AC6MVeDPJX0ObBZfV9UnhORXnQ2AhlXu/z7QOuX5h5W/mNk38ddmGe6Zlpm9QyhtXgkslvSQpHQxb5Iaj5mtInyvaWMCvlmTeNya8WTqfsbM3id0RHUH/pPmkqHAKGAzM1sHuB1Q5curu22m95TUD9gGOCPDZYuA1pKUcmzzlN/nAdeY2bopjyZm9mCae40DNpXUpZr3WgIsJyTo1PdakP7yGn0NNEl5vnHqSTMbamZ7xfczIN0flIWp8cTvYbMsYnI55MnUVecMYD8z+zrNuebAp2b2naRdCKXYSh8DqwjtjLUSO27OA46qrgkgehFYAZwnqULSr4BdUs7fCZwVS86S1DR2ljWveiMzexu4DXgwdgY1lNRI0gmS+sWq+3DgGknNJW0BXMDP249razrQXVILSRsTSqKVn38bSfvFtt3vgG8JVf+qhgM9JO0vqQFwIaEZ44U1jMnlkCdTl5aZvWtmU6o5/TvgKklfAZeT0lkUq7vXAM/Hqna69sqqjid0qLyR0qN/e5qYlgG/IjQ/fBZf95+U81MI7aa3xPPvkLkD5rx47a3A58C7wNGEtkaAcwklyveAiYQS+T21+DzpDAFeJXQsjQWGpZxbC+hPKA1/SOiku7TqDcxsNnAyoWNsCXA4cHj8XlzC5ItDO+dc9rxk6pxzOVAoM1Gccy4xkuYCXxHaqleYWRdJLQjNMW0IzTPHZZq+6yVT55wL9jWzjmZWOcKjHzDOzLYijP7ol+nFnkydcy69IwmTMYg/j8pwrXdAJUkNy4xG9bOlpdPWHZIOwSVg6ivTlpjZhrm6nzZoZCxblfmir5a/RhhyVmmgmQ38yX2kOYQRIEaYRTdQ0udmtm7KNZ+ZWbVLHtbPf8mFolEF7Noy6SgS8fyYiUmH4BLQuKLp+zVftRqWrar539D/FnyXUnWvzp5mtlBSS+ApSW+ubihezXfOFS8B5cr8qAUzWxh/LgZGEiaDfCSpFUD8mXHJQ0+mzrniJmV+1PhyNa2cJSepKXAQMIswZbp3vKw38Gim+3g13zlX3GpX+MxkI2BkXPKhAhhqZmMkTQaGSzoD+ADomekmnkydc8VLta/KV8fM3gN2SnP8E8K6t7XiydQ5V9xqUZXPB0+mzrniVhi51JOpc66IVfbmFwBPps654ubVfOecy5KXTJ1zLkcKI5d6MnXOFbPaDczPB0+mzrni5dV855zLES+ZOudcDhRGLvVk6pwrYl7Nd865HPFqvnPO5UCBLCTqydQ5V7wkKPOSqXPOZc+r+c45lwNezXfOuSwJr+Y751xOeDJ1zrksFVDJtEBaG1xdmzPkRWYM/B/Tbn+Sybc+DsB6zddlbP+hvDXoOcb2H8q6zdZJOMq6N3bMWHZs35Htt9mBf1x7XdLh5FVpfnYhZX7kiyfTemTfP/Zk57MOpmvfHgD0O74v46Y9z9an/ZJx056n3wl9E46wbq1cuZLzz7uAR0ePZNrMVxgxbARvvP5G0mHlRSl/dk+mLnFH7nEQg58aAcDgp0Zw1B4HJxxR3Zo8aQpbbtmOtu3a0rBhQ3oedyyjR41OOqy8KNXPLqC8TBkf+eLJtJ4wM8b2H8qUW//Lb7v3AmCj9Tbgw08XA/Dhp4tpue76SYZY5xYuXMimm236w/PWm7ZmwcJFCUaUPyX72VU4JdOC7YCStNTMmlVz7gUz26OO3vdSM/tbXdw7SXv+4WgWffIRG667Pk/1f5A3572TdEh5Z2Y/O5bPf2xJKuXPXiifo6hKppLKAeoqkUaX1uG9E7Pok48A+PjzTxj5/Bh22aYjH322hI1btARg4xYtWfz5J0mGWOdat27N/Hnzf3i+YP4CNmm1cYIR5U/pfnZRVpb5kS8Fn0wldZP0jKShwMx4bGn82UrSBEnTJc2S9Ms0r99e0qR4zQxJW8XjJ6ccv0NSuaT+QON47IF43QXx3rMknR+PNZX0uKRX4/Hj4/HLJU2OxwaqQP5kNmnUmGaNm/7w+0Gd92bW3NmMevEpeh/YE4DeB/bk0RfGJhlmnevStTPvvPMuc+fMZdmyZYwY/jA9Du+RdFh5UaqfXXg1f3XtAnQwszlVjp8EPGlm18RSa5M0rz0LGGBmD0hqCJRL2g44HtjTzJZLug3oZWb9JJ1jZh0BJHUGfg3sSvj/7WVJzwLtgIVm1iNeVzmm6BYzuyoeGwIcBjyWGoykPkAfABqVZ/GV1N5G627IyCvvAqCivJyhzzzCk1PGM3n2dIZfdjtnHHoCHyxeQM+rz8pLPEmpqKjghgHXc3j3I1m5ciW9TzuV9tu3TzqsvCjZz67CqeYrXVtKIahsM5XUDbjCzPZNc25v4B7gfuARM5ue5j4nAX8G7gP+Y2ZvSzqHUJ1fHC9rDDxoZlemttVK+j2wvpldHp9fDXwMjAGeBIYDo83suXj+GOBiQlJvAdxsZv2r/YxrNzR2bbmmX1FR+3bMW0mH4BLQuKLpK2bWJVf3K2/V1Jqe1iHjNV/1n5TT96xOwVfzo6/THTSzCcDewAJgiKRTJR0dq+nTJXUxs6HAEcC3wJOS9iOUMgebWcf42MbMrkzzFmn/5JnZW0BnQrPD/8XqfSPgNuBYM9sBuBNolNWnds7VSMr8yJdiSaZpSdoCWGxmdwJ3A53MbGRKkpwiqR3wnpndBIwCdgTGAcdKahnv0yLeC2C5pAbx9wnAUZKaSGoKHA08J2kT4Bszux+4DujEj4lziaRmwLF1/gU4V88JUabMj1rdJ/SZTJM0Oj5vIekpSW/Hn+vVdI9iaTOtTjfgIknLgaXAqWmuOR44OV7zIXCVmX0q6S/AWEllwHKgL/A+MBCYIWmqmfWSNAiYFO91l5lNk3Qw8A9Jq+JrzzazzyXdSSitzgUm181Hds79QFBWlpMy4e+BN4C14/N+wDgz6y+pX3x+ScZQCrXNtD7wNlNX3+S6zbRik2a27m92zHjNJ1e/mPE9JW0KDAauAS4ws8MkzQa6mdkiSa2A8Wa2TcZYVj9855wrDJVDo2qwgaQpKc8HmtnAlOc3EjqOm6cc28jMFgHEhFpjqceTqXOuiKk21fwl1ZVMJR1G6Hd5JY4cWmOeTJ1zxSv7caZ7AkdI6k7oRF5b0v3AR5JapVTzF2e8C0Xem++cq98EWU0nNbM/mdmmZtYGOAF42sxOJoz86R0v6w08WlMsXjJ1zhW1OpoB1R8YLukM4AOgZ00v8GTqnCtqtR1LWhMzGw+Mj79/Auy/Oq/3ZOqcK1pSfleGysSTqXOuqCn9rO+882TqnCtqhbJqlCdT51xR82q+c85lSQW0nqknU+dcEcvvavqZeDJ1zhU1r+Y751wOeMnUOeey5G2mzjmXI17Nd865rHkHlHPOZU2527Yka55MnXNFrUAKpp5MnXPFzav5zjmXNSGv5jvnXJbk1XznnMtaLXcnzQtPps65oua9+c45lwMFUjD1ZJqkTlt34PkxE5MOIxHbXnd40iEkZvjpVyYdQumQD9p3zrmsha2evZrvnHNZ85Kpc85ly4dGOedc9oS8mu+cc7ng1XznnMuWfD1T55zLDS+ZOudc9rya75xzWRKivNCr+ZJuBqy682Z2Xp1E5JxztSUoy7JkKqkRMAFYi5ATHzazKyS1AIYBbYC5wHFm9ll198lUMp2SVYTOOVfHcrRq1PfAfma2VFIDYKKkJ4BfAePMrL+kfkA/4JLqblJtMjWzwT8JWmpqZl9nG7VzzuVSRZbJ1MwMWBqfNogPA44EusXjg4HxZEimNY52lbS7pNeBN+LznSTdtqaBO+dcrlSWTDM9gA0kTUl59PnZfaRySdOBxcBTZvYysJGZLQKIP1tmiqU2HVA3AgcDo+JNX5W09+p8YOecqxuqTZvpEjPrkukCM1sJdJS0LjBSUofVjaRWvflmNq9Ku8TK1X0j55zLNQnKczg0ysw+lzQeOAT4SFIrM1skqRWh1Fqt2kxqnSdpD8AkNZT0R2KV3znnklYmZXzURNKGsUSKpMbAAcCbhNp473hZb+DRTPepTcn0LGAA0BpYADwJ9K3F65xzrs7loDe/FTBYUjmhgDnczEZLehEYLukM4AOgZ6ab1JhMzWwJ0CvbaJ1zLtdE9tV8M5sB7Jzm+CfA/rW9T21689tJekzSx5IWS3pUUrvVC9c55+pC5ip+tgP6V0dt2kyHAsMJReFNgBHAg3UZlHPO1YYE5WVlGR/5Upt3kpkNMbMV8XE/GaaZOudcPhVKyTTT3PwW8ddn4lSqhwhJ9Hjg8TzE5pxzGSk+CkGmDqhXCMmzMtYzU84ZcHVdBeWcc7UjKgp92xIza5vPQJxzbnVJRbaeaZxa1R5oVHnMzO6rq6Bc3Ro7Zix/vOBiVq5cyWmn9+aiS/6YdEh1Zq3yBgw76UbWqmhAeVk5T8yewA0TB9O+5ZZcc/D5rFXekBWrVnLZUwN4ddHspMPNuat/35/nn3qB9TZYjwcnhLWLbrryNiaOfYEGDSpo3aY1l93Uj+brNE840jWXz3bRTGozNOoK4Ob42Bf4O3BEHcfl6sjKlSs5/7wLeHT0SKbNfIURw0bwxuulO6Ht+5XLOemhCzn03j50v7cP+7Ttys6bbEe/bn0Y8PwQug86k39OHMSfuv1s7YuScNgJh3DjQ//4ybFd9unC0AmDeODZQWy+5aYMHnB/QtFlr3KcaaZHvtSmseFYwsDVD83s18BOhEVUXRGaPGkKW27Zjrbt2tKwYUN6Hncso0eNTjqsOvXN8u8AqCiroKKsgrDimtGsYRMA1l6rKR8t/STBCOvOzrt3ZO111/7Jsd323YWKilAp7dB5exYv/DiJ0HKm4HvzU3xrZqskrZC0NmGyvw/aL1ILFy5k0802/eF5601bM2lSaa8DXqYyRvf+F1us15ohUx9l+qI3+eu427jvuP5cuu+ZlKmMY+4/N+kwE/HYg//lgCP3SzqMLKhg2kxrUzKdEhcBuJPQwz8VmFTTiyQtzXDuhVpHWAckbSLp4TV87XhJGZfzKmShVPZThfIfY11ZZavoPuhMdr/teHZqtS1bb9CGkzseztXj/sUe/zqRq5++jWsPLd124+rce8N9lJeXc8ixByYdyhqrXDWqKKr5ZvY7M/vczG4HDgR6x+r+aosLCWBme6zJ69fg/dKWvM1soZkdm6cYyvPxPrXVunVr5s+b/8PzBfMXsEmrjROMKH++/P5rXpo3nX3adeWYHQ5izFvPAfD4m8+yU6ttE44uvx5/6Akmjn2Rq/51WdH/MS2Uan61yVRSp6oPoAVQEX+vFUndJD0jaSgwMx5bGn+2kjRB0nRJsyT9sspr15E0V1JZfN5E0jxJDSRtKWmMpFckPSdp23jNIEn/lPQMcK2kfeL9p0uaJqm5pDaSZsXryyVdJ2mmpBmSzo3H94/Xz5R0j6SftRNLOjGenyXp2pTjSyVdJellYPfaflf50KVrZ955513mzpnLsmXLGDH8YXoc3iPpsOpMi8brsPZaTQFYq6Ihe27RmXc/mcfipZ+w22Y7AbDHFjsz97MFSYaZVy8+/TL33TKU64b8H42aNKr5BQVMFE4yzdRmen2GcwasTkPLLkAHM5tT5fhJwJNmdk0swTX5yZuYfSHpVWAf4Bng8Hj9ckkDgbPM7G1JuwK3pcS0NXCAma2U9BjQ18yel9QM+K5KDH2AtsDOZrZCUguF3QoHAfub2VuS7gPOJuw6AISmAuBaoDPwGTBW0lFm9gjQFJhlZpdX/SLilgl9ADbbfLPafn85U1FRwQ0Drufw7keycuVKep92Ku23b5/3OPKlZbP1ub7HxZSpnDKJx998lqfffYkvv1vKFQf0paKsnO9XLONPY/6ZdKh14i9n/pWpz0/j80+/4LCdjqHPxb9m8IAHWLZsGef2vACADp3b0++6Ym3mEOUq/EH7++bwfSalSaQAk4F7FHYEfMTMpqe5ZhhhCuszwAnAbTEp7gGMSKmipJYcR8RtCACeB/4p6QHgP2Y2v0q15gDgdjNbAWBmn0raCZhjZm/FawYT1nC9MeV1XYHxZvYxQLz/3sAjhJ0I/p3uizCzgcBAgM5dOiWyxsEh3Q/hkO6HJPHWeffmx+/RY9BZPzs+ZcEsDh98dgIR5df/u+OKnx07otdhCURSdwqlmSJfKT3trqZmNoGQgBYAQySdKunolGp5F8Jq14fGtQI6A0/HuD83s44pj+3SvZ+Z9Qd+AzQGXqpsDkghfr5wS23+38l0zXcpydw5V0dCB1RZxke+JFo+lrQFsNjM7gTuBjqZ2ciUBDnFzJYSRg8MAEab2Uoz+xKYI6lnvI9iaTLde2xpZjPN7FpgClA1mY4FzqrsrIpJ+02gjaRfxGtOAZ6t8rqXgX0kbRCbKE5Mc41zro6p5t1J8yLpxoZuwHRJ04BjCAkznWHAyfFnpV7AGbFN9TXCHtfpnB87iF4FvgWeqHL+LsKWBDPiNSeZ2XfArwnNCDOBVcDtqS+KW7/+idD88Cow1cwy7hHjnMstFdDi0DUO2ldI7b2AdmZ2laTNgY3NLONYUzNrFn+OB8ZXc24woT0yIzN7mCrV6tgG+7OGPzM7rcrzdKOx5wId4vkVwAXxkfq6caTfyqBbyu9DCYtnV72mWTUfxTmXY+UFMvqwNiXT2wjDe06Mz78Cbq2ziJxzbjUUSjW/NtNJdzWzTrEqjpl9JqlhHcflnHM1UvxfIahNMl0eO1gMwh7ThDZE55xLVtwDqhDUJpneBIwEWkq6hrCK1F/qNCrnnKsFAWWJ96MHNSZTM3tA0iuEZfgEHGVmpbsApnOuiBTOqlG16c3fHPgGeCz1mJl9UJeBOedcbRT8dNIUj/PjxnqNCPPYZwPb12FczjlXI1E400lrU83fIfV5XDHqzGoud865/FHh7AFVqw31UpnZVEld6yIY55xbHUIFM2i/Nm2mqTODyoBOQHFvGuOcKxnZVvMlbQbcB2xMGPY50MwGxHU6hgFtCLMmjzOzz6q7T21abpunPNYitKFWNw/eOefySjX8rxZWABfGled2A/pKag/0A8aZ2VbAuPi8WhlLpnGwfjMzu6g2ETnnXD4pB4tDx0WLFsXfv5L0BtCaUGjsFi8bTFhj5JLq7lNtMpVUEVeer/UWJc45l2+57ICS1IawwNHLwEYx0WJmiyS1zPTaTCXTSYT20emSRgEj+Omiy//JLmznnMuSRFnNHVAbSErdz3xg3PGiyq3UjLBDxvlm9uXqtsXWpje/BfAJYX+lyvGmBngydc4lqnJDvRosMbOM27PHrZP+DTyQUlD8SFKrWCptBSzOdI9MybRl7MmfxY9JtFIiexc551xVOejNF2GnjzfMLHVnxVFAb6B//Jlx8fdMybQcaEb6vY48mTrnEidyMp10T8LWRDMlVW7qeSkhiQ6XdAZhN46emW6SKZkuMrOrso3SOefqjlD2vfkTqX6DzP1re59MybQw5mg551wGZQWSqjIl01pnZOecS4IEZYW+apSZfZrPQJxzbvUV0XqmzjlXyIqhmu+ccwVN1GrQfl54MnXOFTWv5jvnXA4U01bPzuXcaxeMTDqExDQ7qkPSIZSMHA3azwlPps654iXvzXfOuZzwDijnnMuS8DZT55zLARXv7qTOOVdICn46qXPOFbpQzfdk6pxzWfLefOecy558nKlzzmXNe/Odcy5HvJrvnHNZE+U+aN8557Lj1XznnMsRr+Y751zW5IP2nXMuWwLKfNC+c85lSV4ydc65nPA2U+ecywHvzXfOuSwJkFfznXMuW/KSqXPO5UKhtJkWRvnYOefWUFkN/6uJpHskLZY0K+VYC0lPSXo7/lyv5jicc65IhTZTZXzUwiDgkCrH+gHjzGwrYFx8npEnU+dcEVON/6uJmU0APq1y+EhgcPx9MHBUTffxZFoPjR0zlh3bd2T7bXbgH9del3Q4eXX2b/vSpvWWdO24W9Kh5E1ZWRlTb/ovj11xLwA7tt2OF64byYxbxzLq8nto3rhZwhFmQWEPqEwPYANJU1IefWpx543MbBFA/Nmyphd4Mq1nVq5cyfnnXcCjo0cybeYrjBg2gjdefyPpsPKm16kn8cjofycdRl79/ojTeWPeOz88v+u8v9NvUH927HsQI18cw0XHnJlgdNmrRcl0iZl1SXkMrIs4PJnWM5MnTWHLLdvRtl1bGjZsSM/jjmX0qNFJh5U3e/1yT9Zbr8a+hJLRev2N6dF1f+568qEfjm2zaTsmzHoZgKemPccxe3ZPKrys5ajNNJ2PJLUi3L8VsLimF3gyrWcWLlzIpptt+sPz1pu2ZsHCRQlG5OrSjX2u5OJ7/8YqW/XDsVnvz+aI3Q4EoOdePdhsg1ZJhZcDyro3vxqjgN7x997AozW9IO/JVNLSDOdeyMH9r5J0wGq+5ghJGXvrJG0i6eHsokuemf3sWKGM03O51aPr/iz+YglT35n5k+On33gRfXv0ZsqAx2neuBnLVixPKMLcyLZkKulB4EVgG0nzJZ0B9AcOlPQ2cGB8nlFBDNqXVG5mK81sj2zvZWaXZ3qPal4zivCXKNN9FwLHZhtf0lq3bs38efN/eL5g/gI2abVxghG5urJn+y4cseuBdO+yL40arsXajZsz5I83csp153PwZScDsNUmbenRdb+EI11zuViCz8xOrObU/qtzn8Sq+ZK6SXpG0lBgZjy2NP5sJWmCpOmSZkn6ZZXXriNpruKkXElNJM2T1EDSIEnHxuNzJV0uaSLQU1J3SW9KmijpJkmj43WnSbol/j4onntB0nsp92pTOahXUrmk6yTNlDRD0rnx+OWSJseYB6oAi3xdunbmnXfeZe6cuSxbtowRwx+mx+E9kg7L1YFLB1/LZr13pe3pe3LCtefw9IwXOOW689lwnfWBUKL7ywnncfsT9yccaTYyl0rz+U8w6TbTXYA/m1n7KsdPAp40s47ATsD01JNm9gXwKrBPPHR4vD5dfeU7M9sLeAS4Azg0Pt8wQ1ytgL2Aw0hfvO8DtAV2NrMdgQfi8VvMrKuZdQAax9f/hKQ+lUM0Pv54SYYQ6kZFRQU3DLiew7sfSccOnTjm2GNov33Vr790nXby6ey394G8/dbbbN12Owbfe1/SIeXdifscyeyB43nzjmdY+OlH3PvU8KRDypJqeORH0tX8SWY2J83xycA9khoAj5jZ9DTXDAOOB54BTgBuq+Y9hsWf2wLvpbzfg4SkmM4jZrYKeF3SRmnOHwDcbmYrAMyscsDvvpIuBpoALYDXgMdSXxiHZQwE6Nyl088bMPPgkO6HcEj3qhM+6odB99+TdAiJeHbmSzw78yUAbhp1DzeNKpHvIY4zLQRJR/F1uoNxRsLewAJgiKRTJR0dq/3TJXUhtHEeKqkF0Bl4uob3WJ0/Ud+n/J7udQJ+kgglNSIk9GPNbAfgTqDRarync24NZDsDKleSTqZpSdoCWGxmdwJ3A53MbKSZdYyPKWa2FJgEDABGV9e5lOJNoJ2kNvH58VmEOBY4S1JFjLcFPybOJZKaUQKdVc4VOhVQm2nS1fzqdAMukrQcWAqcWs11w4AR8fqMzOxbSb8DxkhaQkjEa+ouYGtgRozxTjO7RdKdhM60uYSmCudcHSuUDfWUbtxhqZLUzMyWxl72W4G3zeyGpOLp3KWTPf/yxKTePlErV61IOoTENDuqQ9IhJOe/814xsy65ut2OnXawx5/LPJ5+82Zb5uRoFVgAABNvSURBVPQ9q1MYKT1/fitpOqFjaB1C775zrogVSptpoVbz60QshSZWEnXO5ZbwrZ6dcy5HCmNujCdT51xRK4xU6snUOVfkfKtn55zLAS+ZOudc1vI7/z4TT6bOuaIlFc56vJ5MnXNFLZ9jSTPxZOqcK2qeTJ1zLgcKpZpfGGMKnHOuyHnJ1DlXxIQKpEzoydQ5V7QKZ2CUJ1PnXJErlDZTT6bOuaLmvfnOOZcTnkydcy5L+d3nKRNPps65ohU6oDyZOudcDngydc65rBVGKvVk6pwraiqYxaELIwrnnFtDquFRq3tIh0iaLekdSf3WJA4vmTrnipey37ZEUjlwK3AgMB+YLGmUmb2+Ovfxkqlzrr7bBXjHzN4zs2XAQ8CRq3sTL5kmaOor05Y0rmj6foIhbAAsSfD9k+SfPRlb5PJm016Z9mSTimYb1HBZI0lTUp4PNLOBKc9bA/NSns8Hdl3dWDyZJsjMNkzy/SVNMbMuScaQFP/spfHZzeyQHNwmXdOqre5NvJrvnKvv5gObpTzfFFi4ujfxZOqcq+8mA1tJaiupIXACMGp1b+LV/PptYM2XlCz/7A4AM1sh6RzgSaAcuMfMXlvd+8hstZsGnHPOVeHVfOecywFPps45lwOeTJ1zLgc8mTqXBaWsTBynJbp6ypOpqzVVWdK86vP6RlKZxR5cSScDu9eH76Q+fMY14cnU1YokpSSObSWtBayVcFiJMrNVAJL2Ag4B3rQSHx5T+d+BpG6SLpO0n6TWScdVCDyZulpJSaR/AG4DbgJOl7RJooElSFKZpJ2BCcAcM1siqVHScdWlmEi7A7cAnwBXAedI6pBsZMnzZOpqTdIJwBHAAYTFIY4HfiOpVaKB5VFqFdfMVpnZNOBPwAWStjCz70q57VTSRsBJwOHA28CGQFPgVEnbJhlb0nwGlKtWatU+WgWcCpwDNASuAy4Gmkq6xczmpblNyajS1HEY0A6YambXSvqOsA7m7mb2rqRyM1uZaMA5klK1b2lmH0n6E9AEuAbYE9gBuDtee7WZfZVguInxZOrSqpI4DgTmmtlwSc2B3YDDzex7Sb8m/Hf0bYLh5kWVpo6TgReAXST1BP5IWH3oTUlbmdncxALNsZhIewC9JP3ZzObEduIvY9PGYuAVwjTMeplIwZOpq0ZK4jifkDiOi6e+Iaywc7OkccA6wD/NrGTXBpW0BfCZmX0pqQlhMeHusZS2I9ATONXMbowdcw2TjDfXJHUF/gGcYWZzAMxsoqQW8b+BzYHzzOzNJONMmreZumpJ2gfoBexpZu9J2gXoAhxNaCc7CfiDmS1IMMw6o6AZcC/QN/7+PbAJcAqAmc0A5gL7xufXmtlbyUScG3H1pCNSDnUEnjKzF2OnW4N4fDfgBuBEM3si74EWGC+Zuh+kaSOdB7wK/FPSt0BX4DvgdkJptbGZfZP/SPMjfhdLY7X+WmAF8E9CD/axknqZ2QPAUqBcUpMS+T7WBz6U1MLMPiX8d7CnpPWBT81seazmNzaz0YlGWkC8ZOqAn7WRbiXpF8BXwGigETAMOAx4BljfglJIHGlV9trH7+VVQkfbQcDvgfeBZ4ELJY0A/gpcUyrfh5lNAV4DnpV0BmHo17qE0Ru/jDWUO4Gvk4uy8PgSfO4nJF0MdAfWA0YC481sfDzXC7gQOKmU28eq/GHZFfgC+ICwd9JdhD8wdxMmLWwOfGRmixIKN2eq1kwkHUrosb8CmEr4/34LoBlwm5k9mkigBcqr+fVclcSxM6E9dE9gS6AHcLCk94FWQG+gdyknUvhZr/1xwGJCx9u9wK8JCbUFcK2ZTU8qzlxKGf60K7ARYbfOJyStBK4HLjezCyRVAC3MbHGaZqF6zav59ZikZimJozmh2iagwszeBh4hJNb2wMtATzObmVS8+SRpK+BEwl7qfYBBQF9Cx9vFwE5A46Tiy6XKNQYk7Qv8hzA19nlJe5vZWOAC4DpJfcxshZkthh//6LjAS6b1lMJeN+dKmk3YqmF/4FJgOnCKpIfNbK6k54FWcQD6F8lFXLeqlNDLCJ1NK8xsKaETahJwKLCrmQ2RdIKZfZ9gyFmTtLaZfWlmqyS1Jwzx6mVm4yW9DDwm6TAze0rS2dSDscTZ8GRaT5nZMklDgfcIuzO2jf+oJgI7A/dJeo6wudiBCYZa56ok0uOA5WY2UtJ8STeb2blm9lkc0fCL+LLliQWcA5LWJvwxvQf4kFAK3w2YLukFMxsc++CelbRvLKGmG/HhIq/m1zMpvdTlhCEvDwLNCe2hAEMJQ59GAw2AQ83snQRCrXOV30XK83MIpfM34qFLCVNlJ0q6iNCefD/8uGJUETNCJ1o54XNdAfwb6AB0idNhBwO/IWV1ME+k1fPe/HqkSglsD8Jg868IyXQGcJWZ3STpIOBVM/sosWDzQFI7M3sv/t6a0LHUJ3WNgVjlP4swvvaFUuh8i22klcsHnkJYtOQBwh/QvxLahR8BJlauL+Al0pp5Nb8eSUmkFxL+Ab1GGKDdhzCG8llJHQnVve5JxZkPsZp7paRzzewLYBGhx/4XkhaZ2Yp46bbAwJTnRS8253QmjBceEnvsjyJ0Pl4B9CeMYpgBfBZf44m0Bl7Nr2ckdQIONLNuhCpeOfCdmU0lTBucBhxRSgt1pGNmXxKqsDtL+mssqb0HdAY2hh+WHLyMMK6y6KU08XQCzgT6x/bQocA4wqSMo4F+wAAz+yyxYIuQl0xLXDXVs7ck/Zkw4PxXsTPqAGCcmd2c/yjzJ/X7iJ/7W+BoSR8QBqjfCnSNpbUdCBMUPk8u4tyJw58OJSxacguwErhIUmMzuz/OuT8CeK7Y1xdIgifTElaljfRgYArwLtAG2AvYw8JixmcRpgpOpv4MfzobeMvMxknqTZgeaWbWS9JOhO+on5l9kFzEuaWwotWOwNVmNkxSS8JIjYslfW1m90p6stTbyuuKd0DVA5L6Ar8jLBv3fux06ESYHjmdsOBzLzOblWCYeaMflxU81cxej8c6E7ZiecLM/l+S8eVSysym/QhTQXcCtjGzQ+P5rYABhDGk/xfn5bs14G2mJU5hdZ/fAN1iIt0BmE0YEjWdsKTc8aWcSFOHQEnagDD4/ihgrqRfSfoNsICwiMm+ktavOmyq2FTGHxPp1oTlEl8ws/OBxZJuiJc2Aj4mjDX9RdqbuVrxkmmJU9jo7ERCT20DQiJ5DfhX5QImpaxK1f4EwndwGGEzuBaEwfdbAfea2R2SGpnZd4kFnAMKmxx2AMYTVnuaBkwkjCVeQfi8fyd0rLUmfB/HAE3M7LIEQi4JXjItfR8CnxPGkj5M6LH/kDDfvuSlJNJuhGr9EEL76HTgr2Z2CjCEsP1IBaGkXux2JkzIaBzn0Z9DWIt2nzi3/g0zOxz4LWHthcpFbB5IKuBS4B1QJSJdr72kBhb26LnNzL6Ox44h/AO6NYk4k6CwY8A5wOsAZva/lHOnEQblH1cqY0nN7HFJjYHBkh6PU0MbALdJ+q2ZPR2ve0dSG8Kc/GNLYUJCkrxkWgKqVGXbKu7dbmFF9G7AZZKaxx79M4Bfl/LQlzTtnW8RBuW3i5MSKq/bDtiH0Gb8BiUi/vFYjzA1+ABJJ5rZcOAS4KE4DA6AOJ74klJuM88XL5kWuSqJ9EJgb8KMpu8kbUkYT3iFmX2lsALUFDP7JLmI61aV7+MkQhvht4TOpZsI242sMrMZZvaGpL5WIivkp9gLON/MjpbUFOgRP/Ow2JTxkymilbUWlx3vgCoRcaxkH+AwCyscrU+YCjnPzD5QCe3jXhsKi5acAPyNsEbnQYQZTpcQksndViJrs1Zt4lFYZ6AfcIOFjRB7Ewbj/zvOdvK59nXAq/lFSlJHhSX0KjUjLE6xl6RL4+8nE0pmlHoiTZkqWR4Ho+9NWKN1W8JUyZfMbD5wNaEH/8OkYs21OPxpH0m/k7SWhd1ivyPM6MLC6k+PEduMK1+TTLSly0umRUxSK2B7QrLoApxNGCs4gLBoR3fgOjN7P7Eg80xxR01JgwmLdGwOnGxm38TJC2OAOVb8S+j9hMImdwMIQ6AWEpp3hgM3V3Y4ubrlJdMioyg+XUIY0jLNzCab2emEWU7/Joyn3IvYPlbq4tfSEZgmqQlhweu+/JhITyJ0vi0rwURaZmaTgIMJJdBdCcvpbUnoYHN54B1QRSJ1Rkt8fjawn5n1lDRM0lSgs5ktlXQyoc3shFi1LUkpUyUr1+ecLulh4GAz+7OkzQjLCk4mjLM81VLWKi12Ke2elX9cVwGzzewESUcTkusriQVYz3g1v0hIqqgcBympO3A6ocd2fjz2INCOsBZpO8LWGyWzSEcmkrarHNoUx40eZGYnxeeHAp8StmOem1iQOZLyB2QfQu3jOTP7XtKmwP+AC83s8XhtuZmt9M6m/PBqfhFQmE/+jqQW8VBXwrTQTSuvMbMTCVMknzKzd+tRIt0I+J+kqyTtZ2aDgOaSLgEwsyfM7OUSSaQVMZEeQthy5PuYSMsJq37dXplI4cdOR0+k+eEl0yIh6XDCOpS7mtkXkv6P0Pl0aeqAa0mtY29uvSGpHWHhkp0IW26MI/zB6WtmRb+jpqS2ZjYn/t4SeBz4g5lNlLQbYX79JxbXWkhp9nB55Mm0iMTq/U2EttEv4hCoToT1KV9NNrq6l666qjBldrmkZrG9+CpCIt0daGdmnyYSbA5J+hsw0swmx+f9CbsBGGEZxW+B183sSq/SJ8eTaZGJCXUAYSjUl4Rxk22A081sWYKh1amqU2aBRRZXd4pTZrsTFnNeFav+ZmGRj6JV5TO3BMabWXtJOwM9gGfN7DmF7amPAHqX+njiQua9+UXGzP4bO/ZfAnY3s79IWr8eJdJMU2ZXAViJrBSf8pn3IyynN1vSROCXZjYtntsL+BPwF0+kyfKSaZGSdBRhs7cu9aVaV4spsyVTxa0cvaGwA8BA4DQzmynpIWAbQvPOusC1wGgzG1VKn78YeTItYpXthEnHUVfiIPyLU4Y59QWaAG8SNrs7FJhFaDNemFigORSbMD6NbeJtgcHAk2Z2Tco19wOdYpW/sq3YE2nCfGhUESvlRApgZtOBCyUdECctTAK2Ay4ibL3yN8LaAw2SizLntgTel7Qu8AFhlfzTFXZMAMDMTgbekLRn5X8DnkiT5yVTV3BSZ3spLGp8D7CDmXWM5ytLY0cAfwUOL6WZXnEc6c2EJpwvJP0F6ExoF30t2ehcdbxk6gpG5boDFsUps0MtbC0yW9LUeL5yyuzfgFNKKZECmNkYwvqrr0hah7D608vADaklVFdYvDffFZLyKlNm9wfOBzCz4+OU2ZfiQPUXCYu6lORMr5RRG1MIw+CuJTRnrJVkXK56Xs13BSFOmZ1C6Fj5VNIVwMXA/mb2Usp1/wUamtkB1dyqpMQq/73Atmb2RdLxuOp5MnUFw6fMpiepB/C11YOtuYuZJ1NXUOr7lNlMfPhTYfNk6gpOfZ0y64qbJ1NXkGJCvZ4wZfbzOGW2ZHdVdcXPe/NdQYq92Q2BcZK6eCJ1hc5Lpq6glfqUWVc6PJk651wO+Awo55zLAU+mzjmXA55MnXMuBzyZOudcDngydXkhaaWk6ZJmSRohqUkW9xok6dj4+12S2me4tpukPdbgPebG9QJqdbzKNas1+kDSlZL+uLoxusLiydTly7dm1tHMOgDLgLNST8a931ebmf3GzF7PcEk3YLWTqXOry5OpS8JzwC9iqfEZSUOBmZLKJf1D0mRJMySdCT+sc3qLpNclPQ60rLyRpPGSusTfD4lrnr4qaZykNoSk/YdYKv6lpA0l/Tu+x2RJe8bXri9prKRpku4AVNOHkPSIpFckvSapT5Vz18dYxknaMB7bUtKY+JrnJG2biy/TFQafAeXySlIFYe+mMfHQLkAHM5sTE9IXZtZV0lrA85LGAjsTNpHbAdgIeJ2w+n7qfTcE7gT2jvdqEZfyux1YambXxeuGAjeY2URJmwNPErZCuQKYaGZXxVWafpIcq3F6fI/GwGRJ/44ztZoCU83sQkmXx3ufQ9gY7ywze1vSrsBtwH5r8DW6AuTJ1OVLY0nT4+/PAXcTqt+TzGxOPH4QsGNleyiwDrAVYWvnB+NWxgslPZ3m/rsBEyrvZWafVhPHAUD7uPAywNqSmsf3+FV87eOSPqvFZzpP0tHx981irJ8Aq4Bh8fj9wH8kNYufd0TKe/tCzyXEk6nLl28r93CqFJPK16mHgHPN7Mkq13UHapqqp1pcA6Fpa3cz+zZNLLWeDiipGyEx725m30gaDzSq5nKL7/t51e/AlQ5vM3WF5EngbIVN9JC0taSmwATghNim2grYN81rXwT2UdgeGUkt4vGvgOYp140lVLmJ11UmtwlAr3jsUGC9GmJdB/gsJtJtCSXjSmVAZen6JELzwZfAHEk943tI0k41vIcrIp5MXSG5i9AeOlXSLOAOQu1pJPA2MBP4F/Bs1Rea2ceEds7/SHqVH6vZjwFHV3ZAAecBXWIH1+v8OKrgr8DekqYSmhtq2ltqDFAhaQZhvdWXUs59DWwv6RVCm+hV8Xgv4IwY32vAkbX4TlyR8IVOnHMuB7xk6pxzOeDJ1DnncsCTqXPO5YAnU+ecywFPps45lwOeTJ1zLgc8mTrnXA78fxmDg9usnULnAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_confusion_matrix(cm, classes, title, cmap=plt.cm.Greens):\n",
    "    \"\"\"\n",
    "    This function prints and plots the confusion matrix.\n",
    "    \"\"\"\n",
    "    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    plt.title(title)\n",
    "    plt.colorbar()\n",
    "    tick_marks = np.arange(len(classes))\n",
    "    plt.xticks(tick_marks, classes, rotation=45)\n",
    "    plt.yticks(tick_marks, classes)\n",
    "\n",
    "    thresh = cm.max() / 2.\n",
    "    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "        plt.text(j, i, format(cm[i, j], 'd'), horizontalalignment=\"center\",\n",
    "                 color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "\n",
    "    plt.tight_layout()\n",
    "    plt.ylabel('True label')\n",
    "    plt.xlabel('Predicted label')\n",
    "    \n",
    "\n",
    "confusion_mx = confusion_matrix(y_true=y,y_pred=logreg.predict(X_latente))\n",
    "plt.figure()\n",
    "plot_confusion_matrix(confusion_mx, classes=le.classes_, title='Matriz de Confusión')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<hr>\n",
    "\n",
    "\n",
    "# <a name=\"M7\">7. Visualización de Resultados</a>\n",
    "\n",
    "\n",
    "* Dado que estamos trabajando con 2 Componentes, podemos representar en un espacio de 2 dimensiones como clasificaría nuestro modelo en ese espacio 2D en función del valor de las componentes de cada flor:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "x_min, x_max = X_latente[:, 0].min() - .5, X_latente[:, 0].max() + .5\n",
    "y_min, y_max = X_latente[:, 1].min() - .5, X_latente[:, 1].max() + .5\n",
    "h = .01\n",
    "xx, yy = np.meshgrid(np.arange(x_min, x_max, h), np.arange(y_min, y_max, h))\n",
    "\n",
    "# Predicción de cada uno de los puntos de la malla\n",
    "Z = logreg.predict(np.c_[xx.ravel(), yy.ravel()])\n",
    "\n",
    "Z = Z.reshape(xx.shape)\n",
    "plt.figure(1, figsize=(8, 6))\n",
    "\n",
    "# Dibujamos la malla con los diferentes colores\n",
    "plt.pcolormesh(xx, yy, Z, cmap=plt.cm.Pastel2)\n",
    "\n",
    "\n",
    "# Dibujamos los puntos (las componentes de las flores) del dataset\n",
    "plt.scatter(X_latente[:, 0], X_latente[:, 1], c=y, edgecolors='k', cmap=plt.cm.Pastel2)\n",
    "\n",
    "plt.title('Clasificación de las Flores')\n",
    "plt.xlabel('Componente 1')\n",
    "plt.ylabel('Componente 2')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<hr>\n",
    "\n",
    "*Este Notebook ha sido desarrollado por **Ricardo Moya García** y registrado en Safe Creative como ***Atribución-NoComercial-CompartirIgual***.*\n",
    "\n",
    "<img src=\"./imgs/CC_BY-NC-SA.png\" alt=\"CC BY-NC\">"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}