{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n",
"
\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "1HEGrQDdckKM"
},
"outputs": [],
"source": [
"import sklearn \n",
"import numpy as np\n",
"import pandas as pd\n",
"import matplotlib.pylab as plt\n",
"import seaborn as sns\n",
"from scipy import stats\n",
"from scipy import optimize\n",
"from scipy.stats import norm\n",
"from matplotlib import cm\n",
"from scipy.stats import multivariate_normal\n",
"from sklearn.model_selection import train_test_split\n",
"#Machinig learning models\n",
"from sklearn.linear_model import LogisticRegression\n",
"from sklearn.svm import SVC\n",
"from sklearn.tree import DecisionTreeClassifier\n",
"from sklearn.ensemble import RandomForestClassifier\n",
"from sklearn.model_selection import ShuffleSplit"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "Uc6frHmHdvsL"
},
"outputs": [],
"source": [
"def make_meshgrid(x, y, h=0.02):\n",
" \"\"\"Create a mesh of points to plot in\n",
"\n",
" Parameters\n",
" ----------\n",
" x: data to base x-axis meshgrid on\n",
" y: data to base y-axis meshgrid on\n",
" h: stepsize for meshgrid, optional\n",
"\n",
" Returns\n",
" -------\n",
" xx, yy : ndarray\n",
" \"\"\"\n",
" x_min, x_max = x.min() - 1, x.max() + 1\n",
" y_min, y_max = y.min() - 1, y.max() + 1\n",
" xx, yy = np.meshgrid(np.arange(x_min, x_max, h), np.arange(y_min, y_max, h))\n",
" return xx, yy\n",
"\n",
"\n",
"def plot_contours(ax, clf, xx, yy, **params):\n",
" \"\"\"Plot the decision boundaries for a classifier.\n",
"\n",
" Parameters\n",
" ----------\n",
" ax: matplotlib axes object\n",
" clf: a classifier\n",
" xx: meshgrid ndarray\n",
" yy: meshgrid ndarray\n",
" params: dictionary of params to pass to contourf, optional\n",
" \"\"\"\n",
" Z = clf.predict(np.c_[xx.ravel(), yy.ravel()])\n",
" Z = Z.reshape(xx.shape)\n",
" out = ax.contourf(xx, yy, Z, **params)\n",
" return out\n",
"\n",
"\n",
"def plot_contoursExact(ax, xx, yy, **params):\n",
" \"\"\"Plot the decision boundaries for a classifier.\n",
"\n",
" Parameters\n",
" ----------\n",
" ax: matplotlib axes object\n",
" clf: a classifier\n",
" xx: meshgrid ndarray\n",
" yy: meshgrid ndarray\n",
" params: dictionary of params to pass to contourf, optional\n",
" \"\"\"\n",
" Z = clf.predict(np.c_[xx.ravel(), yy.ravel()])\n",
" Z = Z.reshape(xx.shape)\n",
" out = ax.contourf(xx, yy, Z, **params)\n",
" return out\n",
"\n",
"\n",
"import numpy as np\n",
"import matplotlib.pyplot as plt\n",
"from sklearn.naive_bayes import GaussianNB\n",
"from sklearn.svm import SVC\n",
"from sklearn.datasets import load_digits\n",
"from sklearn.model_selection import learning_curve\n",
"from sklearn.model_selection import ShuffleSplit\n",
"\n",
"# https://scikit-learn.org/stable/auto_examples/model_selection/plot_learning_curve.html\n",
"def plot_learning_curve(\n",
" estimator,\n",
" title,\n",
" X,\n",
" y,\n",
" axes=None,\n",
" ylim=None,\n",
" cv=None,\n",
" n_jobs=None,\n",
" train_sizes=np.linspace(0.1, 1.0, 5),\n",
"):\n",
" \"\"\"\n",
" Generate 3 plots: the test and training learning curve, the training\n",
" samples vs fit times curve, the fit times vs score curve.\n",
"\n",
" Parameters\n",
" ----------\n",
" estimator : estimator instance\n",
" An estimator instance implementing `fit` and `predict` methods which\n",
" will be cloned for each validation.\n",
"\n",
" title : str\n",
" Title for the chart.\n",
"\n",
" X : array-like of shape (n_samples, n_features)\n",
" Training vector, where ``n_samples`` is the number of samples and\n",
" ``n_features`` is the number of features.\n",
"\n",
" y : array-like of shape (n_samples) or (n_samples, n_features)\n",
" Target relative to ``X`` for classification or regression;\n",
" None for unsupervised learning.\n",
"\n",
" axes : array-like of shape (3,), default=None\n",
" Axes to use for plotting the curves.\n",
"\n",
" ylim : tuple of shape (2,), default=None\n",
" Defines minimum and maximum y-values plotted, e.g. (ymin, ymax).\n",
"\n",
" cv : int, cross-validation generator or an iterable, default=None\n",
" Determines the cross-validation splitting strategy.\n",
" Possible inputs for cv are:\n",
"\n",
" - None, to use the default 5-fold cross-validation,\n",
" - integer, to specify the number of folds.\n",
" - :term:`CV splitter`,\n",
" - An iterable yielding (train, test) splits as arrays of indices.\n",
"\n",
" For integer/None inputs, if ``y`` is binary or multiclass,\n",
" :class:`StratifiedKFold` used. If the estimator is not a classifier\n",
" or if ``y`` is neither binary nor multiclass, :class:`KFold` is used.\n",
"\n",
" Refer :ref:`User Guide ` for the various\n",
" cross-validators that can be used here.\n",
"\n",
" n_jobs : int or None, default=None\n",
" Number of jobs to run in parallel.\n",
" ``None`` means 1 unless in a :obj:`joblib.parallel_backend` context.\n",
" ``-1`` means using all processors. See :term:`Glossary `\n",
" for more details.\n",
"\n",
" train_sizes : array-like of shape (n_ticks,)\n",
" Relative or absolute numbers of training examples that will be used to\n",
" generate the learning curve. If the ``dtype`` is float, it is regarded\n",
" as a fraction of the maximum size of the training set (that is\n",
" determined by the selected validation method), i.e. it has to be within\n",
" (0, 1]. Otherwise it is interpreted as absolute sizes of the training\n",
" sets. Note that for classification the number of samples usually have\n",
" to be big enough to contain at least one sample from each class.\n",
" (default: np.linspace(0.1, 1.0, 5))\n",
" \"\"\"\n",
" if axes is None:\n",
" _, axes = plt.subplots(1, 3, figsize=(20, 5))\n",
"\n",
" axes[0].set_title(title)\n",
" if ylim is not None:\n",
" axes[0].set_ylim(*ylim)\n",
" axes[0].set_xlabel(\"Training examples\")\n",
" axes[0].set_ylabel(\"Score\")\n",
"\n",
" train_sizes, train_scores, test_scores, fit_times, _ = learning_curve(\n",
" estimator,\n",
" X,\n",
" y,\n",
" cv=cv,\n",
" n_jobs=n_jobs,\n",
" train_sizes=train_sizes,\n",
" return_times=True,\n",
" )\n",
" train_scores_mean = np.mean(train_scores, axis=1)\n",
" train_scores_std = np.std(train_scores, axis=1)\n",
" test_scores_mean = np.mean(test_scores, axis=1)\n",
" test_scores_std = np.std(test_scores, axis=1)\n",
" fit_times_mean = np.mean(fit_times, axis=1)\n",
" fit_times_std = np.std(fit_times, axis=1)\n",
"\n",
" # Plot learning curve\n",
" axes[0].grid()\n",
" axes[0].fill_between(\n",
" train_sizes,\n",
" train_scores_mean - train_scores_std,\n",
" train_scores_mean + train_scores_std,\n",
" alpha=0.1,\n",
" color=\"r\",\n",
" )\n",
" axes[0].fill_between(\n",
" train_sizes,\n",
" test_scores_mean - test_scores_std,\n",
" test_scores_mean + test_scores_std,\n",
" alpha=0.1,\n",
" color=\"g\",\n",
" )\n",
" axes[0].plot(\n",
" train_sizes, train_scores_mean, \"o-\", color=\"r\", label=\"Training score\"\n",
" )\n",
" axes[0].plot(\n",
" train_sizes, test_scores_mean, \"o-\", color=\"g\", label=\"Cross-validation score\"\n",
" )\n",
" axes[0].legend(loc=\"best\")\n",
"\n",
" # Plot n_samples vs fit_times\n",
" axes[1].grid()\n",
" axes[1].plot(train_sizes, fit_times_mean, \"o-\")\n",
" axes[1].fill_between(\n",
" train_sizes,\n",
" fit_times_mean - fit_times_std,\n",
" fit_times_mean + fit_times_std,\n",
" alpha=0.1,\n",
" )\n",
" axes[1].set_xlabel(\"Training examples\")\n",
" axes[1].set_ylabel(\"fit_times\")\n",
" axes[1].set_title(\"Scalability of the model\")\n",
"\n",
" # Plot fit_time vs score\n",
" axes[2].grid()\n",
" axes[2].plot(fit_times_mean, test_scores_mean, \"o-\")\n",
" axes[2].fill_between(\n",
" fit_times_mean,\n",
" test_scores_mean - test_scores_std,\n",
" test_scores_mean + test_scores_std,\n",
" alpha=0.1,\n",
" )\n",
" axes[2].set_xlabel(\"fit_times\")\n",
" axes[2].set_ylabel(\"Score\")\n",
" axes[2].set_title(\"Performance of the model\")\n",
"\n",
" return plt"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "0hmfSYdXeYQV"
},
"source": [
"# 2 Dimensiones\n",
"\n",
"Caso Bivariante\n",
"\n",
"\\begin{equation}\n",
"f(x, y) = \\frac{1}{2\\pi \\sigma_x \\sigma_y \\sqrt{1-\\rho^2}} \\exp \\left( -\\frac{1}{2(1-\\rho^2)} \\left(\\frac{x^2}{\\sigma_x^2}+\\frac{y^2}{\\sigma_y^2}-\\frac{2\\rho x y}{(\\sigma_x \\sigma_y)} \\right) \\right) \n",
"\\end{equation}\n",
"\n",
"\n",
"$\\rho$ coeficiente de correlación, media es $(0, 0)$\n",
"\n",
"\n",
"La matriz de covarianzas viene dada por:\n",
"\n",
"\\begin{equation}\n",
"\\Sigma = \n",
"\\begin{bmatrix}\n",
"\\sigma_x^2 & \\rho \\sigma_x \\sigma_y \\\\\n",
"\\rho \\sigma_x \\sigma_y & \\sigma_y^2\\\\\n",
"\\end{bmatrix}\n",
"\\end{equation}\n",
"\n",
"Otra forma de escribirla puede ser como sigue\n",
"\n",
"\n",
"\n",
"### El caso más general viene dado por:\n",
"\n",
"En el caso más general tenemos que :\n",
"\\begin{equation}\n",
"f_X(x_1,x_2, ... x_n) = \\frac{1}{2\\pi^{n/2}|\\Sigma|^{1/2}} \\exp\\left( -\\frac{1}{2}(x-\\mu)^T\\Sigma^{-1}(x-\\mu)\\right)\n",
"\\end{equation}\n",
"\n",
"\n",
"$|\\Sigma|$ es el determinante de la matrix de covarianza, \n",
"\n",
"\n",
"\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "jK0i1XJseYWs"
},
"source": [
"\n",
"# Dataset \n",
"\n",
"|Y | X_1 |X_2 |\n",
"|----------|----------|-----------|\n",
"|$Y^{1}$ | $X_1^{1}$| $X_2^{1}$| \n",
"|$Y^{2}$ | $X_1^{2}$|$X_2^{1}$| \n",
"|. | . |. |\n",
"|. | . |. |\n",
"|. | . |. |\n",
"|$Y^{m}$ | $X_1^{m}$ |$X_2^{1}$| \n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "z3DDUD2ad97U"
},
"outputs": [],
"source": [
"def data( mu=[1,1], mu1=[-2,2], cov=[[1.0, 0.0], [0.0, 1.0]] , cov1= [[1.0, -0.5], [-0.5, 1.0]] ):# Caso mas visible \n",
" rv = multivariate_normal(mu, cov)\n",
" rv1 = multivariate_normal(mu1, cov1)\n",
" return rv, rv1\n",
"\n",
"def sample(N1= 1000, N2 = 100, r = 0.2):\n",
" X_t = np.concatenate([rv.rvs(N1, random_state = r ), rv1.rvs(N2,random_state = r)]) # Todos los datos en la misma distribución, \n",
" y = np.concatenate([np.zeros(N1), np.ones(N2) ] )\n",
" return X_t, y\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "uRmV5KJBdERN"
},
"outputs": [],
"source": [
"def graph_gaussian_contour(x, y, pos, rv, rv2):\n",
" fig = plt.figure(figsize = (12,6))\n",
" ax = fig.add_subplot(1, 2, 1, projection='3d')\n",
" surf1 = ax.plot_surface(x,y, rv.pdf(pos) + rv1.pdf(pos), cmap = cm.coolwarm,\n",
" linewidth = 0, antialiased = False)\n",
" ax.set_xlabel(\"$X_1$\")\n",
" ax.set_ylabel(\"$X_2$\")\n",
" ax.set_zlabel(\"$PDF(X_1,X_2)$\")\n",
" #=============================================================\n",
" ax = fig.add_subplot(1, 2, 2)\n",
" cs1 = ax.contourf(x, y, rv.pdf(pos) + rv1.pdf(pos) )\n",
" #cs2 = ax.contourf(x1, y1, rv1.pdf(pos1) )\n",
" ax.set_xlabel(\"$X_1$\")\n",
" ax.set_ylabel(\"$X_2$\")\n",
" #ax.set_xlim(-2.5,2.5)\n",
" #ax.set_ylim(-2.5,2.5)\n",
" cbar = fig.colorbar(cs1)\n",
"# Histograma de los datos sinteticos a estudiar\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 396
},
"executionInfo": {
"elapsed": 1413,
"status": "ok",
"timestamp": 1638335089848,
"user": {
"displayName": "HERNAN DAVID SALINAS JIMENEZ",
"photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GhBC-5LZ7dZrJhGJeMzy6pOzsnDM7tcAwf58cHvon83rnG0ZgZHkYjtYed5MVVWyx0YIsBH3d5Rt-u8LhImipAJ47T_GJxdC3sX03gpJEpqdzXtiUn_QWj-eIgyQ0WzKukBfSahChtfEfM7zQgZom4BpTN1S0izHwqrALhT_uYIj7r47B9trr7ZfpDvK_xrdRhxfJRfUppIJCCV0E9jOVEyvzPuIQSbmxEKRZwPGz__-DY2kTV3XHPKlj7m9DPxx0pI3Vj34PoH6w0D8qs0nIz0Z_viNFTzydCbqfyl8irfbh40YasBQngmGl-_LSBU6_3hoh39ssLX4euV1xdBpiO8KsQgvDvLZF35hIogSmu4BE17JjVJcJGK0DuSPhh-Y0JxedJn0nskjm1FCQjhLi4FBh5JPAf3R2uJ-it0BFCRci00xUyduL0lelHwRF3abBC6OvNzE6uhi8uNRJHZet7rvNPC9Noi2U0asD-CBUa0tfIU1FExuDvfd1c2zLDWFCsQc_QbBlVvwVirgo0WC_S4TQ8I3pBHFydmquoBXzShhJDT7TixT4cz2Vgl_dXzBlELkYgEeHYMhvF8fRRsHrf7sd4yj8bqfBYj22GKqPSkdBk5A_f08DZBRcegVxK2oi9rWpLLRvSi_kvVe5whaMrPRlqb-uNbk9mDEQh7M8m9LAxMWA84blOlniAtinYP_Q3EOXnQUrCJoF0eyXdNZiYdyM4PeZpc1CmqbKLTBZ_4md9ObCgo2i8fdtUwNughaIcPHrc=s64",
"userId": "00408651407692255291"
},
"user_tz": 300
},
"id": "bgIos_H2dHkn",
"outputId": "9c1ecfd7-7f4e-495a-af27-dc158a1a6913"
},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAqsAAAF7CAYAAAD1zcJDAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOy9eXCkWXnm+zup1L6npNokVamqpNpLJZWUamjGhsbQDR24zUyYnvZc2zMBHW7j5gZ38MXT9851+9ITzPTYHsYLBtNjCMD20DYMhh4HNE3PcMGDgZJqk1SLpJJKay0qKbUvuZ77h+rLTkm5fJn55Sa9v4iKUGae73wnM6XKJ5/vOe+rtNYIgiAIgiAIQjZiy/QCBEEQBEEQBCESIlYFQRAEQRCErEXEqiAIgiAIgpC1iFgVBEEQBEEQshYRq4IgCIIgCELWImJVEARBEARByFpErAqCIAiCIAhxoZR6n1JqQCl1Syn1QpjHP6GUuq6U6lVK/Q+l1KGQx35fKXVNKXVDKfUnSikV7VwiVgVBEARBEATTKKXygD8D3g+cAn5FKXVqy7DLQKfWuhX4BvD7D499FHgH0AqcAZzAO6OdT8SqIAiCIAiCEA9dwC2t9YjW2gO8CvxS6ACt9Q+01qsPb/4UaDAeAoqAAqAQyAfuRzuZiFVBEARBEAQhHuqBiZDbkw/vi8RHgO8CaK1/AvwAuPvw3/e01jeincweYzHSi1UQhFwmag5qp1FgK9bF9opML0MQBAtY9E7PaK3rEjn2He8q0vOuQMLnvt7nvQash9z1itb6lUTmUkr9KtDJw0v9Sqlm4CRvOa3fV0r9nNb6HyLNEUusCoIgCDlCsb2CR/f880wvQxAEC3h96k/HEj123hXgv/793oTP3XZocl1r3RllyBTQGHK74eF9m1BKvQf4t8A7tdbuh3f/U+CnWuvlh2O+C7wdiChWJQYgCIIgCIIgxEM30KKUOqyUKgCeAV4LHaCUage+ADyltZ4OeWgceKdSyq6UymfDcY0aAxCxKgiCIAiCIJhGa+0DPgZ8jw2h+bda62tKqZeUUk89HPYHQBnwdaXUFaWUIWa/AQwDfcBV4KrW+r9HO5/EAARBEARBEIS40Fp/B/jOlvteDPn5PRGO8wPPxXMucVYFQRAEQRCErEXEqiAIgiAIgpC1iFgVBEEQBEEQshYRq4IgCIIgCELWImJVEARBEARByFpErAqCIAiCIAhZi4hVQRAEQRAEIWsRsSoIgiAIgiBkLSJWBUEQBEEQhKxFxKogCIIgCIKQtYhYFQRBEARBELIWEauCIAiCIAhC1iJiVRAEQRAEQchaRKwKgiAIgiAIWYuIVUEQBEEQBCFrEbEqbEJrjc/nQ2ud6aUIgiAIgiBgz/QChOwhEAjg9XpZXV1FKYXdbic/Px+73Y7NZkMpleklCoIgCIKwyxCxKgDg8/nwer0A5OXlAeD3+/H5fAAopcjPzyc/P5+8vDxsNjHlBUEQrGDtTH3Y+4v7p9K8EkHITkSs7nKMy/4+n2+Tc6qU2nRba43H48Hj8eDz+VheXmbv3r1B8SquqyAIgnkiCdRoY0S8CrsVEau7GK019+/fp7S0lPz8fJRSEbOqSqmg4+p2u5mamqKqqgq32w1suLGG8yqRAUEQhPCYEalmjhXhKuwm5FruLiUQCOB2u7l9+/Y2VzUWhhjNy8sLRgICgQDr6+ssLS2xuLjIysoKHo+HQCAgm7UEQRBITqiGm8vK+QQhmxFndZehtcbv9+P1elFKJZw9DRWg4SIDXq8Xj8cTPEfoZi1xXQVBEKzBEKzitAo7GRGruwhDRPr9/k0CM17nM5bYDI0MGHMbeVeDvLw8iouLJe8qCMKuINUuqIhWYScjYnWXYJSlCgQCm4RqtJxqNMweY5wnVLzOzMwwPz/PkSNHAKREliAIO5p0Xq4X0SrsRESs7gJCy1JtveyfqFhNFCMWYLPZyMvLC8YSpESWIAiCdYhoFXYSIlZ3MFvLUoVzLBNxMa10PqOVyIK3qgzY7XaJDAiCkHNkehPU2pl6EaxCziNidYdiiL6tl/23kuoYQLxszbsaVQYMJDIgCIIQH+KyCrmOXGPdgQQCAR48eMDg4GBUoWoQKjy11qytrUUVo+mKDhiRgdASWX6/n/v373Pz5k0WFxdZXV0NZnEFQRCEyGTa5RWERBFndQcRWpbK2FBlZue+ITx9Ph99fX243W68Xi/l5eU4HA4cDgcFBQXpeApRMYR36HMLjQzYbLag6yolsoSdgFKqCPgRUMjG/9ff0Fr/XmZXJeQyEgsQchERqzuErflUm81myv00xOrKygpXr17l0KFD1NbWArC0tITL5aK/v59AIEBVVRUOh4OysjJL1pssWyMDWmvcbve2rlqSdxVyGDfwbq31slIqH/hfSqnvaq1/mumFCbmLxAKEXEPE6g4gXD7V7KV6pRRzc3PcvXuXs2fPUl5eHizmX1FRQUVFBU1NTfh8Pubn55mZmeHWrVusr68zMTGBw+GgpKQkLiGYrGjUWm+bI9xGrUAgwNraGkDYxgQiXoVsR2/8ES8/vJn/8J+0hMsRsv2yu7isQq4gYjXHCQQCeDyeoICLp36q1pr5+XkAnE4nBQUFEY+x2+3U1tZSW1uLz+fjypUr2Gw2RkZGWFtbC0YGqqursyoyYJS+MhoiGP+WlpbYt2+flMgSsh6lVB5wEWgG/kxr/bMML0nYQYhgFXIBEas5SrjL/qHEEqter5fe3l4AWlpa4hKYhhCsr6+nvr4erTWLi4u4XC6mpqYIBAJUV1fjcDiorKzMCiEYGhlwu93cvXsXh8MhJbKErEdr7QfalFJVwN8ppc5orfuNx5VSvwH8BkBRXnmGVinkMiJYhWxHxGoOYqYsVTSxurS0RF9fH0eOHGFhYSHp9SilqKyspLKyksOHD+Pz+Zibm2N6epqhoSEKCwuDG7VKSkqSPp8V6w1XImttbS34WkqJLCHb0FrPK6V+ALwP6A+5/xXgFYDKgr0SERASQnKsQjYjYjXHCAQC3L9/n6qqqoTqp969e5fbt2/T2tpKWVkZi4uLcW92iuXa2u126urqqKurA2BtbQ2XyxWMDBQWFqKUwuv1kp+fH9e5rWBr5jVcZEC6agnZgFKqDvA+FKrFwHuB/5jhZQlpZu64uStf1QOepM8lLquQjYhYzSGMtqk3b97k0UcfjTp2q6AMBAIMDQ2xsrJCV1cXdrs9OC7VFBcXByMDgUCAqakp7t+/H4whGJGBiooKU0Iw3AYrK4nVVSvcZi1BSBH7ga88zK3agL/VWv99htckpAmzIjXc+GSEqwhWIdsQsZoDbM2nmiFUrHo8Hq5evYrD4aC9vX3bHIk4q4lis9koLS2lqqqK5uZmvF4vc3Nz3Lt3j8HBQYqKilIeGYhX7IYrkfWTn/yE8+fPo5SSyICQMrTWvUB7ptch5B6GcE1UtIpgFbIJEatZjtm2qVsxxOrCwgL9/f0cO3YseFk+3LhE1mUF+fn57Nmzhz179gS7Z7lcrmB5rMrKShwOB1VVVRmJDGwltDSY3W6PGBkwGhNIZEAQhESI11WNNU8iolUEq5AtiFjNYiKVpTKDUoq1tTWuX79Oe3t7RJcynFhNhzMYTuwqpSgpKaGkpISGhgYCgUCwysD4+DgADocj6XavVsYIIkUG3G53MAcrXbUEQYgHq4RquDnjFa0iWIVIKKXeB/wxkAf8hdb65S2PfwJ4FvABD4APa63HQh6vAK4D39JafyzauUSsZiGhbVPDlaWKRSAQYHBwEK/Xy9vf/vZgPjUciYinZAWX2eNtNhtVVVVUVVUBBCMDU1NTrKyssLKyEowMFBcXmz5/KjOv0lVLEASD4v6prGsMMHe8QASrkDQPc/R/xsamz0mgWyn1mtb6esiwy0Cn1npVKfVR4PeBfx7y+L9jo510TESsZhlG8Xq/359Ql6X19XWuXr1KbW0tJSUlUYUqJB4DyARGZABgZWWFPXv24HK5GBwcxO12ByMD1dXVMZ93OojWVWtlZQWv18uePXs2RQZEvArC7iYVrmqkc8QjWkWwClvoAm5prUcAlFKvAr/EhlMKgNb6ByHjfwr8qnFDKdUB7AVeBzpjnSzzn+hCkEAggNfrjTufajA3N8f169c5ceIEVVVVPHjwwNRxuSJWQ1FKUVpaSmlpKY2NjQQCARYWFnC5XIyNjWGz2aiurqampoby8vJtojETojC0RJbb7WZ1dZXq6mq8Xi9aa/Ly8oKbtaREliAIqSZe0Sq1WHcVtUqpnpDbrzys6WxQD0yE3J4EHoky30eA7wIopWzAf2JDvL7HzGJErGYJPp+PwcFBDh8+HLdI0VozMTHBnTt3OH/+PMXFxQQCAVMiNJec1WgY4rS6uhrYqIBgRAYWFxcpLS0NRgaygUAggM1m2xQZADaVyJLIgCAI6SDeaIC4rNnPrL+Mv5yLXuIyOn87o7WO6XiaQSn1q2y4p+98eNdvAd/RWk+a/VwTsZphQstS3b9/nyNHjpg+TimF3+/n+vUN193pdAbFj1kRGm5cppxHKykoKGDv3r3s3bsXrTWrq6u4XC4GBgaCnapmZmaoqqrKSGQgXGMCQLpqCYKQEUSwCnEyBTSG3G54eN8mlFLvAf4t8E6ttfvh3W8Hfk4p9VtAGVCglFrWWr8Q6WQiVjNIuLJUZoSiMc7Ipx44cIDGxsaw4icWuSh4EqmTGhoZmJubY3x8nPn5eUZHR7HZbEHXdWtkIFXEeg5mumoppSgqKpISWYIgWEIisQARrLuWbqBFKXWYDZH6DPAvQgcopdqBLwDv01pPG/drrf+3kDH/io1NWBGFKohYzRjhylLZbLbg5eFo2Gw2ZmZmGBoa4tSpU8FL34myE2IA8aCUori4mObmZmDj0rvL5WJycpKlpaVNkYGioqKUrCERwb01d9vX18fRo0cpKirCZrNRUFAgkQFByDKysSJALOJxWUWw7k601j6l1MeA77FRuupLWutrSqmXgB6t9WvAH7DhnH794WfSuNb6qUTOJ2I1zUQrS2Xm0r3hqI6MjNDZ2UlhYWFS69kpmdV42CoUCwoK2LdvH/v27UNrzcrKCi6Xi5s3b+L1eqmqqgo2JjAu0ydLIBBIKn5gvG+Gq2r8Xhjk5eUFxatEBgRhZ7N0NBBzTPlwfFdfRLAKsdBafwf4zpb7Xgz5OebmKa31l4EvxxonYjWNxCpLZYiOSPh8Pvr7+9Fa09bWlrRQhQ3REwgENt021robBY5SirKyMsrKyjh48CB+v5/5+XlcLhcjIyPY7XYcDgd+vz+p18iK19dw4aOVyDKeU2jeNZFKE4IgZB9mRGq4sWaFqwhWIVsQsZomzJSl2iocQ1lZWaG3t5eDBw/i8/ks7cAUroNVqt3WTHagiuf4vLw8ampqqKmpAcDtduNyufB4PFy4cIGysrJgZCCeLw9WiNVIc4QTr16vF4/HI121BGGHEI9QjXSsGdGaSBMBQbAa2ZWRBnw+H263e5MTFg4js7qV6elprly5wqlTp6ivr4/pwMZDpuqN5iqFhYXs37+f4uJiurq6OHjwIB6Ph+vXr9Pd3c2tW7dwuVz4/f6o85jJJsfC7BxGV63QyIDb7WZlZYXr16+ztLTE+vo6Pp9v10VCBGE3Y1bwmm1UkGvZXCF3EGc1hRhlqfr6+jh16pTpXf6hxw8PDzM/P4/T6aSgoCA4LpIDm+g6dxNWRRyUUpSXl1NeXs6hQ4eCkYHZ2VmGh4fJz88Puq6lpaWWNyZIZI6truu9e/doaGgIlsgyIgPSVUsQrMPqTVbJuKqR5orlspqtFCBxACEViFhNEaFlqRYWFkx94Ic6pl6vl76+PkpLS+no6NhWlspKZ3W3idVkifR6bY0MrK+v43K5GB0dZWVlhfLy8qB4tVIwJ4vNZttUIsvn8+H1eoPz5+fnS1ctQdjhLB0NWBYLEMEqWI2I1RSwtSyVWQzHdHl5md7eXo4cOcK+ffvCjtvNYjWdmdVkji8qKuLAgQMcOHAArTVLS0u4XC76+/tZXV0NuplVVVVZIwLD5V2lq5Yg7A7icVlFsArpRMSqhUQrS2W22P/MzAz37t2jtbWVsrKyiON2s1jNNIlefq+oqKCiooKmpiYGBgYoLCxkZmaGW7duUVhYGHRdS0pKTM1vxftm5ndya1et0BJZ0lVLEHYeZlxWEaxCOhGxahGhbVO3ulPGxqloNTq11iwuLrK8vIzT6SQ/Pz/i2EgbsRJhN4qLTDuzBtXV1VRWVgKwtrbG3Nwct2/fZnV1NRgZqK6uDmaVrSZesRvOdfX7/fT09NDW1iaRAUGIQq41BzDjsopgFdKFiFULCNc2NZRYu/c9Hg9Xr15FKUVLS0tUoQrWu6HpdlZzXSBb8XptFbzFxcUUFxdvigzMzs4yNTVFIBCguroah8NBZWVlUARaVac1UYzfdeOL2NbIgNFV6/79+zQ0NFjWUEEQdjNWbq4yez4RrEKmEbGaJMZl/9C2qVuJtnt/cXGRvr4+WlpamJubMyWEUhkD8Pl83Lp1i6KiImpqaiguLrbkPNmEFc5oKnfyh0YGDh8+jM/nY25ujunpaYaGhoKRgUAgkNRzMb5cWcXWyIDRVevXf/3X+eY3vxnceCYIQmSqBzymS0WZpezwwrb7lm9Xmj5eBKuQaUSsJsjWy/7RHKpIl+2npqYYHx+nra2N0tJS5ufnMypWV1dXuXr1Kvv27SMQCDA4OIjb7Q66ela2G81kU4BkSXfZKbvdTl1dHXV1dcBGZMDlcuF2u+nu7qa8vJyamhqqq6tjuvJb15Ds+xnpfQz94ra2tkZJSUlS5xGEnUC6owDhRGq4x8wIVysEqyAkiojVBDD6xw8PD3P8+PGYomOrWA0EAty8eROPx4PT6Qz2iDdb7D+ezGosUWSI1dnZWW7evMmZM2coLi5Ga01jY+OmdqPDw8OWuXqZJBsyq8lcgjfiAnfu3KGjoyNYZWBiYgKtdXCjVkVFRdRzGG1/k8GMO+vxeFKWuxWE3Ub5sM1UFCCaUI00NpZoTVawirsqJIqI1TgxylIFAgEWFxdNfdiHupdut5urV69SV1fHyZMnt9VPNSNCzTqrxrhoazQ2di0sLNDR0UFRUVEwcwjba4dudfUqKyuDrl6uZBKzQaxatQabzUZlZSWVlZUcPnwYr9fL/Pw89+7dY3BwkKKioqB4LS4u3rZByoouWmbed9lsJQjpIx6hGu64aKJVBKuQCUSsmmRrWSq73W7a3TSc0Lm5Oa5fv86JEyfC5vfMOqvxitVIBAIBxsbGcLvdvOMd7zAlKIqLi6mvrw+6egsLC8zOznL79m3sdjs1NTVhOzbtJLJJrG4lPz8/GBnQWge/XNy6dYv19XUqKiqCVQbS0fJVyqIJwmZyoSpA2eGFpARrLESwCvEiYtUE4cpSxXMpXinF3bt3cblcnD9/PuKmpVQ5q+EwHN7y8nKKioo2CQ6z89tsNqqrq6murg7OaQjX1dVVKisrg66eEXXIBnaCs2pGaCqlKCkpoaSkhIaGhuDVACMy4Pf7AVhYWKC8vDwh4er3+2Mel6txEUHIFKnYZBUvsVzWaIJVNlwJVpM9CiJLiVSWyuyHr5H5LCkpwel0Rr1kGk9mNRmxalQgOH78OHa7nakpa/7DKCwsDHZsMoTR7Ows4+Pj2Gw2HA4HNTU1SbttmRY/mc6sGmuI93ibzUZVVRVVVVUAzMzMMDk5yZ07d1haWqK4uHhTZMAMVrizgiBYR6IRgGjziWAVMo2I1ShsbZsar0BZW1vj6tWrFBYW0tTUFDPbF4+zmqgDe/fuXW7fvh2sQLCwsJCSS7VbhZHH4wkK14WFBWw2G/fv38fhcMS1g90KdoqzmuwabDYbZWVlNDc3o7VmdXUVl8sVrAJhOOPV1dURnfFYmVURs4KwnVyIAoQiglXINCJWwxCtbapZjN31p06dYnp62rQTmqoYgNaaoaEhlpaW6OrqSvtl+YKCAvbv38/+/fuZn59nYmKC1dVVJicnATbtYE+1a7pTxKqVeVOlFKWlpZSWltLY2EggEGBhYQGXy8XY2Fgw8lFTU0N5eXlw7bHWsb6+TlFRUVLrFITdSDZEAUKJFgtIVrAKQixErG5Ba43X6w2W9YlXUGitGRsb4/79+8Hd9TMzM6Yv73u93pjj4hWrPp+P3t5eysrKOH/+/LYKBOneBGN0Njp8+HBwB7vL5WJqaoqbN29SVlYWjAzs1JJH2SCYownNrXlkj8fD3NwcU1NTLC4uUlpaisPhiPllbm1tbUc2lhCEZMk1d9UgksuajGAVd1WIhYjVEAKBAFNTUxQUFFBZWRm3GPD5fFy7do38/HycTucm1ypdG6e2jltdXWVgYIBDhw5x4MCBhOeymtBz5ufns3fvXvbu3YvWmuXlZWZnZ+nv7w+2Gq2pqQnWDbVC6CWbF91pzmosCgoKNr1HRmTg7t27wUy30Tgi1LVfXV0VsSoIFmO21mqqEMEqpBsRqw/x+Xx4vV6WlpYoKSkJZi3NYnR/OnjwIPX1m78tm60cEM84Yyd3NDweD9euXaO1tZXKyvB5o0yJ1UgopSgvL6e8vJympqZgq9F79+4xMDBASUlJMGuZKaza4JVJwQ2JC97QyEBBQQFra2tUVFTgcrkYHR0NbqYrLi5meXk5aveq119/nY9//OP4/X6effZZXnjhhU2Pf+Yzn+Ev/uIvgl28vvSlL3Ho0CFgowbw2bNnATh48CCvvfZa3M9FELKZVEcBHj84EPb+N8aPxzxWBKuQTna9WN1aliovL890SSqDBw8eMDg4yJkzZ8KKwnTXTzWiCCsrK5w/fz6iUI3nnJkitNWo4eiNjIxw9+5dpqenN7WCNSu8suESfLJYscHKCnfW7/eTl5cXzBzDxpckl8vFN7/5TT73uc+Rn5/Pl770JR5//HEaGho2Hfv888/z/e9/n4aGBpxOJ0899RSnTp0Kjmlvb6enp4eSkhI+//nP8zu/8zv8zd/8DbBR8/fKlStJrV8QMkkqogDLtytjVgSIJFLDPR5NuCYiWAUhEXb1b5NRlirR+qlaa27dusXo6ChOpzOqe2mlsxpNYAYCAa5du8by8jI1NTUp3WmfbpFrOHrV1dUcPHiQ8+fPU1VVxYMHD+jp6aG3t5fJyUnW1taSOs+PHJ38yNEZ8fFsEauZcla3zrG1GkBBQQH79u3jox/9KH/+539OW1sbc3NzfOQjH2F+fj447sKFCzQ3N3PkyBEKCgp45pln+Pa3v71prsceeyzozL7tbW8LbsgThN1OusRgLGEbSRhHiinEcopzMccrpJ5dK1YDgQBut3tb/VSzzqrX62V9fR2v10tHR0fUjUBWiFAz44wWqOXl5Zw+fdqUo5uos5ppsQYb71VtbS3Hjx+nq6uL5uZmAAYHB7lw4QKDg4PMzs5ui0xEE5uhIjWSYM0GsZoNtV7NzOF2u2loaOC3f/u3+d73vrcpXjM1NUVjY2PwdkNDQ9Sav1/84hd5//vfH7y9vr5OZ2cnb3vb2/jWt76V1PPIVpRSjUqpHyilriulrimlPp7pNQnWEuuydzbspH/84EBU0SqCVUg1uy4GEKsslRlhuby8TG9vLwUFBTQ3N8f8wDe7yz+ZpgALCwv09/dvauVqRogmEwPIlGiLdN7Qbk1+vz/YCnZkZIT8/HxqamqiNiUIJ05/5Ojk5109ps6fTrLFWY3VwcqqDVZ/9Vd/RU9PDz/84Q+D942NjVFfX8/IyAjvfve7OXv2LEePHk36XFmGD/htrfUlpVQ5cFEp9X2t9fVML0zITWI5pbGOjRQLiDcSIPlVIR52lbNqlKUyhGo4wRFr89K9e/fo7e3l7NmzFBcXW+6YJtIU4O7du1y7do329vagUDV73mzPrCaKkaNsaWnB6XRy/PhxbDYbt27dYmJigrt37/LgwQN8Pl/cc2eDWM0mZzVaU4C1tbWIG6zq6+uZmJgI3p6cnNy2ORHgzTff5NOf/jSvvfYahYWFm44HOHLkCO9617u4fPlyok8ja9Fa39VaX3r48xJwAxDbaYeRqCjLRC40nQ6rIBjsGrFqdKOKVT81krOqtWZwcJCpqSmcTmewl7rVu/zjiQEYa7p79y5dXV3bREEqhWiyc6dbIBcXF1NfX09rayv19fVUVlaysLDA5cuXuXz5MgMfeNz0XNkgVq0SmqkWvOvr6xHFqtPpZGhoiNu3b+PxeHj11Vd56qmnNo25fPkyzz33HK+99hp79uwJ3j83N4fb7QY22sb++Mc/3rQxayeilGoC2oGfbbn/N5RSPUqpHk8guby2kJ1kQxQglGixgHgFazQkDiAY7Aqx6na7uXv3bvCDNdoHdDhh6fF4uHjxIkopzp8/H9y0ZLVYjcdZ9fv9XLp0CdjYMR2uI1WizmqqhVimhZ5SKthm1Ol0bhI5ladKYx6fDU50OlxRK9YRLQZgt9v57Gc/yxNPPMHJkyd5+umnOX36NC+++GKwDNUnP/lJlpeX+dCHPkRbW1tQzN64cYPOzk7OnTvHY489xgsvvLCjxapSqgz4b8D/obVeDH1Ma/2K1rpTa91ZYJOatkL4LlOpIF7BGg7Jrwpm2NGZVaMsldvtZnh4mNra2pjHbBWWi4uL9PX10dLSssnZCTc2ElYX+zfE98mTJ9m/f39S8yWzwSobRFsibHVGQy8tL1xfCf5cdWajluvABx7nwNf+G2VlZcHjMi24syUGECuzur6+vu3vJpQnn3ySJ598ctN9L730UvDnN998M+xxjz76KH19fXGuNjdRSuWzIVT/Wmv9zUyvR0gNscpYRaq5mukGAZFyrOEyrJJfFRJlx4pVoyxVIBDAbrebLkcVWg3gzp07jI2N0dbWRmnpdsctFTGAWONmZmYYHh7G4XBEFaqwczOrqbgMf/8fXZtu733UgXvRE3zMNz7O8vIy5eXl5OXlpbQkmBmM3+tk57BC8MbKrEoHq8RRG2/QF4EbWuvPZHo9ghCOdAhWYXezI2MAW8tSmc2CwlsbrG7cuMH09DROpzOsUDXGpqt+qtaa0dFRhoeHOXHihCmxtFPFarKEE7uVp0qjRgBOnz5NV1cXDUqY2ocAACAASURBVA0NeDwe7t27x8WLFxkZGWFhYSGu19CK19uqlq+priiwuroatYOVEJN3AL8GvFspdeXhvydjHSTkJlaWsYoUBTDTnSoR4qkyIPlVIV52lLMaqyyVGXw+HzMzMxw5coQTJ05EFQTxFPs3W5Iq3HxGoX+lFE6nk6WlJUtjBYmwk0RurM1VRhxAKUVFRQW1tbVUVVWxb98+XC4Xd+7c4ebNm5SWllJTU4PD4dgULdhKtlzCT1cMQMRq4mit/xeQ+aLGQlaT6SiAQTiHNVJJq3BIHECIxI5xVo18arSyVLGYn5+nr6+PkpISDh8+HHOOVGyc2ioAjUL/FRUVwUL/VrZvjTRmpwjRcMQSi6ERgGjH5+fns3fvXk6ePElXVxdNTU14PB6uXbtGT08Pw8PDzM/Pb3vvs0Vopqt0lcQABME8uS7Gwjms4TZcScMAIR52hLMamk9NVKhOTEwwOTlJa2srg4ODpo5J1jGNNS5coX9ITvwmMsbK46wglefd+6jD1Pm3/o4ZFQbKyso4dOgQPp+Pubk57t27x+DgIMXFxUHXNS8vz5JL+LngzkarsyoIQvzEs9Fq+XZlXDvzQ/m16n+M+Nhfzj0a9VizDqvkVwWz5LxYNS77Gx/e8X6ABwIBrl+/TiAQoKurK3ifGVJVPxXe2tzV3t6ecP3UnZxZTUaoJSL0Bj7wOMf//o3g8bFEnt1up66ujrq6OrTWrK6uMjs7y82bN/F4PPh8PlwuF1VVVQkJxmxxViH6eyFiVRDiJ1ZlgExjCNlYonUr8QjWaEgcYPeRszEAoxuVx7Px7StW/dRwrK+v093dTXl5OWfPniUvL8+0ADXOaWUMwJhvYGCAe/fu4XQ6w37QWy1WEyHXmgJEIvDn/7epcaGxgHjFrlKK0tJSDh48SFtbG2fOnKGgoICZmRl6enro7e1lcnKStTXzBd2t2mCV6hJcssFKEKwnWdfRqk1W0dzXZGuwShxACCUnxapx2f+nP/2paTd1q2B0uVxcvHiRY8eOcejQoU31M+OpHGBlNQCfz8fq6io2my1iof941pitrmk21Sgd+vue4P1FVZE3RUU6PhGUUhQWFnLs2DG6urpobm4GYHBwkAsXLjA4OMjs7GzUtr9WuaKpZn19PWI1DUEQIpOIcxjOoUx1g4BEBOtWpB2rEIvs/7TbQmhZqnj6uhuC0SgBNTQ0REdHB9XV1QmvxcoYwMrKCt3d3eTn59PS0hKzCkEqxerdu3djOn2ZFJzJikUrjk+GrecvKSmhoaGBc+fO0dHRQW1tLS6Xi0uXLnHlyhUmJiZYWVnZdN5cEauywUoQUkM2ZTrjFazxbLiKhrirmUUp9T6l1IBS6pZS6oUwj39CKXVdKdWrlPofSqlDIY/9S6XU0MN//zLWubL/0+4hod2o4K3L/vG4oF6vl97eXlZWVnA6nRQVFSW1JqtiADMzM1y5coUzZ86Yqp9qdecsAyO/Ozs7i1KKwcFBuru7GRoawuVybTtnNrq2ibA+745rfLL1SaOJ5by8PBwOBy0tLTidTk6cOIHNZmNkZITu7m4GBgZ48OABfr8/4w61Gfx+f8YbKAhCrpJKd9XqequpEqwSB8hOlFJ5wJ8B7wdOAb+ilNra8/oy0Km1bgW+Afz+w2MdwO8BjwBdwO8ppaI6hzmxwcoQqj6fb9Nl/7y8vJh1HkO5dOkShw4doqGhwZJ1JRsD0FozNjbG9PQ0nZ2dUWtzhpIKZ9Xr9XL16lWqq6tpaWnB7/fT2NiI3+9nfn6eBw8ecOvWreCudrO53mwk25zVaBQVFVFfX099fT2BQICFhQVcLhdzc3PcuHGDuro6ampqKC0tjfs5WZF5FQQhc0SqDJApfq36HyNuukqmBqtUB8hKuoBbWusRAKXUq8AvAdeNAVrrH4SM/ynwqw9/fgL4vtba9fDY7wPvA74W6WRZ76wa+dStQhXeEquxePDgAQsLCxw9etQyoQrJOZyBQID+/n6Wl5fjEqqR5ktmXCAQoLu7m4aGBo4ePbpJ/Ofl5VFTU8Px48dxOp0cOXIEn8/H0tISvb29CXVwynaMzVT7z21uZ9v78+8K/pyM0Ev0Er7NZqO6upqjR49SVVXFsWPHKCgoYHR0lAsXLgS7rnm93oTXFg+xnsdO+p0QhEyRS+4qRHdYzZBIflXc1YxQD0yE3J58eF8kPgJ8N8Fjs99ZNQRquI1UeXl5UcWi1pqRkRFcLhe1tbWW70qOR6yGsr6+ztWrV9m/fz+NjY1xC594SmHFWt/c3Byrq6s88sgjVFZG/4Zr7GovLS1lYWGBpqYm1tbWgh2cysrKqKmpoaamJqsv/RrOpplKAKGC1XV7dtPxyZ4/GQKBAEVFRVRUVLB//3601iwuLjI7O8vExMb/AUZd1/Ly8m3ns0JImr2qkQtxBUHIVbLNXYXIDmuy9VejIeWsLKdWKdUTcvsVrfUriUyklPpVoBN4Z6KLyXqxCoQVqrAh2iI5qz6fj76+PoqLi+no6ODGjRtxXbo2IyjiKXNlEKnQfzxY1RTAqOVaUlISU6iGmzsvL489e/awZ88etNYsLy8zMzNDb28vQFC4lpWVWSpYrBB7Silmr90O+1hhRQGOw5Hfm2wRq6FCUSlFZWVl8H30eDy4XC4mJydZWlqivLwch8OBw+GgoKAgLd2rBEGwhkTqrpptEvDG+PGwmdK/nHs0KZc0WcEaDokDmGfRU5Sscz6jte6M8vgU0Bhyu+HhfZtQSr0H+LfAO7XW7pBj37Xl2P8v2mJyQqxGIlIMYHl5md7eXg4fPsz+/RvOWDRhuxVDhMb6II5XrEYr9B8PZoVOtFaqQ0NDLC8v43Q6uXDhQsJrCT1XeXk55eXlHD58GI/Hw+zsLGNjY6ysrFBZWUlNTU1S1ResIp2Z01Qcb2aOgoIC9u3bx759+9Bas7S0hMvlor+/n0AgQFVVVbA6RqJriSV4fT6fiFlBsIhogjUb3VWInmGNhbirWU830KKUOsyG+HwG+BehA5RS7cAXgPdpradDHvoe8O9DNlU9Dvxf0U6W02I1nAC9f/8+w8PDnD17lvLy8uD9sSIDW+e1UqxqrVlfX+f+/fs4nc6I9VOtJpxY9fv99Pb2UlJSQnt7e8qaAhQUFLB//372799PIBAIXqIeHR0NOsOZKhhvCDTjsn6ixydKurtPKaWoqKigoqKCpqYmvF4v09PT3L9/nwsXLlBaWhqMDMSTnY4VA5DuVYKQPsIJ1mTd1VSRTBwglrsqgjU9aK19SqmPsSE884Avaa2vKaVeAnq01q8BfwCUAV9/+Jk5rrV+SmvtUkr9OzYEL8BLxmarSOSEWI1W5scQi4ZbuLS0hNPp3JaZTEVnKjPjjHJZAG1tbWnN720VlOvr61y5coWGhgZLN5rFwmazUVVVRVVVFUePHsXlcjE8PMytW7dYX1+nqqoq6LpmqnbosSfPMP6PQwBRIwCQHc4qJJ4Fzc/Px+Fw4HK5OHPmDCsrK8zOznL9+nX8fj/V1dU4HA4qKyujvh+xvtCtrq4mXR5OEIS3SHcb1mSjACBxgJ2M1vo7wHe23PdiyM/viXLsl4AvmT1X1lcDiIYRA/B4PFy8eBGlFOfPnw+7ucds5QBjrBX1U41C//X19RQVFZnutGXVLurQuRYXF4Mdu6wQqsmIrcLCQkpKSmhtbaWzs5OamhpmZ2eDrUenpqZYX1+PeLzWmhvH38+PHNHiNJHJtNi0KgaQDIFAIJgFLysr49ChQ7S3t9PW1kZFRQXT09P09PTQ19cX8f2I5e6KsyoI6SWcgMtkZQCDSILXjJsrzQIEyBFnNRJ5eXmsrKwwMjLC0aNH2bt3b8Sx8TqrZoRttF35Dx48YHBwkLNnz1JRUcHIyIipcxsC2IqsnyFWjWhEW1ubpa0vrRDVNpstuBELNty42dlZbty4gc/nw+FwUFNTQ0VFRVAYTXY+nfR5V194Nu5jHIdrGPjA4/Cf/jTjYjVZIglNu91OXV0ddXV1aK1ZXV3F5XJx8+ZNvF5v0AU3Mq+xxKp0rxIEa7Fqs5VZrHBX48GqOICws8hpZ3VpaYmJiQlaW1ujClWIz1k1WxoqnADWWnP79m1u376N0+mkoqJi02OxsNJZhY3SVOPj4zidzpzo0V5SUkJjYyPt7e20t7dTUVHBvXv36O7upr+/n7t3724an4i7Gvr6Hny0BSAYATBDsg0RrMisWuHMxlqDUaqssbGRtrY2zp8/j8PhYGZmhp6eHkZGRlheXmZ1dTXs8evr6yJWBSHNmBVw2eyuhutuFQ6pvbp7yAmxuvWDORAIcOPGDZaWlmhsbKSsrCzmHKlyVkPn9Pv99PX1sbq6SmdnJwUFb/0hmRWhiZTDCkcgEGBsbAyv10tHR4fldU+tFtXhMFy+EydO0NXVxaFDhxg6/YvBxytPbYhvrTU/cnQG/8XD4Hf6N91ufrLD1HG57qzG0/nNwGgQcezYMbq6uti7dy82m41bt25x4cIFBgcHmZmZCf7trKysRI0BvP766xw/fpzm5mZefvnlbY9/5jOf4dSpU7S2tvILv/ALjI2NBR/7yle+QktLCy0tLXzlK1+J63kIQq5jVaMAsyS6o38ryTi0EgfY3eSEWA3F7XbT09NDYWEhhw8fNn1cqjdYra+v09PTQ1VVFadPn94mBOJxa5MVgR6Ph56eHkpLS3E4HBnbtGQVP3J08g81zk3VHQyhCvAPNc645tsJpauSxYyzGgu73U51dXUwe1xbW8v8/DyXLl3ij/7oj/j617+O2+0O+3r7/X6ef/55vvvd73L9+nW+9rWvcf369U1j2tvbgznmX/7lX+Z3fud3AHC5XHzqU5/iZz/7GRcuXOBTn/oUc3NzST0XQcg1ognWbHVXI2HWXQ0nWLOxZJdgPTmlYubn5+np6eHIkSMcOXIkrnJU8cYA4hGr8/PzXLx4kZaWFg4ePBh2bDyF/JNxVpeXl+np6eHw4cPs27cv4XlikQ5ndStmXNM9j1Rz8/3vZXx8nJWVFVNrPPbkmeDPrhujMcfvBLFqRRQh1J212Ww4HA6am5txOp188IMfpKCggB//+Me0tbXxW7/1W5t+ry9cuEBzczNHjhyhoKCAZ555hm9/+9ub5n/ssceCzuzb3vY2JicnAfje977He9/7XhwOB9XV1bz3ve/l9ddfT+q5CMJuwOxmq3Ck2l1NVekscVd3BjkjVicnJ7lx4wbt7e3U1tYCqRGg8YxVSuF2u4PrcjgclsyZqAicmZnh6tWrnD17lrq6uoTnyrSQCqX359/F3kcjv66RsNvtjIyM0N3dzcDAALOzs8HfFeM1OfhoyyahagbH4RpL6qxmQzWAVHawampqorOzk1//9V+np6eHZ599dtP5pqamaGx8q/lJQ0MDU1ORnaIvfvGLvP/970/oWEHYqVjhroYjkrtqlWA1i1XuqgjW3CcnxOrMzAwul4uurq5NGbhUOatm5tVaMzAwgM/nw+l0xizRk8ymLTOMj48zPDxMZ2dn8HJ5PGI1XvGTLrFVWLHxn1C8gnXxI/+Ss2fP0tnZSV1dHS6Xi0uXLnH16lXW1tbiXm8oVjiryQjFVLRrTcUc6+vrlJSUkJ+fz/nz5xM+z1/91V/R09PDJz/5yYTnEITdyE4oZWXFZish98kJsVpbW8vZs2e3uTiJtFA1OzbavF6vl4sXL2K32ykuLjbVkSqeGEA8AlNrzY0bN5ibm6Ozs3NTByKzcyUqfBJ198yer/fn3xVzTGh2dc8j21u5GpenW1pacDqdHDt2DOeV/x52LrObqzIdA0h3B6xIxNqktbq6GrECRX19PRMTE8Hbk5OT1Ndvdz/efPNNPv3pT/Paa68Ff7fNHisIuwHZbGUOcVdzm5wQq0bx8q3EGwOwQtguLy/T3d1NQ0MDzc3NpkWHlV2xYOM18Xq9XLp0iYKCAlpbW7eJ+Wx2VhMhkrsaKlhjUVxczGz/cFLr2P9f/kPONxVIdQwAotdZdTqdDA0Ncfv2bTweD6+++ipPPfXUpjGXL1/mueee47XXXmPPnj3B+5944gneeOMN5ubmmJub44033uCJJ55I6rkIQi6T7s1WqYwDpNJdFcGau+SEWI1EKrpSQWTB+ODBg2AmNN7NS2ZjAGYFZiAQoKenhwMHDnD06NGw4iUeZzURlzTdG6zi5eo/eSdX/8k7wz5Wtt9crKD6eOQNc4mSrFDMFmfVbAwgHHa7nc9+9rM88cQTnDx5kqeffprTp0/z4osv8tprrwHwyU9+kuXlZT70oQ/R1tYWFLMOh4Pf/d3fxel04nQ6efHFF6PmxQVht5NMHCBbSbTJgZCb5EQHq0jCIBWbpoyxXq83eFtrzejoKA8ePMDpdG6qn2oWszEAM6J2bm6O1dVVzp8/H+z8FOmcmWhEkAmiuatX/8k7Ofe/fhj3nIZQrT5+kLmBcRwnm4LVAsRZNRcDiNYU4Mknn+TJJ5/cdN9LL70U/PnNN9+MeOyHP/xhPvzhD0ddXzZUXRCEdJGqzlZvjB8P63Za0dnq16r/MaxL+/jBgW2ubrjOVuGI1dlq7Ux9QtEJIbPsKmc1ERfWKPS/tra2rdB/PMRzeT/auDt37nDz5k3Kyso2dceKNFeqMqupFriR8qp7H3WYvuxfWB7+vao5vb0+b6y8as2ZowA4TjYBMDExwfLycsKO9E7IrCbjrKaK0L8dEarCbmMnxQHMEEloy2arnUdOi9VUOquBQID19XW6u7upqqri1KlTET+YrdzlH0msaq0ZGhri3r17OJ1O7HZ7zPPmqrNqZmNVPOw/t5/p559Jeh4Vks8sKChgdHQ0WBortHNTLLKh9FW6MqvpFqs2m4379+/zP//n/+Tv/u7vUEr9klLqMaXUGaVUbVoXIwg5QLLVAZIVrMlWBpDNVruDnI4BxPOBHU9nKJvNxurqKhcvXuTkyZNR83CG0Iu1lnjarW4dZ7i7RUVFtLe3o5QyJX5TvcEqFQLXaqGaKJGyqgb79++noaGBQCDAwsICs7Oz3L59m/z8fGpqaqipqYko1KwoXZUNzmqyMYBU8OUvf5k7d+4wPDzMyMgIwCeACsALXFBKfQ3o0Vq707owQUgT0eIA1QOesK6jmThAriFxgJ1FTjurqWJ2dpbZ2dmYhf4hNbv8Q0Wg2+2mu7ub2tpaTpw4ERTFZsRirjqrBkaNVVNjI1zy30rgC//PtvsKqqLHKWz5+dvuM94Hm81GdXV1sHPT8ePHUUoxNDTEhQsXGBoawuVybXrvk3VGcykGEKl0Vaq4efMmp06d4o//+I/5wQ9+gNb6nVrrdq11F/DfgI8Cv5zWRQlCFpFLcYBUuqtCbiFiNYRAIMDNmzeZn5+ntrbW1CVMq8Vq6LjFxUV6enpoaWmhoaFh07hMi9VkBW64Yx/8778CbFy2t4rQuayIAhhEEpvFxcU0NDRw7tw5Ojo6cDgczMzM0NPTQ19fH3fu3MHn8+2IDVaxHN5MxABefvllPvjBD1JWVrYtlqG1/oHW+le11n+d1kUJQpqJ5RimollAMoI12Y1akZBSVjuHnBGrqd4sYdQstdvtnDhxwvRx8YjQeITj9PQ0/f39tLW1hd3xb7VYTSfRzmelUI0XY6f//vdE/k/XyK2aec3y8vKoqanZaETgdHLkyBF8Ph/z8/Ncv36d4eFhFhYW4hb92eKsQvTXIRNiNRAIBF/PvLw8lFLlSqkW43GlVHoXJAgZwqpmAZl2WFPtropgzQ1yRqxGI9lL2Eah/8bGRpqbm7Hb7ZZv3Iqng9X09DRjY2M4nc6Il1GtzKxmYwwgk6I1FSilKC0t5eDBg1RXV3PixAnKy8u5c+cOFy5c4Nq1a9y7d29TybRIWLXBKtVfUtbX1ykqKkrpObYS2kDE7XYD/CvgN5VS71BKlQMfSeuCkkQp9SWl1LRSqj/TaxF2FmbjAPGS6QoBWzFTGUAEa/aT82I1no1TsF3Yhhb637t3b3BOq8WqmXGBQIB79+7hdrvp6OggP0xW0sBKZxXeel0CgQD9/f387Gc/49q1a9y/fx+fz5fU3GYwIgC5QjLrDQQCFBQUsGfPHk6ePElXVxcHDx5kbW2N3t5eLl68yOjoKEtLS2FfYys2WGmto+7kt4J0nGMrSqng5f+ZmRmALq31bwPvBf4Z0PJwXK7Utfoy8L5ML0LITTIRB4D0CNZ43FURrLlPzojVaI0BzJYMChVYWmtGRkYYHR3F6XRSXl6e0JxWxQA8Hg8XL16kqKiIvXv3xhQjZsRivNEDr9fLxYsXKS8v55FHHqGxsZGVlRWuXLnC5cuXGR8fZ3V1NeZ8WY1OPohv1FpNeAlbMqdKKcrLyzl8+DAdHR2cPXuWoqIixsfHuXDhAjdu3ODBgwfBLw1W7eTPHb1mjsXFRX7yk58EBfLD53cbQGv9/wJOwPhEyoknr7X+EeDK9DqE3CUTcQCwVrCGiwIIu4ucEauRiKfYvyEs/X4/vb29rK+v09HRsa3Qfyqc1WgxgJWVFXp6ejh06BB1dXWmzmvW2TQrVtfW1uju7ubgwYMcOnQIpRQVFRUcOXKEzs5OTp06RV5eXnCX+9zcHMvLy6Zfp0RwHI7cnStRZvpGuH9pKOLj0fKqoSQjWGNtkCooKGDfvn2cPn0ap9PJ/v37WVxc5PLly1y+fJkHDx7g9XqTcratyqxGIhOxkgcPHvCpT32Kf/iHfwCgqqoK4GsASikb8PvAXz9cn2wfFgQixwGsEqypdFnFXd097CqxmpeXx8rKCt3d3TgcjoiF/tMZA5idneXKlSucPXuWPXv2JFWPdStm5/J6vVy/fp1Tp04FoxBbKSwspL6+PrjLvaioiPn5ebq7u4NxATN5y3RSdShy2bH50elt94UToGHLViV5aTuevKjNZqOqqoqjR4/idDo5efIkSilcLhfd3d0MDg4yOztr+m/AINkogdnnkE73tr6+ng984AN8+tOf5o033jA2d00ppbqATwK/qbX+5kPhumNQSv2GUqpHKdXjCaxlejlCFpJIHCAeoglWiO2yZlPOVQRrdpITTQGiEY+wNBzV06dPR62fGm8DgURjABMTE9y5c4fOzk4KCwvjms/Mhi0zQmF6epqFhQVaW1sNJwqI7v7l5eVRWlpKSUkJdXV1LC8vMzMzw8TEBDabjdraWtOlvyC+/OfBR1sY/E5q95qsT0ymdP5khGJRURGVlZVUVFRw4MAB5ufnmZ2dZXh4mKKiomBDglgbm1Ld8jUTzmpRUREf+9jHaGpq4vnnn+eZZ54BeBE4z0Yc4DMP17ajXFWt9SvAKwCVBXuza6ekkDVEaxYQiXDNApZvV4Z1NN8YPx71cr0hSBMtU/X4wYGYothg6WggrDMcq1GAgTQMyD5yRqxGE05mXKXJyUmWl5c5d+5czEL/8ZCIuNRaMzAwwPr6Op2dnZs2ocSbM02GsbEx7t+/T21tbVAsh85vBiNvaWQu3W43MzMz3Lp1i/X1daqrq6mtraWystLSy87HnjyTlGitObnRocq/tm7VkkyTrFA0xK7NZsPhcAR/n1dXV5mdneXmzZt4vd7ga19RUbHttU91nVaPx7MtXpNqXC4Xb775JmNjY7z97W/nc5/7HMAjwAe11rNpXYwgZCFWdbdKVLCC9S5q2eGFsPGESIhgzU1y/nJYLLEaCAS4ceMGMzMz1NbWWv4BGm8MwOfzBeu5njt3bttu6XhKXCUqVrXW3Lhxg4WFhaBYDp0rGSFjxAVaW1vp6Oigurqa6enpYFzg3r17YasLJMOxJ89YOl+qSZWrWVJSQmNjI21tbbS3t1NZWcm9e/fo7u6mv7+fu3fv4vFYU67G7/dH3em/traW9lart27d4qtf/SpjY2O88MIL9Pf3A9wCfk8pZX0AOg08bA/7E+C4UmpSKZVTpbeE3CLZ/CrEjgSkCyu6WkkkIHvIGWc1EtHEotfr5erVq8G6ljdu3Ig725fM+beO83g8dHd309TUxP794euIpro2qhGFKCsrC7ZvTUQ4mTl/Xl5eMBKgtQ7GBcbHx3G73YyPj4dteGDgup28GRbwa2x5m59fXn54kVVYfyDp85nBCmc11vF2u526ujrq6urQWrOyssLMzAz9/f1orXG73SwuLlJeXp7QWmI5q6urq2lvCNDY2MgXvvAF6us3fcB8lI3L/68opT6mtb6b1kUlidY6t2q6CVlPrDhAJIc1HMk4rFaSKncVxGHNFnassxpa6P/o0aMopeLKt5rFrBO6srLC9PQ0J0+ejChUwfr2raG43W56enqoq6ujpaUlKFISbbca73gjKtDa2kp5eTl2u53VF56Nax4rqGp+6z/qvOLo2c6A17tNxGoLvvCkMi8a7lxlZWU0NTVx/vx5WltbsdlsTExMBEtjTU9Px+V4x1rD+vp62p3V/fv3U19fv6nEl9Z6XWv9W2w4rM8DKKXSW/xVELIMq8pZQXod1ngFcDR31awghw3BKi5rZskZsRpPZnV6enpboX9ITODFwsycd+/eZWxsjJqamk2bmMKRKmd1eXmZnp4ejh49SkNDw7bHE3Fpk8nM2mw28v/Dbyd0bKTL/od+7ljC64lEcdMhYLPrWnTksOXniYdkndn8/Hzy8/M5ffo0XV1dHDhwgOXlZa5cucKlS5cYHx9nZWUl6vvr9/tjOqvpFqsGdvvGBaPQ9Wmt/w3wqYc/W3t5RRB2GPHEAaKRLZGASMQjWEFiAZkkZ8RqJELFolHo32hVGlroH+Irc2XMF8/5wx1/69Yt7ty5w+nTp025YfFs2DIrFl0uF1evXqW1tZXa2tqk5trNFDbWU9j41n9W9m//aUbWkWyN1K355MrKymA93dOnT2O32xkZGaG7u5uBgQFmZma2/d0EAoGsy6zGQmudXbXVBCGDJFrOKp78KqRPsIaLI4A12dVQRLBmhpwXq4YAjVXoH+JzVpOtGFDgVQAAIABJREFUd2qsx+fzcf78efLz8y11TOOpoTo4OEhHR8c28R7vXMkeEy8tH+g0PdZwVRN1V71LK9vuK+56W9Rj3FN3EjpXssRTpzXS8ZHEbmFhIQcOHODs2bN0dnayZ88e5ubmuHTpElevXmVycpK1tbWYgnltbS3tmVVBEOJjpwnWSFgVBzCQWED62RFidX19PWahf2NsPA0EEs2OGtlQh8MR3MRkdRY11jitNcPDw3i9Xjo7O6PW3cxE2839/+U/AFBz5mjUcS0f6IwoWq2qAuB3b/yHvHwzcmerbMKKgv5mXf7q6mpaWlpwOp0cO7bxRWBwcJDBwUHm5uaYm5sL+3uYiQ1WkVBKnVJK/bxSqkHyqoKwmZ0kWCO5q7FIRLCCuKzpJGfEaiRBtbq6yp07dzh+/DiNjY1R54i3M5UZYbt1zqWlJXp6emhubt60Hqsv70cbFwgE6O/vx+12U1xcHMzwRSMTzmosoZoMLb9o3pUNZZNgnY6wcTzgz2huNdkYQKLHFxcX09DQwLlz5zh06BClpaVMT0/T09NDX18fd+7cwe12AxsbrCKJ1ddff53jx4/T3NzMyy+/vO3xH/3oR5w/fx673c43vvGNTY/l5eXR1tZGW1sbTz31VNT1hvz9vgv4c+DPgGYAlYlvaIKwC8i0YI1ErDiACNbsJqdLV01OTjI1NcWePXuorq6OOd5ms5luCZqIwzk9Pc2tW7c4d+4cZWVl28ZZHQMItz6v18uVK1eora2lqamJn/zkJ5ad0yqUUpYJ1Xjd1f2/8AhrY5E7VJUebsS3tEzB2bZNYtW/uAiA1gHsJn7XUkmqu0+Zpby8nPr6+mBprNnZWa5du8Y3vvENJicnaWxs3HYuv9/P888/z/e//30aGhpwOp089dRTnDp1Kjjm4MGDfPnLX+YP//APt52zuLiYK1eumFqfkanVWn8O+JxSaj+w+vA+CWkLwkMSLWcVrmEARC5pBW8J1lSVtoq3jFUo8ZS0CkXKW6WenHFWQwkt9H/q1CnTH9zxxADiEat+v5/R0VHGxsbo7OzcJlTBfImrZBzYtbU1enp6aGxs5PDhw3F1oUrnZ3fh1/5j2s4FUHN6swtafGh7NYStePqu4Ll/H8/9+3inp4P3Z0Nb+Uw5q5HmMEpjHTp0iPPnz/OJT3yCuro6Lly4QFtbGx//+MeDx124cIHm5maOHDlCQUEBzzzzDN/+9rc3zd3U1BQsr5UMo6OjKKXalFJnlFJ7gGeByHXjBGEXY2UcAKI7rJAZl9XMZitxWLOTzH/yxonH4+HixYsUFBRw7tw5CgoKLL+0b4w1O+/s7CzLy8sRN3bFM1+iDuzCwgKXLl3i5MmT7Nu3z9S6I82VqmMiYXUcINFNVqWHN2Ib9vLNXza2Pk///IZjkKkoQCo3WJklWgcrw9V/7rnnuHLlCv/6X//r4GNTU1Ob4jENDQ1MTZl3JIwWxW9729v41re+tekx430y/s4+/elPA/wbNuqrvshGg4ASkBiAIIQjE4I1HtFqhcC1ujpAKCJYU0fOxACUUiwtLdHX10dzczN79uwB4hOgZjdNmR3r9XoZGBggLy+P06dPRxUQ8YjQROMH7e3tYXOCsS4b79TP7brWI6bH2gry45rbNzuDvWZ7GbB0kK4NVsnMYVQDsNlsNDU1JXWuUMbGxqivr2dkZIR3v/vdnD17lqNHN77sbP09PnPmDMCHtdZrDx+/DTwAiQEIQqJYGQkwsDoaECsKsHQ0ELVmbKJxAJBIQKrIGWd1bW2Nvr4+zp49GxSqkJpL+2bGrqys0N3dzYEDBygtLY0p+KwuSWWMGx8fZ3R0lM7OzqR2X6fTWQ244/9PoOUDnRx8tCWh86UK3+xM3MdYoZGyLQYQjkh1Vuvr65mYmAjenpyc3NoeNSrG2CNHjvCud72Ly5cvBx978803gbeaATz77LMA/hAX9U+11hMIghCRZIRWog6rQTSnNd2xgUTjACAOayrIGbFaXFzMI488sq1WaDwCNF5hG2ns7OwsV65c4cyZM9TV1VnaFSue5/PgwQPm5ubo7OyMGD8wIyrTmln98ksRH9r/+DvSswYThEYBCg4c2PZ4otnVZDdHWTFHsjECiB4DgMjVAJxOJ0NDQ9y+fRuPx8Orr74ac1e/wdzcXLDawMzMDD/+8Y83bcz66le/yh/+4R8Gf5dLS0vRWnu01loptQ8w309WEHYxicYBIHnBCm+J1tB/8RLLzU1lfnW3oJR6n1JqQCl1Syn1QpjHf14pdUkp5VNK/fKWxw4qpd5QSt1QSl1XSjVFO1fOiFUg7Idjup3VyclJhoaG6OjooKKiwvIWrmaEo9/v5/bt2yilYm5ESZVY3Wldr4rr48v5Gqi//vdxjbdCKFrRwSrVzmqkdqt2u53PfvazPPHEE5w8eZKnn36a06dP8+KLL/Laa68B0N3dTUNDA1//+td57rnnOH36NAA3btygs7OTc+fO8dhjj/HCCy9sEqu/+Zu/yQ9/+ENefvlllpeXcbvdKKXOKaX+GfDHPMyrCoIQm0wLVivIpGDd6e7qw5rVfwa8HzgF/IpS6tSWYePAvwL+a5gpvgr8gdb6JNAFTIcZEySnMqvhSKWzGjqv1prBwUHW1tZwOp1B4ZwKsRoNj8fD5cuXqaqqMjU+65zVCGx1Vbfu4N8pWCEUs6F0Vaw5otVZffLJJ3nyySc33ffSS2857k6nk8nJ7eXFHn30Ufr6+iKe89FHH+Vb3/oWnZ2dDA0N0dDQAPC7QDnweWB7mzJBECKSaEkriJ5hhcQL+FtNrPwqJJdh3cF0Abe01iMASqlXgV8CrhsDtNajDx/b9IvwUNTatdbffzhuOdbJcspZDUc8H9qJOqs+n4/Lly9js9k4d+7cJofXarEajeXlZXp6ejh69Ch79uxJqh2skDz2AxslsEZHR1leXjb1OlsRA4DkNsVZVQ0g29qtvvnmm3z2s5/lscce48qVK/zJn/wJwJe11k9orb8lm6oEwXoScVghfS6rGVGcKod1h7ur9UDoHoDJh/eZ4Rgwr5T6plLqslLqD2J1F8wZZxWSdwDjdWEDgQBra2tcuXKFQ4cOcSBMdjFdYtXlcnHjxg1aW1spLy9nbm7O9GshMYANp7Zg31sb84oPNURtDgAbuVVbeQUAedXV+OfmNg8oLMJ+oIHCwkJGR0dZXV2lqqqKmpoaqqurw4o5q8RqMljlrEbLrGZCrL766qu43W5+8Rd/kc985jP85//8n/nEJz7xG0opl9b6H9O6GEHYIcRyVyExhxXMVQrIJnLJYfW785L9QlCrlOoJuf2K1vqVJJdlYAd+DmhnIyrwN2zEBb4Y6YCcd1bjId4YwPLyMpcuXeLEiRNhhSqkR7TduXOHgYEBOjo6ghvMzJ7XjLOaNuEZZXOVQaY2WfmWYl6F2ERe01tlsQ7e/D5nzpyhs7OT2tpaZmdnN7Ug9Xje+s/NisxqsqSzdFW60Frz0Y9+lL/8y7/k6aefBjDqu/5X4I+VUo+nbTGCsMMwUyEgmx1Wq9xViN9hzWF3dUZr3Rnyb6tQnQJCe9w3PLzPDJPAFa31iNbaB3wLOB/tgJxyVpMlHpGwuLjI9PQ0XV1dUT90Uyk8tNaMjIwwPz+P0+nEbn/r7Yrn8v5uc1YLqioycl6bzYbD4cDhcKC1ZnV1lZmZGfr7+9FaU1NTY6rMWaoJBAKbfpcSwUwMoLS0NKlzxINSio6ODmDj+WmtjX+vKqVGgD8C3kjbggRhh5HrDquZNqxm8qtCkG6gRSl1mA2R+gzwL+I4tkopVae1fgC8G+iJdsCOeVesEk5aa4aHh5mfn+fAgQOWu0Nm1xkIBLh27Rrr6+u0t7dvExfxNA8wI1YFa1FKUVpaGmxB2traSlFREVNTU8zNzTEwMMDs7Kxpp99KrHBWY20U83g85OfH12ghGVZWVvj85z/P6OgoNpuNvLy84N+M1voC8D6lVI1SSioCCEKCpLLYfS45rFLSCh46oh8DvgfcAP5Wa31NKfWSUuopAKWUUyk1CXwI+IJS6trDY/3A/wn8D6VUH6CA/xLtfDnlrEZy84zcaLQMnRn8fj/Xrl0jPz+fo0ePsri4mNR8WzHWH0scaq25dOkSNTU1NDU1hR0fj7NpduOPwb179xgcHKSyspLa2locDkdYsRzXFwQTEQCDdFcCKO3qAiAwNf7/s/fm4W1d57nvb2PiIFKcwJmiJQ4iJVLiLNu5bhorcZ24J2rcOLGTnLSpM7SncW9zetLr3Dp1WjfnOMlNT5xbp03Tpreuk0aO3SdHrp0qduN7M9i1JcoiRYoixXmeZxAEMa37BwUIIDFsABskQa3f8/gxAay99wIFAi++9a73CzkuoG9VJUajkYKCAtLS0hgeHiY3N5f5+Xn6+/tJSUkhJycHs9kcNC/Xl1i/mGkhVtWwE9fw4HA46O7uZn5+HrPZTF5eHhkZGbznPe+5i80v5TXAHcAPgZd3bGISyT4jHgkBHvaKh1XrhID92tVKCPFj4Mdb7nvc5+eLbNoDAh37KnBS7bUSSqwGw7MZKhaxurGxQVtbG4WFhZSWljI3N6f5ximPqA63fLq2tkZ5eTkFBcGzP7XsiOU7Znh4mJmZGZqamrDZbMzOzjI8PIzJZMJsNmM2m0lKSgp73WBE073KQ9p778Ny/sfhB+4kScmwYVM93FORDGUXyM7Oxmw2k5aWFpeqd7zF6m7YQzIzM/nmN7/Jj370I1566SUGBwc97wd/CdiBF4HPCSGi+7YhkUhUs5cFqxo7gGTvsS/EqqfbVLTLjqurq1y5coWqqirM5s1+7x4BrCXhluRXVlbo6OggOTk5pFAF7TdYud1uenp6sNlsNDY24nQ6MZlMHDx4kPLycq+gunr1KkIIkpOTMRgMmuxuNxX5P1dTcRH28QnVx6ffcw+rr77qf4477sL+5i9jmpc4Vg9jA9vu991c5cH4//4Tjrt/K+w5twpFj13AYxlwOBzMz88zPDyM1Wr1VrczMzPR6/Wa/L61EKtq5rCT9pL5+Xl6enq4//77uf/++30fun3HJiGR3CLE078Ke0OwyvzVvcW+8KxGsssf/Cs/s7OzdHR0UFdX5xWqELrdarSEirmanZ2ls7OT+vp6rwiM9lxbCXcuIQSTk5MIIYJ2xEpNTaW0tNTrvzQajSwuLnLx4kV6e3tZXFwMPp8ILACRkn7PPZqdS1dcGnaMPisrpmuEE5seu4AnXSAvL4+FhQUuXbrElStXmJiYSBgbwE7S1dXFN77xDQDGx8e5fPkyAIqi6JRoe+NKJJKgxDMhAOLvYdUyIUASfxLqTVyLLlaeiqQQguHhYQYHB2lubiYtLc1vXDzyU4NtihoZGWFgYIDm5mbvbnGtdvCHq6w6nU6GhoZITk6murra+zsOJ6iys7PJy8ujqamJrKwspqenuXjxIl1dXczMzOB0bm/DbiyvDDvfnSTltoBWmrgSSWVUp9ORlZVFZWUlp06dory8HIfDwfr6Oq2trQwODrK6uhqxeNWiXWu48+/0pj2LxeL1+7a2tvIXf/EXnof0Qgj5iSORxIFbQbCGQ+1mqwSOsNoT7AsbQCSVVb1ej8PhoK+vD7fbTXNzc8AP7kgFsBoBsFU4+rZwbW5u3tbCNdT5tNhgtbGxwdtvv012draqjT2B0Ov1Xi+rEILV1dVtPtdibgrVpOM1bHRdjepa+4FYlvEPHDiAyWRicXGR2tpa5ufnGRkZYW1tjYyMDG8zgnDe7VjFajgxarPZSElJifr80bC8vMzExKZ1ZGNjg8LCQgCEEI4dnYhEcouR6JaAcMg4q73BLSdWFUWhra2N3Nxcjhw5okm1Vo243HpOl8tFR0cHKSkp1NXV+c1DjRBVm7MabJzFYqG9vZ3q6mqcTierq6thzxUORVE4ePDgNp+rFhXVjYVlYg1CctWcQn/1gt99WlZXDa0v4mw+E3KMFkJRp9N57QIFBQW43W6Wl5eZm5tjYGCA5ORkzGYzOTk5ATfDaTGHcN2rkpOToz5/NGRnZ+Nyufj617/OG2+8wcrKCl1dXdTU1BwFZoDVG3EpEolkj7GbglVuuEoM9sXXBbWboaxWK8vLyxQWFlJWVhayOhTJBiu1wtZTgbXb7bS2tpKTk0NVVdW2eag5n9qcVdheWV1aWqK9vZ2TJ0+Sk5MTt6YAqamplM1e2Xa/8fa7IrqWL2nvvS+q4/Tv/k9RXzMsSeqFWawbpAIdv9UuUFFRgdPppKurK6BdQCvBHAyr1brjrVZvv/12vv71r5OXl0dxcTG5ubl85StfAfhD4E+AOwEUGSoskWhOrHaA3SacEA7nXZW5q/EnoSqroaqg4SqrCwsLXLt2jczMTLKzs8NeK5rKqppxVquVzs5OKisryc3NDThOS8/q1nEzMzP09fXR2NjoXardy92o4o0uP3AbXQ+ukjL0PokA9oo6TLOB81iV1aWw14tVrKoRmp7NcKWlpTgcDhYWFhgZGcFisZCRkcHGxkZM/95qulfttFjNyMigqKiInp4e6urqSEtLw2g08uyzz/4UOAHk3RiqALfmi10iiSOJbgeQFda9TUKJ1WCEswGMjY0xNjZGU1MTfX19qiwDkaQBqBWrdrud69evU19fz8GDwVuCqq2sRmoDGB0dZXJykpaWlrAxX+FElWqxbA1sLzDefheOt4JHSymn/xPitZe8twNVVIMlAZjuiL5y60Ecq4/5HIGIdfNRpGLXaDSSn59Pfn6+n12gvb2d5ORkbzOCSLJz1dgAdtqz2tPTw1e/+lX6+/spKSlhamqK5eVlgGbga0KIJUVRFLnZSiKJH2oEayjCCdbdRHpXd5d9IVaDCUvPBiar1UpLSwt6vT6iJXu11Sc1YycnJ1lcXOTYsWMhhara80UiWFwuF729vVgsFpqamrYJjXhVVg2XXgo/KAD28QmUqkaNZ7M3CNemNByxLOF77AJJSUk0Nzezvr7O3NwcXV1duFwucnJyyMnJIT09PeTrS01ji50Wqz/84Q/JysriZz/7mfc+u91OUlJSOfCkoij/VQihvnuDRCKJC6Gqq+GQ1dVbl33xNSGQv9TpdNLW1oaiKNTX13sFmtrNWJGIwVACWAjBwMAA4+PjFBUVbWtbGuzaWolHRVEYHh7G4XD4/R5ivd5+tQ6oyVq15wYfY2h9MeSxWnhWtchIVRTFaxdoaGigrq6O1NRURkdHuXDhAt3d3czNzQX8WwlnA7BarTsuVufm5jh1o23u+vo6NpsNk8mEEOJBwAX8Omzmru7oxCSSW4xE969K9iYJ9cYd7EN+qwC12Wy0traSl5fH0aNH/Y6LR35qsHO63W6uXr2K1WqlsbHR24Eo2vNFitPpZGpqiqSkJI4dOxb097dbwjPYRivldOSboTyWAF+7gBZ2gEgQ6Znhx2jgWY3HHiGPXaCmpoaWlhYKCgpYWlri7bffpr29nfHxcTY2Nrxz2Gue1fn5ea5fvw5ASkrK1jQCAeysepZIJFGx20vtoSq3oSwKcpNVfNkXNgDfyury8jKdnZ0cP36crADdhnZKrHoqu9nZ2d6IrHhYEIKxsbHB5cuXSU9PJz8/Pyr/qRatPcOxtdXqTrNSWs/BkbaYz7Na3kzaTF/YcbtpA1CLTqcjMzOTzMxN8W21Wpmfn/faBcK12t0NsVpRUcHbb7/NF7/4RTIzM7ntttsoLCzkV37lV1KAfMCyoxPSEEVR3gt8E9ADfy+E+MouT0kiCUmsm60kkq3sC7Hq8axOTU0xODhIQ0ND0A/LSFuzqr2+rwi12WxcvnyZw4cPe8PJA41Te75IsVqttLW1cfToUZaXl6Na4g8nmKMV1O702NqV7gS64lJ8XyFbEwECYTemYsmrCCtYw21OCsdOfIHYSmpqKqmpqRw6dAin08nAwABLS0tcuHDB24wgOzvb+7x2oynA7//+7zMwMEB3dzevv/46//qv/+ppdPEV4A7g73Z0QhqhKIoe+BZwDzAGXFQU5UUhRNfuzkwiiR+hNlrtdpMAye6QUDaAYOh0OhYXFxkbG6O5uTlkVSfeldWVlRUuXbpEdXW1n1CF6OOmImF5eZnLly9TW1uL2WxW1TwgHjaAaDdX7WXsFXUxnyMRKquhMBgMpKWlUVhYyKlTpygoKGB5edlrF3jllVeYnp4O+jd4/vx5qqqqqKio8OSg+vHzn/+cxsZGDAYDL7zwgt9jzzzzDJWVlVRWVvLMM8/4PZaens709DQGg4E//uM/5mc/+xnPPvsswCvAPwN9APFKA1AU5R5FUf5OUZT6G7c/o9GpTwF9QogBIYQdOAv8hkbnlkjihhrv6l5FiuG9R0JVVgNVlNxuNwMDAzidTm6//fawH+SedqtqUVPJ8ojV2dlZent7qa+v58CBA0HHhSNaQe25/tbKcjzE6k74XLfGV3lIe+99WM7/OKpzBupitZPEoylANOeIBU91WFEUP7vA+vo6Fy9e5F/+5V9YX19nfHycD3zgAzQ1NQGbG7M++9nP8uqrr1JSUkJLSwtnzpzh+PHj3nOXlpbyj//4j3z961/3u+bCwgJ//ud/TmtrK4qi0NTUxJkzZ7xWn//+3/87P/vZzzAajfzN3/wNn//857n//vsRQrwMvBzTE1bHw8B/Ab6oKEo2oFX2WTEw6nN7DLhdo3NLJBJJQpBQYnUrdrudtrY2bz90NRWnSISgR5CpEauzs7Osra3R3NzsWX4MOE5tEkGkgmJsbIzx8fFt11fT6Wq/7uzfKeaPnPK7vTI5xMHCwwHH7kRTgFBoIXbdbnfAVIuUlBQ+9alPMTc3x9GjR0lNTeWNN97witULFy5QUVFBWVkZAA899BDnzp3zE6uHDx8G2PYcf/KTn3DPPfd4G3rcc889nD9/no985CPex//xH/+Rmpoa3nzzTZ544gkaGho4fPgwiqIYhBDOmJ50eFaFEEvA5xVF+QrQEufreblRxf0MQLI+facuK5HEjPStStSSsDYAi8VCa2srR44c4dChQ6p9qJF4VtUIWyEEc3NzrK6uhhSqnvNpbQMQQtDf38/MzEzA66sRNtFWVsOOCdIQwJfFpl+P6LoeImkSEAxLS+DWra6D2zucuUrK/G4Hi6+y5FWEvGa8W53G+3hQF12Vm5vLhz70If7gD/7Ae//4+DiHDh3y3i4pKWF8XN1SYbhj7XY7MzMzWCwW7rjjDhwOB2lpaQDsgFAFn+qtEOILwD9pdN5x4JDP7ZIb93kRQnxHCNEshGg26WTogUQST/Zq04L9TkJVVj0CybPcffLkSdLS0nA4HJq3RlUz1uVy0dHRgaIoFBcXh904o6bKGckchRB0dW3us6ivrw8qILT26PpePxCJ7ledyT9J3vSVuJx7t20AWojVcJvEdmODVXV1NT/4wQ945ZVXKC4uprOzk0uXLvHe9763ALAIITRNA1AU5Rng0zd8pAghzvk+LoT4K40udRGoVBTlCJsi9SHgoxqdWyKRSBKChKusDg8PMzg4SHNzs7dyEo/WqBC42YAHu91Oa2sr2dnZFBYWapqfqmacy+VifX2d5ORkjh8/HlSAxLOymijWAf27I89tjRe7LTa1EqvR5KwWFxczOnrTfjk2NkZxsbrWjOGO/fKXv8w999yDyWTi/PnzHD16lL/6q78C+HPgc4qihO4vHDmjwH8oinLY905FUU4qivIPWl3kRlX4EeAnwDXgh0KIq1qdXyKJJ4m8yUqyt0ioyqrL5WJjY4Pm5ma/D8tIBWgkwjbQ2LW1Ndrb26msrCQ3N5eJiQlVm7a0sgHY7XYuX76MwWCgrKwsrPi5lT2rpjvuQtugMhXXdFiDPrZfKqvRdLBqaWmht7eXwcFBiouLOXv2LP/8z/+s6pr33nsvf/Inf8Li4iIAr7zyCk8++aT3cU9KwIc+9CHvfdevX+fll19+jc3oKk1f4EKILyqK8ibw74qi/CFgBD4HpLOZiarltX4MRLejUCLZRcJlrUZDpDv1f620B4BXRqo0n4tk50ioyqper6eqqmrbB6VWrVHVjF1cXKStrY3a2lpyc3MjOqdaG0Ao8Wi1WmltbaWsrIykpKSwIjNeldVgRGMBiNa3qhUrpdps3LYbwwfh73Z0lVae1XA2gEBpGAaDgaeffpp7772XY8eO8eEPf5iamhoef/xxXnxxs03txYsXKSkp4fnnn+d3f/d3qampASA7O5s//dM/paWlhZaWFh5//HHvZivY/L0KIXC73d4vmEePHkUI8ZwQ4r/Gybf6c+A88K/At4HHhRBNQgit/KoSyb4m3purPEJVkvgkVGVVC2LZYOVpOtDY2OhXOVIr9mK1AaysrNDR0UFtbS0ZGRn09/erum48qqZqnvPq4XrSh2LrDmU89b8Fvr5LffxYPNmaBOAhWCJArO1S94JYVVNZDZazet9993Hfff4b25544gnvzy0tLYyNjQU89uGHH+bhhx8O+Jjnd7pTDRMURflr4NeBHwDHgC8B/7uiKK1CiOCldYlEsiv8WmmPJtXV3W4He6tyy/3Wo6msCiEYHBxkbGyMlpaWbUuckYjQaG0Ac3NzdHZ2Ul9fT0ZGRtBxwZ5DpNeLFUteOQDzFXfGdB5n2t7oeDWWvz2JKFgigIdAX4puBRuAzWbb8Xaru0A7UC2E+IIQokcI8VHgP4A3FUU5ustzk0gSGi123Muq6v4i4cRqrJWTSMWqy+Wiq6uLtbU1b2edaM8ZbRrAxMQEfX19NDU1+S2vqhWiaual5QYrj1CNJ/2lkcVU7Qaerk4TExPY7XZg98XmTtgA1tfXdzwNYKcRQvytEGJ9y31/yaZvVfpLJbc8e8GvuhUpYBOXhBOroVAjuCKxASiKQm9vL8nJydTU1ITcca9lJJVHCAohGBgYYHJykubmZpKSkgKOC3euaCqry8vLbGxshJ2r1jiqGuN+DVdN4KV7X2byT8aLy4kYAAAgAElEQVR8nZaWFiorK3E6nd4oJYvFgs1mi7qSvdueVzXncLlcGI1ab75PDIQQrwF37/Y8JJK9Tjz9qsFEaTw3WWX12ON2bsk+8qx6hGC4rFO1gtFmszE5OUl+fj7l5aErhVqLVU9Ft7u7G6fTSUNDQ0BxoFashsP3PB7Lw+zsLLApTMxmM2az2a+qG+jattFrEKCqOl9xJzl9/xF2HrHSX3oPZeP/n6qxwQTrZOW7tJsQkJqaSmlpKaWlpd6Oa6Ojo/T395OVlUVubi4ZGRmqq61aeF616GAVq+DdzwghRsOPkkgkgZCh+5JA7Bux6qmYqgnmD8fq6ipXrlzxColwRFoxDYcQgqmpKQoKCqiurg46Z7UeWLW2ByEEPT09OBwO6us3d8g7nU7m5ubo7+9nY2OD7OxscnNzw/6eA84jPbj/dLHp18m6tBMt3KNnMqOawuXuqI83mUwkJSVRXV2NwWBgcXGRqakpenp6SE9PJzc3l+zs7JC/271gAwhlZdivEWgSiUQ94SwA0VRV1VoAtFjqtwyG/9yPFJk5GxsJJ1ZDCTctOjXNzc1x/fp16urqmJ+f17xiGm6cw+Ggv7+f5ORkKisrQ45Vs8SvNrrK7XbT0dFBUlISNTU1OBwOr/gvLCyksLAQl8vF/Pw8o6OjrK6uIoRgcXGRjIyMsNeZr7iTrOnohZ7WrKea43r+YFmrHqGn1+u9FWshBCsrK8zNzTE0NITJZCI3Nxez2bytfe5ue17VslO78iUSyd4iFq/qXq+qyiSA3SPhxGowIvGiBmNsbIzx8XGam5sxmUwsLi7uqFhdX1+nra2N/Px8nM7wsZBqK7Xh5uZ2u7FYLOTl5XH48GFvAgJs+g9dLheKoqDT6cjLyyMvL4+1tTW6u7uZmZnZFPdFB7ed127Y35tshtLryBHTfvfZjaleoRoovirQMryiKGRkZJCRkUF5eTlWq9Wb/iCEICcnh9zcXFJTU/dEZVUikUiiZbeqqrIpQGKzr8RqtJVVIQS9vb2sra3R3NzsXYbV6/XeXdyh0CK6ymM9OH78OACTk5Nhz6c2uirUGIfDQVtbGwaDwStU3W43BoPBW3F1uVze/3uEq6IomEwmqqqqEEKwMbY3qqYDxe9S7Vv1ZaW0noMjsWXCqkHNBqmtPtf5+XkGBgZYX1/H5XJhsVhITk6OqnoZ6watcDidzoCJGRKJZP8Tj6rqTi7/Q3QWALm5Kv7smxJLsNao4XC73Vy5cgW32019fb2fX1DtOWONrpqfn+fKlSvU1dWRlZWlumIa6wYrm81Ga2srt912GyaTyStUPWIUNp+b0WgkKSkJk8mEXq9HURSsVitCCBwOh2Y+xWBJAN2FscVU9R7cnpG6W0S6jG8ymSgsLOTEiRM0NTWh1+uZmZnhwoULdHV1MTs7G9HrPt6V1VshtkoikURHsKrqXl/+l+w+CVcCCfZBH6kNwCO02traKCgooLR0e8B7PHb5bx03OTnJ0NAQTU1NJCcnA+qX92OJrlpbW6OtrY1jx46RkZFBT08Py8vLpKenh/QF63Q6FhcXGRgYoKamBkVRVP/e+wvfSfnkz723l/LivywTq9ANx7ySv80KEIpYPKd6vR6DwUBVVRV6vd7P55qUlOT1wG71ufoSq1gNN3+r1SrFqkRyC5Lom6rCIf2qu0vCidVgRNpG1WKx0NHRQUVFBXl5eUHHaZ2f6svQ0BBzc3O0tLT4LZ1qed1AwmJ5eZnOzk5OnDhBeno6LpeLyspKhoeHWVtbIzs7m7y8PDIzM1EUhV978KL32O8/fZiBgQEaGhq84nptqDPsXONBf+k9lI+8qtn5Vkrr/W7P5J8kb/pKwO5VsaBFB6twPtdAcWMQ/zQBWVmVSG49YhGqO7H8r9avGo8UAJBJAFqwb8RqJGkAbreb9vZ2Tpw4ETKaKloRGg5PRJTdbqexsXHbh388K6vz8/P09PTQ0NBASkoKLpcLIQT5+fnk5+fjdruZn59nYmKC7u5uvvSU/0vkY48Mseke6eCV57QVcZa0QtIs4b26u4na+Cq7MT7tRoOJxVA+1+zsbMxmMxkZGTGLVZfLFVas3gKtViUSyQ12c/f/TnWkClVVlX7VnWHfiFW1ldWpqSmsVivNzc1hM1S1isPyRQjBlStXSE5O5sSJEwGFrpa5rb7nn5qaYnBwkKamJkwmkzdxwFd86HQ6cnNzyc3N9auoBsLz+I++mniVtF6qqSR+m8IGnWUcMQzE5dzhvhx5fK6euLHFxUUmJyfp6elBCIHBYCAjIyO6rNwwjTdkZVUikfgSjU9VTVVVrVDd7aqqRBsSTqxG61kVQniX3bOyskL6+nzPqaVYdTgcWK1WSkpKuO2224KO07KyCpsCY3R01Nu2Va/X43Q6/TZSbSWcUI2Wrb7V/UiPUuv9OVB81U6yNc+1o6MDq9XK22+/rdrn6ku4yqz0rEoktw675VPdqYoqyKrqXiHhxGowdDodDocj4GNut5vu7m5cLhdNTU10dnaq3uUfa3arB5vNxuXLlzGZTCGFque6WnpWLRYLQgiampoA/OKnAhGJUP3+NwrArs5btNdZS8rcdt9M/smAY4fS64KeZ91hIMUYPid3p1EUBYPBQGlpKWlpaap9rr5IG4BEIoHd86lGIlT3QlVV+lW1Yd9sbwtWWXU6nVy+fJmkpCRqa2u9O9q13DgVDovFwqVLl6iqqsJoNIYdr1VlVQjBwMAATqeTkyc3RdfWaKqtxKuiqiUDxe/a7SmoYtBZtttT2IZvZdTjc21sbOTEiROYTCb6+/u5cOECfX19LC0tbXt9hbMB2Gw2WVmVSPY5u5WnupMVVZBV1XAoivJeRVF6FEXpUxTlCwEef6eiKG8riuJUFOUBn/vrFUX5D0VRriqKckVRlAfDXWvfVFYDiVWbzUZbWxulpaUUFRWFHBsILcTq4uIiXV1dnDx5kvT0dFXHaOFZ9bRPNRqNpKWlee+Lh1C1mjJIjVN1tS+9GdiMoKqe1G7nfyB6LaVUpo1EfFyo+KpZpYDtvb12j2DL+KF8rgcPHsRsNpOdna3KBiArqxLJrU2kPtXdFKrSqxodiqLogW8B9wBjwEVFUV4UQnT5DBsBPgF8fsvhVuC3hBC9iqIUAZcURfmJEGIp2PUSTqyGygD1FXiejlDHjh0jOzs75NhgxCpWp6enGRgY8MtQVUOslVWn00l7ezs5OTkUFBTQ0dERF6H6/W8URHzMThCui9WV1LtiOv9kRjUEeVl0rR+N6dzxRk0awFafq2+eK4DRaMRutwf0uUobgESyv4l2+T/RhKoWuar73AJwCugTQgwAKIpyFvgNwCtWhRBDNx7z+8cXQlz3+XlCUZQZIBcIKlb3pQ1gbm6Ojo4O6urqtglViJ9Y9RWOIyMjjIyM0NzcvE2oRhvkH2h+W89lt9u5dOkSBQUF3Hbbbej1evR6PRcuXKC/v5/V1VW/Y37twYtxWfq3G4IvBfcXvjPoY23J0QnJ/tLQ4f97qYvVbhJpdJUnz7W8vJyWlhYKCja/oHR2dnLp0iVvNq+HUGL1/PnzVFVVUVFRwVe+8pVtj29sbPDggw9SUVHB7bff7hXHQ0NDpKSkUF9fT319Pb/3e78XwTOWSCRaobVQDcevlfZELVTV+lWj4RaxAJgVRWn1+e8zWx4vBkZ9bo/duC8iFEU5BZiA/lDjEq6yGgyPWB0bG2N8fJympiaSkpJCjg1HJGLVM1an09Hb24vVaqWpqWmbMPAIzFDxQ2pzW7eKWpvNxttvv01FRQW5ubnex06ePInL5WJ+fp7BwUHW1tbIycnhkT/Vbum+X7f5xlDuDvzGsqAL3HjBlz5neeTX1bgxQDzon7JRXqC+sh5PYs1ZNRqNZGdn++W59vf3Y7PZ6OjoYGRkhOLi7e9XLpeLz372s7z66quUlJTQ0tLCmTNnOH78uHfMd7/7XbKysujr6+Ps2bM8+uijPPfccwCUl5fT1tYW9bwlEkls7PTO/3j7U6OtqqoVqrtdVdVvxFwdnhNCNGs1n0AoilIIPAv8thAipNjaN2JVURSWl5dxuVzeeKZgxCPsX6fT4XQ66enpwWg0UldXF/B4j8AMJRgiEaueKqnFYqG9vZ3jx4+TmZnpDfv3LP0bDAZv8L/L5eJ9H31b9XMLx4wrP+JjdqLVqhZcXT5MZYY2jQrUWDviebznHLF00PJNA9jqc52enub555/n5Zdf5rXXXuNjH/sY7373uwG4cOECFRUVlJVtbjp76KGHOHfunJ9YPXfuHH/2Z38GwAMPPMAjjzyiyXOWSCS7RzTL/7EK1ViqqrKtqmrGgUM+t0tu3KcKRVEOAi8Djwkh3gw3PuH+VQJ90Lrdbnp7e3G73dTX14cNO4+kNWsktLe3k56ezrFjx1R7a2PBI1aXl5dpb2/n5MmTAYXqVrQWqi5x8xqeCmsi0Et13K9hcdxcEo9VKMZaFfUQjzno9Xre//7309TUxN/8zd/w6U9/GqvV6n18fHycQ4duvq+VlJQwPu7/vuY7xtO4YH5+HoDBwUEaGhr41V/9VX7xi19EPX+JRBI5O+lT3Ykd/9FuqkqUquoOcRGoVBTliKIoJuAh4EU1B94Y/yPgn4QQL6g5JuErqw6Hg7a2NnJycnA4HKo+iENlskbDxsYGKysrVFZWapahqgZFUVhZWWFsbGxb+9RgokYrf+o3vl4HBN4BH8qvup/x3VwVKGs1VrEa6/Fa4Ha7QzYQWF9fJy0tjTvvvFOzaxYWFjIyMkJOTg6XLl3iAx/4AFevXuXgwb2UsyCR7E/2m1ANhYyqUo8QwqkoyiPATwA98A9CiKuKojwBtAohXlQUpYVNUZoFvF9RlD8XQtQAHwbeCeQoivKJG6f8hBAiqNcrIcWqp6JotVppa2ujoqICs9nMzMyMquP1ej02m02TuaytrXkrqmazOez4QJuiomVlZYXZ2VnuvPNOb/tURVHiLlQ9BFv+H3WXckgXeQSU1nQXht50pSXzSngrhCeRIVq0qqzGQrimAMFyVouLixkdvenFHxsb2+Zt9YwpKSnB6XSyvLxMTk4OiqJ4/edNTU2Ul5dz/fp1mpvjaqeSSG55ElGohrMABKuqarX8f4tUVQEQQvwY+PGW+x73+fkim/aArcd9D/heJNdKOBuAh6WlJS5fvkxNTQ15eXmqd9CDdtXNpaUl2traqK2tJSUlRbUPVotrj4yMsLCwQElJiZ9QTZSw/3l95D7XYOIzXBJAJPRaSsOOubZQFHZMIEJVvHfieC2INme1paWF3t5eBgcHsdvtnD17ljNnzviNOXPmDM888wwAL7zwAqdPn0ZRFGZnZ722nYGBAXp7e73eV4lEsjtoJVRj2fEfKfFe/pfEj4SsrE5NTdHf309jY6O3ihPpZqhYBePMzAx9fX3eOajNRo312kIIbwRVWVkZa2trOy5UNy0AsTG4ZCYnPbCNoC35Luptv4zp/APF7wqahRrwmo7AbVUBepcLNdlkpYVnNVYbQKxVfTUdrAKJVYPBwNNPP829996Ly+Xi4Ycfpqamhscff5zm5mbOnDnDJz/5ST7+8Y9TUVFBdnY2Z8+eBeDnP/85jz/+OEajEZ1Ox7e//e2AkXQSiUQ7YulStZVQQnUvoNXy/61UVd1pEk6sekLKW1paVLUuDUQkG6wC7d4fHR1lYmKC5uZmr38vkuzWaAWDEIJr164hhKC+vp7Z2Vnm5uZIS0vDbDYHFDK7VVHtsFZxInVvvBHFg4GlHMoy51WNnbelMz8EJwtsCW8DCDeHUDmr9913H/fdd5/ffU888YT35+TkZJ5//vltx33wgx/kgx/8YJQzlkgkkaLl8v9OCtVQFoB4d6qSQjW+JJxYVRSFqqqqmKqT0ean+lY1t8ZjRRKHFc3c3W43V65c4cCBA5SXl+N2u8nMzKSyspKZmRn6+/tJS0sjLy8Ps9mMXq/fc0v/WjBQ/K7dnkLUxFoZ3QuV2XCeVavVyoEDB2K6hkQi2T0SVahGi9xUlRgknFjVgkgqqx4R6na76erqQqfTUV9fv+1DPx5dsTzixOl00tbWRm5uLqWlpd756HQ6srKyyMrK8lacZ2ZmGBwc5PFvhI7vipYvP3kK2IjLueONt4uVM/Q4rUnSu9hw6WP2nMZaWdXC8xrOBrCxsRFRa2GJRJI4JKpQjWZTlVz+31sk7AarYGjtG/XEXF2+fJmUlJSgGarxsAEIIbDb7bS2tlJUVOQnVLd6VD1tMSsrK+MmVAEOGNUL1Q5r6F2Zras1sU4nYtadgbuaxUrndG7YMUNDQzgcjphsILttIwh3jr2wCUwikURHND7VSFqpxlOoatleVQrVvce++lSJR3UTNsP+CwoKKC8vjznsX60NQKfTYbVaaW1tpby8nMLCwqBC1Zf9uPQfT0I1Bri6fDimc687/BcuDh48iM1m48KFC1y7do25ubmImlPEKja1EKvhbAC7nQMrkUiiQ8t2qoGqqru19B/PqCopVHeOhLQBBPtA9Czva9XBymq1srS0xNGjRwP2O/dFa6Hsdru9sVgZGRlhu1LtJZG6ZLs1mwKEo2u1msNZVo4ePcry8jKzs7MMDAyQmpqK2WzGbDZjMAT/k9wLYjWUDUAIIdujaoiiKB8C/gw4BpwSQrTu7owk+5V4+1SjEaofz3qDZxffEfFxatBi+V8K1Z0lIcVqMLQUjMvLy3R2dpKVlaWqU46WNoClpSXW1tZobm7eU0I1UGSVb6tVLWlLvos04/bGDRdGCjhVOuV3X+tYAc0lU9vGRkuvpZTKNP+mBoHiq7YmAsyumshND/1G5xF0mZmZZGZmIoRgbW2N2dlZ2traMBgM5ObmYjabvUH4vsfuBRtAuDnI6qpmdAK/Cfztbk9EcuuihU81HB/PeiOq4zwEswBEmgAghereZV+JVbUV03DCcnZ2lt7eXhoaGhgaGtK0YhrOBuC59sGDB3e0fWq86LBWUZy2uNvTCEkoK8BWom0IAJCVaifnQA4ulwuXy+XtNpaWlkZaWhpHjhxhfX2d2dlZrl69ihDCK1xTU1P3RGUVgotRWVXVFiHENZDiXxJftPKpRruhKpRQjUd1VatOVZKd5ZYUq6He/MfHxxkbG/NmqEZSMXU4HKrGBTvfxMQEIyMjNDc309HRgcPhQK/X7xmhuubYrPRFsslK4k9BQYHXe+wRrZ7XrF6vJyUlhdLSUkpLS7Hb7czNzdHb24vdbicpKYnk5OSoK6zxzmn1zFGysyiK8hngMwDJ+vRdno0kkdDSpxqIUEI11mpqOGRVdX+RkGI1lGc12vxVIQQDAwMsLS3R1NTk9Q5q7UUNNm54eJjZ2Vlvfmt2djbt7e1kZWWRn59PZmam3/PeaaG6GVmlPa2rNTSnX6XPWa75udvGcqgvCR/cv+aI7Q057DyGUnG5oPbwpsjX6XTodDoMBgNutxshhJ9wFUKg1+sxGAwUFRVRVFSE0+mkv7+fpaUlLl68SFZWFmazedvrIhTxFqvr6+vejnISdSiK8u9AQYCHHhNCnFNzDiHEd4DvAGSY8mV5W6IJ8fSpailUI0kBCFZVlUJ175OQYjUYOp0uot3VHjydodxuNw0NDX4f6FqL1a02ACEEfX19rK2t0djYiBACp9PJoUOHKCkpYXFxkcnJSbq7u8nMzCQ/P5+H/kt/xM9xv7PVt3phJNDnP3SMpnHikCXs+XotpaqvrSa2Sq+HRauJH3fAU1/+Ba88t5n56nmteb5oeYSr52ePVcDXLlBYWMji4iLT09Ncv36dgwcPkpubS1ZWVsjNhfEWq1arVYrVCBFCvGe35yC5NYnn8v9uVlQhcFVVCtXEZl+J1UjC/j24XC6uXLlCenp6wGgqtdVatfmpvnYBIQRdXV0oikJdXd22aCpFUcjJySEnJwe3283i4uKeEKr985tvBOU56sz0raP5NB+ajueUdoTe5cKIj3nreipJAQoUvpXxUMLVYxcAcDgc3oq/53UhhFCdLKBFU4FQhGq1KpFI9g7RLP/vRaGqZbZqKKRQ3X32lViNND/Vbrdz+fJlioqKOHToUNBzqvXBRlKBdbvdtLe3k56eTllZWdgMVZ1Ox0d+fyD8k4ozUys3uxP1z2dwOHtlF2ez+wws5ageOz6jUJy3XfCFE65Go5G1tTWmp6c5duzYtg1avskCFouFubk5v2SB3NxcTCaTJu1eQ4nd9fV12b1KQxRFuR/4KyAXeFlRlDYhxL27PC3JPketT3W3hWowtK6qSqG6N0hIsRouZ1UNbreb1tZWKioqyMvLCzouHp5Vl8vFpUuXyM/P59ChQwkT9j+5nEw8NiYH86u+NZDD7WXhPaex0jVs5Pht4TfHqUFNfFUoAglXm81GR0cHNTU1ZGRkeKutvpu0PMI1PT2d9PR0v2SBzs5OAEwmU0zL9OEqs7Kyqi1CiB8BP9rteUj2F6GqqpH4VCMhHkI1UFU1kk1VUqgmFgkpVoOhVqyurKywvr5OY2MjOTmhK2Na7PL3xel0Mj4+TnV1tXdnuKcr0F4VqqE2Vw0tHIy4ujq4ZI5pPoGyVrXmSr+B6sOxvUFfGYxxJ63Pv/sP/7aajIzNN2KPhxXwvn58N2lB4GSB3t5eZmdnWVhYICcnh9zcXNLS0jTboCXFqkSyt9kNn2osQlWL2CoZVbU/2Hdi1W4P/W1pfn6e7u5ubwUqHFpWVq1WK319fWRmZvpFGO1loQqbFdVQdE1kcLwotH81lG91aT2ZzJTtDQC05sKg+iX7ePK5L/4KT335FxEd8+Hf7fW77WsX8BWuWyOxfJMFDh48SFZWFnl5eczPzzM8PIzVaiUrK4vc3FwyMjJCCtdw3eHkBiuJJHGJx/J/vJb+1VZV5fL//mFfidVw/lLfHNPOzs4diaTysLq6ypUrVzhy5Airq6thu1LB3hCqn/jcrwIyDQege0hHxaHYfhfjMwrW9dgqthDc5xooEsvzunQ6nRgMBnQ6Hfn5+eTn5+N2u1lYWGBycpKenp6QyQJqKqsHDhyI+blJJBLt2Y3l/0RFCtW9R0KK1Wg8q4ODg8zPz9Pc3IzBYIio21WsXbEWFxfp6uqirq4Oh8PBzMyMVzjsZvtUyc6i10MUyWphUZMs4PGvVldXe6uunoq+Jz0gULJAbm4uOTk5XgEsN1hJJIlHoi3/Q3ALQLyrqlKo7k0SUqwGI5CwFELQ3d2Nw+GgsbHR+2GrtmIaSXRVoHGe9qlNTU0kJSVht9tJTU2ltbWVtLQ08vPzycnJ8VaxEk2oWm2bYttjBViy7d4y8NasVVDfGCBaro8olORvv9/Xr7phJ2B8VTwIJFydTicdHR0cO3aMzMxMP6sA4CdctyYLzM7OMjIygslkCls1lZ5ViSTxSKTl/1hRm6kq2XvsK7G6VVi63W6uXLlCamoq1dXVflXMSCqr0YrViYkJRkdHaW5uxmg04nQ60el0HD16FCEEq6urTE9P09/fT1paGv/ty/H3bUbCJz73qxxI3jsWgLYxbTynahsDBKJvVNlmBRibFpTk33xtza6GfvNPTdHx6x+/i5ef/WVUc1CLr3D9wV9XkpmZCdy0CxiNRr9kAc/PW5MFysrKsFqtjIyMsLCwwKVLl7yRWL4e1fX1dbKysuL6nCQSSWRoVVXdym4I1VirquGQVdW9y77aJucrQB0OB5cuXSI7O5ujR49uW26PdxvVoaEhJicnaWpq8gpVjwiATSvDwYMHqays5I477thzQjUQ4TaNh6uqto4GKEEG4eroZhXvrYHQArV1LHCnKrVc7Nu8Ttewcdtj3UP758/jI78/wK89eHFb5d4jWpOSkjCZTOj1ehRFweVy4XA4cDgcuN1ub7OBwsJCTpw4gV6v5/r161y8eJGBgQFGRkZCbrA6f/48VVVVVFRU8JWvfGXb4xsbGzz44INUVFRw++23MzQ05H3sySefpKKigqqqKn7yk59o+nuRSG5lYgn/jydapABsRS7/JzYJWVkN5Vl1u93YbDYuX75MWVkZ+fmBBZIWXtRA44QQ9Pb2sr6+TkNDg7d9aqiNVPc+1Br2/InAxIKJouzollku9qbSUmlVPX4n4qv2K4GsAhA+WcBut6MoCkajkeLiYoqLi3E6nczPz/Pkk0/y6quv0tjYSHFxMXfeeafX2uJyufjsZz/Lq6++SklJCS0tLZw5c4bjx497r/3d736XrKws+vr6OHv2LI8++ijPPfccXV1dnD17lqtXrzIxMcF73vMerl+/HjKVQCKRbBLNpqqt7JXl/1iqqlKoJj77p3TE5oftxsYGly5dorq6OqhQhdi9qFvxdLC6evUqLpeLkydPettl7vUd/4HYTAGInclF/zfEWDNW1XBhJLZq662Ep9q6terqSRXwrbhOT09z4MABnE4ndrvd63XNz8/nqaee4r777qOlpYVnnnmGL33pS95zXbhwgYqKCsrKyjCZTDz00EOcO3fObx7nzp3jt3/7twF44IEH+OlPf4oQgnPnzvHQQw+RlJTEkSNHqKio4MKFCzvzy5FIEph4Lf8HQyuhGktVVWaq7l8SsrIKm+Jwa6/ylZUVVlZWuPPOO0lLSwt5vNY2AM+O64KCAo4cOZIwXan2Ar1j++sNprvfQUbGdlsBxC8RQCs8r0nfiquiKHR1dXnjrrZGYnlEq91u5+677+Yd7/D/sBkfH/drZ1xSUsJbb70VdIzBYCAjI4P5+XnGx8e54447/I4dH5dVEIkkFmJZ/g/VUjVWYk0ACISsqu4P9o1KmJqaoq+vj9TU1LBCFdRvsFIzzuOPNRgMlJWVJXRFNVba+3b3+0+sHlbY7F4Vir7R2HvOejZZ7XWEEFy7do309HQOHTqETqdDr9djMplITk72Vl1dLhevv/46KyuRdTOTSCTao8XyfyD2yu5/rTZVSaGaOOwLsTo8POzdda+2daRWldWNjQ1aW67SaewAACAASURBVFspLS3FZDL57ahOVKGqlQUgEPOr2nkN1S73t43laJYkEIyx6dCpCRsJlJjy5f8mGBwcZG1tjb6+PvR6PUeOHAk41mMX+PKXv8zp06c5ffr0tjHFxcWMjo56b4+NjVFcXBx0jNPpZHl5mZycHFXHSiSS2IhlU1W8l/8DVVXVIqOq9g8JK1Y9NoCenh6WlpZoamrCZFL/jVGL6Cqr1cqlS5c4evQo+fn5uFwulpeXE1qo+rLTsVUXewNndIZLBFCD3eF/u2M0fPUdbiYCBEoGuD4S/IvR8rIj6GN7mVeea6GhoYGkpCTa29sZGxtDp9Oxurq6zXYDm5XXb37zm0xPT/P0008H/BtsaWmht7eXwcFB7HY7Z8+e5cyZM35jzpw5wzPPPAPACy+8wOnTp1EUhTNnznD27Fk2NjYYHBykt7eXU6dOxefJSyT7gJ3eVBVP5KYqiYeE9ay63W46OjowmUycPHlSdUXVQyQbpwLhaZ9aW1vLwYMHcblcHD161Ntv3Ww2k5+fT1pamvcciSRUAUan/MVJaaG633F7n4G6Cmc8phRXuoaNHL8t/iKzr3+VoqLNyKy7ztzOL198K8wRO4PHp2o0GhFCkJKSQktLCwsLCwwPD2OxWMjOziYvL4/MzEwUReF73/sev/jFLzh37lzQHfoGg4Gnn36ae++9F5fLxcMPP0xNTQ2PP/44zc3NnDlzhk9+8pN8/OMfp6KiguzsbM6ePQtATU0NH/7whzl+/DgGg4FvfetbMglAIgmC3FSlDilUE4+EFatXrlzh4MGDHD58OKrj1YrVQCwsLHDt2jXq6+tJTU3F5XIhhCAnJwez2YzL5WJubo7BwUGvcP39xxajutZOE8oC0DvsovK22ITCyLiDnOrI31zisYy/WV2Nrnocq291YmKNoqIDQTdi7TS+G6pmZmaYnJykoaEBvV7v3VjldrtZWFhgcnKSRx55BIvFwsTEBD//+c/Drmrcd9993HfffX73PfHEE96fk5OTef755wMe+9hjj/HYY4/F8OwkEsleraruxqYqSeKRsDaAmpqaqIUqqLcBbGVmZoaenh6ampr8hKqnXaXn3Pn5+Zw8eZLHv6FPGKHqy+y8/+9mw75d2HtarUbK5e7ov8lriW1ju1CNpTnA2LSguz90Zbbn+qr354mJNUaGV7nrzO2qzh8vfIXqwsICQ0ND1NXVbatg6nQ6zGYzx48f5zOf+QzLy8u8853v5PTp03zta1/b6WlLJBIfErWqqha5/H9rk7CVVZPJFFRsCiHC2gKiqayOjY0xPj5Oc3MzBoMBp9PpPVcgEm3ZH7aL1K2Eqq7OLuwNEbpX8RWqe5GVlRWuX79OQ0MDRmPwim9nZyePPvooL730EqWlpQghWF7e2Q43EolEPfshqmorcvn/1iJhK6vBUCtCI62sDg4OMj097SdUfdunbiURherpB+5UNa53OPzvbScirLp61He82g22brJybAT+vS0v2VCCvI7ijaequra2xtWrV6mrqyMpKSno+OHhYT71qU/xgx/8gNLSUmDT152Zmbkj85VIJNuJpqoaC1pUVeXyvyQS9p1Y1WKXvy9CCGw2G6urqzQ0NACEbZ+aiEI1HmztXqWGi72pXB09sO3+ji5LTHPp6Y3t+HhztLl6x6/pEao2m42Ojg5qa2tJSUkJOn5mZoaPfvSj/O3f/i01NTU7NU2JRBIDe7GqGqtQlcv/tx77Uqxq1ZnK7XbT2dkJQG1t7b4O+z/9wJ2kpNxc3tfrQ9so1FRXd4POa1bVFdf+ge3L8r3X47Oc3duzFPB+g3F3K6p2u5329naqq6tJT08POn5lZYWHHnqI//E//gd33qmuAi+RSOJPvBoA7FXk8v+tScKK1WBiUafTadKZyuVy0dbWxoEDB0hJScHpdO5boRovOq+tBX1sr2yyUsPQ0HZROzq6xvhY8OcXDVWnjmt6vmB8/bEkpqamsNlstLe3U15eHnIZ32az8bGPfYw/+IM/4H3ve9+OzFEikYRHq01ViVRVDYRc/t//JKxYDUYkNoBg4zztU/Py8jhy5Ah6vZ6lpc3K2H4Uqmq9qvGgdyy2l2DntfBV1FgtBGoZHbWwuGiL6lhT8s5EWP3kbDNHjhzBYrHwxhtv4HK52NjYwG4P/GbvdDr59Kc/zfvf/34++tGP7sgcJRJJ7MSrqhqLX1Uu/+8vFEV5r6IoPYqi9CmK8oUAjycpivLcjcffUhTl8I37jYqiPKMoSoeiKNcURfk/w13rlhWrwewCNpuN1tZWbrvtNoqLi3G73ZSVlTExMcFbb71Fb2/vtm4+iSxUY6H7+vpuT2HHCFRdVcPysoOr7bPe28E2WXmIZ3X1ledaUBSFtLQ01tbWqKys5MSJE9jtdtra2rh06RKjo6NsbGwAmzaYz3/+81RVVfGHf/iHETfekEgk8SPRqqrxEKrhkEI1fiiKoge+BbwPOA58RFGUrR9gnwQWhRAVwDeAr964/0NAkhDiBNAE/K5HyAYjYaOrghFJZ6qt7SOtVittbW1UV1eTlZWF2+3G7XaTmZlJVlbWtrD/3Nxc/sufLMTrqewIu1lVDcbVrmVqjqtb/tltxsfWKC7ZviEsEgxGHU7HzddsaU05o9cGEVE2rQiEx6MqhODatWukpaVx6NAhAI4cOcKRI0ew2WzMzMxw+fJlvvCFL5Cfn09qairf/va3pVCVSBKIveZVjaVDVSjk8v+ucgroE0IMACiKchb4DaDLZ8xvAH924+cXgKeVzQ8TARxQFMUApAB2YCXUxRK2shrsw1NtZXXr8SsrK1y+fJna2lo/oerrUfUN+29ubk54oeqL7+aqrQRqCOAhXHU1lG81UguAlokAV68F/7sYHAj5NxM3tloBFJ3O+18s+Ib+9/f3o9PpKCsr2zYuOTmZ0tJS7rjjDs6cOcPCwgLLy8u84x3v4PXXX49pDhKJRDviWVXVmlBCVS7/JzTFwKjP7bEb9wUcI4RwAstADpvCdQ2YBEaArwshQgqqfVdZVZsG4MvCwgLd3d00NDSQkpLi7UoVajPVfR+7rMV0d5WGdzdEfezqcuhOTYEYGfc/xmJxkpYWn5dgV4+V41WpYcf1D6xSXhZ8F3w0LC7ayMpKBmBsKHAKQCDmp5ZIzzyAfd3OoWNHGL8+gvvGFy9fweqpuAYSsVursb5CdXh4mPX1dWpra0NWSl944QVee+01fvrTn5KcnMzS0tK2VQiJRLI3ibWqqqUFIF5CNRxSqILeJmKtPJsVRWn1uf0dIcR3YpyWh1OACygCsoBfKIry754qbSD2nVhVmwbgYXp6moGBARobG0lKSvJrnxqM/eJR9QgqgOnpmxVSl8vNoUNpuzGlqOi8ZqX2WHhhGitDQ6scPhydsLXbHN7KqWPDhTHJv5I9P7UpaleXNivR65Z1sotycWw4WJ3ffMxXuCq6m2JTuG8KSY+AFW63n1CdmJhgYWGBurq6kEL1tdde46//+q85f/48ycmbrw8Z+C+R7B12uq1qtMRTqMrl/x1hTgjRHOLxceCQz+2SG/cFGjN2Y8k/A5gHPgqcF0I4gBlFUV4HmoFbR6xG0pnKbrczPDzs7UoVTqjuF5G6lcVFG3q9/3MeHbWoFqzd19fJMQfvepQoqM1YHR31tzaMj63hjqHyOD26ffXD8zo2JhlJz8nEYbfjsjtxbNx8kxbCU2HdLlx9herMzAwTExM0NDSE/BJ26dIlvvjFL/Jv//ZvUqBKJAlIJFXVeFkAwvlT1bRSDYVc/t8zXAQqFUU5wqYofYhNEerLi8BvA/8BPAC8JoQQiqKMAKeBZxVFOQDcATwV6mK3pGdVCMHAwABOp5OmpiZv+1TglhGqHgtAqKil0VFtIp9C+VaDcbXr5hvp9b7t89ipVqvRpgCoZWp4Puhjev1m9dWYZMRoMqE3GUhOT8VgMmAwGVAUnfc/D4pO4Sc/aPLeXlhYYHBwkLq6Ou/5AtHT08NnP/tZXnjhBQoLCzV4ZhKJRGvi3VZVjQUgmBh9dvEdUQtVufyfeNzwoD4C/AS4BvxQCHFVUZQnFEU5c2PYd4EcRVH6gD8CPPFW3wLSFEW5yqbo/X+EEFdCXW/fVVbDpQEIIejp6cHhcJCamooQ4pZsn2qzOlhUMW501EJefvyW2Lu7Fqg+nq16fEeXhRPHA1d81WSuqmVwYIUjZQe33T80tLqtCh2IxUUb81MrJKcEr3QEE6o6gw630/817HI40Sk6nA4HOr0et8uFwbT55+u0O1EUHV96xEZqaipTU1OYzWasVivXr1+noaEBozF4juvExASf+MQneOaZZ6ioqAj73CLF5XLR3NxMcXExL730kubnl0gkwYnHxiotd/fL5f/ERQjxY+DHW+573OdnG5sxVVuPswS6PxT7TqyGqqx62qcmJSVRW1vLW2+9hc1mIzk5+ZYSqsfurCU59aZ4CSe+ZqatYQVr3/VFKo5m3Txmxj8lYOvmqmBEuxM/kFDt6rHicvm/Uff0Wqiq9Be7WmyyWpy1kJUb2jbh61tVi+f17BGwhhui08lND+u/fa8e2PwiZrFYmJmZYWBgAJvNRllZWcil/4WFBR588EGeeuopGhqi33AXim9+85scO3aMlZXdSVmQSPYD4aqqey2uaivx9qnKqur+5pYRq572qdnZ2Rw+fBi3201RURHt7e2kpKRQUFCA2Wz2Wyrdr0L1Vsc3wspD/8D25f7BgRU/P2g02NbtAaur4wMz3teacItt1/GtrtptG9vuO/d3x7adU1EU0tPTMRqNzM7OUltbi8Vi4fLlyxgMBvLz88nNzcVk2pzP2toaH/nIR3jssce4++67Y3qewRgbG+Pll1/mscce43/+z/8Zl2tIJJLA7MbGqkDE0koVpFCVJLBYDVYJDWQDcDgcvP3225SUlFBUVOTNUC0qKqKoqAiLxcLU1BQDAwOkp6eTn5/Pxx4Z2oFnkTiMDC5TeiT0m8vW6qovKyt2Dh70F21jI9FV2jq6LKqW4oPhdMb2Bj47uUxu4c3fxfxMfHyt62vbq8X/8tehl+kdDgft7e1UVVWRlZVFXl4eZWVlWK1WZmZmaG9vp729ndnZWS5cuMBv/dZv8Zu/+ZtxmT/A5z73Ob72ta+xuhpf7+9+RFGU/wt4P5uB2f3A7wgh1GehSfYNWlZVdyJb1ZdINlRJn6okGAm7wSoYWyurnvapR44c8ROqHo+qpxpVWVnJHXfcwaFDh/atUFVTVd26bA6b/lZfoslYDUV3V+AsYN9NVnuFhRDCdHH2ZsV2fiqwELfbHMyMB99U5WFteXv19/n/e3uQvy9Op5O2tjbKysrIyvL/0pCamsrhw4dpaWnh9OnTXLx4kd7eXr7//e/z1FNPRZxNrIaXXnqJvLw8mpqawg+WBOJVoFYIcRK4DoTtny2RREs8WqxqsaFK+lQlkOBiNVB11Vesrq2tcenSJaqrq8nNzQ3YlWrr+T70metxn/ducFvtzYqcr181EkYGd148xkOwBoqKCrTZaXZSmyKWbT3wm63vlyrfrNRAhBOqbrebK1euUFJSQm5ubtBxQgj+/u//nurqavr6+jh79ix5eXkhfa3R8vrrr/Piiy9y+PBhHnroIV577TX+83/+z5pfZ78ihHjlxo5bgDfZzDGUSPwIVlXdbQvATghVWVW9dUhosRoIjw1geXmZtrY2Tp48SWZmZlihCvvToxqOSJfTw1VV+64HzhiYm1pl4Hr4iqLWTAzF55qzk9GJ6IXp8ALYsuhfvQ0nVIUQdHZ2YjabQ8ZOCSH4xje+wdzcHH/5l3+JTqejoKCAj350azSeNjz55JOMjY0xNDTE2bNnOX36NN/73vficq1bgIeBfwv0gKIon1EUpVVRlFa7O3T7Y0niEe+4Kq15ZaRKClWJ5iSsZzUYer0em83G1atXI2qfup+F6m21FaSmJ2Nd3cxU3fCp9LndAnPB9oimYCwuWMnKTt12ny9bkwC0ZmJonkPl26uH44NzFB8xa3qtmfFF8ooD+3AD4WsFCIbb6UZn2HyDdrlcfpv6ohGq165d48CBA5SWloYc++yzz/LGG2/wv/7X/wqZuSrZORRF+XegIMBDjwkhzt0Y8xibARDfD3SOGy0QvwOQYcqXfXElQYm3XzWUPzWSDVXhkEL11mPfidXZ2VmsVivvfOc7MZlMUqjWVuCwO7CG2N8yd8NfGYlo1ZrurgWSkre/HGfGF1UJ0PHBuXhMKySzk8voAlSm15Y3myCkpKV477Ot27GuhM6BXZ2/mT4g3IJHf3uBoSEd+fn5pKSkBDymv78fnU5HWVloUfvyyy/z/e9/n/Pnz3vTAHaSd73rXbzrXe/a8evudYQQ7wn1uKIonwD+E/BuIWJokyZJSKLdWBWNBeCVkaqYfKvRCFXpU5WoJaFtAFvF58jICGNjY6SmpvoJVZ1Od0sKVQCH3YHR5JOpagj+Tz4XZFPQVsaHA2+I8rCyFLwrllZWgNH+2ZjGBPKtzowHtjAEuz/e1NXVYTAYuHbtGhcuXGBwcBCr9abgHR4eZn19naqqqqCvb4Bf/vKXfPWrX+VHP/oRBw4c0Gx+NpuNU6dOUVdXR01NDV/60pc0O7cEFEV5L/B/AGeEEDvTsk0iiZBQy/6gvVCVVdVbk31RWfW0T11ZWaGpqYm33nrL25Uq1MaR/SxUFZ2OwopDER3jcjiZm1jCXBRbX/iVBQsHM5NjOoeWTAzNU3Q4B4itAhsoCcCyaCEtK3AzgHXLure6ujK/+UXAYNz8k9tqBVi/UXX1ZK6+8HQ5ACUlJZSUlGC325mdnaW7uxun00lSUhIOh4PGxsaQQrWjo4M//uM/5qWXXsJs1tYikZSUxGuvvUZaWhoOh4O77rqL973vfdxxxx2aXucW5mkgCXj1xr/xm0KI39vdKUl2ikRoAhAulkoKVYlWJHRlFW569tbX16mvr/dWUbu7u1lcXCTYytl+EqqKTuf3n86gVyVU3UF2oM9N3NwEtDW2ykO46mo4AmWsTgwFroRGIjCjEaNqkgDUVlc9FoBIWN9iD/AIVV9MJhPFxcU0NjZy6NAhLBYLer2eixcvMjAwgMWy3Ss7NDTEpz/9aX7wgx9w6FBkX1zUoCgKaWmbQt3hcOBwOEIKZ0lkCCEqhBCHhBD1N/6TQlUSllhSANRmooarpoIUqhJtSWix6onrMRgM1NTU4Ha7cTqdNDc3U1BQwNTUFG+++SY9PT0sLy/7CddXnmvZxZlrh7KlcqzoFLIKgkcXxYpnk1Y4xgaCi8adtAJESjRL/pbF4Juq1i3r3qoqgNPh9P7sDtCcIFzo/+LiIiMjI5w6dYqGhgYaGxtJSUmhr6+PN998k/7+fq5cucLk5CQf+9jH+M53vsPx48cjfk5qcblc1NfXk5eXxz333MPtt98et2tJJJL4E0qEqhGpID2qEu1JaBvAtWvXyMjI4PDhw7hcLr9oquzsbLKzs3G73czPzzMyMsLa2hpms5mCggJMJhNf/m+C0tJSCgoKEqLSquh0CJ/wdl+hqruxuzszP4ek1CTv/Wr9qluZm1giLTO0v3F8eIHU9L2z3B8KXyuAh+nRBfIPZW8bOzO+qKrNargNU2rYWlUNxerqKj09PdTX12M0bv67Go1GCgsLKSwsxOl0MjExwRe+8AWuXbvG3XffTXJysneDYTzQ6/W0tbWxtLTE/fffT2dnJ7W1sqWvRBIL8bAARJIEEEnXqa1EKlTVIKuqkoQWq8eOHUMIsU2o+qLT6cjNzSU3NxeXy8Xs7CzXr19ncXGRgoICMjI2/7B8K617Ubh6hOnmUr//c9QbjSSnpmBKuSlSN6yb/eQdGzeX8dODeCuDYVlaCytYQzE2MEdJmZm5Ke1abQbKTd269B/MChDo2EAbrYKxPLdEhjmwn3dh8ua5k1L9BbzL4URvvPmn5nQ4MRgN3naqOmXz3zZUVXVtbY3Ozk7q6upITg78BcFgMHgD/p988kkyMjL46le/ygMPPMAHP/hBdU8ySjIzM7n77rs5f/68FKsSyS1KNEJVLv9L1JDQYlWn0+FwOMKG/XvQ6/Wkp6djt9upq6tjY2ODrq4u3G43BQUF5OfnYzKZ9pRwDbTM78GYtPnt2pScjCklCafjhjANUqxbvbFcHUi0unyWpyNhdmye3JKc8ANVHjcxNEvR4U0bg++SfLAM1dH+WXQBqqCB/LixJP942p+GEqweNqw2r2C1WQJnznqEqodQQtVms9HR0UFtbS2pqalBxzmdTj71qU/xgQ98gN/5nd9BUZS4itTZ2VmMRiOZmZmsr6/z6quv8uijj8btehLJrUAsTQB2s2uVFKqSeJLQYvUf/uEfaGpqorq6WtUy58rKClevXqW2tpb09HQAiouL2djYYHp6mvb2dvR6PQUFBeTm5mI0GndVuG4VqjqfIHeDafOfzpScjFu4vULVYDT6jA/8JrG6aMHtcpOenR52DnMTC5iLti+Vq0VtVXV+Kv7xUKvzK6Tn+GfJLs8tcTDb/03WE8y/dWwsbK2uenb9Q2ih6nA4aG9vp6qqyvuaDYTb7eaP/uiPOH78OP9/e3ceHnV19///ebJAAlmAhGSAAEEWZQkJIEsrIqiAIgoIiSgVFe4qvV2g1gW1KlqRStVWa2u/Vr259cfmwhplU6gtvVBESBChIUhymwQmCyH7Npk5vz+SGSbJZLLOkvB+XBfXldk+5wx4wcv355z3efjhh92y0en8+fPcc889tjsbCQkJzJ492+XjCnE5a20XgJK0UJccCuCs2b8EVdFeOnRYDQ0N5YUXXsBoNDJnzhzi4+OJiopy+A/1xYsXSUlJITY2tkF1qmvXrgwYMIABAwZQXl6O0Wjk6NGjBAQEYDAYCA8Px9fX163BtTlB1a9rFyzaYmuH1FLF+cVOA2tF7VKCpgKrfZW0KL/uZqOKskoC7NbQNod9ddVeVlpes0NYUV4hIeHtd2KKvbysSxu7ugV3p6qissF7Si4W1fkfB7gUWK3v1xbNtv/X+Nows9lMUlISV1xxBT17Nn6Kltaa1atXo5TixRdfdNuO/NGjR3Ps2DG3jCWE8D6uCqpC1NehuwEsXLiQnTt3snv3bsLDw1m2bBk33XQT77zzDrm5lwJFZmYmp0+fZsyYMU5vowIEBgYyaNAgJk6cyODBgykuLubw4cOcOHGCvLw8LLUbnPZsupo3X2i6Mtkc9VtPNRZU/br41QmqQJ2gWj8cNUdxfvPXkza3E0BTCvNqdsfnZrZPV4CCnIKm3+REUb7jakPxhaYPSSgrbrxVlW1Zht3jyrK6ywKOHj1KZmYmVVV1//K2WCwkJycTFRVF797Ouzu88847pKSk8PbbbzvtKyyE8G5tWQLgbq4MqlJVFfV16MqqVVhYGMuWLeOBBx4gKyuLjRs3cscdd9g6BaSlpfHxxx/bdlA3V1BQEEOGDGHw4MEUFRVhNBpJTU0lNDSUiooKgoKC2LPpalslqzXVVmdrUpW69Jo1pMKloGplbYdk3wopIMjx8ZyOFOcX0y3Y+fvzzuXTLbjxoO9s7WreuTzC+7asIf259NxWV4ydcbQUwJkLWTkABATVfHf7DWtWVeU1lVLrBjdT5aW/jKtNJvz8/esEV22xoHx82Pb/rqS8vJzs7GySkpJsG6R69+5NSkoK4eHh9OnTx+n8Pv74YxITE/nss8/w82vf36+MjAwWL15MdnY2Sinuv/9+li9f3q5jCCE6FmchFSSoCtfoVGUYpRRRUVE8/vjjHDp0iPHjx3PgwAHy8vJYunQpiYmJVFY2vGXbnOuGhoZy5ZVXMmHCBEpLSzGZTFy4cIHU1FSKiorQWrN383jbr2Zdt4VB1WLR+Pj6Yqk2Y6k2A5eCqo+qe62KknLbr/os5rqL8E2VVRTmNb2WqSC37rrS+g3w6y8BACgpcF65bUl1tfhCEUX15ln/MVyqtNq/5uh9hXmXKrLW6mpebTitr6LE8a41658DXAqt9dWvsAK22/+BgYFER0czYcIEhg8fTnV1NV9//TWFhYX4+Pg0qLja+/LLL3n77bfZsmVLox0C2sLPz4/XXnuNkydP8vXXX/OXv/yFkydPtvs4QojmaeupVU0FzbZ+XoKqcJVOUVl15PDhw6SmpnLixAn8/Pz47rvvWL9+PS+++CLjx48nISGByZMn42u3FrQpZrOZ5ORkIiMjGTBgABaLhby8PNLT0ykrK6N3794YDAa6d+/eILDWr7o2N6hC4ydNNZf1VnW3YOdtqArzCgltwzrPyrLKOj1e7TmrruZmXrAdPWpVWljS5K57ewU5BfSIaN77G6uu1l8OUFVRd9mDNbD6OqnQV5aV11lfDGAx1wRa6/Pb/z7c4WcDAwMxm81EREQwcOBAcnJySE5OxsfHh8jISCIiIujSpeYfq2+//ZbnnnuOXbt22dqvtTdr/1aA4OBghg8fTlZWlksPGRDicuWuJQCt2WjVnJArQVW4UqcNqxMnTmTTpk22W/Tjx49n/PjxmM1mvvrqKzZs2MDjjz/OtGnTSEhIYMyYMU7X+1VXV5OUlETfvn3p27cvUNM6KyIigoiICKqrq209XE0mExERERgMBlvFy9nmLGdB1cfPbmOV3W1x++ebyxpaA7o1fsvfGlitm6vqK8i9SI/ejjf7lJeWNRpWnSktLCY4rOFfho21iWrJ5ilrRbWirLzO9y6+UERlaTlduzf8vaj/Xivr/zRYKqvw79qlTlW1zvvswqn1Z+vz9YOsPevBFTExMSiliI6OJjo62rZUIDk5mfXr1xMSEsJnn31GYmIiBoOhGb8LbZeens6xY8fkhCohLhMtqcJKUBWu1mnDKuBwV7Svry/XX389119/PZWVlezevZs///nPnD59mlmzZhEfH8+wYcPqfLaqqoqkpCQGDhxIZGSkw7H8/PxslSiTyUR2djY//PADWmsi25XLVAAAIABJREFUIyNtPVzhUnAtKSnh+++/JyYmhgUPnG5wzcaCaoP3qZat5igrKqFbSOMHBFw4f4HuoXVft9/xnvOTkYgBjkOSszCbdy4P/y5tu43ljPX2f/2wWVHWcClEeUlNcLcPrPa38SvKyh32b7Wyrku1r8xrXXd5hX1QtWqsqnr+/Hny8vKIi4tr8N+tdalAdHQ0JpOJJ598koCAAJYtW8Yvf/lLEhISGp1neygpKWH+/Pn86U9/IiSk/dp5CSE8wz6I2ldZW7NMoC0nUwnRXJ06rDala9euzJkzhzlz5lBUVMS2bdt4+umnyc/PZ+7cuSxYsACTyURiYiILFy4kPLx5m4T8/f2JiooiKiqKiooK2wYaf39/2+3csrIyTp48yejRo+nevTu7149t8fzn/aphwK3PUj9A1VYDy4pq1pf6d21deKy/XtWes8DamOILhU6rq/V35hflFTqsgDoKpvavOaqYOqqwWqrNWKi7Xrg+bbFQXdsdwlEnBvvP+Pgodr7v+GSn3NxcMjMzGTt2rNPq/oULF3jyySd57bXXmDp1Kj/99BNpaWmNvr89mEwm5s+fz6JFi7j99ttdOpYQl6vmLAFo63rVxrRlHWtTQVWqqqK9XNZh1V5ISAiLFy9m8eLF5Obm8tFHH7Fo0SJ++uknfvGLX7S6d2VAQAADBw5k4MCBlJWVYTQa+eabb6iqqmLIkCFt2hiz9e1hdR7P/+8zrb6WlXXDVmlhSYPqqr3SwmK6hzpv3ZWXmW3bRW9VcrGIoJ7Oq3PW06KsCvMKqCwtJzCo4ZrbxgJoS563mM2UF5U4XItaXVW7ga3esgttabhJDep2bbDXWFC9ePEiZ8+eZezYsU7XT5eWlnLnnXfy29/+lqlTpwLYegO7itaapUuXMnz4cB599FGXjSOE6FiaU02VoCrak9TvHejduzc33HADVVVVvPPOO4SHhzN//nzi4+P56KOPKC1tvKroTLdu3ejRowc+Pj7ExMRQUVHB4cOH+eGHH7hw4YKth2trnDt3jqeXFrL5zegWfa6yrLxB70+r0sISCnLyKcjJb+T1Szv96x8fmpeZDTjeRV9y8VKV1P4axRdqbkdVljZyRGmJ49/3xqqpFWXlDV4zm0yUFjrun2qu3bVffy2q1hbMJpPtdWeqq6qprqpu1qa44uJi20EVztqqVVVVsXjxYu69917mzZvX5HXby7///W8+/PBD9u/fT1xcHHFxcXz++eduG18I4X0kqApPUE2cl962begd2KFDhwgODmbUqJqKmNaaH374gQ0bNvDZZ59x1VVXkZCQwA033GBbi9qUvLw8fvzxR+Li4ujatavtuoWFhRiNRi5evEjPnj0xGAyEhoY2u5qblZWF0WgkLi6uTnWu/jKB+iGs/prKrt0CbZVVK/ueod1Cujtsz2R9j/VWenVl3b+orNXVytJLu+uDeoZw0Zhn601af6z6t+XtA2xgUHent/utrOHSWjGtHzbrb4CyZ7/Rrf5aVPvPQ81JVI3x8fPls3UxDZ4vKysjOTnZtgyk0e9gNvPAAw8QExPDypUr3XY6VSdyWf2GhXaJ1D+PuMPT0xAt0NwuAM1ZBlA8uPUFj+Zoj9v+IEG1uXZn/fk7rfXVrflscGiUHvezh1s99ld7VrZ6bFeQsNoKFouFw4cPs379eg4cOMDPfvYzFixYwM9//vNGb+Xm5OSQnp5OXFxco+HWYrFw8eJFjEYjRUVFhIeHYzAYCAoKajSkZGZmkpOTQ2xsbKNjW0NrU2HVXG2mS2DdZQmmesHT0UavpsIq1ARW+7Bqf11HzfTtr1e/0mo2Vdc5GME+hDYWTH39/ZtVGa0fPJWTTVZ13tfIJrcXl9d0hoiMjLQt+aisrOTYsWOMGDHC6YYlrTVPP/00Pj4+vP766+0eVJcsWUJiYiIRERGcOHGiXa/tRSSsCq/WEcJqe1VTQYJqS0hYvUTWrLaCj48PkyZNYtKkSVRXV7N//342bNjAY489xo033khCQgIxMTG2zTLnz58nMzOTMWPGOL3d6+PjQ1hYGGFhYZjNZi5cuEBaWhplZWW2Vlj2x8VmZGSQl5fnNKjCpbWtc355qsnvVlVeEyi7BAY0CI81r1c2Wg1trBWUqbIKU2UVXRpZn1v/ms6uZ66t/FZXVjU4yavmdZPDtaem8ooG607tw7uPn6/DCqn1PU21CrNWX+uH1tjYWFtnCIvFQnh4OEajkauuuqrJoPrHP/6R/Px81q1b55KK6r333stDDz3E4sWL2/3aQgj3C/7Rp90DqwRV4Q0krLaRn58fM2bMYMaMGVRUVJCYmMirr75KWloas2fPth0csHbt2hYdh+nr61unh2tOTg4pKSmYTCYiIyMxm80UFRURGxvb7PPgt/99uC2wOqqq2qsqr2hwcIE1uDUWLgHKi0oa7TBQVVHhNLDWr2KaTdWU1Z6A5eugoltdWeWw8ukomNrP38fPt2GVufax/Xe230RlqTbbXnNWbbVfMmDt8NC/f3/69+9PWVkZR48exdfXlx9//NFWcbUuCbH3wQcfcOjQIbZu3dqigytaYsqUKaSnp7vk2kKIprnrIIDWaG5LKgmqwh0krLajgIAAFixYwIIFCygoKGD58uV8+eWXDBgwgL/97W8sWLCAyMjIFlfJ/Pz8bIcRVFVVcfLkSQoKCggKCuLcuXNERkY6rdjas/b5vHWJ89u+FrMZzOZGT2uyrl11GBZrq63+Xbs0qM5aT4VqcEqX2QzmS6HUXG/tbP3Hts/Vq3xaHzcnmDpSf5e/Vd0Qq+2ed/xnWb8VmcViISUlhSuuuIK+fftSWVlJdnY233//PQCRkZGEh4cTGBjIzp072bhxI7t27Wr2emghhIC2V1db0jdVgqpwFwmrLvLFF19w8eJFzpw5Q0FBAZs3b2bx4sV06dKF+Ph45syZQ48ezT9O1CorKwsfHx+mTJlCVVUVRqORY8eO4e/vj8FgoHfv3s2q4O58f5QtsNavqtqzrvN0tCazfsCsX611tIzAGvQ0l05zsv9cY6HUPkRag2P9yqcjzoJpe2hOcLVuzuvVq5ft9LOuXbvaWk9VVFSQk5PDE088wXfffUdRURG7du1yuvFKCCEa05rA2tLm/hJUhTvJBisXKS0txd/fv05lTGvN2bNn2bhxI9u2baNfv37Ex8dz00031VmL2pgff/yRsrIyRo4c2eDWf2lpKUajkdzcXLp164bBYCA8PLzJJQJaa26553id5xztkNcWXeeo0Prv8fX3a/CcoyDnbOf8pc85D6aeVn95RH17No6z/ay1JiUlBT8/P4YMGeL0c8ePH+ehhx5i1qxZ7N+/nwEDBrBhw4Z2mXNj0tPTmT17tmyw6iRkg1XH0pJlAK05FKCpwNra06ckqLqHbLC6RMKqh2itOX78OBs2bGDXrl3ExMSQkJDA1KlTG9zS11rz448/UlFRwciRI50uI9BaU1xcjNFo5MKFC4SEhGAwGOjVq1eDz1nDc0VFBSNGjLCFVmehExy3fLKGSfu1oi0Jpt4YSpujfnC1D6pQ8z8YlZWVDB8+3OmfW1paGnfeeSebN29m+PCapRqFhYWEhrb+dJnmkLDauUhY7VhcHVZdQYKq+0hYvUQOBfAQpRSxsbG88sorJCUlsWzZMnbv3s0111zDr3/9aw4dOoTFYsFisfDPf/6TysrKJoOq9bohISEMGzaMSZMm0bdvX3Jycvj6669JSUmhsLAQrXWDoKqU4vMPYhtcz1HgNJtMdUJl/dvxlmpzg89pi8X2y9Hn6r/WUTj6XlY//fQTpaWlTQbV7OxsfvGLX/D3v//dFlQBlwfVO++8k5/97GekpKQQFRXFe++959LxhBCXePPmKkd6plRJUBUeI2tWvYCPjw/XXnst1157LSaTiS+++IJ169axYsUKwsLCCA8Pb1X7IqUUPXv2pGfPnlgsFvLz88nIyKC4uBg/Pz/8/f2JjY2tc11rYL1p0VGH12wsbDb2Pke3zDtiKG0O+6rq+fPnycvLIy4uzumfW2FhIXfccQdr1qxh4sSJ7pimzcaNG906nhCiY2puSAUJqsI1OmVl9bXXXkMpRV5enqen0mL+/v7cfPPNfPDBB0yZMoWAgABMJhPXXXcdf/jDH0hLS6OJpRsO+fj4EB4ezsiRIwkPD7cFqMOHD5OWlkZ5ed3G+7vXj22wo701IbOjVkxbau/m8bafc3NzyczMZPTo0U7XDFdUVLBo0SJWrFjBTTfd5I5pCiFEi0hQFd6g01VWMzIy2Lt3LwMGDPD0VNokKSmJbt26sWvXLpRS5Ofn8+mnn/LII49QUVHB7bffzu23305kZGSzr2m99V9VVcW4ceNQSmEymcjNzeXUqVOYzWYiIyPr9P60rxbOuOPbdv+encHGv16B1hqlFAUFBZw9e5YxY8Y47cpQXV3N0qVLmTdvHnfddZdL5rV7926WL1+O2Wzmv/7rv1i5cqVLxhFCdD4tCakgQVW4VqfbYLVgwQKeffZZ5syZw5EjRwgPD/f0lNqV1prz58+zadMmPvnkE4KCgpg/fz633Xab0zWOjtao1ldVVUV2djZGoxEfHx8MBgMRERENNnxJaL1k09uDyc7OpqCggO7du1NUVMS4ceMIDGx4kpeVxWJh+fLl9O3blxdffNElp1OZzWaGDRvGvn37iIqKYvz48WzcuJERI0a0+1heTjZYCa/UmjWr7tpkJUHVO3j7Biul1E3AG4Av8K7W+vf1Xu8KfACMAy4Ad2it0+1eHwCcBFZprV91Nlanqqxu376dfv36ERvbcKNQZ6GUom/fvjz66KP8+te/JjU1lY0bN3LLLbcQHR1NfHw8M2fOtJ1DD80LqgBdunSxnbZUXl6O0Wjk6NGjdO3a1dbD1dfXt84t78s5uFp/H3r16kVpaSlHjx4lJCSEpKQkevXqRWRkJKGhoXV+v7XWvPTSS/j7+/PCCy+4JKhCzfKOIUOGcMUVVwCwcOFCtm/ffjmGVSFEM7U0pIIE1cuVUsoX+AswHcgEvlVK7dBan7R721LgotZ6iFJqIfAKYP9/068Du5ozXocLqzfeeCNGo7HB86tXr+bll19m7969HpiVZyilGDZsGM8//zzPPvssSUlJrF+/npdffpkxY8aQkJDA5MmT+etf/8qUKVMYM2ZMs8NRYGAggwYNYtCgQZSUlGA0GklPT6d79+4YDAbCwsI4f/48v3/Sl9jYWHx9fS+r4Gof2CsrK/n++++JjY0lJCQEi8XChQsXyMzM5NSpU7aTqfr06cM777xDamoqmzdvbvYxua2RlZVF//79bY+joqL45ptvXDaeEKJjk6AqWmgCcEZrfRZAKbUJmENNpdRqDrCq9udPgLeUUkprrZVSc4E0oLQ5g3W4sPrFF184fP77778nLS3NVlXNzMxk7NixHD58GIPB4M4peoSPjw9jx45l7NixmM1m/vWvf7FhwwYefPBB+vTpw4QJE2zrKlsqKCiIIUOGMHjwYIqKijAajZw6dQqAESNG2ELX5VhxNZlMJCcnM2zYMEJCQoCaP4vevXvTu3dvzGYzFy5c4NVXX+Xzzz8H4OOPP27WKWNCCGGvZ0pVuy8FaE1IBQmql4FwpdQRu8fvaK3fsXvcD8iwe5wJ1G9pY3uP1rpaKVUIhCmlKoAnqanKPtacyXSafzFjYmLIycmxPY6Ojm7xmtVnn32W7du34+PjQ0REBOvWrbMdj9mR+Pr6ct1113HgwAGmTp3KvHnzeO+99/jNb37DTTfdRHx8fJO9Px1RShEaGkpZWRnFxcVER0eTk5NDamoqvXr1wmAwEBISglLKFlw7Y2i1fjez2UxycjLR0dH06tXL4Xt9fX2JiIhgxowZHDlyhF/96lesWbOG6upqtm3b5rI59uvXj4yMS3+PZGZm0q9fx+rrKERn1ZYeq+0VWFsbUkGCakfgU25q659TngsPBVgF/FFrXdLcHNJpwmp7ePzxx/nd734HwJtvvsmLL77I3/72Nw/PqnUsFgsRERE899xz+Pr6MnfuXIqLi9mxYwerVq0iOzubOXPmEB8fT1RUVLODa3Z2NllZWcTFxeHn50d4eLjttre1CX54eDgGg4GgoKBOV221fh+LxcLx48fp27cvERERTj/z7bff8vzzz7N7924iIyO55557MDs40rY9jR8/ntTUVNLS0ujXrx+bNm1y+dGton0ppX5HzW00C5AD3Ku1PufZWQlv0NrA2paAaiVBVdTKAvrbPY6qfc7RezKVUn5AKDUbrSYCC5RSa4EegEUpVaG1fquxwTptWE1PT2/xZ6y3cQFKS0tdtvnFHXx9fXnwwQfrPBccHMyiRYtYtGgReXl5fPzxxzzwwANYLBbmz5/PvHnznFaic3Jy+Omnnxq0Zap/2zs3N5czZ85QWVlJREQEBoOBwMDADh9crfPXWvPDDz/Qq1evJivvp06d4uGHH2bbtm112oz5+vo6+VTb+fn58dZbbzFz5kzMZjNLlixh5MiRLh1TtLs/aK2fBVBKPQI8Byzz7JSEt7AGz6ZCa3sEVCsJqsLOt8BQpdQgakLpQqB+H8YdwD3AIWABsF/XtKC61voGpdQqoMRZUIVOHFZb65lnnuGDDz4gNDSUAwcOeHo6LhMeHs6vfvUrli1bRmZmJhs3biQhIYEePXoQHx/P7NmzCQ4Otr0/NzeX//u//7NVVBvj6+uLwWDAYDBgMpnIycnh5MmTWCwWDAYDkZGRdOnSxRb8SktLmbfkZKPX8xb2QTUlJYWAgAAGDhzo9DOZmZncd999fPjhh7Zd+e40a9YsZs2a5fZxRfvQWhfZPexOB2wlKFyvPcOoMxJUhb3aNagPAXuoaV31vtb6B6XUi8ARrfUO4D3gQ6XUGSCfmkDbKp2uz2pTnHUTmDNnju3xmjVrqKio4IUXXnDn9DxKa81//vMfNmzYwM6dOxk6dCjx8fGUlJSQmprKypUrG/Rcba7Kykqys7PJzs62Bdrg4GBOnDjBqFGjbMHYGyuu9hVhawuwptb8Xrhwgblz5/L6669z3XXXuWOawrGOe3sEUEqtBhYDhcA0rXWus/dLn9WOoS1rVt1NQqrntKXPalv/LmjL2K5w2YXV5vrpp5+YNWsWJ06c8PRUPMJisXDkyBHWrFnDv//9b26++WbuvPNOrrnmmjbfwi4rKyMzM5OMjAxCQ0Pp378/4eHhDa7rDcH103evIigoCKUUGRkZXLx4kZiYGKdBtbS0lHnz5vHYY48xd+5cl87v448/ZtWqVZw6dYrDhw9z9dVe83eLt/DqsKqU+gJw1K7kGa31drv3PQUEaK2fd3CN+4H7AQJ8g8dNNdzrotmK9tJRwqoEVc+SsHqJLAOwk5qaytChQ4GaAwauuuqqFn3+8ccfZ+fOnXTp0oXBgwfzP//zP/To0cMVU3U5Hx8fysvLycrKIjk5mZMnT7J+/Xoee+wxrr/+ehISEoiLi2tVr1A/Pz/y8/MZO3Ys/v7+GI1Gzp49S3BwMAaDgV69euHj4+PxNa7/358HkpaWRllZGYGBgXWOqW1MVVUVd999N0uWLHF5UAUYNWoUW7Zs4YEHHmjTdaxtzVrb3ky0jtb6xma+dT3wOdAgrNa2k3kHav6Bar/ZCVcJPJHl9YFVgqrwJhJW7axcuZKUlBR8fHwYOHBgizsBTJ8+nTVr1uDn58eTTz7JmjVreOWVV1w0W9cbMGAAO3fuJDIykj59+nDDDTdQWVnJ559/zhtvvEFqaiqzZs0iISGBoUOHNivkmEwmkpKSGDp0KD179gSw9XAtLCzEaDSSmppKjx49MBgM9OjRo04rLHBPcLWOFxERQXZ2NmfOnKFbt258++23dTaN2TObzSxbtoxp06Zx3333uXyOAMOHD2/T5w8cOMCYMWPo0aMHFosFHx8fKisrMRqN9O7dm27durXTTEVLKaWGaq1Tax/OAf7jyfmIy4cEVeFtJKza+fTTT9v0+RkzZth+njRpEp988klbp+RRgwYNavBc165dmTdvHvPmzaOoqIitW7eycuVKLl68yLx581iwYAF9+vRxGFytQXXQoEGEhYXVeU0pRY8ePejRowdaa/Lz8zl37hz/+c9/CAsLs61xdUcPV/tgXFBQQHp6OuPHj6dLly62TWOnTp3CbDYTGRlJSEgIISEhPP300/Tt25cnnniiQ1Qnc3Nzef3113n33XeBmmq6NXAPHTqU0tJSfve737n0pC3h1O+VUldS07rq/5BOAMLFJKQKbyVh1UXef/997rijc290CAkJ4Z577uGee+4hJyeHjz76iPvuuw8/Pz/mz5/P3Llzbc3yq6urSU5OZsCAAfTu3dvpdZVShIWFERYWhsViIS8vj/T0dMrKyujduzcGg4Hu3bvbQmV6ejr3P+l030mz2QfV4uJiTp06xZgxY+jSpaY9jL+/P/369aNfv35UVVWRnZ3NvffeS2ZmJiEhIezcubPdg2pzNwW2VHh4OBUVFSQlJTFjxgyUUqxfv56rr76aBx98kJdeegkfHx+qqqps31+4j9Z6vqfnIC4fElSFN5Ow2kLNCQ6rV6/Gz8+PRYsWuXt6HhMREcFDDz3Egw8+SHp6Ops2bWLevHlERkZy2223sW7dOl599dU6vUabw3qaWEREBNXV1eTm5nL69GlMJhMRERForSksLGT3xnG2CmBrK672QbW8vJwTJ04wevRoAgICHL6/S5cu9O/fn7lz57Jt2zZuvPFG5syZw9SpU3n55ZdbNQdHGjtiuLVycnI4e/YskyZNYsiQIRQWFtoC9jXXXMPgwYMB6NWrF1999RVnzpzhnnvukSNiheiEJKSKjkD+9WmhpoLDunXrSExM5Msvv2xxha0z7OxWSjFo0CCeeuopVq5cybFjx0hISCAkJIQ333yThIQErr/++lZV6vz8/OjTpw99+vShqqqK1NRUcnJyCAoKIisrq0EPV2h+cLX/TGVlJcnJyYwcOZLu3bs7/dzOnTvZvHkzn3/+Od27d+eJJ56grKysxd/NXaqqqli3bh3vvfce8+fPJz09nbNnz9o2Vg0ePNh2utYXX3zB7t27+eMf/yhBVYh25g2brCSoio5CFqO1o927d7N27Vp27NjRqo0p1p3dU6ZMccHs3E8pxZ/+9CdWrFjBkSNHeOSRR9i/fz+TJ09m+fLlHDx4sNXHjhYWFlJWVsa1115LTEwMFouFpKQkjh07xrlz56iurgZqQqj1V3OYTCaSk5MZNmxYnRPNHPnnP//Jq6++ypYtW+qEWnduStq6dStRUVEcOnSIW265hZkzZzp9f5cuXXjkkUc4ePAgBoOBQYMG8dvf/paXXnqJv/zlLwBUVFTg6+vLsGHDeOKJJxg8eDBNtLgTQnQggSeyJKiKDkX6rLajIUOGUFlZads8NGnSpBZ3FACYOnUqr776aoesrNZ37ty5BkeSVldX8+WXX7JhwwaOHj3K9OnTSUhIYNSoUc3azJOfn8+ZM2cYM2ZMg0MKysrKMBqN5OTk0K1bNwwGA2FhYU57uFqDrNls5tixYwwYMICIiAinczh+/DgPPPAAn332GVFRUU3O2ZuYzWZ8fX3RWnP69GluvfVWXn31VdsyjUGDBrF27Vry8vKcHr/bQXj/Trd2JIcCdCyeqKxKSO04pM/qJXJvrx2dOXPG01PwOvWDKtTczp85cyYzZ86kvLycxMRE1q5dS3p6OrfeeisLFizgiiuucLiMoqCggNOnTzsMqlBT1bziiisYNGgQJSUlGI1GfvzxR0JCQjAYDPTs2bNBD1eoOQTh+PHj9OnTp8mgmpaWxv3338/mzZs7XFAFbMFdKUV0dDRXXXUV1113HbfddhtHjhwhOjoaoDMEVSG8mjuXAkhIFR2ZhFU3c9XO7o4qMDCQ+Ph44uPjKSgoYMuWLTz66KO2U6Dmz59PZGQkSiny8/NJSUkhLi6Orl27Or2uUorg4GCCg4Ntm4iMRiOnT5+mZ8+eGAwGQkNDbY3wT548Sc+ePenXz/k/HEajkUWLFvHuu++2ucdpU1x9yITWGovFQmlpKd999x3XX3+9rZovhwMI4R6uDqwSUkVnIGHVzdp7Z3dn0qNHD5YsWcKSJUs4f/48mzdv5u677yYgIIBp06axadMm9u7d26AZf1Pse7haLBYuXrxIVlYWp06dIiwsjMrKSrp27WqrKDamsLCQhQsXsnbtWiZMmNCGb9o8rj5kQilFYGAgY8eOxWQyNXhNCOEergisElJFZyIbrDqB3bt3c+WVVzJkyBB+//vfe3o67aJPnz6sWLGCgwcPsmrVKv76178SHBzMf//3f7N161bKy8tbdV0fHx/CwsIYOXIkEyZMoKKigoKCAvLz8zl79myjO/nLy8u56667ePTRR+sc/uBKM2bMsO3CnzRpEpmZmS4Zx9/fn5KSEpdcWwjRPO0RLq0bpySois5GKqteZOvWrTz88MPk5uZyyy23EBcXx549e5x+xmw28+CDD7Jv3z6ioqIYP348t912GyNGjHDTrF3LYrHw5JNPsmXLFiZOnEhycjIbNmzglVdeYfTo0SQkJHDdddc5XL/alHPnzqG1ZvLkyZjNZnJyckhJScFkMhEZGUlkZCQBAQFUV1ezdOlS5s+fz8KFC13wLZvmykMm2rMnrBCi9awhs6VVVgmnorOTbgAd3KFDh1i1apUt1K5ZswaAp556ypPTaleFhYWEhobWec5isXDw4EE2bNjAwYMHmTx5MgkJCUyYMKFZHQWMRiNZWVmMGTOmwfutJ1MlJyezdu1awsPDiYmJYe3atR45nWr16tUcOXKELVu2yO35lrusfsOkG4AQnYd0A7hEKqsdXFZWFv3797c9joqK4ptvvvHgjNpf/aAKNbfzp0yZwpQpUzCZTOzbt4/333+fFStWMHPmTBISEhgxYoTDcJeXl0dGRobDoAqXTqaKiori4MGHeRzpAAAN5ElEQVSDJCUlcfDgQW677TZeeuklYmNj2+27ufKQCSGEEKIzkLAqOjx/f39mzZrFrFmzKCsrY8eOHbz00ktkZWVx2223sWDBAgYOHIhSCqPRSHp6OmPHjnV6KpPWmrfffpvz589z4MABfH19SUlJITg42G3fy3rIxFdffeXWgwaEEEIIbyJhtYPr168fGRkZtseZmZlNtl/qzLp168bChQtZuHAh+fn5fPLJJzz00ENUVVUxZcoUtm7dyldffdXkca+bN29mz549JCYm2vqSXnnlle74CjYPPfQQlZWVTJ8+HWj9IRNCCCFERyZhtYMbP348qamppKWl0a9fPzZt2sSGDRtadI0lS5aQmJhIREQEJ06ccNFM3a9Xr17cf//9/PKXv+Sbb75h/vz5DBo0iLvvvpsFCxZw6623OjxSdd++ffz9739nz549TfZzdSU5ZEIIIYSQ1lUdnp+fH2+99RYzZ85k+PDhJCQkMHLkyBZd495772X37t0umqHnWSwWHnvsMbZv386//vUv3nzzTbKysrj55pu5++672bFjBxUVFQAcPnyYVatWsW3bNodBVgghhBDuJd0ABADp6enMnj27U1VW7ZWUlBAUFFTnOYvFwrFjx1i/fj379u1j8ODBnDx5kj179jBo0CCXzufZZ59l+/bt+Pj4EBERwbp16xweTSva7LLalSbdAIToPKQbwCVSWRWXhfpBFWo6CowbN47XX3+dpKQkbr/9dl555RWXB1WoOUr1+PHjJCUlMXv2bF588UWXjymEEEJ0RLJmVQjA19eXxYsXu208+yUGpaWl0pZKCCGEaISEVSE85JlnnuGDDz4gNDSUAwcOeHo6QgghhFeSZQCizTIyMpg2bRojRoxg5MiRvPHGG56ekle48cYbGTVqVINf27dvB2pOpsrIyGDRokW89dZbHp6tEEII4Z2ksiq48847+cc//kFeXh5RUVG88MILLF26tNmf9/Pz47XXXmPs2LEUFxczbtw4pk+fzogRI1w4a+/X1OlUVosWLWLWrFm88MILLp6REEII0fFIWPVyR48eJTo6ml69eqG1dsnaxo0bN7bp83369KFPnz4ABAcHM3z4cLKysi77sOpMamoqQ4cOBWD79u1cddVVHp6REEII4Z1kGYCXsrYU+/bbb3niiScAUEpRWVnpyWk1KT09nWPHjjFx4kRPT8WrrVy5klGjRjF69Gj27t0rSyeEEEKIRkhl1cuVlpbaft61axf/+7//y9VXX81jjz3mwVk5VlJSwvz58/nTn/4kDfWb8Omnn3p6CkIIIUSHIGHVSyml0FoTGxtLZmYmCxYsoKKigjvuuIN58+YBNU3tfXy8ozhuMpmYP38+ixYt4vbbb/f0dIQQQgjRSXhH0hEOKaXIyMjgrbfeYujQoXzyySfcfffdBAUFeVVQ1VqzdOlShg8fzqOPPtqqa1RUVDBhwgRiY2MZOXIkzz//fDvP0v1ee+01lFLk5eV5eipCCCFEhyWVVS91/Phx1q1bR2JiItOmTeP5558nICDAtsnKW4IqwL///W8+/PBDYmJiiIuLA+Dll19m1qxZzb5G165d2b9/P0FBQZhMJiZPnszNN9/MpEmTXDVtl8rIyGDv3r0MGDDA01MRQgghOjTvSTwCgNzcXD744APuuusuoqKi2L59Oz179uT8+fMAVFVV8a9//YuMjAwPz/SSyZMno7W2HR+alJTUoqAKNVVk65GoJpMJk8nUoU91+vWvf83atWs79HcQQgghvIGEVS/zzTff8Mknn7BhwwYeffRRQkNDSU5OJjo6GqgJdefOneOuu+4iKyvLs5NtZ2azmbi4OCIiIpg+fXqH7Siwfft2+vXrR2xsrKenIoQQQnR4sgzAy8yePZvZs2fbHldVVTFmzBgOHTrEz3/+c7p06cL06dNJTEykX79+Hpxp+/P19SUpKYmCggLmzZvHiRMnGDVqlKen5dCNN96I0Whs8Pzq1at5+eWX2bt3rwdmJYQQQnQ+Ela9jNlsxtfX1/Y4Ojqa5cuXEx4ebntuy5YttrWc3rTRqr306NGDadOmsXv3bq8Nq42dTvX999+TlpZmq6pmZmYyduxYDh8+jMFgcOcUhRBCiE6hc6WcTsA+qFpNnDiRwYMHU1hYyHvvvcf69eu59dZbATrNmsjc3FwKCgoAKC8vZ9++fa0+1clsNjNmzJg6FWp3iYmJIScnh/T0dNLT04mKiuLo0aMSVIUQQohWkrDagXTv3p3AwEBCQ0PZv38/0HnC6vnz55k2bRqjR49m/PjxTJ8+vdVh84033mD48OHtPEMhPEMp9RullFZKhTf9biGEcA+l1E1KqRSl1Bml1EoHr3dVSm2uff0bpVS03WtP1T6fopSa2dRYsgygA/Hz8+Ouu+7irrvuoqKiAsDWyqqjGz16NMeOHWvzdTIzM/nss8945plneP3119thZm2Tnp7u6SmIDkwp1R+YAfzk6bkIIYSVUsoX+AswHcgEvlVK7dBan7R721LgotZ6iFJqIfAKcIdSagSwEBgJ9AW+UEoN01qbGxtPKqsdjNlc82cZEBAAdJ7KantZsWIFa9eu7XTreMVl64/AE4D29ESEEMLOBOCM1vqs1roK2ATMqfeeOcD/1v78CXCDqgktc4BNWutKrXUacKb2eo2Sf9E7GEdrWkWNxMREIiIiGDdunKenIkSbKaXmAFla62RPz0UIIerpB9g3fM+sfc7he7TW1UAhENbMz9bR1DIAKduJDuPWW29dA9z99ttvVwMBQIhSaovW+hcenpoQDimlvgAc7b57BniamiUATV3jfuD+2oeVu7P+fKL9Ztgi4YCnzhaWsS+fsS+n7zywtR8sMuXs2Z3157ascw9QSh2xe/yO1vqdNlyvTWTNqug0tNZPAU8BKKWmAo9JUBXeTGt9o6PnlVIxwCAguXapTxRwVCk1QWtdp8Fv7T8g79R+7ojW+mrXztoxGVvG7szjenrsltJa3+TiIbKA/naPo2qfc/SeTKWUHxAKXGjmZ+uQZQBCCOFltNbfa60jtNbRWutoam6Tja0fVIUQwkO+BYYqpQYppbpQs2FqR7337ADuqf15AbBfa61rn19Y2y1gEDAUOOxsMKmsik5Ja/0P4B8enoYQQgjR6Witq5VSDwF7AF/gfa31D0qpF4EjWusdwHvAh0qpM0A+NYGW2vd9BJwEqoEHnXUCAAmrQgjh9Wqrq83hsTVlMraM3cnH9fTYXkdr/Tnweb3nnrP7uQKIb+Szq4HVzR1L1VRkhRBCCCGE8D6yZlUIIYQQQngtCatCCNEJeeKYVqXU75RSx5VSSUqpvUqpvm4c+w9Kqf/Ujr9VKdXDjWPHK6V+UEpZlFIu3y3e1DGXLhz3faVUjlLK7e3RlFL9lVIHlFIna3+vl7tx7ACl1GGlVHLt2C+4a2xRQ8KqEEJ0Mh48pvUPWuvRWus4IBF4rqkPtKN9wCit9WjgNLVt7NzkBHA78E9XD2R3zOXNwAjgztrjK91hHeDqlkiNqQZ+o7UeAUwCHnTj964ErtdaxwJxwE1KqUluGlsgYVUIITojjxzTqrUusnvY3Z3ja6331p6SA/A1Nb0b3TX2Ka11ipuGa84xly6htf4nNbu63U5rfV5rfbT252LgFE2cetSOY2utdUntQ//aX7Lhx40krAohRCfi6WNalVKrlVIZwCLcW1m1twTY5aGxXa3FR1V2NkqpaGAM8I0bx/RVSiUBOcA+rbXbxhbSukoIITqc9jim1RVja623a62fAZ5RSj0FPAQ8766xa9/zDDW3jNe317jNHVu4nlIqCPgUWFGvku9StX1A42rXQm9VSo3SWnvqaOPLjoRVIYToYNrjmNb2HtuB9dT0YGy3sNrU2Eqpe4HZwA26nfsytuB7u1qLj6rsLJRS/tQE1fVa6y2emIPWukApdYCatbsSVt1ElgEIIUQn4eljWpVSQ+0ezgH+445xa8e+iZp1urdprcvcNa4HNOeYy05H1fzf13vAKa31624eu7e1u4RSKhCYjhv/2xYSVoUQQrSf3yulTiiljlOzFMFt7YWAt4BgYF9t66y/uWtgpdQ8pVQm8DPgM6XUHleNVbuJzHrM5SngI631D64az55SaiNwCLhSKZWplFrqjnFrXQPcDVxf++ebpJSa5aax+wAHav+7/paaNauJbhpbICdYCSGEEEIILyaVVSGEEEII4bUkrAohhBBCCK8lYVUIIYQQQngtCatCCCGEEMJrSVgVQgghhBBeS8KqEEIIIYTwWhJWhRBCCCGE15KwKoQQQngxpdQypdTbdo9fUkp96Mk5CeFOciiAEEII4cWUUt2AFCAGmAz8Dvi51rrcoxMTwk0krAohhBBeTim1FugO3AxM11r/6OEpCeE2ElaFEEIIL6eUugo4BczRWu/w9HyEcCdZsyqEEEJ4v+eAXMDP+oRS6gql1HtKqU88Ny0hXE/CqhBCCOHFlFK/AQKABGC59Xmt9Vmt9VKPTUwIN/Fr+i1CCCGE8ASl1PXAfcDPtNbFSqkQpVSc1jrJ03MTwl2ksiqEEEJ4IaXUAOBdIF5rXVz79BvACs/NSgj3kw1WQgghRAeklAoDVgPTgXe11ms8PCUhXELCqhBCCCGE8FqyDEAIIYQQQngtCatCCCGEEMJrSVgVQgghhBBeS8KqEEIIIYTwWhJWhRBCCCGE15KwKoQQQgghvJaEVSGEEEII4bUkrAohhBBCCK8lYVUIIYQQQnit/x/A6oK4uGc1qgAAAABJRU5ErkJggg==\n",
"text/plain": [
""
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"rv, rv1 = data( mu=[1.2,1.4], mu1=[1.4,-1.4], \n",
" cov=[[1.0, -0.8], [-0.8, 1.0]] , \n",
" cov1= [[1.0, 0.8], [0.8, 1.0]])\n",
"#Mesh para la grafica en 3D\n",
"x, y = np.mgrid[-4:4:.1, -4:4:.1]\n",
"pos = np.dstack((x, y))\n",
"graph_gaussian_contour(x, y, pos, rv, rv1)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 279
},
"executionInfo": {
"elapsed": 388,
"status": "ok",
"timestamp": 1638335092113,
"user": {
"displayName": "HERNAN DAVID SALINAS JIMENEZ",
"photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GhBC-5LZ7dZrJhGJeMzy6pOzsnDM7tcAwf58cHvon83rnG0ZgZHkYjtYed5MVVWyx0YIsBH3d5Rt-u8LhImipAJ47T_GJxdC3sX03gpJEpqdzXtiUn_QWj-eIgyQ0WzKukBfSahChtfEfM7zQgZom4BpTN1S0izHwqrALhT_uYIj7r47B9trr7ZfpDvK_xrdRhxfJRfUppIJCCV0E9jOVEyvzPuIQSbmxEKRZwPGz__-DY2kTV3XHPKlj7m9DPxx0pI3Vj34PoH6w0D8qs0nIz0Z_viNFTzydCbqfyl8irfbh40YasBQngmGl-_LSBU6_3hoh39ssLX4euV1xdBpiO8KsQgvDvLZF35hIogSmu4BE17JjVJcJGK0DuSPhh-Y0JxedJn0nskjm1FCQjhLi4FBh5JPAf3R2uJ-it0BFCRci00xUyduL0lelHwRF3abBC6OvNzE6uhi8uNRJHZet7rvNPC9Noi2U0asD-CBUa0tfIU1FExuDvfd1c2zLDWFCsQc_QbBlVvwVirgo0WC_S4TQ8I3pBHFydmquoBXzShhJDT7TixT4cz2Vgl_dXzBlELkYgEeHYMhvF8fRRsHrf7sd4yj8bqfBYj22GKqPSkdBk5A_f08DZBRcegVxK2oi9rWpLLRvSi_kvVe5whaMrPRlqb-uNbk9mDEQh7M8m9LAxMWA84blOlniAtinYP_Q3EOXnQUrCJoF0eyXdNZiYdyM4PeZpc1CmqbKLTBZ_4md9ObCgo2i8fdtUwNughaIcPHrc=s64",
"userId": "00408651407692255291"
},
"user_tz": 300
},
"id": "r8e5aSU6fhjA",
"outputId": "14820c73-e73c-449f-b2aa-c4b4c86be1b3"
},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
""
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"rv, rv1 = data( mu=[1.2,1.4], mu1=[1.4,-1.4], \n",
" cov=[[1.0, -0.8], [-0.8, 1.0]] , \n",
" cov1= [[1.0, 0.8], [0.8, 1.0]])\n",
"X_t, y = sample(N1 = 100, N2 = 100, r = 10)\n",
"\n",
"plt.plot(X_t[y==0][:,0],X_t[y==0][:,1],\"ro\", alpha=0.4)\n",
"plt.plot(X_t[y==1][:,0],X_t[y==1][:,1],\"bo\", alpha=0.4)\n",
"plt.xlabel(\"X1\")\n",
"plt.ylabel(\"X2\")\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "00QFRKiM_jSP"
},
"source": [
"# Curvas de aprendizaje\n",
"\n",
"\n",
"Permite definir las metricas del modelo, define que también es sepración realizada por el estimador. Permite entender cuando se presenta sobreajuste y bajoajuste en un modelo de machining learning.\n",
"\n",
"Su construccion puede ser realizada como sigue:\n",
"\n",
"1. El datasets se divide K veces en entrenamiento y validación.\n",
"```\n",
"x_t_index = [0, 1, 2, 3 4, 5, 6, 7, 8, 9]\n",
"```\n",
"\n",
"|div|train_index |test_index |\n",
"|---|----------------|-----------|\n",
"| 1 |[9 8 6 7 0 4 2] |[3 1 5]|\n",
"| 2 |[0 7 9 3 2 1 4] | [5 8 6]|\n",
"| 3 |[8 6 5 2 9 7 3] | [0 1 4]|\n",
"| 4 |[5 9 4 6 3 0 2] |[7 8 1]|\n",
"|k=5|[3 5 8 1 7 6 0] | [9 2 4]|\n",
"\n",
"\n",
"2. Se construyen subconjunto del dataset de entrenamiento para entrenar el estimador.\n",
"\n",
"|split| train_index(div 1)|train_index(div 2)| ... |execution|\n",
"|-----|--------------------|-----------------|-----|---------| \n",
"| 1 |[9, 8] | [0, 7] | ... | i |\n",
"| 2 |[9, 8, 6] | [0, 7, 9] | ... | i |\n",
"| 3 |[9,8,6, 7, 0] | [0, 7, 9,3] | ... | i |\n",
"| 4 |[9,8,6, 7, 0, 4] | [0, 7, 9,3, 2] | ... | i |\n",
"| 5 |[9, 8, 6] |[0, 7, 9,3, 2,1] | ... | i |\n",
"| 6 |[9,8,6, 7, 0, 4,2]| [0, 7, 9,3, 2,1,4]| ... | i |\n",
"\n",
"\n",
"\n",
"\n",
"3. Se computan los scores para los datos de entrenamiento y validación.\n",
"\n",
"4. Se promedian los scores de los datos de entrenamiento y validación \n"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "8m33Wm_1tRW5"
},
"source": [
"# Overfitting:\n",
"\n",
"Modelo con una complejidad alta, por ejemplo querer ajutar una funcion lineal con un polinomio de grado superior. En este caso decimos que el sistema se esta aprendiendo los datos de memoria.\n",
"\n",
"1. Para las covarianzas mostradas, realiza las siguientes pruebas, ¿la Separacion de los datos mejora con el aumento de datos?\n",
"\n",
"2. El hiperparametro gamma, por ahora, permitira definir la complejidad de la frontera. Que pasa si se disminuye el valor de gamma?\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"executionInfo": {
"elapsed": 359,
"status": "ok",
"timestamp": 1638335103824,
"user": {
"displayName": "HERNAN DAVID SALINAS JIMENEZ",
"photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GhBC-5LZ7dZrJhGJeMzy6pOzsnDM7tcAwf58cHvon83rnG0ZgZHkYjtYed5MVVWyx0YIsBH3d5Rt-u8LhImipAJ47T_GJxdC3sX03gpJEpqdzXtiUn_QWj-eIgyQ0WzKukBfSahChtfEfM7zQgZom4BpTN1S0izHwqrALhT_uYIj7r47B9trr7ZfpDvK_xrdRhxfJRfUppIJCCV0E9jOVEyvzPuIQSbmxEKRZwPGz__-DY2kTV3XHPKlj7m9DPxx0pI3Vj34PoH6w0D8qs0nIz0Z_viNFTzydCbqfyl8irfbh40YasBQngmGl-_LSBU6_3hoh39ssLX4euV1xdBpiO8KsQgvDvLZF35hIogSmu4BE17JjVJcJGK0DuSPhh-Y0JxedJn0nskjm1FCQjhLi4FBh5JPAf3R2uJ-it0BFCRci00xUyduL0lelHwRF3abBC6OvNzE6uhi8uNRJHZet7rvNPC9Noi2U0asD-CBUa0tfIU1FExuDvfd1c2zLDWFCsQc_QbBlVvwVirgo0WC_S4TQ8I3pBHFydmquoBXzShhJDT7TixT4cz2Vgl_dXzBlELkYgEeHYMhvF8fRRsHrf7sd4yj8bqfBYj22GKqPSkdBk5A_f08DZBRcegVxK2oi9rWpLLRvSi_kvVe5whaMrPRlqb-uNbk9mDEQh7M8m9LAxMWA84blOlniAtinYP_Q3EOXnQUrCJoF0eyXdNZiYdyM4PeZpc1CmqbKLTBZ_4md9ObCgo2i8fdtUwNughaIcPHrc=s64",
"userId": "00408651407692255291"
},
"user_tz": 300
},
"id": "v9rFHfB4AcQm",
"outputId": "66a9777a-9521-40db-c8ac-616665afed0c"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[0 4 1 3 7 5 6] [8 2 9]\n",
"[4 0 6 7 5 3 9] [8 2 1]\n",
"[4 9 7 3 1 8 6] [5 2 0]\n",
"[3 7 4 6 5 0 2] [8 9 1]\n",
"[3 4 1 8 6 2 7] [9 0 5]\n"
]
}
],
"source": [
"q = X_t[0:10]\n",
"cv = ShuffleSplit(n_splits = 5, test_size=0.3, random_state=None)\n",
"for train_index,test_index in cv.split(q):\n",
" print(train_index,test_index)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"executionInfo": {
"elapsed": 12,
"status": "ok",
"timestamp": 1638335104347,
"user": {
"displayName": "HERNAN DAVID SALINAS JIMENEZ",
"photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GhBC-5LZ7dZrJhGJeMzy6pOzsnDM7tcAwf58cHvon83rnG0ZgZHkYjtYed5MVVWyx0YIsBH3d5Rt-u8LhImipAJ47T_GJxdC3sX03gpJEpqdzXtiUn_QWj-eIgyQ0WzKukBfSahChtfEfM7zQgZom4BpTN1S0izHwqrALhT_uYIj7r47B9trr7ZfpDvK_xrdRhxfJRfUppIJCCV0E9jOVEyvzPuIQSbmxEKRZwPGz__-DY2kTV3XHPKlj7m9DPxx0pI3Vj34PoH6w0D8qs0nIz0Z_viNFTzydCbqfyl8irfbh40YasBQngmGl-_LSBU6_3hoh39ssLX4euV1xdBpiO8KsQgvDvLZF35hIogSmu4BE17JjVJcJGK0DuSPhh-Y0JxedJn0nskjm1FCQjhLi4FBh5JPAf3R2uJ-it0BFCRci00xUyduL0lelHwRF3abBC6OvNzE6uhi8uNRJHZet7rvNPC9Noi2U0asD-CBUa0tfIU1FExuDvfd1c2zLDWFCsQc_QbBlVvwVirgo0WC_S4TQ8I3pBHFydmquoBXzShhJDT7TixT4cz2Vgl_dXzBlELkYgEeHYMhvF8fRRsHrf7sd4yj8bqfBYj22GKqPSkdBk5A_f08DZBRcegVxK2oi9rWpLLRvSi_kvVe5whaMrPRlqb-uNbk9mDEQh7M8m9LAxMWA84blOlniAtinYP_Q3EOXnQUrCJoF0eyXdNZiYdyM4PeZpc1CmqbKLTBZ_4md9ObCgo2i8fdtUwNughaIcPHrc=s64",
"userId": "00408651407692255291"
},
"user_tz": 300
},
"id": "Ci6kfqXhES6c",
"outputId": "f1cc8c28-1386-4124-ef75-e108a61fe410"
},
"outputs": [
{
"data": {
"text/plain": [
"200"
]
},
"execution_count": 10,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"len(X_t)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 1000
},
"executionInfo": {
"elapsed": 6896,
"status": "ok",
"timestamp": 1638337181909,
"user": {
"displayName": "HERNAN DAVID SALINAS JIMENEZ",
"photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GhBC-5LZ7dZrJhGJeMzy6pOzsnDM7tcAwf58cHvon83rnG0ZgZHkYjtYed5MVVWyx0YIsBH3d5Rt-u8LhImipAJ47T_GJxdC3sX03gpJEpqdzXtiUn_QWj-eIgyQ0WzKukBfSahChtfEfM7zQgZom4BpTN1S0izHwqrALhT_uYIj7r47B9trr7ZfpDvK_xrdRhxfJRfUppIJCCV0E9jOVEyvzPuIQSbmxEKRZwPGz__-DY2kTV3XHPKlj7m9DPxx0pI3Vj34PoH6w0D8qs0nIz0Z_viNFTzydCbqfyl8irfbh40YasBQngmGl-_LSBU6_3hoh39ssLX4euV1xdBpiO8KsQgvDvLZF35hIogSmu4BE17JjVJcJGK0DuSPhh-Y0JxedJn0nskjm1FCQjhLi4FBh5JPAf3R2uJ-it0BFCRci00xUyduL0lelHwRF3abBC6OvNzE6uhi8uNRJHZet7rvNPC9Noi2U0asD-CBUa0tfIU1FExuDvfd1c2zLDWFCsQc_QbBlVvwVirgo0WC_S4TQ8I3pBHFydmquoBXzShhJDT7TixT4cz2Vgl_dXzBlELkYgEeHYMhvF8fRRsHrf7sd4yj8bqfBYj22GKqPSkdBk5A_f08DZBRcegVxK2oi9rWpLLRvSi_kvVe5whaMrPRlqb-uNbk9mDEQh7M8m9LAxMWA84blOlniAtinYP_Q3EOXnQUrCJoF0eyXdNZiYdyM4PeZpc1CmqbKLTBZ_4md9ObCgo2i8fdtUwNughaIcPHrc=s64",
"userId": "00408651407692255291"
},
"user_tz": 300
},
"id": "8UnvI75ntQzj",
"outputId": "3d2c2774-4b6d-4d11-b80f-33cb9d3e74ac"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Training error:0.979\n",
"Test error: 0.896\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
""
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
""
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"rv, rv1 = data( mu=[1.2,1.4], mu1=[1.4,-1.4], \n",
" cov=[[1.0, -0.8], [-0.8, 1.0]] , \n",
" cov1= [[1.0, 0.8], [0.8, 1.0]])\n",
"X_t, y = sample(N1 = 120, N2 = 120, r = 10)\n",
"X_train, X_test, y_train, y_test = train_test_split(X_t, y, test_size = 0.2, \n",
" random_state=1)\n",
"clf = SVC(gamma = 10.0)\n",
"clf.fit(X_train, y_train)\n",
"fig, ax = plt.subplots()\n",
"X0, X1 = X_train[:, 0], X_train[:, 1]\n",
"xx, yy = make_meshgrid(X0, X1)\n",
"plot_contours(ax, clf, xx, yy, cmap=plt.cm.coolwarm, alpha=0.8)\n",
"plt.plot(X_train[y_train==0][:,0],X_train[y_train==0][:,1],\"bo\", alpha=1)\n",
"plt.plot(X_train[y_train==1][:,0],X_train[y_train==1][:,1],\"ro\", alpha=1)\n",
"print(f\"Training error:{clf.score(X_train, y_train):.3f}\")\n",
"print(f\"Test error: {clf.score(X_test, y_test):.3f}\" )\n",
"\n",
"fig, axes = plt.subplots(3, 1, figsize=(8, 14))\n",
"title = \"Learning Curves (Naive Bayes)\"\n",
"# Cross validation with 100 iterations to get smoother mean test and train\n",
"# score curves, each time with 20% data randomly selected as a validation set.\n",
"cv = ShuffleSplit(n_splits=100, test_size=0.2, random_state=0)\n",
"#clf = SVC(gamma = 20.0)\n",
"#clf.fit(X_t, y)\n",
"plot_learning_curve(clf, title, X_t, y, axes=axes[0:,], ylim=(0.8, 1.01), cv=cv, n_jobs=4)\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "mVYvIguvkyB_"
},
"source": [
"# Underfitting \n",
"No se obtiene la naturaleza subyancente de los datos, el modelo ajustado presentan un error en la estimación de la frontera desfasada respecto a la frontera Bayesiana"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 688
},
"executionInfo": {
"elapsed": 2719,
"status": "ok",
"timestamp": 1638338470122,
"user": {
"displayName": "HERNAN DAVID SALINAS JIMENEZ",
"photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GhBC-5LZ7dZrJhGJeMzy6pOzsnDM7tcAwf58cHvon83rnG0ZgZHkYjtYed5MVVWyx0YIsBH3d5Rt-u8LhImipAJ47T_GJxdC3sX03gpJEpqdzXtiUn_QWj-eIgyQ0WzKukBfSahChtfEfM7zQgZom4BpTN1S0izHwqrALhT_uYIj7r47B9trr7ZfpDvK_xrdRhxfJRfUppIJCCV0E9jOVEyvzPuIQSbmxEKRZwPGz__-DY2kTV3XHPKlj7m9DPxx0pI3Vj34PoH6w0D8qs0nIz0Z_viNFTzydCbqfyl8irfbh40YasBQngmGl-_LSBU6_3hoh39ssLX4euV1xdBpiO8KsQgvDvLZF35hIogSmu4BE17JjVJcJGK0DuSPhh-Y0JxedJn0nskjm1FCQjhLi4FBh5JPAf3R2uJ-it0BFCRci00xUyduL0lelHwRF3abBC6OvNzE6uhi8uNRJHZet7rvNPC9Noi2U0asD-CBUa0tfIU1FExuDvfd1c2zLDWFCsQc_QbBlVvwVirgo0WC_S4TQ8I3pBHFydmquoBXzShhJDT7TixT4cz2Vgl_dXzBlELkYgEeHYMhvF8fRRsHrf7sd4yj8bqfBYj22GKqPSkdBk5A_f08DZBRcegVxK2oi9rWpLLRvSi_kvVe5whaMrPRlqb-uNbk9mDEQh7M8m9LAxMWA84blOlniAtinYP_Q3EOXnQUrCJoF0eyXdNZiYdyM4PeZpc1CmqbKLTBZ_4md9ObCgo2i8fdtUwNughaIcPHrc=s64",
"userId": "00408651407692255291"
},
"user_tz": 300
},
"id": "Ulotjb2T2FGH",
"outputId": "76541d36-41d2-4588-a88b-a185451694f2"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Training error:0.547\n",
"Test error: 0.375\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
""
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
""
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"#variar la cantida de datos\n",
"\n",
"\n",
"rv, rv1 = data( mu=[1.4,1.4], mu1=[1.4,-1.4], \n",
" cov=[[1.0, -0.8], [-0.8, 1.0]] , \n",
" cov1= [[1.0, 0.8], [0.8, 1.0]])\n",
"X_t, y = sample(N1 = 40, N2 = 40, r = 10)\n",
"\n",
"\n",
"X_train, X_test, y_train, y_test = train_test_split(X_t, y, test_size = 0.2, \n",
" random_state=1)\n",
"clf = SVC(gamma = 0.001)\n",
"clf.fit(X_train, y_train)\n",
"fig, ax = plt.subplots()\n",
"X0, X1 = X_train[:, 0], X_train[:, 1]\n",
"xx, yy = make_meshgrid(X0, X1)\n",
"plot_contours(ax, clf, xx, yy, cmap=plt.cm.coolwarm, alpha=0.8)\n",
"plt.plot(X_train[y_train==0][:,0],X_train[y_train==0][:,1],\"bo\", alpha=1)\n",
"plt.plot(X_train[y_train==1][:,0],X_train[y_train==1][:,1],\"ro\", alpha=1)\n",
"print(f\"Training error:{clf.score(X_train, y_train):.3f}\")\n",
"print(f\"Test error: {clf.score(X_test, y_test):.3f}\" )\n",
"\n",
"fig, axes = plt.subplots(3, 1, figsize=(8, 6))\n",
"title = \"Learning Curves (Naive Bayes)\"\n",
"# Cross validation with 100 iterations to get smoother mean test and train\n",
"# score curves, each time with 20% data randomly selected as a validation set.\n",
"cv = ShuffleSplit(n_splits=100, test_size=0.2, random_state=0)\n",
"#clf = SVC(gamma = 20.0)\n",
"#clf.fit(X_t, y)\n",
"plot_learning_curve(clf, title, X_t, y, axes=axes, ylim=(0.3, 1.01), cv=cv, n_jobs=4)\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "8YupCLOMy6aM"
},
"source": [
"# Underfiting"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 300
},
"executionInfo": {
"elapsed": 386,
"status": "ok",
"timestamp": 1638335162576,
"user": {
"displayName": "HERNAN DAVID SALINAS JIMENEZ",
"photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GhBC-5LZ7dZrJhGJeMzy6pOzsnDM7tcAwf58cHvon83rnG0ZgZHkYjtYed5MVVWyx0YIsBH3d5Rt-u8LhImipAJ47T_GJxdC3sX03gpJEpqdzXtiUn_QWj-eIgyQ0WzKukBfSahChtfEfM7zQgZom4BpTN1S0izHwqrALhT_uYIj7r47B9trr7ZfpDvK_xrdRhxfJRfUppIJCCV0E9jOVEyvzPuIQSbmxEKRZwPGz__-DY2kTV3XHPKlj7m9DPxx0pI3Vj34PoH6w0D8qs0nIz0Z_viNFTzydCbqfyl8irfbh40YasBQngmGl-_LSBU6_3hoh39ssLX4euV1xdBpiO8KsQgvDvLZF35hIogSmu4BE17JjVJcJGK0DuSPhh-Y0JxedJn0nskjm1FCQjhLi4FBh5JPAf3R2uJ-it0BFCRci00xUyduL0lelHwRF3abBC6OvNzE6uhi8uNRJHZet7rvNPC9Noi2U0asD-CBUa0tfIU1FExuDvfd1c2zLDWFCsQc_QbBlVvwVirgo0WC_S4TQ8I3pBHFydmquoBXzShhJDT7TixT4cz2Vgl_dXzBlELkYgEeHYMhvF8fRRsHrf7sd4yj8bqfBYj22GKqPSkdBk5A_f08DZBRcegVxK2oi9rWpLLRvSi_kvVe5whaMrPRlqb-uNbk9mDEQh7M8m9LAxMWA84blOlniAtinYP_Q3EOXnQUrCJoF0eyXdNZiYdyM4PeZpc1CmqbKLTBZ_4md9ObCgo2i8fdtUwNughaIcPHrc=s64",
"userId": "00408651407692255291"
},
"user_tz": 300
},
"id": "pSJ-OTULy5fW",
"outputId": "b2e62c94-9b87-48b2-d8d1-5bcc01e2f58d"
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Training error:0.875\n",
"Test error: 0.786\n"
]
},
{
"data": {
"image/png": "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\n",
"text/plain": [
""
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"rv, rv1 = data( mu=[1,1], mu1=[-2,2], \n",
" cov=[[1.0, -0.8], [-0.8, 1.0]] , \n",
" cov1= [[1.0, 0.5], [0.5, 1.0]])\n",
"\n",
"X_t, y = sample(N1 = 60, N2 = 10, r = 10)\n",
"\n",
"X_train, X_test, y_train, y_test = train_test_split(X_t, y, test_size = 0.2, \n",
" random_state=1)\n",
"clf = SVC(gamma = 0.001)\n",
"clf.fit(X_train, y_train)\n",
"fig, ax = plt.subplots()\n",
"X0, X1 = X_train[:, 0], X_train[:, 1]\n",
"xx, yy = make_meshgrid(X0, X1)\n",
"plot_contours(ax, clf, xx, yy, cmap=plt.cm.coolwarm, alpha=0.8)\n",
"plt.plot(X_train[y_train==0][:,0],X_train[y_train==0][:,1],\"bo\", alpha=1)\n",
"plt.plot(X_train[y_train==1][:,0],X_train[y_train==1][:,1],\"ro\", alpha=1)\n",
"print(f\"Training error:{clf.score(X_train, y_train):.3f}\")\n",
"print(f\"Test error: {clf.score(X_test, y_test):.3f}\" )\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "zRNQMmyzlxAi"
},
"source": []
},
{
"cell_type": "markdown",
"metadata": {
"id": "8r9nNYOIWGvF"
},
"source": [
"# Model Selection and Train/Validation/Test Sets\n",
"[1] Reference http://cs229.stanford.edu/syllabus.html\n",
"\n",
"\n",
"Just because a learning algorithm fits a training set well, that does not mean it is a good hypothesis. It could over fit and as a result your predictions on the test set would be poor. The error of your hypothesis as measured on the data set with which you trained the parameters will be lower than the error on any other data set. \n",
"\n",
"Given many models with different polynomial degrees, we can use a systematic approach to identify the 'best' function. In order to choose the model of your hypothesis, you can test each degree of polynomial and look at the error result.\n",
"\n",
"One way to break down our dataset into the three sets is:\n",
"\n",
"Training set: 60%\n",
"\n",
"Cross validation set: 20%\n",
"\n",
"Test set: 20%\n",
"\n",
"We can now calculate three separate error values for the three different sets using the following method:\n",
"\n",
"1. Optimize the parameters in Θ using the training set for each polynomial degree.\n",
"\n",
"2. Find the polynomial degree d with the least error using the cross validation set.\n",
"\n",
"3. Estimate the generalization error using the test set with J_{test}\n",
"\n",
"This way, the degree of the polynomial d has not been trained using the test set."
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "Hw6_tmH_l5Do"
},
"source": [
"\n",
"Tarea :\n",
"\n",
"3.1 Para esta situación tomar el 20% como datos para realizar la validacion cruzada y el 80% para realizar el entrenamiento. Construir una curva del score en función del parametro gamma del clasificador que se esta empleando. ¿Cual presenta una mejor solucion al problema?. \n",
"\n",
"3.2 Construir las curvas de aprendizaje estadístico para el modelo descrito, comprobar que la solución es similar a la esperada a través de sklean.\n",
"a\n",
"3.3 Elegir, los párametros adecuados para los centros de medias y las matrices de covarianzas y mostrar Mostrar un ejemplo de Overfitting underfitting con el algoritmo generados construido por ustedes. Analizar los resultados.\n",
"\n"
]
}
],
"metadata": {
"colab": {
"authorship_tag": "ABX9TyPW7zl9NLyDK3kYULoQYOV4",
"collapsed_sections": [],
"name": "Copy of Sesion_03_intuicion_estadistica_II.ipynb",
"provenance": []
},
"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": "3.8.5"
}
},
"nbformat": 4,
"nbformat_minor": 1
}