dj_Beeline_MS01.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Sklearn\n",
    "\n",
    "model selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Populating the interactive namespace from numpy and matplotlib\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "%pylab inline\n",
    "plt.style.use('seaborn-dark')\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\") # отключение варнингов\n",
    "pd.set_option('display.max_columns', None) # pd.options.display.max_columns = None \n",
    "# pd.set_option('display.max_rows', None) # не прятать столбцы при выводе дата-фреймов\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib as mpl\n",
    "plt.rc('font', size=14)\n",
    "\n",
    "from mlxtend.plotting import plot_decision_regions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.162435</td>\n",
       "      <td>-0.061176</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.945128</td>\n",
       "      <td>-0.043227</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.078331</td>\n",
       "      <td>-0.102277</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.156040</td>\n",
       "      <td>0.115038</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.999199</td>\n",
       "      <td>0.228718</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1\n",
       "0  1.162435 -0.061176\n",
       "1  0.945128 -0.043227\n",
       "2  1.078331 -0.102277\n",
       "3  1.156040  0.115038\n",
       "4  0.999199  0.228718"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x216 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.datasets import make_moons\n",
    "\n",
    "n_samples=100\n",
    "\n",
    "X, y = make_moons(n_samples=n_samples, shuffle=False, noise=0.1, random_state=1)\n",
    "\n",
    "# прореживаем, чтобы был дисбаланс\n",
    "i = (y==0) | (np.random.rand(n_samples) > 0.5)\n",
    "X = X[i, :]\n",
    "y = y[i]\n",
    "\n",
    "plt.figure(figsize=(7, 3))\n",
    "plt.scatter(X[:,0], X[:,1], 40, y)\n",
    "pd.DataFrame(X[:5,:])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## train/test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x7f36944230d0>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.33,\n",
    "                                                    random_state=43,\n",
    "                                                    shuffle=False)\n",
    "plt.scatter(X_train[:,0], X_train[:,1], 40, y_train, lw=0.3, edgecolor='b')\n",
    "plt.scatter(X_test[:,0], X_test[:,1], 20, y_test, marker='s', lw=0.3, edgecolor='r')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 [4 0 3 1 7 8 5] [2 9 6]\n",
      "1 [0 8 4 2 1 6 7] [9 5 3]\n",
      "2 [9 0 6 1 7 4 2] [8 3 5]\n",
      "3 [0 6 1 5 8 7 9] [2 4 3]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import ShuffleSplit\n",
    "\n",
    "ss = ShuffleSplit(n_splits=4, test_size=0.3, random_state=1)\n",
    "\n",
    "for i, (train_index, test_index) in enumerate(ss.split(np.arange(10))):\n",
    "    print (i, train_index, test_index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 [8 4 0 2 5 3 6] [7 9 1]\n",
      "1 [9 3 4 7 5 2 1] [6 8 0]\n",
      "2 [8 7 0 3 6 4 2] [5 9 1]\n",
      "3 [0 2 9 5 3 6 8] [7 1 4]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import StratifiedShuffleSplit\n",
    "\n",
    "ss = StratifiedShuffleSplit(n_splits=4, test_size=0.3, random_state=1)\n",
    "\n",
    "for i, (train_index, test_index) in enumerate(ss.split(np.arange(10),\n",
    "                                             [1,1,1,1,0,0,0,0,0,0])):\n",
    "    print (i, train_index, test_index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 [3 4 5 6 7 8 9] [0 1 2]\n",
      "1 [3 4 5 6 7 8 9] [0 1 2]\n",
      "2 [3 4 5 6 7 8 9] [0 1 2]\n",
      "3 [3 4 5 6 7 8 9] [0 1 2]\n",
      "4 [0 1 2 6 7 8 9] [3 4 5]\n",
      "5 [0 1 2 3 4 5 6 7] [8 9]\n",
      "6 [0 1 2 6 7 8 9] [3 4 5]\n",
      "7 [0 1 2 3 4 5 8 9] [6 7]\n",
      "8 [0 1 2 6 7 8 9] [3 4 5]\n",
      "9 [0 1 2 3 4 5 8 9] [6 7]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import GroupShuffleSplit\n",
    "\n",
    "gss = GroupShuffleSplit(n_splits=10)\n",
    "\n",
    "for i, (train_index, test_index) in enumerate(gss.split(np.arange(10),\n",
    "                                                       groups=[1,1,1,2,2,2,3,3,4,4])):\n",
    "    print (i, train_index, test_index)\n",
    "    # if i > 10:\n",
    "    #     print ('...')\n",
    "    #     break"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k-Fold"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x216 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.model_selection import KFold\n",
    "\n",
    "n_splits=3\n",
    "\n",
    "kf = KFold(n_splits=n_splits, random_state=1, shuffle=False)\n",
    "\n",
    "plt.figure(figsize=(10, 3))\n",
    "for i, (train_index, test_index) in enumerate(kf.split(X)):\n",
    "    # print(\"TRAIN:\", train_index, \"TEST:\", test_index)\n",
    "    X_train, X_test = X[train_index], X[test_index]\n",
    "    y_train, y_test = y[train_index], y[test_index]\n",
    "    plt.subplot(1, n_splits, i+1)\n",
    "    plt.scatter(X_train[:,0], X_train[:,1], 40, y_train, lw=0.3, edgecolor='b')\n",
    "    plt.scatter(X_test[:,0], X_test[:,1], 20, y_test, marker='s', lw=0.3, edgecolor='r')\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "kf.n_splits"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 [4 5 6 7 8 9] [0 1 2 3]\n",
      "1 [0 1 2 3 7 8 9] [4 5 6]\n",
      "2 [0 1 2 3 4 5 6] [7 8 9]\n"
     ]
    }
   ],
   "source": [
    "for i, (train_index, test_index) in enumerate(kf.split(np.arange(10))):\n",
    "    print (i, train_index, test_index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkcAAAEvCAYAAABPDGVdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nOydd3hVZbb/P3vvU9IgEEghJKQSQgihd6T3jihKsTA69U51Zn7qODrj9d7r3DvOOOqMzujYKQIqIoggvUmRXkNCSKGGkEbqKXvv3x87ycnJOQkQkpyTsD/Pw6N5T8mbc/a737XWu9Z3Caqqqujo6Ojo6Ojo6AAgenoCOjo6Ojo6OjrehG4c6ejo6Ojo6OjUQjeOdHR0dHR0dHRqoRtHOjo6Ojo6Ojq10I0jHR0dHR0dHZ1a6MaRjo6Ojo6Ojk4tDE31Rnl5JU31Vjo6TUJwcDuP/n59Teh4G/qa0NFx0NB60CNHOjo6Ojo6Ojq10I0jHR0dHR0dHZ1aNNmx2r3GhQsC6ekmzGaBykqV5GQrERG62LjOvYksw/btRiRJQlVBFGVGj7YhCJ6emY6Ojs6doxtHjSA9XaSy0ofFi6WasS+/lIAK3UDSuSdZt87MokUm/Py0n/PzDaxbJzJ9usWzE9PR8SSyjJR1wXkoOhYkqZ4X6HgL+rFaI8jMNDFunPPFPWuWxOnTJg/NSEfHc6SliQwbZqwxjAA6dRLo1s1AXp4eOtK5d5GyLhBUnEdQUID2rzjPxVjS8U5046gR+Pi4v+GbTPpGoHPvkZMjkZzseu0PGCCSnq7fYnTucWJiICFB+xcT4+nZ6Nwm+rFaI7Ba3R+dZWdLpKZKJCbKLTwjHR3P0aWLzPnzKvExCmRk1IxnHFHo4VcBcox+jKCjU8WuXSZi/QS6dNFTMLwZ3a1rBGFhNg4eVJzGdu+GWbNEcnJMKEo9L9TRaYP06qWwbZsdW2oGZGbWjPfvL5Jovq4fI+jo1GL2bBOXL/uSlqZvv96MHjlqBMnJMkuX2sjJMWMygc0GPXtCjx7g7y9x4oRI3766haRz7zB1aiXrV1iZO6fqCKE2BaWemZSOjjdQy2EgMxMhJoapUyWWLTORkFDpuXnpNIhuHNXBZgNRvPUpQHi4ysyZYDY7j6sqevmyTpvBboebNyEwsOE14eMDo0dZAb0oQUenGjk6lq9XmplanWoUEwNxcQD4+uqRI2+mTRpHigLbthkRRe1uLssKY8daMTTw1+bkCJw5Y6ZDBwm7XaWoSGbsWAv+/o7nnDsnkpNjxGAQsFpl1q6VmT/fecfYuVNm4kQ9aqTjXVy7JnDsmAl/fxGrVSUgwM6QIfYGX7NjhxEw0qWLwIkTKqJoY9Qom+MJdcqUxZxsKKjjLWRmQmBwE/4lOjqtCEmioHM8JJhdHrLZ9Jwjb6ZNGkfr15uZP99Iu3ZaCKeiApYuFZgzR9NcSU2VyM424usrYq2wEWE5x418E4/MqTJ04uKQMfHBBwKzZmlhz+++MxASYmbhQs3aLysz8sYbdlaulJk2TcJiUdm4USEmxoKoOwQ6XsTNm3DypC+PPOIw5DMzDezZU8nIkXYUBbZuNWEwSEiSQG6uwqVLAj/8oURUlCMMeuGCyMGDCoMHawUH1WXKNRU4xX4UFZej1D5GCwzWdF10dLyRO9QhUu0y+Qcz8fVVad/e/WsKCuDAATP+/pogamGhzIYNCtOmOTaG3FwVQWjYOdHxLG3OOLp8WaBXL0ONYQTg6wuDBhnJyLCiKAKK4sPixVUXaloGZF6D+6pu8FXnw1JCAomJErm5AiEhKiUlRqeL298fnnhCYvfuCrZsEZAklQkTZL0oR8fr2L/fxEMPOV+YMTECBw4YATsbN5qYO9dE+/bVa0bi9dehSxfn94mNFTh40AjItd/IKcdIKShFjuveLH+Hjk5T42LgZ2ZSkIV2DdcxnNLSJMpOX2b8aCA63vEmtV6jqrBrly9PPmmoSa9QFIlXX7VTXCzj4yNitYIg2BgzplYUVsfruC3j6Ny5c/z617+mvLycbdu2Nfec7oq0NIk5M1VIS3Ma7+MLK052RRWNLFpUJ7QT4yaJFAgPF8nOFpAkiIhwtXqCgwWsVonx462Nm6wbrwV0BdXWQGtaE0ajgMEAOTnw3XdgNGq5dYoicvMmdO5sqGUYaTzxBGzfDpMnO79Xc1+Wly8LnDxpxs9PRFGgvNzO+PFWl9w+HZ0mo+79vyryWddwGjoUCK1aAPUUHRw9KjF1quSUdyqKMHGiRFlZGfHxjT9KUxQ4elTEYhEZONCOSU/va1ZueQC0YcMGnnzySaKiolpiPndNfLxM6lfpzhUCgJCVSU9T2h0JNR45IpOQoNCunUpenmsekdUKqtr4i91FPVVXUG0VtLY1YbernDwJqakwbx7MmgVz50JxMWRmCkREuN4G/P3BUqfzh6JARUXz5dNZrXD8uA+PPGJk3jyJBx+UWLDAzKZNumXUGjh37hwzZsxg3Lhxnp7KXbF/v5Ga23ptAcfbEHHMzxfp2tV1j4mLE7h2rfGexYULAps3+9Knjx9jxvjy7bd+HDumO9DNyS2No7KyMlauXMmwYcNaYj53TWSkSmqq3e1FHRWl1ivgWJcDBxSMRitGo1aRdv68nW+/VZ00jD7/XGbo0LsMjd7h4tPxPK1tTQwfbuXrr2UmTXKMiSL86Ecily4ZSUtzNXiuXIETJ6C8XPu5tBTefdfOiBFalDQ/X+DgQYPmhKSlaf/qOCR3yv79RubNcw5mm0wQHW2goOCu3lqnmWltDkM1V68KLtfwwIGmqmKEerh0yfH8Otd9crKdvXtd19OePTK9ezcux0hV4dw5HxYvNtCli0BgIDz4oERZmZlSXSWj2bjlsdqDDz7YEvNoUsaOrb+kuGdPK19+KTFrVi2ru44OxbpTXQkZWc6wYSpXrwocOeLDpEkS/v7w7rsq5eUKnTopJCZaCAho3r9Fx/tobWvC3x+io13HBQH8/UUMBitnz/rQM0FTuK6shH1fwW9mwfZtcWRfErBabcybZ8Vkgm++MREWZmTME704t83MsZU2hg+34H+XydcWi+B2PXXrJpKbKxIUpFeBeivVDsO2bds4e/asp6dz2xwv7UGvXrVKkmNiCImLQ/20yomua/BnZkJERM2P69bZGT48GCU6FmSZiPIL7PzWxGWbUYsgxcWRfkGksNBGYGDj5nj2rMjw4a5RounTJT77zMj48XruUnPQ5hKyoeG8iMhIFUGo4B//MOPvb6CdXxwDO0IUWgXBsbRwhi6KAElbHEeO+PD4446PKS4O1qyBvn0rGzSMcnMFjhypqoizqnTvbiUmRi/d1PEMsqwArgvDZlMZM8bOsWMVnN9wgZnJl/GJiWHePCAzE1uqld7DYkhI0AyTQ4ckRo401VSx9ZjZg/hp8PHHVqZPt7i8fzV2O2zfbkKSxJr5jBtndVqrXbvaSUszkZDgfCxx/LjMsGG6YeTNtDaHoRqD2eg239RolJGjYynIAsvVUnbuNCMIBgYMiCF+chx2VWL9epnA/haU+KrqzYx0gorzNCFUgMxM1q+34pMSy/jxjW8ppSi4rYAWBLiLrA6dW9AmjSPAvcVfpbcSEaHSoYPKokUC2oahLY7QBCi8bAepAoDsbIE+fVw3lBkzJD7/3Mi4cTaysgTOnTPj6ytis6kYjXZ69LBz9qxWOl2dmLdli0R6eiXdu9e5yTcwTx2dpiIw0EZqqoHERIfhsWePQkyM5nX27SsjtbNCkHNy6sjAUuQ4xzVbVGRwKu8HzRnx969aJ/WURq9f78eiRSb8/LSx8nJYtswhrwGQlKTwxRc2OnY0Ehys/Y5jx1QMBquefKrTLMiygtWKy/VVUaGAJCHHdWfLRhMPP2vGaITTp2Htes3YLyiwM2JEHaOnTnL38MBS5DjHc9LTRbKzRZKSZMLD1duSEkhKUtiyRSY21nm73rRJZvBgPWrUXLRJ46ja4qcBvZX6ErONRsd4ZaVAsBs7xWAAu10gP1/g4kVfFi92XMhXr0r8+98Sv/+9c8XChAmaXHz37g65+NuZp45OUzBokMyBA5UcPWrAx0ekokIhPNxGjx53FpGpT/29etxdafSJPQIDBvStMYwA/Pygb18jWVlWoqMd7u/s2Rb27JGxWCQUBSIibIwYoUeNdJqHESOsfPCBxOOPGzCZtEjMli0ykZEOo8PHR8JYlYLUq5f2D2DlytsXtKushI0bfRgyxMD8+QL79yusX29nVuIpgm7WIyVQhShCVJSV5csFZsyQMJvh669lzGaLk9aSTtPSJo0jJAlLVHf27zdQWaYSL6QRF6fUWOhydCwWi4Ldjotqdu1qnIQEhc2bZXr0cH7Sjh0KffrYOHTIyMKFzpGlLl0EunaV3G4iiiKyZYsPBgNUViqMGGGlna4Jo9NCaIrYdm7cEDh0yMiNGya2blXx87MzbFhVsmid/Lu6UUx/fztXrhgJD3dc4IoCpaX1ax9lr6tkxjzBubWOLNM/IIOvv64kTnYkqsrRsYwcCaAL5Ok0P76+MGVKBZ99ph35Wq0KYWF2OnVyGOxN0Uh8+3YTjz9urDGyRowQSUw0cmi5kcmTXaUEFEWrUAsIgLAwlR49ZGJjy9m61YDFAkOH2p26N+g0PW3SOMrNFTh40Ie5cyU65qVDZq52XAA1lvmIEd354AOJRx4xYDZrHsO6dTLduzvC/IIAERFWVq8WmDVLwmSC7dsVCgutJCerGI2C27NgPz+BTz/VQrVWK/Trp+UqqarIggWaMaUo8N57ElOnVuhHBjrNT1X4vqICzuz3YfH9BoT4OJAM5OQY2LnTwuiRdSKZbqKYQ4bIbNxoJTbWyIgRIllZKps3y4wfX3++kdGo8u67KqGhArYqh3x2zwyknExmzKhVnenGa9bRaW5OnpQAEZtNICvLQIcOJoqK4PBhmeTkShRFpqREchIWLihQEQQ3eUT1OBe+vo7oUzWdOgnY7a4bSOo5iexMX/r2lSgq0pK+R4yoJCgIRo7UnYaW4pbG0eTJk7ly5QqKomC32+nduzcAGzdupGvXrs0+wcZw9KiJ732v6k/Lw63IV0AATJpUwZo1RkRRS5ru189GSIhzhltSkkxUVDlffGFEliElxU7v3tpzVFWhokLzPmpjt8PDDzt+/vJLuHBBdcr3EEVYsMDAhg1GXSm1ldEa10Tt467qZGsEICGBbt00tWxFsMEtDBNBgKlTrVy9amPFCgOhoTJz5igNNls2mUw8+aTjCaWlsPEdmD7djfhqgfva5PJyTVJD10bVaUq+/dZAUpIP3bs7rs9PPoFRo2DECAMffODD5MkVrF4tEB9voHdvkWPHFHJybEyd6nzfdkmTuFWKhCzjfz0Ldl9zGFWyTGVFDAsec2zNQ4ca+eADmDGj0u3b6DQPtzSONm3a1BLzaFL8/G7vDurry22VQfr7w7hxrs8bMcLGxx8b+N73DDXHc3v2qE4LDWDGDPjjH1X+8z+dx/39QVGcx/bvN1BaasTHR6C8XKFrVxu9ejW+0kGn6fGmNXH4sERhoQEQMJvtjBxpr99QqUcJHjQvtrSU285h6NJFpUuXetZOHe95zBhn7a6AALBaFW5DZo1TpySuXTMRFCRSVqZSWmpn4sSGm0jrtDyt0WEAqKw0utyv582DDRtg9myYMkXiyBGJmTMt5OZa2btXJC5OITnZTZlYVQK3OyoqZGw2g1P0qPDQecaMMUDMfdpAZiZbdyqMecl5jQoCxMdL5OUJBAfr5WktReu5xdxBq42mOCO+HXx8tOjTypUmzGYt+lRUJPGTnzjPRxShUyfX11dUADgu9gMHDPTo4VOrlFli926JjIxy4uL0RaHjzNatRgYPNhMbq10veXkG1q2zMXNm/Udc9XHjhlqTaHo31PWe953pyvQJcS7Pc7ce6nL1qoDN5sOCBQ4jqrxc4tNPteiVjvfgTQ6DqsKOHUZU1YDRKFBZqRAZaSMx0dnJtNs1na+6GI2O3LigIIGSEu2H0FCV0NDGOapjx1r54AORYcMMJCUJHDigcGWHjXnznB2W8hArksnVuQ8IEKr2iyrusGGuzp3TaowjlyoYqDdHoaLCjtUqOXJ5bpFkejf4+8OkSY4b9ebNvi7PsVjg5k2ZixcFIiO1haaq8MkndqfIVUmJwUXj5b77RJYtMxEXd+cbnk7bpaQEAgONNYYRaL3++vY1cuGCldhYN8Z03XVQtZbS01WMRqvb/Lk7po73bLliIL9IdDGGiosVyMx2nV/V2lQUOHHCtQ+inx/4+hpQlCaar06bY8cOI6NHmwkNdTiZmzaJXLlSoZXPV2EwQGmpawTTWsvu3rxZZvDgu8/z8fGBOXMqSUsTWbVKpGdP2a1YcbduNg4dUhk40HkfOHNGZuJEx9wbbJir0yS0GuMIcH8s4CZHYfx4Kx9/LBAXZyCmWxxp52WUNAuDBtmavVQ+ONjKgQMiQ4ZUi93B0qV2HnusksOHDezdq3kz+fkKyDayN2eQlKTlbATd8AU5wcX6v5N+cDr3BqmpIkOGuHqJ/fsLLF9uIDa2/nyIkhI4crwr9qJu2A/b6NDBxsiRzXN0O3y4ndWr7SxcaKR9e80p2L5dwa93JAXtK2vWb2GhwJkzXTlrjcf/sIGICJHSUpVdu2D0aOf39PUVsNtdtWl0dABU1VjLMNKYNEli+XIj4eHOEUd/fxtnz0r07OlwWlev1voP7typYLdbm7QLQkKCUiOoSgYuDktkZDAb06woipHBg0WsVq1QKCLCSl6ewIkTBjp3lukfQL0Nc3WahtZlHN0mJhPMmmXhxg0rmTkCibMi8PGBlsjc6dtX5ty5CpYvN2I0ClgsmhKwry+MGKGVUn/zjYmEBCOju2QhZOVCJ836n5yQCRmS0wWvqlrZv45ObcLDVc6fV+jUydnrvXJFpUMHN9dLrYiOHzCyL4Ct6l/zIUkwe3Ylmzfbsdsl7HaVnj1tREaCjDafI0ckFHyY/iuRGQIcPAgFBfCjH8HZsyr79wtaR/Qqiotl3TDSqRd314YggMHgGmocOtTOkSMVHD1qxGQSyMtTEEWVdeugVy8bSUnNl85QXwL3xDgr2dk2ViwVaZ+XweAUG0e/M+EbZuDhQSKXzLFsXerD/PnNNjUdWptxdIdq0p07q3Tu3PK5Oj16KPTo4f4Y7OhRiaFDTcTFCZCGi/W/c6fM6KofFQWWLZMZMkTPr7jXsdvh2DGRzEyRqCiF/v0VvvpKpn9/sSbJU1Vh/XqZOXO8K4HfYIBRozTHoC52O9y8aebBBx0b15AhsHMn5OZCz54Cn3yiMnSogKLAxo0yERF6dadO/WjOpFRnDFTV/bro31+mZVznOjSQwB0VpRJrTyUoOA8iY4iMrHogM5PIGIh40NCs6SI63mgcyTJSxnnEyxdrhpSuEZqVDa1eTTo/38CkSfUfk4WEVLJ0qQWzWcRiURg2zNrohoU6bYOTJyVycsxMmiSSmAi7dsGnn2oG+IoVCr6+BgQByspkxoyxNFhW722cOiUyYoTr8eCoUbBunXa8UVamsHKljN2uyW107qxy9qyIqkLPng3LCOi0bS5dEvjuOyNFRRKKohIfbycszMbWrRLjx2sGt8UCH31kdym9bxXUU2EqxMex6RuZge1Lteu/Fe6F3o7XGUdS1gWCSvOhb1XpTGYmpJ6goAEruy0RGqoyeXhVpKi6IuGG83NqVyXIsnYsIcswcKCslzi3McrLIT/fzOOPOwyImBhYsUIiP9/M+PHlmEytN7LYrp1KQYHqpLgNDl0jVQWTyc64cdrfmJEhcOiQL8OGaZ/Hxo2acGt8vHaUaLHA/v1GrFaB6Gi7ay9DnTbD0aMSguDD978vIghw9Sp89ZWR69dtDBpUztKlRkwmEbtdZsoUGz4+np5xEyJJlIYnIMeV1zgH168LHD1qRJIgMFBm4EBZdxzuAu/cShvQY2nthITYOHvWWJMA2FBo9FYVeufOSWRlmZk4UcRgENi8WSY01EJysncdq+g0ngMHjMye7VBVLyyEwEAYPBjy8iT27zdUHVndGRkZAhcuGDEaVYYNs2E2N/XMb4+4OJX162WSk51vRRs2wNSp8PHHslZIgeYInD/vw6OPOp7bvbuBZcsEYmLKyckRSE/3Ze5cCT8/OH7cxPr1NqZPb13RNJ1bo6qQn29yknno0gXGj4esLAMXL9qZPLn1Og23QlWhpMRh/Jw6JVFRYWbhQglRhIsXVb74wsacORYE5fZlcHQceJ9xJMuuBsOlSxAQ1Oj38yY9iJQUhW3bLGRnmxg9Mo4rGSqHV9kZMqSSdu5Co/VU6MkyZGWZnZrePvywxKpVZiyWco9tdjpNiyxrl+rWrXDzJgQHQ36+pjIdHd04Ta8NG0ykpJhYuFCgvBy++MJIfHwFMTGe0dIaNKiS99/3ISFBa6r57bcqgmDnyy9V7rvPUS106JDE5Mmu63bqVJE9eyRu3jQ6rYc+fQTatzdy9Ki9Kq9Ep61w7ZpAfLz2XV+8CEePauXyY8bAyZMCNpvEnfbnc+gjae8rCDJjxtg8a1i7ybO9oEazdaedsWO1vFZVhatXTU59PiMjBaZM0a79QYHpty2Do+PA64wj8fIlCKzVvrvqCxUvX0JOSLzj9/NGPYhx42wUF9vYsMlASEgEY3+o7XB1b98NbXxHjohMmOBafTFjhsSGDQZGj9Z78LQFBg2y8dZbRqZOlYiPd4xfugRLl8osXHhn3/OJExIjR1YVBKDpBi1cKLF0qZmYGM+0JwgNVZkxo4LcXK332sMPq243JKtVwM/P9QE/P02oLyjI1XCKiRHYv1/CIwm3Os1Gu3YqV66ofPmlQGgozJypHcWuWQPFxeDnd+eG/vr1ZubONdKxo3aNFRYaWLNGbJSoalNQt5rt5k04m9oF/5QuzJnjUIQsLITISNe9oFs3gW+/lSCQ25bB0XHgdcaR0jUSgtu7fJFK3s3Gv+ld6EGoVWusqb2HwEAaPA7JzRVI3eHDAwPcV+jZ7QJHjwr4+cGwYY7yVVFsOYVwneYnMBBEUXUyjAAiIqBTJ9Wlr9+tuH5dYvx414u5c2eJsjJu2enbZoNz5xRCQ6Umb2UQGtrA+8kywzqls/d9XyZNqmUExcXx9dfQr5+dtWtN2O3QrRv07ev0Up02RkAAHD1qZ9o0E336aGP+/vDQQ/DGGwrDht1Z8vWlSwJJSYYawwigY0dt7PJlK127eiCqWifP1h8Y2M/1aRUVAgcOaEaSLGsOz6RJ2t6l7wWNx+uMI6R6ZG/rG28mrFb44x8PsmPHJioqSklJGczTT08jKekOd6NGcviwmcee7gEZjq/o8mWVjIthGP1jKC42MWGCiq+vwFdfaZvloEGwYYPM0KFtL2qkqk1voLYW6jZDriaoESfNsuz+s7TZ1Fsm8y9fnsM//7mU1NRVdOjQk7Fjn+TVV8fj59fw65oCKesCwZV5dJnkfDSwbZtMiV80hw/78YtfaLlG6emwdCksWgR79yr06NEKq5R0bklEhFpjGNUmJUVw6mF2O6SmSjz0kOseM2iQyMqVEl27euc91WKBI0d8+PWvxVqJ2Vqlp6BYGRp0GjEnG4prLdI413Y+Oq54n3HUHDRCD+Lpp/eybNksqkXyLl9+i8zMP7Bp06/v2Fu/U1QV2rUTtWSTWhGvrgmwO9eKPUvkkUccX93cufDFF3DihEJkpKXZ59dSlJfD88/v5dtvt2KzWenffwTPPz+ByMg7vPO1cioqFBQFp3YZqgrl5XfuFvbsaWP3biOjRjneTFHgxg25wTy1kycr+eMfn6Ko6BsAioousGbNRvz91/PXv46443k0CjdHA/FSKVeuGZxyjbp31yJuf/iDQkqKhdGj20hfQi/Ln/Q0ouje0LdaVYzGO/vOY2JkTp1SSUlxfrPTp1ViY28RevTg97Jvn5GHHzY4fQYhIVBcrBJrP0tseJ6j8hu0/a96P2xDukjnz9tYvfosqqrywAM9SUi4e5VY7zSO7lDssSHqUyFtiJIS2Lx5KXXVg1NT/8TSpQ/w/e9HNWoud0J94dC8PIFZs9wlpcKyZQIVFW0nvPKzn33DunXzAe3DyMl5nUuX3mDdusfuqf3gvvus/PvfEgsWGGjXThO0W7HC3ihx0G7dVA4etLB6tYn+/SWuXlU4d05m3LiG8ypWrTpeYxg5kNmz5xsUZYTH+pz5+7tvHhoSAv36iRQVGbjTxFxvRcq6wIJhA6i+O8YAK/YdvmeTalNSbGzbZqzRMwLNWLp0SXY6Vr0d4uJU1qyxk5horElRsFrhwAE7c+Y0bGh58ntRFMFt5LZ7dwi6obh1KG5u2Ix90JA2o4v0/vuZ/OlPz1JYuKHq58k8/fSfePLJu4uQeZ1x5GLMwN0JXDVCH6mgQKCg4KybRyxcv14ANK9xJAhamaaiSE6bzsmTKn5+NkTR9WsTRQgNFbh500hb2Ayys2V27HiTasOomsOHX2T9+pnMnt3I6sVWSEAATJ9ewaZNRux2AVFUGT/e1ugI4eDBdux2O+npIkFBKrNm3drLrqysqGe81CWq1Wy4cZqEwOAGc4qioyUKChp3BOmNZALpnp6ElxAWpnLpUiWrV5sZOlTi+nWVY8fsjBzZuMKCadMqWblSwWzWPC+LRWbatNtzQDz1vZhMCkVFKh06ODvFFy8qxHZxv67bt/OhQBLbRMSxtBT+/vd/1BhGAMXFm/j738OYP/8N2rdv/Ht7nXHUGGOmqYmIUElMnMWpU0ecxk2mHtx3X3w9r2paxoyx8N57Av36GYiKEti3T8ZisTFjhszWrYpT2SbAN99oDTo3bWobkaOLF62UlJx2GVfVG+TkFAJtZLe7BaWlUFYmEBKiMnp00+XOGAyauvTtMnZsH5Yti8Juz3Ya79t3ZIsIjzbkNJWds2O1Sk49tY4cgcREqKwUKC8XCApqI0drOk4MHChjs5Vz9qxIYODtGfr1YTbDlCmtSxtp6FA7K1fKPPmkocbWOX9eRRRt9a/LiIgWm19z89VX+Vy8+KHL+JUrS1m79gUeeaTxR4deYxzt3WugslJr1lpRodC9u4XYWM/c0CQJfvzjh/j97/dRWEbK+AkAACAASURBVLgJAEEIZs6cvzFqVLsWmUNAgNawMytL4MABkd695ZpKoogIC2+/7cOCBSImE2zapOnfBAQ0Lg/FG+nf35fo6AfJyvq703i7dqOYMCGynle1HQoK4NNP/UhJEenQQUssTkjwnMDn1KkdeeSR91i58mnKyw8BfiQnP8dzz01tmQk04DSNH2/lH/8Q6dXLQGSkQGqqFinq3x+WL5eZMKHtGEYx9fz/vUZpqdaXr0MHMBo1/ThP4qnvxWCAyZMrWLHChNksVn0mNkaNkiED9ykqbYguXXyQpDBk2dlpE8UQwsLuThJdUFW1Se4ceXkljX7tnp0CgztdpmtXR9Rj3ZloYuKtDZf3NjMnTlSycuVRKirKGD48hfvvD/FYbkVd8vNh6VI/hg6VGDcOfH3h669lgoIqiItrG5vBv/6Vzp/+9BPKyg4AIEmRLFnyIf/zPwNv6/XBwS1jyNZHY9eEosBbb/nz/POiU+R75UqFfv3K6NixiSbYCI4dq2DLlkzCwjoyf34Xtx3QPcVnn5kIDzcxbZp2H9mwQaZduzakGN8Eib+tdU1UU5Qvc2LNZYKDtUjh9esyUVFWwu+L8dwxkbcmylfNS8zJpkNJgSNiJMsUBIV5/ISmKVBVmD17Kfv3/8RpfNCgv7F+/fduWeHc0HrwCuPo4NKLTE+66CTUqEbHsPRgFFOmWLFaYccOE0aj1mzSaLQzcqT9ni3trqa0FL791oTJpDWp7d3bRnh42zCMqtm3r5Q1a45gt1sZN64/06cH3fb33lo3gm3bJDp29HVpUCzLsHSphWnTrJw7J5GdbcTPT8RmU1FVO2PHeljN1wtITxfJztYC4oMG6U2b69Ja10Q12/+Vw/xBl1zUnnP9ghETW/9m3yx4q/HWRGRmWvn979dx8OAaQGHgwPt56aWZxMffuk1EQ+vBK47VTCbBJateAHx8NGNo3Tofliwx1pQaX7tmYPNmK5Mmta7z4aYmIIA2+xlYrfDyy8fYt28XiqIwcOBwxo27fcOoNZOfLxEb6/qHSpImb3D1qkBpqZlFixw3t+vXJbZv19TX72W6d1fo3r0Nrok2vsHdDpcuCSQmGtxWYB3bVEn/RPTPyR1ekMfbnPj4GBg0KIXevQcwb1403ZvoT/UK48hiUQDni9dmA5tN4cQJiSlTDE4aLGFhAkFBBkpKrLTzrCOk00z8/Ofb+fzzBwFtozt2TCQ7exnLlk337MRagLAwhfPnVQYNcjaQzpxRiY62c+KE0ckwAggJEQAjdeUndNoGehk/FBYK9Ojk3juyWrV8B29sF6XTfHz8cTZ/+tNL5OWtAgSWL3+EF154lgce6HrX7+3RDBq7Hc6dE7ieJ2iJYmlp2r/MTNautTNsmJXr1yWio10XRHKyRFaWlyQA6TQpGRl2Nm9+lWrDSENh584X2bevzFPTajGGD7FQeOAsO99JQz6rrYm0s3ZWrZIZOFDBZBLcRtC8Kf9Hp+mpLhdPr/r/e43ERIXNm2XnvaJqv4iNreUUVEeWEhKcj9902hQlJfDqq69VGUYAKteufcQrr7xNZRO0iWzxyJHNBidOiJw/byAszEBIiEinQT3YddHAqKrr+P0dUdz3eBgBAdCxo8yVKyrh4c67wblzClFRbaMyS8eZkydLKSn5zmXcak1l06brrF17hdLSIvr378Ojj0a0SCl5s1PrOMCYk81PelcnUHaDzBz27LDyk19p+lqKolBRgYvOUUWFvh7uKWQFKcNZXactHiFV51YWF0sI4d25bJLoirY5fv65SmBgV0Y82nbK03VujzVrcrl06QOX8QsX/s2mTb+iV6/2lJbKpKQYG1VI1aLbyrFjEoWFZkwmiZkzVaKiqg0eiZycBHZmalo9M55S2b+/gqFDZQYMkPn8c5nvf9+h41BcrHLpkt1tX517jWPHJPLyDEiSAMiMHm1r9ffG4cMDCQ6eQl7eZ07jPj4jWbbsbxQXvw/AqlX+7NjxER98MNFrqgjvmNoVJYF+mqcbVCX3n5mpbXQxMcwJLEWuWi4jR9r46CMDjz/uOG7etk0hPPweP1Jzk28CbcdgqFsuLl6+yEMP3d8mj9oUBbZvN2KxSOTnS/zsZ2JNv7Rt2xIotELyABgVpHL+fAVItSoSG9Eu6l4gP1/g3DmR6GilTRTudOhgBtoBhU7johjEG2/sJS3tVSyWfFJSHueZZxYxfnznO3r/FjOObt6Eigoz8+dLrFlDLcNIo1s3OHxY+/9OnQRKSkRARhA0HYfly034+EhVnYbtTJ3aPBvBvn3FvP32NjIzzxIaGsmCBROYM6dLs/yuu2X3bgNJST5MnKh9lqWlRpYtE5k719KqE5dDQgQefPDn/Otf3yLLV6tGAwkImMeNG7+q9cwyNm36IV98cYD772+dN8CaHIlqw6hOoqk7zGaYNq2Czz83IooSNptCYqKNqKhmuOG1ogRXl3wTaDM5J3J0LCv2HXYaU2S5zSpmf/WVmQceMLJtm8CiRThFh8eNg9WrITkZYmIE9u+XAM04aky7qMZitcInn1whJ+cafftG31ElbUuiqrBxo4mICCNjxoicOaOwdq2d6dMtrTrqPn16B/r0eZrjx59xGg8IeIQTJxbU/Hz8+PM888x1tm//bwICbv/9W+yjOXjQyIMPajfU+u6r1d7/9u0K/fs7WmAEBLSMcumpUxX86Ee/5erVTwA4cwa++24w8LHXGUgWC4CJxETHagwIgMmTjRw/bqdv39at7fKHP/SjR48dbN26D0VRGDBgCK+++ozL81T1BocPn2u1xhHQcF5EPU0ifXxgwgQbzZ2A3eoSgd0ZmLVVtVsrbiqO6h6ptRWuXxeIjTXQvr2AJOF2A6+OmJaXg8FQyyloocqsK1fsLFnyIUePPgeUI4rhTJjwGu+9N9nrcv/27DEwZYqpqmgDhg0T6dvXxGefqUye3HorOyUJ/vKXx3jhBT8OHfoIUTTSu/ejHD++yeW52dlvsXz5D/jBD26/9VeLGUe1+y9Zra7dlFVVy0c6ckQlP99KSkrLh/3ef/9AjWFUTWnpQT75ZCtz5ixu8fk0RE6OSFKS61lSdLTAvn0OT6q1IgiwYEEXFiy4H9CujY8/juPmzbrPlOjcuRUbRrVxo2ZbVFyO0i3Ko00iW3t0wmbzcOWJzh1x4YLIyJHaN2arx/a3V/nOn34qe0S+4n//dydHjzqi2IpyhW++Wcy//rWfn/3s7hqeNjU2m6HGMKrG1xdMJu+L/t4pKSn+rFnzONnZjyOKUFamMHbsG26eaaei4s4MwRa7Z/TrZ2fbNi1hdPx4+PhjKKsqPCothT//WaGgwEJlZTljx3omd+L69ctux3NzL7XwTG5NaKhCTo6rAXnzJqSmipRUaa3l5gocOKA132zNGI0wZcpctDNmB/Hxv+Z737v1UZRXk5lJTffU3bspOnaagoJSCgKDsY0aq3nCXniM1Vo4etSX775rxecHDRADdK/611bqsuLiFI4d0/aKfv1g507nx7OzVc6fl/noIxspKRVOMi8txfHjh9yMWjhy5Iibcc9S31Ffq83TrIMgQHS0lpqTmCjSt+/jLs8JCprNvHl35mC22B0jOFglPd3Cxo1mJk4UmTkT/vpXmc6dFTp3llm0yObxcGRMTOIdjXuS9u3h0iU75eUm/Py0MVWFtWvhueeMvP8+GI0q8fEGhg0TOXZM4cABG1OmWL3yXPx2eOGFfvj4bOKbbzZy82YhyckDeOqpyQQGttI/iFo5EsUV2tFZdY6EFxlDraqfl5vo25QpMWzLNpC/L5WQEGeHwts+6zvBXR6SJyOMTUVwsMqBA3YKC43ExAiUlsInn2i6dwaDgtlsY8kS+63fqBnx93efvOLndwdJLS2EqsqUlhoJ8JUhIwPQIm/SBStShJUbgbHsO+iLv7+EzaZiMsncd1/rVNsXBPjd7x7i6aevk5HxOmClU6f7+cUvniMi4s7WeYu3DykqgkOHjBiNKkOH2j1i9ddHbq7M/Pl/4+zZF2vGunZdyD//+X8MGdLegzNzj90On3xiJizMiJ+fgNUKo0ZBaCi8+67Kww8LNc1qAfLyVHbtsjBq1L1R1dTaWyV4Ba0oIRtZpvL0BbKz/WqOZQCIi0M9n4GQlemarB0Y7L35U81Aa1kTqqpVq6mqhChqhtHYsdaaijVP8+qrqbz88gTAcc7v79+Pd975ggkTPNj80A2yDOvWmRkZmklPc5bLGlh9KJIHfpdUYwxdu6aybZuViRNbbz5SWRl88sllysosPPBAFOHh7u9XXtU+pEOH6kRS7yM0VGLVql/y5ptjyMw8R2hoBEuWDCYp6e66+zYXBgOEhck89JDJJUTasaOzYQQQHCxgt0voKsqtg4ICOHTIhNEoYLerDB9udflOm53W1HpAkiiP6I5s8IcE5wUhCLTdZO02iCBUt8LxznvVL36RSFHRl6xfv5xr1w4TFzeJRx+d73WGEWh+zJw5Fi5urYBE1zUwLcT56C0sTMDX14DVavX4aU5j8feHJ564O5Vs7ziI9yJ9ktBQiRdfHAAMaNHf21h695bZvVth9GjnzaCkREXrUKfTGsnJEcjM9GXBAs1ztlph6VKJ0aMraO8FQczKSlizJg9ZVrn//pCao11PExQEBw8q1E2nPHtWpWdPfT20BdLSRLKyTPj6ak2XBcHOmDEtewwkivDii/15+un+5OcLhIWpXhPVqo/oaPcise4crshIkfx8gS5dWr8eUmPxCuPIq/VJvMhwc0doqMrRozYuXjQRGandHbKzVbKzZex2g1MZbHk5yHLrrmK7V0hNNTn1TzOZ4PHHDSxbZmoRWYuG2LLlBs8//yYZGX8HFN5448e88MJPmT491KPzqiYhoZKlS32ZM0fC3x/27lUoPG+jp0+dggtdILDVcfWqwM2bPk5r49o1iR078Eghj58f+Pk5DIhvv73JqlUHuHmziF69UvjJT3q4KNl7FDc5ecVB0QTWeVpWlsLw4feuYQReYhwBXhfyvnxZ5epVO/0Dslg8aoBTL6MYYPm3hymP6I7ZXH81QEsxZYqV776T2bvXgKpChw52Jk1SePllPxYuFIiLEzh5UmXvXhuzZ3tnmFrHGR8fV8NbFMHHx7MlJjYb/Od/vktGxis1Y5mZr/Ff/2Vg3Lg/eMVGEBurEhFRzqZNBiwWgeRkOwyMYu1BkZ42FasV0tNlEnuG0Dna61PMdWrhrulyWJiAoni+6fKnn17md7/7MUVFOwBYv97Avn3/ZsWK+70iquQikAnIAcGsORTNE/0dvn5Wloqq2rxizp7Ee4wjL6G0VOUXv/iGnTvf5ubNVCZGTSQLV52Xxx57nwOFRcTEpLBkycwm6QJ8NwwaJFOtbVRaCvv2+fH734scPgxffgmBgRAYqLYaRVRZhr//PZ09e/YAMHTocH7xix6tZv53i82mAK4GktXqAW+uVvR01/ZS5NTXEIHqIL0IiBlvsPOdqUyfruVceDqyajLBqFFaRZPFAmfO+PHwb5NqHk8GPvjAznSpwkMzvENaU2J8M1Jf02VPFPYoCrz+eho7d+7EZrOSne1LUdH+Ws+ws2vXT5k9ux1du5YyeHB/liyJ8tw9zF3+oCwzzXKadX8xYTYLmqJIXDfGjPHEBBtHdrad1avPYzBILFwY76Lp1Fi8Z6txE+7zRMj7d7/bxbp18wFtE8rKfs/t89LS/84N4MYNSEv7kuDgFYweXTc46Rn27zfx4IMSggADB1aPCqxebUCWra3ifvrUU3tZsWIeUA7Azp0+nD+/mrfeGu3ZibUQvr52srMNTm12jhxRCA9v+RJmdyrZ4HAY4oAYrLz0XxN46b+8T0V7/34Ds2a5XvRDhkikpYkkJHhnw15Zhk2bblJQUMkDKcU8OaEVKZU3E4qiYLG4GkPl5S3/Hf6//7efjz6aC1QJ9iEBi4EPaz2rlEOH3ufQofWsXevHgQPLeeedcR4/bahGyrpAhC2PiDlVqzozk4JACzKt47p6770L/PnPfyQ//wtA4L33FvPHPz7H/feH3/V7e4Vx5C7c15w9ceqdhwx7935DtWFUTd3Aeww4HbMVF+9h1ap9jB49pZlneHtIEm5DokFBAuXl0M6z1bzuqeUZX7woc27df9KdcjKojlBUsnHjy5w6NZTkZC/Sf2gmhg+3s3t3Jfv2GfHxESgvVwkKsjJggGdyxm6lku3NKtpWq4CPm4LTwECBixdbfj63w+nT5Tz11FKOHftfVLWINaFzuIb3fsYtxYgRNj78UGu6XF1JtWNHyzddzs1V+eqrN3EYRqBF7rejFfNU609JOApjyvn661+xdesuJkzwDkcacE1p8fIKzhMnyvn440Pk5haxe/dXlJV9UfWIyrVrH/OXv4QxY8bzd11p5xXGkbeUCysK2GzlTmMZVf99YN7b/OY3A3jssXc5l/ZmzXg1xcU3WmSOt4MoqpSUuBpBeXkKvXt7Zk63om50AlwjFGVl33LwYCHJyWEtPDvPcN99dsDu0mrHG0js8RTn0z5BVWUSuj9EWvrrnp5SvfTrZ2frVoWJE53ztfbskbnvPu+LGtnt8Nvffs7Ro7+pGbuWu9qDM/IefHxgypQKPv3UhNEoYrEoJCTYiIlp2ePmU6fKyc/f4+aRHKAPDuNoDrC55lGbLZPvvstiwoQ+zT/JNsjmzTf45S9/RV7eWmA68LXLc9LT32X79l8yefLdRQE8Yxx56fm50Qh9+45m48Z3a8YUIMsYT+95U5HjAvHtM570tDfrvFIkKcl7Lvbhw20sX27giScc1WonT6oEBHih6mnVtSDmZN8y+uDvP5whQ7xPR6SxqCocPChRWiqRnw9+fiLt24soClRW2pkwwYrB4B2GUV2V7HffXcSOK7/FbleZ0O0qi0e+7va53kDnzipnz1rYscPM6NEisgxffy3ToYP3dSX/8MNM/vnPr8nI+NTlsValVN6EWK1w4IABmw0GD7YTEIDHBQr79vUnJGQC16879+I0GJLp3n0qstyLnBw7lZWrqC0UCQFERoa06FxvSe2UFi+u4Ny7t4Snn15RZRgBVAD+gLOwqNEYTFDQ3WeTt5xCdi2DSMzJpkOgn6N034uUatPSKvnJT97nxImXgFICAkayaNF/8dJL/QHIyLCwZMnrpKb+D1oY1YchQ/7G8uULaNfOC3axKsrLYe9ezbuy2VRCQ+2kpMhcvaodrdntAjk5WnVb7942j+lZSBnpLBg2ABVcEt+rrwZtzId58+4s58ib1YDtdli71oe5cw20ayfw5Zfw8MOOx8vLYdUqK9OnW1pgprfgVs6Mlzo7dbl2TeDkSSOCoDJokA1Jgv37jQiKQoQlnZ495RpD1BPz37OnhMcem0dJyUUgDEf0QUt6jwN++IPPWLIkutFz9OY1UU1ZGezaZcLPT+LKFa1Cc8ECEbMZNmyQMZstDBzoeUmSP/zhKP/61wMoSl7ViJnZs1fxzjtjAXjmmYO8994sqvMmAfr1e5X165/wnkqw21i7164JFBYKJCQoHlvSzz57iKVLf4bFEgusqxqVgPsB56jqsGFv8sUXi2/LqWxoPbSYcSRlpDu0jDIznc8509IoKCj1CuMItHLlzz7L5dq1YiZNiiIpyTnHpbhY5d13z3HlymUSEuJ59NEotzkN3sSNGwL79plJSjLQrh0cOgQpKQJ9+8L27QqVlRYGDfJAsm9GOkOHOQQ36xpHgwf9nkt+wQwbNpyf//zOqtW8eSPYvt3ItGk+BATAV1/BhAmuSaZffSWTnFzuNQKLbYmMDIGcHF/mzpUwZqY57kngMWftqae2sHTp/VU/PQSsdHo8JORhtmz5J2FhjZdz8OY1AVo0dc0aX5580oAgwKefwvz5zs9Zt04mKanc42KoqgoffHCRbdu+xW63MGTIMP7jP7rXGD6yDH/+8ym2b99KeXkpffoM4dlnx9C1q5eFK2tTy1gqKYGDB83Exxto1zeOI8cF2re3tHje4549pSxYMBGL5TQwF1hT69HuQD9E8QJmsz9Dhizm5ZfnEBd3e3mp3tM+xJ2WkRdiNMLDD4cC7kXtAgMFnnoqEfC+hrT1sW+fmSVLjDXWdK9esG4dxMXBuHEin35qwmaze9SjcZf4/vrrc73GaG5Kbt6U2L0bunTRokjuSpFDQ0WKigQnkTmdpuH8eTOLF9dyg70gKbW8vHZy735gAbADKCIq6gl+/esf35Vh1Bo4dkxi2jQJSYKjR2HIENfnTJ0qsXq10eNtqAQBliyJZMmSh9w+LknwzDPJPPNMcgvPrPHUFmQOCoKoKDTHQRCIn5/Ali0+XL1a3qInDdu2pVYZRqBF4cKAa1U/pxMVNYG33nqbrl2lJp1XsxtHWVkCqalm2l31Y86c5v5tOu64dk0gMdHgEmacOhU2boQZM2DoUIkzZ0T69PFMgmp1gns0sGrl5yjdooC20WW8Ltu2GQnpJDA+Mo3cXDh+ENIkSJga5xTOzshQGDNGN4yaGpsNAgO968gPYNCg/nz+uT9aBVR21b/BDBr0NqtXj277EURZxnoqC98AM5u/hJs34XR4HOELJSenTVFw6SWp04Q0EMQYP15k+XIjXbq0XM5XQEA7qFFW2wRMBIYhSf5MnDiRX/1qIv36idStMr9bmtU4Ki2FjAxfHnlEgjTRkfjlJZpG9wrl5bgNQUuSFvoFuH5dpUMHz2zEtSNGAqB0i2qT0SKAU6dEevc2k2RIh8xMImNieOwxIDOT0uMQ0F+7KR06pODvb9U3gWZAkjwkpnkLHnsskn37lrFhwy+x27MAP1JSHuQf/xje9g0jtKjFjF550DmGiROBzEzKQuDjjxNYvJia0uz162WGDNGV/j2BILR8gchjjyWwbNkPuHjxn1UjmwGRuXPX8OabY5vt9zarcbR/v4kHHqjy0OLiasZ3ZHajd/tsjpwrY8Bg1SOaRvcSMTEqW7bIJCc7f92HDkGfPtrZ+bFjdmbObPkNQ46OZcW+wy5jbZWrVw2MHStAGi4e2ruv2QhNs6EoKlFRNoYM8b4y87aAKMLNmzJWq+TQQvGCih2DAd55ZxybN+/mu+8uEB4ezMKFER5Rf/YYddaEP7BgAaxerTJtmsDmzTIdO1rcNkvVaX7OnFHp2rVlc1M7dRL461+f4ZVXunPq1AYCAsIZNWoe//d/zWcYwW0aR1evXuXFF1/k6NGj+Pj4MH78eJ599lmMt0hQEYRaYoSSVHPR+xSAHNuT/G0W5DjPlmTeCwgChIdb+fxzgZkzJQwG2LNH5ehRlcRE2L5dZsSISo/M7fgpG6dPd2T8+CBCQ1tPmORu1kR9LFxogQTPfA/3GhMnWli6FMLCDIR0iiX3jJ12FytITpY95qwdOVLOypVHKC0tZuDAvjzySFevkxrwBL6+cPOmnd27bYwYId+1uJ/OLXB3shMTw6FDKmfPWpkypeWdttGjgxg16vvk5f0APz+VgIDm/523tfR++tOf0r17dzZv3kxJSQk//elPee211/jNb37T4Ov8/GSuX1ddep1cuQLh4SodOni+HLM5yM2V+dvfDpKWdooOHTozf/59TJ7c2aNz6tVLprS0nM8+M6IoAomJNiZPVjEaoV+/lp9PWRn8+Mdfs3Pnn6moOE7nzrNYsOCXPP+89+hFNURj10RIiJ3UVBOJIi7RCikwmDa5Iryw1N9shpkzLZSWWigpERi0KBJBwGOf/2efXebZZ39OUZEmGLh6dQC7dn3Ee+9N8Aqdq9uhsQ7D7dCuncKwYW1vdZSVwYcfXuD69Tz69Ytj5szOHj1Kr9utwmKBo2fDKc6LIC6hjClTWvh0oc69owtACci+zX//uGUp/8mTJ5k/fz579+4lKCgIgI0bN/LCCy+wf/9+xKpv0l2JpqLA55/78MgjxhpL7/BhLdHu3Dkb999f2WoW/u1SWgpz5rzJiRPP1Iy1azeYl1/+gPnzIzw4M+/i6af38f77U3G0LwWDoRv/+Md25s5tmiON5ipbvps1AbBli4mQTiJju2Vy7ZrKnj0yAwZYaN8vxuu0gZqCai2re70vWH0oCkyZ8jrHjv3eaVwUw3jnnX3MnNmpyX5Xc5byz5s3j+7du/O73/2uxmEYMWKEk8PQUCm/k9wL1EQsdl6Nx8+vnOho78sTuxvS0ir5/vff5OzZ/wGsCEIIkye/xbvvTvQeHSQPU/feAU17/7irUv7Tp0/TpUuXmk0AoFevXhQXF5OTk0N0dHS9rxVFmDOnkq+/lsnLkygtFfH1VejUSWHWLGubM4wA3n47lRMnnnMaKyk5yEcfrWf+/B81+++/dEngzBkjBoOAICiMGGHzyjD0/v27qG0YAdjtOWzZ8h1z507zzKRuk7tZEwATJljJzxdYeTSS4GCFMT9Q2uRaqI2neq8dOGCgpETrT1dZqdKhg9UrBARrc+OGQFraNpdxRbnGoUPpTWocNRcnT57kzJkzvPPOO7Rv35727dvzwx/+kBdeeIGnnnqqxmFoCDk6lvxM+HadFUmS8PGJ4tr1boSGV7Y5wwjglVe2cfbsH2t+VtXrbNy4mPffP8APfhDtsXl5G566d9zSOCoqKqJ9nVKnwECtaV5hYeEtNwKDAcaMsQFtr7pAlmH9+gJycgoZPz6SpCQTOTlZuAvOX7x4vtl7ZJ0+LSHLPixcKCIIWpXaRx8ZmDWrwsUTuXBB4Px5zYgyGmWGD7e3aNBCENzf7JpIk7RZuds1AdCpk8qECS0vutlcVFbCihXX+Prri9jt6XTs6MusWSOYPdtzrRIOH5aIjfUhKcmx6E6cEDl+vJI+fWqtUTfHftByR3/t26t07tyDnJy6BpKRkJDWUcV7tw4DAJKEEt+dofFakYgsQ29D29s3qjl16js3oxUcPnwMTdSkmbjD6/3cOZGcHInISJnExHunSKRR6X7VG5jQ1t1dd1RdWLnXZZ7//TpOnvw7Y67gHQAAIABJREFUKoW8/uoI5s1/ifDwbjg0GRyEh8c2e3TgyhUjCxc6PDQ/P1i82MCGDUbGjnXcZI4ckQgI8GHRIu25BQUqq1bZW/SYc/DgUZw58zK1tSkkKZyxYwe2zASamHt5TezaVcgzz7zJ+fOvoTlB44FKNm/+X65e/Zj/mOiZvmCFhUamTHH+PlJSNJ2W2g5MbeG7GjIzKciiRY7+tGaqD/L22x9Su9VEjx5P89hjcfW/0ItoCoehNoJAm0hG37WrmFUr9uB75QzR0Qk8+mhPOnQQkKNj8fcPdPua+sabitu93m02WLfOh9GjDYwcKZCaqvL553Zmzqy8J479bnn5BQUFUVhY6DRWXFxc89i9Rr3d40v38uGHj/LKK7tJSnqBM2f+WPO4n19fFi6c3qzzUhTw93e1+gMCQFUdBpOqQkGBicmTHWNBQQKzZhk4fFhqsSOHF14YxqVLS9m9+7+xWM7QseN0HnrotzzwgHtVcm9CXxMOFAVeemkp58//b63Rb4BJVFZeYfny1Tzx+NMekWswGt0bqiaTwPXrAocPG/HxETFlmZgzx43wXQuqZL/44kB8fb9hy5ZNlJbepHfvgfzmN5NapCqnubiXHQaAtWuv8Nvf/pDgop3EAN/ug+UrHDkzEydO4vjxP1O7cWpg4CgeemhQ80/OndBj1fWuqlqi+O7dJh5/3FgjJZGYKBATY2T1aoXJk1uuytxd54SW4JbGUXJyMrm5uVy/fp2QEC1EfuLECTp16kRkZGSzT9Abqe8MVJYvcejQYT788Ge8/vpAzp07QceOwcybN5o5c8KadU6iCBaL65GUqoKt0o6Uoc34xg2BRNEP0gRNe6oqjBoeLrB7t0RL1esEBAgsXz6TgwfHc/p0CePHd6Jbt9bhKuprwsGhQxZOnHjbzSNbgGlkZe3mRuFzdPFA8nVlpfsjgCtXZARBE6cVBFDPeT4pT5LguedSeO65FE9PpVHoDoMr7723gaKinQTjfs/4zW+SKCtbx4YNn5KXd4r4+HE8+eQ8hgzxnEV8/LhEXp6Jzp1FRBHWr4c5cxynbWYzGI0tV07nTgevery5ueVulJSURN++fXnllVd4/vnnKSoq4q233mLRokX3rEfQEAaDiagoI3/5yyhgVIv+brvdTkGBSFCQ43vZuFFmUNA5p345gKOMvMp7sNlAUVo+32fwYD8GD25d8r/6mnBgNosYDP7YXFJDDIBM164DCQryTB5ZSoqVlSslHnxQQhS1HJaVK2V8fQXmzXNEWe+xr6xZ0B0GZ1QVsrPTGnyOKMKLL/bn2Wf7U1wsEBystlwZvxstoytlIRiCfFiwwDGJ0lJYuxbuvx/PIEkeq2q9LVf9tdde48UXX2TChAn4+fkxdepUfvSj5q+8am34/X/2zjwsquv+/687d2YAQVEQBXdEwAV3XHHFJe5GjRo10Zik2ZM2sWmbpE1/SdPmm6ZNmqRJms0ad5PGqNG4E/d931cQVARUBBUYZubO/f1xgWGYERdgFjyv5+F54Mx2gPu5530+57PUiGf06M63f2IVMXCgmVWrwGDQo9dDfr6NJk0shBlUCCm/6e+SJQrdu1ff4MfKRtiERrt2Bjp1eoadO18q88hDQAZjxkz0WIXniAgVP7985s83YjDosFptdOtm5sCBAOcni5ZGFUJsGByRJAgPb0J6uvZzeTF3/v7g7+++DUTZWkYABIdxKCuWyWMd1VnZY93cXBVV9a5sz6rijsRR/fr1+eyzz6p6Lj6DqzPQevVGEvPsn+jVq+rqiNwOSYJBg8xAmfPgsy6fzsqVCqajCvn5Nlq0KPTp+AZ3I2xCQ5Lgvfcm8MQTZs6e/QwoAHoCx6lduw2jR3u2llFICE7xEU591YpaGy1fbqFnz0JtzANVsk0mmDs3jQsXLtGmTSRjx9bzqbJXYsPgyMSJwzl2bDZnTSdKxiRG0KjXcPLCm+PvqYndwhsjp7qWA1arSm6uxMGDNs6csTJ8+P2xib5tEcg7pbziXtWKomw1XVoqr0wcy4WiYRmYV6owVV6edk7r7zELsONUXA207ITgMKzNoz1zrOCG9OmqLHh3J9w3NgGMGzebzZu/BPyBvRTHrj3++EreeiuB+fMvcOnSFbp2jWTAgGCPHmUdPixTu7Y/nTrZd8kHDti4etVEu3ae2RVfvGjlscdmcvDgHwETEEL//p/w7bcjK/UeImzCvbz66nG+/fYjtKBrA7AHSOfpp9fyl7947pTBFZs2GRg0yN+pSfk331ioW9dKVJRC/fruPyLfseM6X365npSUk9Sv34jJkwcyalTlxPBWqAikoAylVPcW7EF2xTr82LEC3nlnBQcOrMNorEGPHsP5v/9LJDjYc6vBrdyoSrOqLy9wK26V9ScqJ/smWVlpwCGn8eTkywwf/hmHDr0FFGAwNGfYsH/xn//085hXpG1bhQMHTCxYoGWrmUw2QkMtdOxY9cKooACOH7fQuLGBsFKndu+/v5mDB0u3nsnml1+m8emn25gxo2WVz0tQNdy8eQaY7zS+c1sS8tlaTuOebKvTs6eFefNkJkzQExoqYbPBTz8ptGxpJibGM/WNDh/O55lnXiU9fQEAR4/C7t3dgDmVJpBuhRBHlYjFAi+++A2HD79eMvbDD1+Tl7eA2bOrNpW/XDwU1JadbeODD/Zy4sRBatUKYcyYXowcaS8K6KnKp4LKp1mzlpw86Tx+4UIeZ87YW+lYLMksXTqFhIRdPPZYQzfO0JEOHRTc3Unt889PM2vWLFJSlhIS0pHBg5/iH//ojdEIhw/vcfEKKwcO7AeEOPJVbDbXoqJBwVWv2xzq9TBmjInNm/UUFspYrSqdO1sIDa1ib1E5vRdnzdpVIoyKuXFjJ/Pnr2fUqClVOi0hjipA2SC71auvcvjwX52et3nzl6SlDaVJE9/pOl9RCgpgypTZ7N1rD9RNSupAdva3TJvmrkoVAnfxq18N4MCBcWRm/lAy1rr1/yMvz5X8vcGuXfs8Ko7cTVJSLv/3f09QUHAAgOzsNBYuXEmdOmt4661OBAY6exEAgoJcjwt8gz59urJkSV1U9UqpUYm2beNZfMb7NoeyDAkJVsB91fvLniKUFolZWRdcviYz0/V4ZXL/rNaVTHH9hR1FXwu27+W0rQ6Q5/TcvLzTZGRUn1YRd8KsWefYu/dVh7H8/APMm7eM4ii3SLTjyOIvIZl8lz596jB37qdMmbKEwYM/5emnk1i06NfUrOk6yt/f//6K/l+2bFeJMLJjZtOm9QA88MAAdDrH7LjAwI6MGdPDTTMUVAWTJ0cwbdpCatXqD4DRGENi4hyef769h2fmXRSfIpwu+r6Ypk1dZ1hHRla9N1V4ju4VF0dVQ4ab+fjT/ly//ovDeEzMZDp08HyhOXdy7tw5oNBp/Pz5w+TnQ6AHi3sJqob27Wvw4YeJDmP9+w/iyJH30bLYNGrV6sn48d7XJiYvD95+ewe7dm1CVbUWN3/+c3cCAyv+3mazqdzx555rQW7ujyxb9j2XLh0gMrI306Y9xODB92cBxeqCJMHf/96VJ5/8no0bL9O6dTA9ewahT/Y2n1H52GywevV1srIKGDWqHnXquCdY9bnnOrF58584fvwvJWMNG07mqaf6V/lnC3FUicTEGJk27S98+eUzFBYeAaBOnWE8//w0zGY4etRKZKSe2rU9PNFiyjnrrSha4TcjZcsKNGjQiho1AMlzxb0E7uO119pz/fpS1qyZT1bWfmJiRvDk9IfoVS8DzmaUPM+TgajFPPvsz6xaNZnivojHjulIT5/H3LkVjxfs0aMz//tfbSDHYbxDh56Atoi+/no7ZsxoR06ORN26qqf/HIJKJCbGSEyM4zHyrdpiqKomRrzl/3/qlInf/GYee/e+i6pe4Z//nMILL7zMU0+1qPLPbtBAz8KFL/PZZ/04d+4k9es35rHHuhIX56JWWSUjUvmrgK1bb7By5REMBiMPP9yWpUtPsWjRLM6fX029egmMGvUk77wT775qqLdAPnv6lme9FSUvD0aP/g+HDv2uZMzPL4Y//WmhW4wKRNqyN3Hzpta6plEjFb/Uqrvu7pX9+02MHv0AJtN+h3F///b8+OMaOneu2M3YZoPf/GYrS5b8AZPpIFCH+Pi3mTnzEcLD3bcKCpvwElxsTC0WeGPmFTZsXkt+fi5t23bnd78bRrt2leC6rACTJn3P+vVPOIwFBw9k+fKFxMZW/ESkKteh2yFS+d1MQkJNEhK0WIFFiy7x0UcPY7GkAZCVlcrXXy8nIuIXXnzR856TqsoYCwyEOXOe5MMP4zh27ADBwSGMHt2PCRMaVMGnCbydoCAICrLvw7wtU/HIkRtFosURk+kQR45cr7A40ung448TePTRVWzefJ6mTUN58MEwr/EOCNyMi7CMGS9vZd68UYBWZPHSpa84d+51Vq/+Q6Uc7d4L2dmwZ8+PTuO5uetYvPg4r71W8dgpV/3TvCG8QoijKmb16q0lwsjOdZKSNniFOKpKIiL0/P3vvYBenp6KQFAuAwaEEBo6gqtXlzmMh4SMYODAupX2OV26BNKli0jNFzhy8yasX7+AYmFUzKlT7zN79kSefdYz6SpaHTzXCr7S2sJ4sH9aeYhstSqmsNB1IKbJVOBy3N2UzhgT2WKC+5UGDWQmTpyBXt+oZEyWGzJx4gwaNhTuHUHVkpsrce2aq0a1Fq5eveb2+RRTpw506zbGxfgIJk5s7YEZuQ/hOapi4uO7snZt2cBkiY4du3lqSiV4qztTUP0prxGnp/jznzsSF/cL69fvBiAxsQsPPVTfw7MS3A9ERKi0bDmcgwd3OIwbjdEkJLgnRvNWvPPOSG7c+Izdu/+J1ZpO48bT+PWvnyMysnrLBxGQXcWYzfCrX61l3brfYbGcRacLJyHh/5g1aww1a/put2qLBZYtyyYnx8TYsRFuS+28G0TwqZdShVmSnkBV4ZtvzrF27SbMZhMdO3blt7/toGVlehnCJryXxYvTeeONV7l69ScAJKke48d/yyefJHi0FyFo1/j27flkZBTywAN1PBYDVdmUZw9CHLkBVYUNG66zf/9FYmPrM2xYiMcv9oqwf/9NZsyYyZEj7wF5hIc/wm9+8wqPP+5dXiexEFR/0tMldDoID3d9G0tNBVnWsuSqirfeOsDnn4/DZrtcNCLTv/9sFi4c6XV2LmzCuzlxwsy8eQfIz79BQkJ7xoyp63XX0B3jhubiFUWII1/nDi4ykwn+85/THDp0kKCgYCZM6E6vXpV7IywshGvXJJ566r/s2PGSw2OhoWNZu/YbGjXyjosexEJQnTlyJJ8331zM7t3zkWU9XbtO4a9/HU10tNbC/tChPN588wf27JmNTqena9ep/PWvY4mNrYQW96XsMS8PJk9+h8ysHzlLcYUkkOVG/Pe/WxkypE7FP68SETZR/Tl3zsrs2UcoKLhBnz5xDBlSxy6w3Oi1LZuiD95RtqM0IpXfx7ldB3urFaZO/YkNGx6jONvh558TeOedr3j44UYu3vHuUFV4552D/PTTIjIyrmAyOQcOXr26mO++e41XXomt8OcJBOWhKPD8899x/PhXwHHAyoYNG3jllUKWLZuKzQYzZnzLwYP2hrebNm1jxgwzP/30WIV34reyR7CXJ1CUCxw9muF14khQvfnpp0xee+0NsrK+A2DWrCZMmvQNH3ygxbiW18esMvj550y+/vpnzp07Rdfa/l5XsuNuEOLI21EUdGmplHbvnS3zlO+/z2DDhucpnQZ6/fpWZs5czIQJL1W42OS//nWCTz4ZAeQCjYB6Lp4lIXuJq1RQfbl5U2X8+BUcPz4fuAEMB7KBzezd+wHbtz9EVpaZgwffc3rt3r0fsGXLBHr3Lj8Y6No1SE21EhOjv2Xc0O1u+rLcmLi4iDv7pQSCSuDaNZU33lhJVlYBMApYj6Kk8d13LzJu3HoSEmqiqlVXY2z79uu88srzZGevASCg6nvDVilCHHk58rlkJk4cy7min11l9hw9dJLoMm0JzgIpKdvJzX2JOuVsXvfuzWPmzJ1kZ2fRvn0czz3Xhlq1HLfWa9asQxNGABfQ6hbtc3hOePhUpkyJuuPfSyC4F157bQN7906Bku3COaAb0Byr9TKXL5uZOfM4mnByRFGucu2axWm85HGzwocv/MjGjf8j+9oBGjboS8LUF/n1K3F3OUuZfv3+zuDB3tInSFDdOXfOzOTJH5OR8Q7a4a4emACswmw+wS+/nKZVqw689eJqp9feNrCm6CiubFuTssdx8+dvLxFGxdyqRYovIMSRD3A7pd/aT+EcOLhKAdSwltQsJ8Tgu+8yeeWVZzGb1wGwfr2OVas+ZMWK6Q7ZCHl518u8ciswCdgFZNOixRP89rdPUreur0YOCnwBiwW2bl0FlL2b7wQeJDKyKd261ea1174DBgCOC0FY2GMMHhx8y/f/9s01HFjyONlFPxvTF7Dkn9uJbLGeUaPCbju/nj3+H+H6WnTq1JVXXmnnu4G0Ap/jX//aypkzb5casQLzgbHAMurUCeattzaxZ+/rTq9dt+4aieVUC9ClJDOp5+1jh65du+zwuuITjtiY5/j2W3v7EV8pFyPEkQ/SDCh93x07tjGffuYooHRSHYaPGoe+nP/wH/84p0QYadg4duy3fPFFL155xX7Rx8XFc+JE6VeeBxbz2GM7SEysR//+NfHzq8AvJBDcAYoChYV5Lh/z92/Byy8/zvXrEleubEI7/u0LbESzlkTi40fiX0489q5dO503IpZzzJx5kGvX2jJiRH1CQ7VhVzviDz4Y6TWBpoL7i5MnD93iEQtRUS8zbVoUgwfPQ8K5xtiOHUdJfKbrLd97/vw0lxv0EycKiS51WNCiRRvWlHIc2Ype0zaun0/ahRBHPsh3ixZja9K0RIG7qjnx7LNfM+n3tz4OOHnSSk7OMYcxHRCFhdMrViOP1saUZs159dV+nDjxN44ceQsoRJLCGDjw37zzThTGivcdFAjuCH9/6NChH2vXznYY1+li+PDDXzNuXDBms0qLFiM5c+YDoD5a7IWKXp/J9Oltyn1/s9lVNftRbNv2Etu2pfP++xN56qlXeOn5Fk7FU8F3dsSC6kedOq5b3NSr151PPnmMoCCQZb2TwEkBwmqVHxu3ceMWl+P/7/+thfn7aNIkluee68RLL7Vn+/YP2b//VTTPFURFvcJLL/W9y9/GOxDiyAcoq/RtTZo6KfGyz3n00WYo5bj1r15VKPvvjyp67aHDr9O9x+slrtPIqGiWL3+BOXNGcOlSBu3aRfHgg2EVDvQWCO6WP/5xBJcu/ZUjR/4G5BEU1ItHHvkr48Zpx2VGI0yf/gh/+9sG8vL2AcuAIEaOXEDfvuWnsbdu3YGTTomY9l5rWVkL+PDDc/TqtYROnXxvJyyovowd258tW9pgMh0tGatZswcffTSF+Hjtuu/Roz8nT/6d0yUFJ8DPryV/HtG93PfOzc12OX4u9c+cTtW+37z59yxY8Dt++OEJvvyyF8nJZ6hfvxFPPdWW+vV9c6EQdY68nTupS3EPtSssFujQ4T9cvvwxWpC11l8N7O7TaGCHF9WkuFtETZfqidkM3313icuXrzN4cDPatHE+001KymXp0l2YTPl06dKJ6dMb37aUy429p3h2aLxD7J6r44RnnlnL2297vv3PvSBsovoya9Y5Fi5cTnr6OZo2jWbq1BGMH9+w5PG8PHj++VVs3PgReXkHiYgYx6OPPsNvf1u+R/WPEz7n3IbfO8UclbWNJ55Yw7vvli+0vA1RBFLgksWL0/ntb//LzZunACut5BpYlUVCHFUSwiZ8DEVBSk5m/fprpKRc44cf5rHv+pJS+2yNZ59dz1tvdfHIFCuKsInqj6pSbjLAiRNmTp8uoFevWnfU9unQ/uv87YnXOX+h+Dh7ALDeoegpwMCB/2b+/KkVmLn7EUUgBS4ZO7YB3bv/gXnzTmI2m3m4k5Gp0xZ5eloCgWeQZdToaBKL9gJbL1uxLV/i8JSgoF6MGdPWA5MTCO6M22VJtmxppGXLOw8WbdexFh+v+oSvv36KK1cy2bPnAidOrHd6Xnh447udqlcjxNF9ToMGMq++2hrQyr17Y7d0gcAT/PWvD5CT8xU7d/4diyWVBg0m8/TTL9GhQyW0IBEIfIh69SRef70dACtXXuall74nN9ceqN2gwSSmT/fNo+ZbIY7VBHaqWbd0cYQgqCiqCjt25JOaWsCwYaHUquXpGVUMYROCymDlysvMnZtERkYazZrF8PTT/ena1feMQ8QcCe5LxEIgEDgibEIgsFOePfhmjp1AIBAIBAJBFSHEkUAgEAgEAkEphDgSCAQCgUAgKIXIVqvG3LwJS5deJjTUj8GDa4mK1oL7l2qWbCAQVBaqChs33uTiRRMjR9b1+aSDykIsl9WU2bNT6dNnBi+/3JJp0zoxbNhXHDtW4OlpCQQeQT6XzKQenele9DWpR2cnsSQQ3G+cO2dh1KjZTJrUjZdfjqVPn1/z5ZdnPT0tr0CIo2pIWpqVd9/9MxcufAVYUNUs9u2bwZtv/ujpqQkEHqO43cHpou8FgvudP/5xOTt3voCinAcspKf/l/ff/x0nTpg9PTWPI8RRNeS7705z9epip/Fdu+Zz6dLty8ULBAKBoHqTmwu7dy91Mb6W//3vmAdm5F0IcVQNkW5RP14ngo4EAoFAILgtYrWshjz8cDR16453Gu/SZRIREZVS81Mg8Dki0ZopRyNa4wgEwcHQpcsoF+ODeOih1h6YkXchstWqIQ0byrzxxp/58MMw0tJmotOF0rnzq7zzzhhPT00g8AhKs+Ys2L7XaUwguJ95550R5OZ+yp4976Eol2jQ4FGee+7Fu2pMW10R7UOqMXl5sHz5VcLC/OjfP+i23ZqrG6JVgkDgiLAJQVnu51R+0VtNcF8iFgKBwBFhEwKBHdFbTSAQCAQCgeAOEeJIIBAIBAKBoBSVdqwmEAgEAoFAUB0QniOBQCAQCASCUghxJBAIBAKBQFAKIY4EAoFAIBAISiHEkUAgEAgEAkEphDi6Ay5dusQzzzxDt27d6Nu3L2+//TYWi8XT06qWXLx4kRdffJFu3brRvXt3fv3rX5OZmenpaQlKIezBfQh78C2EbbgHd9iFEEd3wAsvvEDt2rVZu3Yt8+fPZ//+/Xz00Ueenla15JlnnsHPz4/169ezYsUKcnJyePPNNz09LUEphD24D2EPvoWwDffgFrtQBeVy6NAhtWXLlurVq1dLxlauXKl26dJFVRTFgzOrfuTm5qp/+MMf1IyMjJKx5cuXqx07dvTgrASlEfbgPoQ9+BbCNtyDu+xCeI5uw9GjR4mIiCAkJKRkrE2bNuTm5pKWlubBmVU/atWqxbvvvkv9+vVLxi5duuTws8CzCHtwH8IefAthG+7BXXYhxNFtyMnJoVaZTnzBwcEAXLt2zRNTum9ITk7m888/57nnnvP0VARFCHvwHMIevBthG56hquxCiKN7QC0qKi7db23u3ciRI0d45JFHmD59OiNHjvT0dATlIOyh6hH24JsI26haqtIuhDi6DSEhIU6qPzc3t+QxQeWzefNmpk2bxgsvvMALL7zg6ekISiHswf0Ie/ANhG24l6q2CyGObkNcXByZmZlkZWWVjB06dIjQ0FAaN27swZlVTw4ePMjLL7/Me++9x+TJkz09HUEZhD24F2EPvoOwDffhDrsQ4ug2tG7dmg4dOvCPf/yDGzducP78eT7//HOmTJkiXKWVjNVq5Y033uDFF19k4MCBnp6OwAXCHtyHsAffQtiGe3CXXUhq8aGo4JZkZmby1ltvsXfvXmrUqMHQoUOZMWMGsix7emrVij179jBlyhSMRqPTY6tWraJhw4YemJWgLMIe3IOwB99D2EbV4y67EOJIIBAIBAKBoBTiWE0gEAgEAoGgFEIcCQQCgUAgEJRCiCOBQCAQCASCUghxJBAIBAKBQFAKIY4EAoFAIBAISiHEkUAgEAgEAkEphDgSCAQCgUAgKIUQRwKBQCAQCASlEOJIIBAIBAKBoBT6ynqjy5dvVNZbCQSVQlhYTY9+vrAJgbchbEIgsFOePQjPkUAgEAgEAkEphDgSCAQCgUAgKEWlHavdj9y8CTt2GACJ2rUVOndWkCRPz0og8AwmE2zbZkBVJQICbHTvbkUntl8CgcAHEeLoHklOlkhNDeChh2QMBjh/XuXHHy08+GChWBAE9x2ZmRJ79vgzcaKegADIylL54QcrI0ea8PPz9OwEAg+gKMjnkh2HmjUHWfbQhAR3g1jG75HTp/2YMEETRgCNG0uMG2dg2zahNwX3H/v3G3nsMU0YAdSrJ/H44wY2bTJ6dmICgYeQzyUTknuZkJAg7Sv3spNYEngvYiW/B2w2CApy1pV160oUFsqA1f2TEgg8SECA827Yzw/0erH/EtzHREZCTIz95+ybnpuL4K4Qd657QJLAegv9c+aMjNns3vkIBJ5GUVSX46mpOm6K9UAgAECXlop89jQoiqenIrgNQhzdA5IE+flWTCbH8e3bYexYmTVrRJCF4P7CZlO4ds1RIJ04AX37yiQl+XtoVgKBh0lJgVOntK+UFGoH1xDHaz6COFa7RwYONPP++zJduuipWxfS0iA8HFq31rxHhYWIQFTBfcOAAWY+/1xHXJyBhg01e6hVCwYMAKtVJiNDIjzctXdJIKiOKM2ak7MpldopKdrxWmQkREXB2bPieM0HEOLoHjEYIDpaoU8fPTk50LkzJWn8tWpJmExCHAnuHyQJmjWzMXgwXL0KHTrYk3IiInSkpgpxJLjPkGVsTZpCSJBj3JHAJ6jW4ujcOYmTJw3odBARYSUuzlbu8/PyYPNmI35+Omw2qFHDSo8eVqfn7NxpKDoyVsjJUWnQwLG4UXq6jTZtKvmXEQjczV2mIkdFWTh+3EjHjo72sGePQkJ5+t7sAAAgAElEQVRC+bYnEHiCggI4c0ZHeLhKWNjtxbuiwPbtekwmHXXrKrRv71zbTlW1dSIgAETSvu9SbcXR9u16QkP9eOQRLazq+HEja9aYGTzY7PKmf/q0jqTU1vzuNWPJvf/CBT1JSYUkJloAOHxYJifHjzFjZPR6WL/exhdfKDz7rEx4uISqwpo1CuHhIiJb4GXcRujk5sLOnUb0egmrVaVOHSu2E6mMaHNZOw4ASEkh+xwoUdEu3zMaWHGwJbVqBRAVpa0YW7bYCAgwl5S8EAi8hY0bDRgMBjp2lElJsbF9u5UhQwoxFlWfUFU4cEBHQYFEfLzCzZuweXMAY8fKhIRInDunsnixldGjTeiLVtL9+2WuXjVSt66O69dV9MlGRrdLcfzglBQIDnPvLyu4a6qlOCosBEUxkpBgjzdv1Uri2jUDFy9aaGLS6k+U3PSBbvVSiImpgSzb3Z+NGkkYjQYKCy3o9ZCZ6cekSfa9wKBBOlQVkpIKkGUZqxXat7dQv744PhB4F8U1V1wJnYwMiSNHAnj4YRlZ1jIxZ8828mBPI4TcOhXZ1XsObwl7k1uya5eMokBsrIVu3YQ9CLyLAwdk4uL8aNlSE/FNmujo0cPIwoUqQ4eaSUmROHHCn8REmUB/GzvnnyI11cZTTxmQrgBXoFlUFFOnGli2TGHAAAtnz0oEBfkzeLB93bl2pSU/fQsJwaVijILDtI2JwKupluLo6FEdPXo4OzR79NCxYIGeJs1xrj8B1HHxXs2b68jMlMjNlejRwzm5b+BAHYsW6UhMvEdvkaiiKnAXZa559ap2wz5wwMijj9qvN70epk+XWPNveOCBu3tPsm/SoYMC3HuqckaGxMGDBoxGHYWFNrp0sRAaKgSWoPK4fFnPoEGaMLLZwGwGf38ICNCjqmZOnPBn6tSi5fHUWYbEpsEQ+2aaFM0bVCMmBp1OBiwkJxuYPNlxjahTV09B41YoUfkVnK/EgQN6goJsdO2qiOXBDdyRODp58iQzZswgPz+fpKSkqp5ThQkJUbl0SSU8zKZlBhRx8wYEGBvc8nXXr0OtMmMpKTY6d1YxmSTyXVzfZnPFdEx5O3qB9+JrNqG60Bbr1/vTro5EQICz6JckSo4K3ElqqsTFiwFMmSIjSdq8Fy3SExdXIDyyXo4v2YROp4VBLFumiaPAQLhxAy5c0GKQunQpc1N3sZkuxlYUTnergqcVtaOkJANhYUYeflhHdrbKTz8ptG9vonFjYQ9VyW3rHP388888+eSTNG3a1B3zqRSaNVPZt8+KeuZsicIHqHklhYTwU7d83dq19gsdICNDJT/fir8/REfb2LXLeTe8fLlCt26Wik242PBiYhyO+gTeiS/axKFDslPNlfHj9ezcacRqdX2Ttdlweo0Tt3v8Ljl50sjo0XJJkKskwcMPy+zbJ4KWvBlfswmDQeH771V694YxY2DwYBg3Djp21DIr73TDm5ysEhxcfP9XKChwfk5+/r0nI5w6pSM21o/ERB06ndaFYfp0PUeOiNphVc1tNW1eXh6LFi0iKSmJ48ePu2NOlUKfPiZ+/MHC2LGOil9fHDNR9kaeksIDsbDm32BqGMX5dIkGDcwMHGjh5k3YtMkPPz+J//5XpbAQevaEI0cUGjQwU6OGG38xgcfxRZvIqhkDkQH2gchIpKgogo5AjRqFnDypJzbWnnZz7BjU7hwFIZr35n//szBggBYrUVAAv/xipIZ/HLWyzpJ/wEzHjmZqVUIshb+/6/2aK++WwHvwNZvo1cvCkiVGQkIcU8369ZOYN0/Pzp0KMTGllkcX68XaM03IqVdIv95m5LPJ9I2AH9/zY/hwPcHBEkqzKJb8BNHRhfc8z9RUPZMnS07jERE67aSj7FGHoNK4rTgaP368O+ZR6dSuDf36FQLOO06lWXOyz4HuwFFq38iGRo0ACGrVhCGBKfywz8LgsU0ICdGev26dP088oZUEAG1H/eGHVh55pKBkzBUmE2zfbkBRJEJDFTp0cE77FPgevmgTNkl2eSygqgrx8Qrbt5vYudOAn5+M1Spx8SJ07iyzJCOatDSFoRMKUGprr1m92p/p0w1FxwWtUFX4+msro0cXVPj6Nptde7FESx7vxtdsQqfjlqn7BoNEw4aFzJ8vMWKEjH/jKLYkQc0r4Odn48QJK82b1yd2WAQBQRbks/bQiMmTi94kJYWffzLTblRTate+9TxUFZKTJfz8oFEj5/m4Og4HbQ0qb+0RVJxqGZDtQGnFX5xCKcv2mJ6QNo6LhizTP/gmSpEwOnFCR9++eocLUaeDESNkTp7U0SrG4rIUfKo+iiPHg3joIZmAAEhN1dI+H3zQ5OyydTVHgaASCQuzcuqUgZgYu3pRFLh5Uzsq7tHDyurVOiZM0JcInBs3QK+XWLxYLbnBp6VJdOigd4ijkCQYPlxm3z6Zzp21lOdt24wYDDrMZhuxsRaaNba6tJOyyQehoRaOHNETF2ef586dNho2rODRtUBQBu24y/FmbDJprXBiYmw0a5bPBx/UoGtXmd5TYqhZU3uOZa+K1ZpHQFAp5eIiJikh+GbJhsJkgh079FitEp06WQgJgRMnZFJTjXTsKJOfr4VodO5sIiLC/r4xMRa2bzc4JAOpKqSnK3TqVKl/DkEZqrU4KvYQlaQf34Pb/8IFHT17Om+HW7SQ2L9fR5zRuSwAKSlsPOjHo7+zV4Js2lRi0iQD69Yp9Oljv9FXxhwFgtvRvr1CUlIhZ84Y6NFDJiVFi8sbONDu8vfz0zl4fooXg9JHXenpOnr1craHBg0kNmzQkZenkJQUwLRpdgGVlGQgZd1huoY520nZ5IOOHRX27Cng8GEDfn46TCYb9epZaN9eNOoUVC7R0YUsXqzjwQdldDptMzB3rpWRI7X7c1aWxKhROicR0rmzxPz5Bpo3vzN35vHjMpcu+TFypIy/P6xbZ+TgwULAyCOP2MVZly56Zs3yZ/hwe+BSs2Yq27YVsnSpkb59ZTIzVbZuVeje/d6P6gR3RrUWR8gy18KiyczU0bixDf97iGFr105hyxYbffs6+jA3b7YRF6fATVzuGmqccr6ZBwdDdrbM2rVapkRsrIWmTWWRmSZwC4mJFtLSLHz3nZ6GDW0EBOjZutVIq1YWGjdWb3mkVVhoDyht3Vph506FIUMcd9x79qhER1vZts3I1KmOnqXERB1L39dDVxcZPy56TMXHV6wUgEBwJzRvrlK7dj4LFmjFT202G5GRCteuaa1uAgIgM1MFHDcDyl1cmqoKaWmOImjwYB0//+xP69bOm4z4eJk9e3RYLBKRkTbCw1V69rSSn29l40aZkBCVkSNFtXl3UG3FkaLAypV+NGigJ7KJyollp7BarSWZZSXeGReBdqWPterVU9m3z8KFC0YaNdIu5gsXVM6ft9CmjaqJIxdcuABLlmhHDkFBkJhIUWqyPcBu2zYD27YV0rOn1fWbCASVhMkEa9b4ERmpp29fid27oXFjib59tetw0yYzLVqY+eUXPf372zcC+/fbCAuzezpr1YKcHAvJyTqaN9eu48xMlUOHLIwcqZKSonOuhq0ohN5Ig5RSG4yoKAB0aamOTxU1vgTuQlEIu5bM0CjYs0ePJBlo1TOGtIsSO3cqDB5s4uxZhT59HD2qK1cqdOlSxmt0i3Xk9Gkd8fHO1/OQIRIzZ0K9elp8bK9eWrhGzZpw4UIA48frOHbMxq5dVoYPL6RGDejZU2wY3Em1FUfr1xsZP95IUBBw6hQxLdKKXPp+dnd+2SMtcHmsNWSImR07FDZv1v5cQUFWhgwp/0Lt109H1FDt+ytXYMECiIlR6dHDbmU9e+pYtsxIfr5VZLwJqpT16/2YOtVYIlxatoQ9e7QM/IQEHatXGwkJsWC1FjB3rpGAAK0AY1iYpaioo50BAyzs3Wtjxw49kiTh729lxAhN4JvNLna1Z8/SK0HnUMureDGpfSNbi/srGhc1vgTuonSNucGD0a7JQplGPWPo2lXH3LkqCQkmZs6EVq301KkDBw4ohIWZHYKsy1tHjOe17Oay2GxgsaiMGiVx+TLMnQtDh8KOHfD44zpkWVsf2rc3snSpyqBBIiPB3dxWHD3wwAOkp6djs9mwWq20bdsWgFWrVtGwYcMqn6ArkpMlUlJkWrdWHILXSmMwyJowKsZFJV+HwOzb0L17Oa5+F7uGqCh7bEXdupowOnZMYupUx6f27SuzYYPe7j0SFbO9Hm+0ifJQVQgKkp08OvHx8OOPEBurVXr/7jsDAwdaiIkx3fY9O3d2bQ9t2lhYu1bPoEH26zUjQyW8jP3lHDgKQO2ePW/ZnqSY5GSJs2eNGI0SBQU2EhLMJfFQAu/A22yiOLuxuE/aLblFcUe9XrOZkBAYNcpERoZETg4kJqrOWWLlrCPNmqmsWKHQvr3jUrtihcqUKdpGOSwMHn0U3n9fpX9/xxpLgYGg1zvf+w8d0pGZqcff30aPHlaPFGyt7tz2T7p69Wp3zOOOMJvh55/9iI83MG6cxO7dNn76ycqIEYVOKcQ6nXty5svuGo4e1dGqVSR1i44NiomPlzh1yvn8OjNTpU4d+25bVMz2frzGJu5QSKvqrav0Ft/oc3NVAgMrHsvQqJFKXl4h8+YZCAjQYbGo1LlcSHi4Y8CfrUlRscDbiP5jx2TAv6Qtg6LA7Nky/foVOG5+BB7FW2zi8mWJXbv8CA3VrperV2107Vp4y7T98ii9poSH33s16vbtTcya5U+PHtqGfdUqlXbtdA41iiQJ6tdX6dLFed0qPQ+bDZYs8WPAAAMDBkjk5sIPP1jp1s1EvXqiYnZl4lN685dfjEybZizZDfTsqaNlSwPr19vo188x1Tc/X0FV7ZV2qyxdvsyuIbiGxOm8GtSVHbcXmZkqp04plP6Tqyps2aI4B9i58nIJBGW4UyGt08HNm85pyxkZUKeooeCKFQqDBlVOTENsrEJsrP295LPlxNTdxi4vXnTsVyXL8OijehYuNPLAA+KoQWBHVWHbNn+eeMJejkJVZWbOlBg58u5rcOXlVU7gc6NGKg0bFnDypI60NPDzM9Cli7NLS1UhPx+HEAurFQoK7La0ebOBSZOMJcIqOBimT9czZ46RoUNFBltl4lPiyM9PdnKThoRI2Gxa47/SdO1qZuZMHWPHytSJiiI9XWXjAgsDBhSir8J0+YgIlaVLrcTH2+M7VBWWLlUYMcLEt9/607ChDpsNLl5U6Nq1kI0bDVit2lispCspPikQ3JY7FNKtWhUyZ46OceNkatSAo0dVVq5U6dlTYs4chdhYF/W3KhMXIqis11WtFUZ27eYk75PJztZuTRcu6FBVx92zXg9Go6iAJ3Dk8GEdAwfKDteKJMHAgTKHDulo396F2Cl7XUZGkp2tsnSpQnx85YkNSYKWLbXPP3DAyokTBlq2tE/UYgGwMHeuRM+eBuLiJM6cUVm/3sqwYfZNgKLITlWxJQkCA0XYRWXjU+LobggNVRk2rIANG/SYTDpCQxsz4FmtQnVVx/w/8EAh8+er1KihL9q1K/TpYyY0VKVp0wKuX1Mwnk+mc0tY84MfI0YYtBiKqCgOLxE9cwSVT5MmKvXr5/PzzwbMZolGjawMG6Zis8GQIVXrjr9lLa9SXte9e2WuXTMSlidz/ToYjRJjxmjxSosWwcMP299PVR3LCwgEAJcv6+jTx9k91LChxPbtOsDxmil7Xabm1OPErkYEBJkYOtRaZZuFDh0UNm4s5NQpA926ySQn2zh61MoDD5jx94eTJy3Mni2Tnq4jqpnK4R/OYzbb6NHDTOBFI5wyatmepSZoE+ZQ6fiUOCooULBY9A6BpTk5KpLkWu7o9dCrl/vT5P39KVL7rt3+dbKTCbFehtBIJk0qGizawbRtqxMVswV3zl1cK35+0LevBypN3ybx4dQpHSEh/gwZYvcGXb8Oy5fDqFES0dEqp09LRBe9xZo1CnFxomK2wJHOna0kJdl44AFHr2JSko3OnV2sA2Wuy0ZR0MhN9bX69rWQl2fh4EEd4eEqo0fbNyixsTaOHTPyyisGjOdOQcr5oqNzIyNGYLf5Io+xyQQFBaIcTGXjneJIUZDPnkF38bzD8ODGjZj1TVt69PKjTRstIPvwYSsjRvjgjfIWWRJERXH0qIp0LE8LAhQVs+97rFZISjKSl6dHp5PIzFSpX99Kv95F14WPV1dPTXWMKwKtnpK16H7fubPE228rxMWByWSjWTOzyz5UgvsHmw22bNFz8aKBwkIJsNG6tYW8PDOHDvnRrp3mQTp8WOXaNTMdO3p2vq4IDISOHZ1dPmfOSHTrpreHkLhYK9auVWguQ0qKjbQ0K0OHivi7ysYrxZF8LpmQm1ehg739BikpkHyIcR1kjuXEsmCBTGyslVGjqtlNUpZJ829Bq1b5KIHa0P79MtnZBvR6CZPJRrt2ZqcSBlar5mUVjW2rHytW+DFlipGAAO1niwXmzZNJ2igzalS0z1d50OtdX7TFGXaXL6t07FhYVE4DLl2SWL3aD39/LRsuNNRCx46OO35F0WxBNOesnixZ4sf48QZCQrRr5+hRmSNH9CiKmZycfObP1y6ehg2tJCb61plTSorMww+XfyNv27aA0+l5NGlio32cKP9SFXilOAJu7VlBczvGxt79Ba+q2pdX3DBdVVSNjKSgADIzFeLjteHdu/U0berP4MF2Y1m0SMbPL5+QENizRyYnx0hQkI78fBVZtnjm6ERQJVy8KNGmjb5EGAEYDJCQABkZMtu36+/56Nhb7EFVFfLzDU6FUM1mbX7LlimMGKGJn+xsOHEiwKEdw9GjenbvNtGli5XLlyV27/YjOFgL5M7NVUhIKCy3M7rAtzh8WEdiol0YAbRpo3lcZNlARISZVq1815MSG6uwZ4+Nrl2LDNPFWmEMDqNllLYGymfLyVoVdfPuGe8VR5VIfj788osfNWvKyDJcv26jbdtCj7nmSwcCXrwokZxsxN+/CddzmmHaWeiQonz9uoE2bRx3EePHy8yfb6RBA4WICH+GDrWvbufP69i6VSUhQZxBVwdSUmQGJapaKetSRAPHr0RhMt29ssnNha1b/QgK0hpuXr9uo0uXe6sFUxn06mXh229lJkzQExoqYbXCnDkqhYU25sxR6NHDXCLgdu82Mnmy4429TRuJw4cNqKrVZSr311/rePDBAgTVg8xMPYmJ2nWyYQMUFGhHVLGxcPOmjqNH9Xd3/3MhIMBzIqJJE5Vly6zExhoILqqXd/26ypo1FhITza6Pz2+RtSrq5t0794U4WrfOj2nTjKUK4cnMni1Rv36Bcx8od1AqEDA8CsL7FD/gnDrq5yeRlaV5C4pr0uh0WirzpUs6EhMdF8fGjSW2btUDQhxVB2JiFI4uO023eqlOHe0j8lWuhDS+6/fcsCHASUB8843EqFF3XwumMjAaYcwYE5s26Sks1Epa9O3rugq20agjL0/roB4RYT9G9vOT2LdPZuhQ51Tu3r1ljh/X0aqVbx2vCFxjNNq4cAHWrYMxY7RaP9euUdTmQ6Vhw7sLqnYSEOBxETFiRCGrV9tQVRmIRJIU+j5hQbkXL6+om3dPeKU4OndOR0iui2OnCxfI04dxN8nuWVkSMTF6pwrBY8boWbNGT58+3isikpMlUlMl6tfXjhjS02HQIAgN1XpY+fnpsFi02IzSC4LB4A3nhoLKoF49lZMXrdDN+Zj5zA4LA569u4Xg5EkdCQnOAmLoUJl9++SitiDuR5Zvn1lqMsHJkxLBwRASAtu2QdOm0KULmEwqBQU66teXUBTHDX+TJlJR5XohjqoDPXpY+eorhTfftF/HderAb34Db71l48UX78ED6iqMw4MiQqeD/v0tlK3f50CRx0uXlgq5NRzGBRXHK8XRSWs0nZr4OZ21Xo3tybHrDel0F+91+bJE06bOYqFmTSgsLF9EqCrMmXOBpKRtWCxmunfvzrPPtnBLHxuzGc6eDeA3v7HPUVVh3jyQZYXAQIWLF/WsXastGkYjDB+uLXQmkzCO6kTfvmbAuaJuYqJz25zbkZWlo2NH5xc1aCCxcaOOcquAeTh+Yf16P157zb7R6dIFkpJg3TobsmwmO9vA4sUqAQESeXkwcKDW13DjRoUOHbx3EyS4OwwGiI5Wna59WYY2bdwrgM1mmD37PMeOnSA0tD5PPhlH/fru2ZyWeLzKJi5duAAt2yGfPe0onMq0tBKUj1eKo1p1JDLqtCS8VSuH8VPbbTRonA/c+c4gOtrGzp0KkZGON/DDh1WaNi3/hvn22wf4z38moCgZAKxd68+RI9/xxRf97vjz75UdO/SMGuU4Z63Kqsq+fYU0b+7PmDF2I8zKgp9/hhs3FFq39t1gxLL8738ZrFq1DZOpgPj4Ljz3XMztm0lWM24lgAyGuy9o2r69lc2bFYYMcby2Nm+20a5d+fYgn0tmUo/OFG9ZIoEF2/e65ejBbIaQENlpY9K/P/zzn1Zq1zby9JMSUvJpQNtILJkJrVup3Mxr5FRV2FdRVZg16zwbN25HVW0kJHTnySebeTyo3t1Ikus1QHVj2JzJBFOmLGbz5mcBLaZt6dIX+OKL1+nY0U2N/1x4vHJy8wHJUTilpNidDdWpbl4Vbti8UhzFxyssW6bw5JP2uAizGY4du/vUfaMRZNnMnj1+xMdrd5CLF1V277Y49zQrRXa2yvfff14ijDRMrFz5J3bvXkWXLoF3+2vdFYWFEoEuPiIiQuLqVQPdujneDevVg4sX4cYNG4mJ1aO8wT//eYwPP5yM2axd/GvWGNm/fwGzZg26/0oWuMpuvIebXK1acP26haNHdSWB/mfOqJw/b6FVq9tfNynA6bv+1IpjNkNAgOumnP7+EomJsiaMirI+JQnGPAiknCMsLA+F6hGA+sYbe/jmm/Go6lUAVq6sxfHj3/Phhz08PDP3YjRaycjQEx5uvyYuXFCpUeMePYT3YF9ffnmWzZufonSx33Pn/s1HH8Uza9bYe5tHJVDS1LmMcFKTU9h0vglENaK1h+ZW2VTlhs0rxZFOB337FjB7th+BgTKqqlUAvddGkz17Wjl92sa8eXpkWSIoyFqSGnwr9u/PJytrrdO4yXSQXbsy6dKlaovtxcVZ2bzZRp8+jiJo2zaFZs1cvyY8HIYN07N3r0x8vG8freXnw8KFs0uEkYaZdeteJSlpIwMGBHtsbu7Gqf0GVKjgY//+Fo4cUTh0SMtGCA21Mniwd18vQUGQlaUAjvZw/rxKXp6NJk0kSMbrYkcqk4sXFRYv/leJMNK4zrJlf+OZZ34gNvb+can27m1l3TozNWroadpUR0qKDYvFQmLi3Yuje7Wvo0cP46oLwokTBwHPiaNbITWPpO+QGObPt9C6rcnT06k0qmrD5pXiCLQAu2HDKq/xX3S0jejoOxdXrVsHEBKSQHb2UodxgyGKNm2q3i0ZEaFy7JiZffv86NRJQlVh/XobQUGFXLsmY7HglGlnsUCjRhJbtshUfQe5qiU52UZq6iqncYslmf37L95X4uh27Tfuhbg4G3FxvtXFu2lTM4sX6xg1SjteS05WWb/ewoMPmvnlFwODmnp6hlXLtm25ZGevdhq/cWMjmzdfJTY2wgOz8gySBIMGmSksNHP5skS3buq9H7ffo33Vru16Hahdu+49TuQeuAePV506MgUFONROEzjjPeJIUdClJHP8uI7cXB1t21oJCvJcrYmICB3Dhz/HnDlrgfyiUZm+ff9C374ucoyrgAEDLJw5Y2X+fE0FtWtnoX59lfx8ha+/lpk2TU+NGlop/eXLoX17rSeVweD7WTlNmugID+9JRobjebIkhdGiRT0Pzcq9nDkjkZJiRK+XsFoVeve24O/hvsSRt/jeHcTG2oiIyOf777VVsF49Kw8+WNzp3MwZi0qLFtX3vLVDh5oEBXXj5s1NDuP+/h3o2PH+qHKZliaxZYuRwECIjLTQrp3NY/XqpkzpxPLlo7l82b6B1unqMWLEELd8fnkeL/lcsnPfxaJSBfn5FRCS9xGSqlZOCNvlyzcq9PrsHWeINWY515oIDvNYrQlFgX/96wSbNm3CYjHTpUtPfv/7Tk6VfD1BYSF89ZU/bdtqR4V9+mjpzTNnWhk61EP1myqZN97Yw1dfPQhcLxqR6NnzaxYvHn9HAahhYe4RsbeiIjZxYC/Uyb5YEidnNsN/NzVlxGiz525sXl5t9+zKs3Svn+lV95DK5plnNrB48UPYj3P0DBu2kFmzBt/R633ZJr77zkiLFkYGD5aQJDh+XGXLFjMPPui5BJTVq7P44ovlnD59iLp1GzJq1Ah+85tWno+JVBSu7UkhM9Of3r2L7DMqigKzzHffFRY1Rvd95LOnKxRzVJ49eI042jIzjTFjDI7xAqdOcfBgHg36tajg7KonNhusX29Er5fR6STy8hS6djVTt271CMi22eCTT06RlLQBkymfjh2789pr3QgOvrM7jy8vBDvnnGdEm/OUrmybVy+SFaebFNU/gaNHdaSn61FV6NjR4rEK116Dl1U6rgosFnj//cNs27YZm81Gt269+P3vO9yxR9FXbeL8eYkjRwKZOtXR9nfvVlGUPKKitGv/xAkd58/rqVlToWtXxW1ZfGVra3kLR47IZGQYCQ3VcfOmyo0bVgYPNrulHI1bqOCGrTx78Io/0c2bEBbm+pdJTTXQALx+1+oJdDrt3L26smXLDVJTL9CkSRP69OnIQw/V9/yOzA1YLBAYqHMKLg4EVFW7269Y4UePHgb69ZOw2WD1aj0XLxbSoYNvx5pViCqIzfIqFAX/tGT+NNEfJg7Shpo1h/vgFrhunZExY5yNv0sXiW+/NRAZaWbpUj969TLQu7dEVpbKkiUK/foVlHQWqEq8dRmKi1OIiysgLw/8/Kg+oght87zsp2vs3p1JnTqhPPFES+rUqbwFwmf+VKJHzP3F55+f4b33njZAcLUAACAASURBVCQ/fx8A339fn5075/KPf3Tz8MyqHr1eq/hcFpsNLBaVU6d0xMcbiI3VbgQ6nVbhesECI6rqmRYggqrHk3WmPE1QkMqVKxBd5lctKABQ2bTJwIQJBmrX1i7+evUkfvUrPbNn+1VqYo+v4qosjC+jKPCrXyWxYsVTqGoWoOP771/k009/R3x85XhHvUIcBQXB4cM2ejVMdXwgJYWmTUsF34oeMfcF+fkwc+bMEmEEYLNl8r///YmpU5fRrp2Ho5KrGMmm4J9xFjZn2oMqFYX1KZF0TrRw4ICBSZOcFVDr1jJpaRJNm97nx2vVGE/VmfI0UVFWfvrJSGamdt3XrQu9esGcOTaGDrWwbZt/iTAqRpIgMNBLXTqCCrFoUQbLl08DcotGbKSkfMSHH8Yxb97ESvkMz4qjoqOyvXv1DHtAXzKGLLNwoYq/fwMSpt59Y02Bb3PggInU1B+cxvPzd7B+/QVOn65NRsYNRoxoTNOmXqHvK4UjR2TS0w3UzT7NA/0MENlbeyAlhfUbVWqNbkDduiqSpLos5XD1quqxzB2B5zCb4YcfMsnJMTFmTGPCw6tPuezDh2UuXTKQk6OjQweJ3r213pLJyfD66zY6dy7A3//W7cTcWTFb4D527z6EXRjZOXJkL6o6sVK8525fWSwW7fw4IEBPcOYpBkdfZtCg4qMytMPbmBjGvQbff18IcqmYmrKpidWpDHoFKA5C9POz0aOH1ecz1Zo2NRIc3J7c3EsO45JUl/nz95KaOgO4zkcfDeXxx//EH/4Q55mJVgZFG4RTp2RCa/jRv4sOFAlkRy9pxzE3UaK0Ra9bNwvLlhkYN86+K7bZICVFoW1bt/8GXsf167B7twGbTSopf1FdOXYsn6ef+JBjx/4O5PHxx2N44YU3eP75mNu+1ltJTZU4ftyPK1d0JCTomDJFW+lUFRYsgBEjoHlziI9X6dpVK+VQp46F5GQ9zZvbV8XCQq14sKD6ERTkunRF7ZrB6JMdfav3GpvsdnG0erUfkyYZteyKUzLgoqIt2q5Yr7df6E41HSpQIbhcfCjjRVVh2TItMLd3b4nr12HxYiudOpmIiPDdBaFhQx2Jic/w44/rAPvNrVatF0lNfRrQbog5OSv57LMM+vf/mW7dfPNQvTiWrnt3eywd4JiOXoagIKhXr5B584w0aSKTl6eSnm4lMbFqYiuys2188cVR0tKSiYhoytNPt3Nbc8275eBBmbw8P8aN0wpFbthg4NSpQnr3rh6LZNk6U59+uoZjx/5SMnb16o/86183GDRoITExvlfMJicHkpMDePRRmR9+gK5d7Y9JEkyYACtWwOjREBOjIz1dIjJSJT5eISmpkGPHDHTqJHPmjI3Tp60MHVo1CSsHDxbw/fcHsFjMDBrUgYEDvbQorQ+tZ3fDo4924Mcfx5KVtbjUaA0e7tKu0uLy3CqObt6E8HD5jtJOCwvBai1VzNBNmShlgx7BewMft23TM3Kkgfr1NRFZqxY89pieb7/1IyLCt8vDf/RRIiEhK9myJQmz2US7dgls2rSWYmFUjMm0nxUrDtOtW3fPTLQyKBtLV7bqrQtatlRo2bKAK1ckGjVS6dy5aqaWnm5lypR/c/Tonylu+Lx69avMmvUq0dHeFfulKHD1qh8TJ9pv+v3761i3zo/LlxWfL3WgNGvOgu17S36+fFmi84SXnZ6Xm7uOH388ye9/73tuxN27jSX/P1cecL3e3og5NdVGmzb2/2liooW8PK1vYIMG6h31CrwXZs06x1//+mtyc38BYN68aB5//GvefrtjlXxeRXBKZIJqkcwUE2Pk73//O59/3pMTJzZSp05zBg8ezbNTg5k7r3Li8twqjnJypDvacaoqzJ9vLann4m58JejRZJJLhFFpAgO9c1d/N/j7w7vvdgO07DSTCbp2XeHyubLs4+eIrrjDI+Sqrmn18ce7OHr0TYex06ff59//7slHHw2o0s++W44e1dG9u/O1n5ioY+FCPYMGeeZ+UmmU2SDagiVkYyA47YMk9HrftAlZlkocGhYX/y6bTRPB2dkqV69anbKwAgOhQ4eq6xBgMsF//jOzRBgBmM2nmTfvTSZP/oGWLb3QW1dN+w0OG1aPoUOf4vr1p6hRQxPT8tnKW7nduoo2aKBy5kyZCzclBU6d0r5SUli50sLcuYX07FlQ7dIPK5tbBRveuCFx4UL1yuf294fu3Yc7jdeuPYSJE31vh1wuTZpAZCRJSQrHjuVpFZ6r4gj5Djh79pjL8dOnj7p5JrenRg2Vmy7u+WYzpKb6/oahLHXrqnTr9rDTeP36k3nkEd+MOZIkG3l52vchIXD2rOPjc+eqXLliZe3aQoYMcX+Nt337CklOXuQ0fuPGRtasOe/2+dwrurTUW0ex+xCSBMHBjl7GSCC66KsiLY7c6jnS6SAw0MyuXX507aqDqChyclRWLTQzaJAZKTiM+IkNHYOwBbekYUMLBw8aaN/eLoSKCwhevBjAlSsmLl0yEhioQ1UhL0+hX79Cr2h/ci+8++5A8vLmsmXLpxQUnCUqagrPPfe4d+7W7oYyXqJVJxtxJSSK1l3NhDVWPdpCOCQk/K7GPUmLFiorVii0aeN4W/v5Zxg/3sDq1So1atiwWAwYjRImk42wMAvt2/vuIvHee8Mxm79h587PMJkuEhs7nZdfnua1MWG3IyHBwty5eh5/XE///rB5M2zcqKLX29DpFOLizDRu7Lnj0YYN9QQFxXLz5qUyj9QmIqKWR+Z0W1w0p619I5vss6ddxh35cjxS2aPn4rF7wSPtQ86c0ZGcrPUEMxoVEhKsbivzfjvK9moB7405Ati6Vc+NG3706qXj4kVIS4OxY8FohA8+sDFjhq7kjF5R4L//tTB6tG/HI506ZSU93Uz37jXKjV/ziVYJXl75fcOGazz11DRycjaUjAUGdubjjxcxcqT3NQBOT5fYutWf7t1lAgMlDh6EVq2gbVv45BMbo0ZJNG1q30xs22YDCoiO9u1mzceOWbl6VbOJ8rJVfcEm8vNh61YjBoMOi0WlUSMLrVp5z/9n+vSVrFjxMMUxeAAdOvyTn3/+lfdVoC66v+jSUqkdXMMee6Qo5Bw+4TgG1a4X4e3wid5qpcnN1Uqde6QDuQ9G969fH0Dr1nrCw6Fe0Xq1eTO0bKkSFuZ4vLZzp4rRmHdf1MPxhYXAF1i2LINvv11FWtppGjSIZMqUIUyY0MjT07ol69YZ6NHDH5MJYmPtAbyLFsFEF/Xh5s838//bO/O4qOruj79nAdlCQYVQUBQRE1PEXcsVXBOt1DL3VjXNn0+WWT0+aYv1qC2mYZvm2mNZbpnmklvuC7gmIKKAISr7PszM/f0xbAOjosC9A37fr9e8hO+9M3Nw5tzvud/vOZ8THPxgqCjXBJ8wGEzK2I6OKKIGn5Eh8c47+zl0aAf5+bm0bfs4s2cPpGlT613B1kRH4erqVJx7FBlJavh56gT4l+lnmpycKYIjlBaBLMXff2uIi7OhQQNTefLNmwb69MnD3l5GI6phfyajUaJ1a/OxlBTKBEYATZqoOHlS9UAER9UaKwrSQ0IeJiRkvKzvWRHq1jWiVku0aGH+/TcaJaCsT9ja1qz8vJqKJMHOnbbY2WmpU0fFjRtGXFx0tGsn77boQw+pWLSoB0ZjD4zGmtWvTFCM1XysqamQnFyLUaOKL/x6vYaVK2HwYIXv6qxoorKEp2c+R45ozCp1JMnIsWMqOnYsvXJk4NFHrWeJWmAZay/BPXAgjZUr93PjRiyNGvkxaVI3WraU8y7m9gQEGNiwwcBLL2mLtutzcyEuzkjpLq2mVQjhD9WBP/+0YcAAW1xdC69pao4dUxMRkYOfn/yfoVpN0fdLp4PPP/+bY8cOo1Zr6dGjG5Mm+VhNughQtgJWcEesJjgqqW9RiFYLDz+sJTMzDycnhQzD+rWPHnnEwLlzuaxda0OtWmry8oy4ueVz6JCWhg1taNjQdDG5cEEiNzdfVAFWF6y0BHfPnmQmT55EUtI2AA4fhiNHXmHt2g/w9a2lsHWmrZa+fXNYtcoWR0ctkiSRl62jj1cEu0Pt6N1bg0plmtDWHGlMzz6iAKQ6oFZrSwRGJjp2VLN2rQ1+fsreQE+atJstW54BTN+lvXudiY39hU8+sY5G2RZFlB2MFpO1RecJE1YTHKlUlhdhXFxUZGWpcHKScRuo1EqROvYqVymrffT22zv527APP7/WTJ/ennr1lLtNaNXKQKtWxcvLp05paN9eS3S0iuPHTRPBuXN6Jk+uPlov58/n8fPPZ9Dr8xk4sDVduyoYIT/AGI2wcOEFdu7cQUZGKllZ7iQlmZf4X736NaNHB+DpmU3Tpo8wZUonGjdWTmvHyQkGDtRROFlpoqNMK3HtilfibK/FENw0k1pOyt/gCO5OyY4J5RmvSs6fz+HTT3dw/vwp1OraxMTEU/hdA1CTzvkNH5P85CdF6Q2K7jSUSheRJNj4i5ZnO8RRWGN34YKRhAQPWj/ZSBkb7xGDAX766Trh4RG4uNTjhRdaWkwluV+sJjiqV8/A5cuSWW8cMPWLCgqSNz+m9EpRE6AREFHqvD/3vEUUsH8/nDjxGRs2vGAVZfKSBMnJtvTrZx6sBQVp2bdPy2OPWX8rhZUrr/D++6+Tlraz4Pc2TJr0FbNm1TBNozthJXd1778fzpIlTwDpJUaHA7eArKKRmJjfiYnZwoEDcPToO/z88wzc3ZXfdi7CwkqcQ3KmolIJ94SVVzZWNdnZZbdF8/LAILNeT3o6vPzyZ0RFfVxi1BkYCmwEwAeok7qbwSHtAevaaQAIC9PQf5Adzo2K/aFlczj/swG9lG09gcFtyM+HCRO2s2PHy0AqoGLDhql89dVM2rWrnKIDq9kRbdPGwJ9/5nP1qikQkiTYvt2Au7tOkYqEQpXsqIKf70ZY2EyWL4+++4kykJkJbm5lP1pXVxU6ndV85OYYDGiio9BER5F/PorfP5+PW9rOoi9obu5pli//kKtXrT+wqwwM3k1Jrl2f5OTM4ocCYpB5efD77+sxD4zANAn0KjVWvFJ08eKHLF16poqtq/ns3JnMyy9vZsSIZXz5f5sZ2aUdnQseI7u0s5gLWVPx88vjp58M6AsuASkpEsuW6Xn8cXlXw5ctiyQqan6p0XTAQMng7V7nEDlJTtbQqFHZidXPT221AsJGI4SGxjBy5Cp69lzPjh0TMAVGABIxMYv47LNtlfZ+VhMgqlQQEpJHWJieQ4c0GI0QEJCPm5t1VFV5YlLcLKQJpb/wOmJirmC6Z1AWe3tISyv7/2YwWK8o6u3yuqB4OzM19Xe2bo1n8mRveY1TAiupmszIUHHzpiWV7HzMK7/6AsfNzoiPvwxYX7+p6sK6dXG8/fbzZGQcBeBawXh1aG1UFXh7S7i6ZvPTT7ZoNCq0WgNDh+plrxa7eTMRLK43pgJOQBrgSMlVVWtDo5HIyaFMJXhCgkTz5tYx55bmrbeO8sMPwzD9/4Zg6f/3woUwYESlvJ/VBEdgCpACAw1Y/uIpy6frfsXYqDGSBBOeX8XFi59jvk6kxdPTSyHrzNFqIStLT1qamtq1iyewjRsNdOhgfcmnu3enc2b9yXL0tHPBw0NZnZaq5Pp1FWfOaNFqoXPnfKvYonV1lWjWrAfh4bvMxlWq+rRrNx5b28eIjlaTmLgEMG+f0KBBYxktLQdWsk15Vwq2z/Z8vZaHM47yMGAda9LyExGhJjZWi0Yj0aVLPs7OEBys7DXs0UebAy5Aitm4h0dvfHyeQqu14alWnny5eKgi9pWHLl3y+flnG8aOLV7pyskxNZoODFTQMAtkZMD8+RdZu3YBpsAISm+vFlK7tmulva+swVFODuzZY4uDgxaV0YDmSiTt2umKKtGsaf+8SamfjY0aF93Jdx03gZ1z/sCY+3fROa1azeHFF62nn1Hfvjq2bgWt1jTZ5uQYadYsjzp1is+5ckVFQoKaVq0MPKRQzDFnTjjffvsSjXWlM7rKEhDwLiEhdWWwSh5MSvG2ODiouXjRVHnz3HMqdDrYvNkGN7c8WrZU9kZBrYYXX3yad9/dXUIl247g4K9YuTIYtTqYjRsTmDFjA+npV4qe5+MznZdfDlDCZIuUqdYBU8VOiW1KnQ7CwzU4Okq0bGlUZDsfLOc8cpefawQlcqokyTRXtG5tw2PPNCM3X8PGjTY0aZJD06bKrmwMH+7O5s2h7No1nsKuvy4uwbz33kiefLIBYCoA+G1x8XOs7bOys4NHH81h1apaODioMRohJ0evSL+6O5GYaGDUqK85c2YXcKjEkdNA11JjzvTrN6DS3ltWheyNG+2YMMHGtAxa0Gi2SMfFmmTLy5H4uH59Ahs27CUtLYkWLdowfXpnGja04k7YJf6m7Gw4dKgWLVtq8ezRjMPHID09n9695d27v3QpnwEDRpCWtrtoy7LkypEv4O42nAjJjsDAIP7zn4E0a1b+UnFrVgNOTFRx5Yo9TzyhITYWrl2DLl3Mz1mzRk9wcI5ik3RJjh3LZN2642RmphMY2Ibnn/c2a1OxfftNfvxxP7du/YOPT0tefbUrfn7WoXt0O3Q62L3bFnt7Lbm5cO2aimHD1EiSxF9/GQgMzKVBA/knYk10FJ27tCvyhULfiAZ8UPPyS+t4/nlTUHevN5TW7BNFFYWltL1KJtKvWqWnf/+cqjbzruh08P33Vzh9+jQPPeTCyJHtCQwssdRbDZPn9XrYtcsksqmWDGivRtGsmQ4PD5MPKGH/228f5bvvgoEGmJJbjpU42h612h8np1zc3LwYOHAIb7/d5p60paxCIfvKFRWBgVrz/eHS1SNWoOEClCvfY9gwD4YNGymTQRWnpKigqyuMGIHpwpOkYsCA5kRFqQkPNxIQIN9KxR9/xJOWtrvo99J3V02AFT+/hc7bV16VdBkIC7Nh9GjThebUKRgypOw5bdpoiIlRKX6nDNCxoxMdO5ZOwC6mf//69O//tIwWVZxt22oxZowttQribUmCFStg3DgVvr5afvjBjgYNlJ+ICzEC2kdmM/TNvhhqW0HEXBVY0vYqQb16arKzUXzL2dYWJk3yBrwtn2AlOYP3wvbttXj2WVvTtTYyGmrHmy9eXJFfgDYq6lzBT/8APYBzQHbB2BkGDnyHL7/sg51d5cdt8gRHBgNpx2Px87Njx2LTH2HUGwjuZ71RdI3kDhceX18VJ05okTPfy5Q/ZNq7L5lT4WDfnhUrQvHyUmPwbop9DfuaZGVBbAz8uTQSoxGyb4KhOWj9fMw8PCtLof6CDwCJiSpatNAWBUZgynns398katm1K7RooSEhQVV056wkrfyn0TGgD9OmdTPLI6wp3LihImq/LUPvkqaj00l3bKwruD8yM8HdXWN+E2oFixd16pTMCVwHDALUuLj4MWHCk/zrX/7YVlFLO1mCI82VywT53ASvJnh5ATExGIyYl3tZa3LkA4Tc2zdDhtTjm29mc+rUdIwUbqmp6N/jdRr09LPCtPyKk5cHO3bY89bwy2hizbeV0WJ2MbpwwcATTyg/MddErl1TERBQdv394YfheEHRna2tSU9FCUrnFn333dhqtxJRXlJT4fRpe54beudZzmCApCSDCI6qgNRUFR4e1ifz8uyzj7FnT1fS0w9hWjvdgoNDAPPmzeeppx6u0veWLyG7VBSqMRiQ1BoWf2lg1KicMsmRAnm5dk3C0VFeDSGNBkJDxzB3bj1OntyHjY0dXbsG8+GHPWW1Q04OHbJh9GgtmnjK+MQvv0gEjJDIyICwMAMdOuQqZ2gNp3lzI6dOGRgwwHxZ8sIF8ClQ4zh/3kD//vIHpwbvpvx4+GSZsZrKkSO2pp6al7BYUXj0hjc5/xi5fFlP794K99msoXh4SBw4YKRt2xL+ULoXmwKLF336uPLJJ9+zYsVW4uIu4eHhzciRg6o8MAIlS/k1GlTNm/Nw93wMPmISkAULFx7JuwlHjxiJiMhnwAD512qaNLFl+fIh5OcPQa226nzFSkGS1LfNl2jYMJu4uGycnCSxYlTFODlBaqqeq1fVNG5sWjJNTZU4cULF0KGwdq0BX19lrkuSWkO2p6/Zll9Nxs5ObUqi9THXiPvzT+jZswnH/m5IcIds/P2FT1QVGg3Y2ek4daoWgYFq8PEhMxO2/JhP3755qBRcvOjVqyHdur2Cu7sk6+5GuYKjhIQE5syZQ1hYGHZ2dvTp04dZs2ZhUwnrmzrrqhysNHJzYfXqWK5cicPHx5tRoxpW2d5oeShdyixJcC72YRJuNcS3RTYDB8p84SlVzVEYE1l7RUch9+sTBoOEwWBZpcPXV8LYTHSIl4ugIB3Hjxs4dMgGtRpycw1otbB7t0TPnnrZt2+MRpg37wzbtv1GSso/NG/enldffYKgoHryGnKf3K9P5OQYkSRQaTRmK6kpZ+GiUaJt+2zq1q15gdHevels3nwSo1FP796BDB5cV9HK1M6d9Vy8aGTtWhvUahVabSN6vZKPUaHLcWKigTff3MqhQ+vQ6TJo2/Zp3n33Sdq3d777kyuBcpXyP/300/j6+vL222+TkZHBlClT6NatGzNmzCg6555KNAvKM09lNiMtLdesYWpN4MYNA2PHruTUqbeAHMCRjh0XsmbNyBqZTHk/aKKjLCpiV2b/oaosW75fn0hLg5MnHXimbbT1SllUNhbKmqH6BMJy8cknZ1m4cAAlW7W4u4/g11+/wte3cu6srNEnbt1Sce6cPcOHF38XLl6E69chMjKfJ5+seTsLixZFsHDhS+TkhAOg1TbixRdXMneulSkwKsjIkRvYvXuc2ViLFv9mx443Kq1Q5U7+cNfg6OzZs4wYMYKDBw/i6mpSn9y+fTuzZ8/myJEjqAtEBe6oc1RwcQwL05Kbq8XWVkVSnSa41jfStm3NCowAZs48zPLl/UqNqpg8eQ/vvVf1X/4bN1ScPq3F3l6ic2f55fXLQ2ktFzDpuRypBsFRRX0iLk7FxfNa6qbGoNdLaLUGOnTIr7HBghyBcE2gZ8/5XLjwfpnxV17Zzfvvd6iU97BWn7h2TUV4uC35+Rpyc1Xk50vUraunVy9djZPxyMiAHj2mEx//vdl47drBbNv2P5o1q9olS50ODh+2IT9fhbOzgQ4dDFahpVaSy5eN9Ow5gNzcw6WO2LNgQThjx3pUyvtUSOfo/PnzeHh4FH3hAfz9/UlLSyM2NhZvb++7W1Cg+dDabEtZoTKQKiAlBbKzVTRoYNoTPX/+pIWzJM6fDwOqNjj6808b3NxsGTlSTVqaxJYtBlq0yMXbu2wMHBOjIjpaS716Btq0UU4RuLpRUZ/w8pLw8srHJGpmoibdImRkmKpfGjaUigTZ7t4apoq4BzG+/Hw4elRLbq6Kdu3ycXGRy0jTNndaWpLFY7cbtyYq6hMNG0o0bPhgJFsfPJhBfPwvZcbT0naya1cCzZo1qrL3TkxUcfy4HSNGaHF0NPVS++UXPUOG5FpVFWBysoHc3H8sHMkhLS3bwnjlc9favdTUVJydzff4ateuDUBKSoqlp9R8CjrIZ56K5N/PfM+oDk8xplN3hg5eyv79qTz0UB2LT3NysjxeWVy6pKJZs1r06qVGpYI6dVSMGaPl77/N1yAlCbZsqYVW68jIkXb4+TmwcaM9GXcXORcgfMISUVF5TJ9+gDZtPsPf/wM6dnyO/v2/YNOmBEXtKhQ/dXV1Mj3Sblrc4ouOVrF3rwNBQfaMGGHHlSuO/PWXfEuuKhU88kgbC0ccadu2lWx23C/CJ8zJy4MP5oQxsfd7TOz9HqHTt2D4OwpNdBQ+3rY4Opb9TDWahnh7V+0cER5uy4QJpsAIwMNDxfjxNuzbVyoyKpjjSj7k7FoeEGBDy5Zjy4y7uAwiJESeno335f2FO3GqB3Sp4XYd5GOOhTFjxjUmTnyFAweaotMVX4Tt7FoyZMhjVWrX5cs2jBpV9jNp3FhNcjK4upoSxQ8f1jJkiA1165rObdRIxYsvalm1qhYDB8p392ZJEbu68iD7RERELmPHziEmZkmJ0T6Eh6/g7bfDaPflTGU/6zuI2RkMpoksKsqOMWOKV5P69lWzd28tEhIMsolATps2iIiIicTFLS0YcaZv328YPdrzjs+zVh5kn3j11d38vXk4TdATA5w9B6vXFG8n9+z5Glu3Fmr3mOjUaTb9+lVtsrGdXdkVU3t7UKs1lNzNKdlRAZBdIVurhX/9awzvvnuV69dXARJOTt155ZXZNG4sz03LXd/F1dW1TOSflpZWdOxB5XbbBFeuLCE3dwKzZv3ETz/9yrVrYXh5tWfkyKcYOtS9Sm2SJNOj9LUoPx/i4tQcP26Lq6sGvV7F7t0qQkKKFZjVanBwkC/fxZKWS+G4tSN8wpylSw+XCowAdgNPc/PmLyzb96rVfdZGI+zYYWqCbWenQq9XcfIktGtXfE6PHmp+/NEGDw95Smo7dXLm998/Ztmy0SQn3yIg4BGeecbDKnMGSyN8opizZ/PYtetjGhYERpbmicWL+1O37lYOH96DXp9Pu3aP85//9Kry1Aaj0XKgb8zXm1aHClDHXoXaDiZ5hcItaJkVskNCHqZTp0WsXj0FnU7HkCF+tGwpn77FXd2uVatWJCYmcuPGDdzc3AA4c+YMdevWxcvLq8oNrH7oyczM5s03WzFx4lukp4Ozszx5tv7++ezfb0OPHsW7pZIEkZEG3Nzsinp5gSlgWr8eRirVHq4a9h4qRPiEObGxl25zxBRUZOflYPCpnITi+8KCmN2uXbYMGWKLs3PxbHTggHkBYeW05L433N3VzJrVWv43riDCJ4oJC0snO/vYHc9xdIQFC7oB3eQxqgC93kB6usbse3/xokSj/EjzlSJXf5MzREffsdddVePurub111so8t53zTlq2bIlAQEBLFiwgIyMDOLi4ggNDWXUqFEP5HLp3XByQWydNQAAGP9JREFUeoyBA02TvkYDLi7yFSB5ekqkp+fx668Gbt2CCxckli3TY2trYMgQcyNsbKBePYryjPLzITtbXoXs6orwCXPc3W+37aNFo/Hi8ccfldWekhi8m5Jcuz7JyZmmR4GYna2t1myCAHj8cQgLK/59xw4jgYE1p3CkKhE+UUz37i7Urh0EmLbRfEs8lE4d6NNHx4YNOrZuNXD6tMTPPxv4++9cHnnEULz9XPho0sQUIEVGlhUQfgAo14LtF198wZw5cwgKCsLBwYEBAwYwceLEqrbNqrGUQ3FVNYbRoyfTqpVy0radO+vJzdVz9KiGOnUkQkKM7Npla7ESwdlZIi5ORVaWRHh4PgMG1FBFzipA+EQxEyY8xoEDT3IzcQPFBamt0GpSaf/MtwwcKGPZV2lus0J5u62qrCyJW7dU7N1rwMkpr0aKD1YVwidMeHtrGTr0TVavOA6kFow6odU8TbuhvdE3bopS4aJaDQMH6sjK0pGUpKJbt4JGvtGWz09Ny8aYnKloey9JMm2Dy61yUi4RyPJwR52jmobBQOS2i0yd+hqZWceLhqOBp4dvZMmS3srZZoHTpzX4+NjTpIm5S65ercfJSYeXl5FGjZRVyC4arkStn6oUvCsPD5JP/PVXKtsX/ciVvTOLChWsWctoxw47Ro0yv2PIyIDly3Np2tSkv6aEov2RI+l8990e4uOjadDAm/Hje9G9e+UFl8In5EGSIChoG2fPri0YyQO2o9F48vXXBwgJqaukeWUoI9QMigvTGo0wf/45tm7dSkpKAn5+gbz66hP06lV5OWwV0jkSWECjYeWRDMJKBEaFHD20HSnCq8ydqZICf61bG9i0KZ9Bg2xo0ECFJMHevUbq19cRGKiMws7tKv6sdTIV3JnHHqtDD48gOncpm4CamQnz54dx+vQx7O0d6du3O+PHN1JUV8vbW8emTWoGD9agVkNKisS6dXqGD89XLAH6xIkMXnrp/0hMXA/AqVNw9OgTLF36FY8/XrUl3oLKRaWCtLSDwCazcYMhnj17wrh40YPt27eRmprEI4+0Zfr0/rK1xbBE6fZSgOLN4BcsOM+nnw5CkkyJ/omJy4iMPMSGDV/i41P1okwiOLpP9HrL+TmeebcY87h1TfoqFQwZksfx43r279ei10v4++fj6Vn1q0UnTmTz22/n0GhseOYZf5o3L74dV0wYUCAbRiOMG7eBAwcmUFi2vG9fQ27c+ImZM5XLRWre3Iiraw5r19pgY6NCqzUwdKheUYHy5csPFQVGhdy8+RsrVjzD448PUcgqwf1iY2N56fHMmTzWrOkPmKr54uMhMvJVtm79CDc3he4YFCqQiYvLZ+nSU8TFRePh0Zjnn2+Hn5+phHrbtt+LAqNCrl9fy8qVLzBnTtUXeIjg6D7p378Na9a0JC/vgtm4v38n9u5bb3WTvkoFHTsakFOLecGCCyxePI3s7KMArF49iHff/ZgxY+QR8RLIT5NSP+/alcJff82gpJ6LXn+N9evXMHXqxzg4yG1hMfXqSfTrJ2Oe3V22kq9fj7X4tOvX46raMkEV0K1bENHRn1Hymuvg0IWUlFMUBkaFXL36Fd9//xyzZil3wyA3MTE6Ro/+mKioBUVju3e/wLJlH/Doo46kpt6y+LyUFHkU4+9arSawTM+eD/Hyy19Tu3bfghFX2rX7jClTuitql7UQH2/g228/LAqMAFJStvLVV8vIrXl9JAUUa1cdKXj8ePgkx5O1SNLNMufGxu7l+vUHq4qpcCu5c4nHyC7tigImL69mFp/n5eVjcVxg3cyd25UhQ37C2TkY8KBJk6m88cZX5OZaUo2XSE62HAzUVEJDj5kFRgCxsd8TGnoAlQpatAiw8CwHWrf2l8U+sXJUAf797zaMGfM/tm2Lp0EDZ554whXbK9a2ZqQMW7bEk5Kypcx4dPRqDh58k77eNUshW4DFpfkW/rdQqeqXCZAaN+4pm/L0vaLTwf79mbi42BAYWKtSc6PutJU8cWIXDh6cyNWrS4vGPD3H8sorj1eeAQLZcHCAb78NJj6+D7GxOtq2tcPeHnbuDOTWrdWlznamTRtl9HzuxqVL+fz6ayS2trUYM6YZdSspl/zKlQiL4zExfwP9ee21gUREvEh8/HcFR5zo0+dbxo2TRzdLBEcVxNtby6RJ3mZj1WbSr8KKMTc3J6AOxaWsJuzsmtCggW21VsgWlJ+QkHqsXr2Q/fvHU7i1ptU2ZNiw0VbZbX3jxgTmz/+eqKhl2NjUp0OH1/jss+E0aVL1pWstWtixevVcvvnmCeLiLtGwoTcvvNCZVq0U3HsUVBhPTzWensX9LSdPHsqlSwe5caOw+awDwcHf8MwzHly9qsLODtzdrePGYfHiSL744m3S0nYAan74YTwffjiLQYMq3u3Bzc2yPpq7uyn46dKlNlu2zGfZsrGkpt6kdWt/nnuugWwNckUpf2VjIeCQJAjdYWTz1j9JSblJixatmTYtmDZtHBUy0oQmOqrKKsb0ehgwYCmnT79pNt6790r+97+hFXrt8iLKlq2DrCxYsCCc06ePY2/vSP/grox/vKy4opIVnQA3b0r06TOV69dXmo336vU969YNr/DrV6W/lRfhE9ZBREQeq1aFkZaWTJs2rWjWzJn5838iPPx/2NnVpWvX0SxY8ATu7sr5Q3y8gaCgcSQnbzYb9/efza5dMyrsqseOZfD885O5caO4oq9One4sWbKC4GB5pA7u5A8iOJKBxYsj+eijoej18UVjTZq8xubNc3F3Vy7tSxMdRecu7cyW+X2BI5V0sY6IyGXOnM2cOvUHWq0dnTs/wbx5/ahfX56/WUwE1ok1BAmW+PzzSD76qANgfkl0cOjJ4cNbKr4NKIO2190QPmF9ZGVBUNAcoqMXmo0HB69gzZonFbIKFi++zNy5lvJ+GvPbbyfp2LHiq6n796fw3Xe7iY29RIMGjRkzJogBA+pX+HXLi9A5UpgNG34zC4wAYmIW8d13I3jnnerXR6m8+PnZsXbtCDIyRqBWm/oJCQRgnTIOer2B0oERgMGQy22UO+6NatxPUFB1rF0bS3T0ojLjBw8u59q1J2nYUAGjAGdne8AOMK+gsbN7GBeXygnmu3d3oXv3YZXyWpWNqFarYoxGuHnzH4vHbt60VLUgL3L0/nnoIREYCayf4cN9cXEZXGY8MPA5WTTBBA8mGRk5QNlt5pycBNLTjWWfIBMjRnjg5/d6mfEuXV7B11dBQTCZECtHVYxaDT4+/ly/XvqIHS1aKNftGBBJ0QJBCRo31jJjxly++OIhbtz4H+CIv/+b/Oc/Tyuq5i2o2QwZ0pTQ0H6kpf1hNt669Rj8/JRbv7Czg4ULJ/H++/UJD/8RW9vadO48iv/+9wnFbJITERzJwCuvDCIychc3bxaWtmvp2nUx48crXMdWwWV+nQ7S0lS4ukqKKgsLqh/WWtH50ks+PPlkKL/++i4uLnYMHVrvnqpjMjNNxQh1RLcPQTnx8bFh8uS5LFmST3r6n4AaL6+XeOON51ArvLfTsaMzmzc/T3z8C9jaWk8VnRyIhGyZOHMmm1WrjpOaeotHHmnFxIl+iqoDVwRJgv/+9ywbN67n+vWzNGnSlfHjhzF2rLfSppkhkk+tFCtITK5sbt408tZbf3DkyEZ0umwCAwcye/YQ/P2ty8mFT1gvERE6Nm2Kwt7ejjFjmlKnTjVerrTg49bo36Ja7QEhLw/WrfuHxMQUevb0pkOHqkn0CQ2NZs6cvhiNxcJ+jo7t+eabXwgOrrwO4hVFTAQ1nyNHsjh4MBYvr3o89VR9xZrGjhy5kd27x5qNPfroR2zfPkU2XZbyIHyi5mIwwPffX+HYsRPY2tozaFBnBg0qLonX6+Hbb2M4dux4wfEuhITUq1QbMjNNuwme2ZGM6lZckWoN1aiWENVqNYG73G1HROQyadJ3nDv3AZDNl1+2Zfjw+SxY0LFS8iWio/NYtOgvIiLOcvlypllgBJCVdYINGw4THDyw4m8mENwFoxGmTTvIpk0zyc09A9Thhx/m8t13o2nQQFv1q1MlXj8mxkDigYX4AtEUd5E7e/YjfvllBM8+61bx9xMI7sLkyfvYsOFZIAuA335rwYwZq3jtNT8AJk3ay6ZNzwA5APz+e3MuX17N//1fxZW5dTqYNesQO3euJSUlkr7eHa2yIvVeEMFRNUCS4NDKcEJn9rqtNsy8eds5d+7tomO5uWGsWTOWHj0OEhJSMUGtpCSJ8eM/IyLio4KRshU9AJmZaRbHBYLKZtGif1i3bjnQGIgHkjlx4jU++MCTr74KKupjVlVaSrd7fSg5IWRy61ZWhd9LILgjBgPhv5wl8rdp+BYERtFAbu5FVq9ewYsvfsSJExls2/YGhYERQF5eJGvW/MDLL1e8AfR77x1j1aoQwNTI+WLEkYq9oBUggiMrR6eDF174g6s7p2DAciSu18O+ffvKjBuN/7Bv32lCQnrf8T2OH89iz54Y3NxcePbZhtjZmR//9tvzRETMKzFSy8KraGnduu3d/hyBoMIsWRLF/PlvAH8CKiAYU5fzo4SHHwKCOHQo5b7vXHNzYe3aeG7dSqVPn6a0a2d55rjb67u4DGTwYHn6QAkeXDRXLrNgSveCsMQ8SL9yZQuRkXM5eewqjXXmvcyigatXtxIZ+QEBAbcJBQpWSG/cUKHTSTRsCCqV+QqsJMHevTugyAJzO0r/XF0QwZGVExoaxR9/jMK31BevJEuWXCIrK8XiMbvSkU4JJL2BBZM2sW3bx+TpLgK12PvNRN769i1a+BfnKyUkxGEujvcXMBL4FcgDXOjRYxGvvmqdjRMFNYeEBCOLF39Ifv6fBSMSsAMYCmioVcuejAxYsOCH+3r98PBsRo78nKSkz4FcPv88gOeeW8j8+R3uaXvawaETL700m8aNxSVWUPXcLlCvV68TjRppCXgom5MF50FxsJLm1hlv79t/R68fjGTGsE53XIGVJMjOTjd7XnTBv0NCFjNrVheg+knECM+1ck6ePElhRH678mfTqpG24FEs5avVPs6IEbdfzdm//BSnN40nr2gkD/2lL1gxuwHzfplUNNq4cdNSr30d2EDjxosYMMAbf/8mDBv2sLUVIghqIL/+epWkpF8tHDkOtKN79yDWrLnCPwk/WvSXhARw87n9648du5KkpI+Lftfrw1m58gWCg/fTr9/d6/PHj1tDwkMeDB7sT9u2VthZV/AAYc/AgS/g6gq9e7vw7r9LB1B2DBz4wh1lJ+bN22Ex8NLpoPByr1bDo4925J9/lhYdNwKXVPVp2r8/Bp/qmXMngiMrx87OtKQfXWq8Tes3+PrrZzF4N0Wn2wlsAoZjyr9IAFri6/sEbdrc/gJ9/PhFi1/8m6cSeOONPXh6evLCC768/LIf27bN4/TpN4rOcXZux6xZPXjqqQYV/hsFgvJSu7Y9YE/J3AkTdRkw4EPeeactX30VSTSGMs+NVTcj2aUJt7tUX7iQz/XrBywcucJPP52iXz/z7WlLwdfEiS2sriJH8ODh33I6g/sPZ8aMVgAW9ZJeeH4l4z7sfNvXuHBBz5kzP1o8NmrUKpLq1iYgoBMzZ3Zl5syBxMS8S2TkJ0A+NjY+DB26mKeeqp6BEYjgyOoZPLgL27e3IDf3YokgpjYjBj9TdBFu27YTx47lAz8CdYH6wE5CQv59x9e2tbUcOGVmfcWKFZ8DWjZunMPKlRNZu/YlFi1qR0TEWVxd6zNixGP07i2U7gTyMmzYw3zzzQwuXnzfbLxr16msWNEJgCefbEpo6GCiUraYnRMY8DJ+LW9/yTtzpmyZuRrwAZz+uY4m2uSBBu+mQl1eYFWUzu/5/vvRZYL00ue89JIPhjus9qemGtDl37J4LC7+c6Li4fRp+PvvpWza9Bx//PEmK1cOJykphW7dmtGzp3O1VpYXOkfVgCVLIlm9eg2XL/9BgwZdGDx4DO+9F1h0N5CZCc8/v4kDB2ZhMMRja+tLUNDHfPNNMLZ3aJx86ffzjBvfxWzlqNCdSo6NGrWFzz7rUcl/VdUjNF1qJseOpfPBB+s5depHbGweonPn0fz3v4Px8ioWFAoNvcSiRXNJStoI2OLj8xqffDKV7t1vr8N17ZqBTp3modMthIKVp8J+g9au11JehE/UQMojuHgfoox6PTz/2CyMl5eUyTky33FwYfHi44wYUf1WiYQIZA0gLw+uXVPh5ibh5FT2uCTBjh2pXLiQSGBgA7p3f+iuUbsmOspiOXLprbbAwLls3/5/Ff8jZEZMBDUXSYK4OBW1at2+pcG1awZ++eUKjo61ePZZz3I1P37jjZOsWLESuATcwJdGwM4if/AFjojg6L4RPlG9+H3LP3z/9r+5nrge8MZUAHHFTM8L4PXX/2LmzNZKmFghRHAksEzB3YReD/HxKkJDd7Nv/xtlvvi9ei1h3boxSll534iJQHCvSBL88EM8O3acJDfXhmcDHVj0ZYgIjioJ4RPVj+RkiTVrLpOdreenn1YTF/eF2XGttjHLlx8oV8GCtSEUsgWWKWg8qwK8/KCH3oUfT67DmHWi6JRatZoREtJLORsFAhlRqWDCBE8mTPAETKurW74sPl4d9VoEgorg6qpi6lRTiae9/Wjmz99GXl5kwVFbgoIW0rdv9QuM7oZYORKYsWZNLCtXbiY29jQeHi0YPnwIkyY1U9qs+0LcJQsqTDVpoFlehE8IKsrmzTf57beD5OTk0L59eyZN8r1jbqs1I7bVBPeEJEFODtjbU62rDcREIBCYI3xCIChGbKsJ7gmVigr32hEIBAKBoLpiQRpKIBAIBAKB4MFFBEcCgUAgEAgEJRDBUQ1HrzeJRFZOZplAUP3JyoL8fKWtEAgE1ozIOaqhGI3w4Yen2bZtI0lJV2nevBOTJg1l4EB3pU0TCBTh0KE0Pv10E+fP78fR0Y3evUN4//3O1KqltGUCgQLUsErMykYERzWU+fPP8eWXg4B0AI4eXU909AG8vb+nZUsxGwgeLJKSJKZN+5irV5cU/A4//PA1BsMWFi7sqrB1AoH8aK5cNuuQUN1b41Q2YluthrJ9+zYKA6NCbt3azKpVp5QxSCBQkOXLL3L1amip0Xz27NlATo4iJgkEilPYKioKzNpICURwVGNJT0+1OJ6RkSKzJQKB8qSnp2HeFMdEWtpVsrOrsZiXQCCoEkRwVENp2bKthVFHAgIeld0WgUBpunVrgVbrXWa8ZcsgXF1FtYLgwaQJpn6BvojWOKUROUc1lH/9qz9RUdO5fPkLTHfMLvTt+zXjxnkpbZpAIDt9+9bh6aeXsn79ixgM8QB4eIzhtdeeqtYq8ALB/WLwbsqPh0+WGROYEO1DajDJyRLLll0kKekmAQF+DBvm/kAVIohWCYKSSBJs25bCX3+dw9GxNuPG+ePp+QA5BMInBIKSiN5qggcSMREIBOYInxAIirmTP4icI4FAIBAIBIISiOBIIBAIBAKBoASVtq0mEAgEAoFAUBMQK0cCgUAgEAgEJRDBkUAgEAgEAkEJRHAkEAgEAoFAUAIRHAkEAoFAIBCUQARH5SQhIYGJEyfSqVMnevTowdy5c8nPz1farBrHtWvXmDp1Kp06daJz585MmzaNxMREpc0SlEL4g3wIn7B+hD/Ih1z+IIKjcjJlyhTq1KnDzp07Wbt2LWFhYXzxxRdKm1XjmDhxIrVq1WL37t1s3bqV1NRUZs+erbRZglIIf5AP4RPWj/AH+ZDNHyTBXTlz5ozUokULKSkpqWhs27ZtUocOHSSDwaCgZTWLtLQ06a233pKuX79eNPbbb79Jbdu2VdAqQWmEP8iH8AnrR/iDfMjpD2LlqBycP38eDw8PXF1di8b8/f1JS0sjNjZWQctqFs7OzsybNw93d/eisYSEBLPfBcoj/EE+hE9YP8If5ENOfxDBUTlITU3F2dnZbKx27doApKSkKGHSA8Hly5cJDQ1l8uTJSpsiKIHwB+UQPmF9CH9Qjqr0BxEc3SdSgbC4SqVS2JKayblz5xg9ejQTJkxg8ODBSpsjuAvCH6oe4RPVB+EPVU9V+4MIjsqBq6trmTuAtLS0omOCyuXAgQOMGzeOKVOmMGXKFKXNEZRC+IP8CJ+wXoQ/yI8c/iCCo3LQqlUrEhMTuXHjRtHYmTNnqFu3Ll5eXgpaVvM4ffo006dP55NPPuG5555T2hyBBYQ/yIvwCetG+IO8yOUPIjgqBy1btiQgIIAFCxaQkZFBXFwcoaGhjBo1SiybViJ6vZ533nmHqVOnEhQUpLQ5gtsg/EE+hE9YP8If5ENOf1BJhZujgjuSmJjInDlzOHnyJA4ODgwYMIDXX38djUajtGk1hhMnTjBq1ChsbW3LHNu+fTsNGzZUwCqBJYQ/yIPwieqB8Ad5kNMfRHAkEAgEAoFAUAKxrSYQCAQCgUBQAhEcCQQCgUAgEJRABEcCgUAgEAgEJRDBkUAgEAgEAkEJRHAkEAgEAoFAUAIRHAkEAoFAIBCUQARHAoFAIBAIBCUQwZFAIBAIBAJBCURwJBAIBAKBQFCC/wdsNhrYlYCgrgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 720x360 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.model_selection import RepeatedKFold\n",
    "\n",
    "n_splits = 3\n",
    "n_repeats = 2\n",
    "\n",
    "kf = RepeatedKFold(n_splits=n_splits, n_repeats=2, random_state=1)\n",
    "\n",
    "plt.figure(figsize=(10, 5))\n",
    "for i, (train_index, test_index) in enumerate(kf.split(X)):\n",
    "    # print(\"TRAIN:\", train_index, \"TEST:\", test_index)\n",
    "    X_train, X_test = X[train_index], X[test_index]\n",
    "    y_train, y_test = y[train_index], y[test_index]\n",
    "    plt.subplot(n_repeats, n_splits, i+1)\n",
    "    plt.scatter(X_train[:,0], X_train[:,1], 40, y_train, lw=0.3, edgecolor='b')\n",
    "    plt.scatter(X_test[:,0], X_test[:,1], 20, y_test, marker='s', lw=0.3, edgecolor='r')\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 [0 1 3 5 7 8] [2 4 6 9]\n",
      "1 [2 4 5 6 7 8 9] [0 1 3]\n",
      "2 [0 1 2 3 4 6 9] [5 7 8]\n",
      "3 [1 2 4 6 7 8] [0 3 5 9]\n",
      "4 [0 1 3 5 6 7 9] [2 4 8]\n",
      "5 [0 2 3 4 5 8 9] [1 6 7]\n"
     ]
    }
   ],
   "source": [
    "for i, (train_index, test_index) in enumerate(kf.split(np.arange(10))):\n",
    "    print (i, train_index, test_index)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## LOO"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 [1 2 3 4 5 6 7 8 9] [0]\n",
      "1 [0 2 3 4 5 6 7 8 9] [1]\n",
      "2 [0 1 3 4 5 6 7 8 9] [2]\n",
      "3 [0 1 2 4 5 6 7 8 9] [3]\n",
      "4 [0 1 2 3 5 6 7 8 9] [4]\n",
      "5 [0 1 2 3 4 6 7 8 9] [5]\n",
      "6 [0 1 2 3 4 5 7 8 9] [6]\n",
      "7 [0 1 2 3 4 5 6 8 9] [7]\n",
      "8 [0 1 2 3 4 5 6 7 9] [8]\n",
      "9 [0 1 2 3 4 5 6 7 8] [9]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import LeaveOneOut\n",
    "\n",
    "loo = LeaveOneOut()\n",
    "\n",
    "for i, (train_index, test_index) in enumerate(loo.split(np.arange(10))):\n",
    "    print (i, train_index, test_index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 [2 3 4 5 6 7 8 9] [0 1]\n",
      "1 [1 3 4 5 6 7 8 9] [0 2]\n",
      "2 [1 2 4 5 6 7 8 9] [0 3]\n",
      "3 [1 2 3 5 6 7 8 9] [0 4]\n",
      "4 [1 2 3 4 6 7 8 9] [0 5]\n",
      "5 [1 2 3 4 5 7 8 9] [0 6]\n",
      "6 [1 2 3 4 5 6 8 9] [0 7]\n",
      "7 [1 2 3 4 5 6 7 9] [0 8]\n",
      "8 [1 2 3 4 5 6 7 8] [0 9]\n",
      "9 [0 3 4 5 6 7 8 9] [1 2]\n",
      "10 [0 2 4 5 6 7 8 9] [1 3]\n",
      "11 [0 2 3 5 6 7 8 9] [1 4]\n",
      "...\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import LeavePOut\n",
    "\n",
    "loo = LeavePOut(2)\n",
    "\n",
    "for i, (train_index, test_index) in enumerate(loo.split(np.arange(10))):\n",
    "    print (i, train_index, test_index)\n",
    "    if i > 10:\n",
    "        print ('...')\n",
    "        break\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## StratifiedKFold"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x216 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.model_selection import StratifiedKFold\n",
    "\n",
    "n_splits=3\n",
    "\n",
    "skf = StratifiedKFold(n_splits=n_splits, random_state=1, shuffle=False)\n",
    "\n",
    "plt.figure(figsize=(10, 3))\n",
    "for i, (train_index, test_index) in enumerate(skf.split(X, y)): # + y!!!\n",
    "    # print(\"TRAIN:\", train_index, \"TEST:\", test_index)\n",
    "    X_train, X_test = X[train_index], X[test_index]\n",
    "    y_train, y_test = y[train_index], y[test_index]\n",
    "    plt.subplot(1, n_splits, i+1)\n",
    "    plt.scatter(X_train[:,0], X_train[:,1], 40, y_train, lw=0.3, edgecolor='b')\n",
    "    plt.scatter(X_test[:,0], X_test[:,1], 20, y_test, marker='s', lw=0.3, edgecolor='r') "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 [1 2 5 6 7 8 9] [0 3 4]\n",
      "1 [0 2 3 4 7 8 9] [1 5 6]\n",
      "2 [0 1 3 4 5 6 8 9] [2 7]\n",
      "3 [0 1 2 3 4 5 6 7] [8 9]\n"
     ]
    }
   ],
   "source": [
    "skf = StratifiedKFold(n_splits=4, random_state=1, shuffle=False)\n",
    "\n",
    "for i, (train_index, test_index) in enumerate(skf.split(np.arange(10),\n",
    "                                                       [1,1,1,0,0,0,0,0,0,0])):\n",
    "    print (i, train_index, test_index)\n",
    "    # if i > 10:\n",
    "    #     print ('...')\n",
    "    #     break"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 [0 1 2 6 7 8 9] [3 4 5]\n",
      "1 [3 4 5 6 7 8 9] [0 1 2]\n",
      "2 [0 1 2 3 4 5 6 7] [8 9]\n",
      "3 [0 1 2 3 4 5 8 9] [6 7]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import GroupKFold\n",
    "\n",
    "\n",
    "gkf = GroupKFold(n_splits=4)\n",
    "\n",
    "for i, (train_index, test_index) in enumerate(gkf.split(np.arange(10),\n",
    "                                                       groups=[1,1,1,2,2,2,3,3,4,4])):\n",
    "    print (i, train_index, test_index)\n",
    "    # if i > 10:\n",
    "    #     print ('...')\n",
    "    #     break\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 [3 4 5 6 7 8 9] [0 1 2]\n",
      "1 [0 1 2 6 7 8 9] [3 4 5]\n",
      "2 [0 1 2 3 4 5 8 9] [6 7]\n",
      "3 [0 1 2 3 4 5 6 7] [8 9]\n"
     ]
    }
   ],
   "source": [
    " from sklearn.model_selection import LeaveOneGroupOut\n",
    "\n",
    "logo = LeaveOneGroupOut()\n",
    "\n",
    "for i, (train_index, test_index) in enumerate(logo.split(np.arange(10),\n",
    "                                                       groups=[1,1,1,2,2,2,3,3,4,4])):\n",
    "    print (i, train_index, test_index)\n",
    "    # if i > 10:\n",
    "    #     print ('...')\n",
    "    #     break"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Time Series"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 [0 1 2 3] [4 5]\n",
      "1 [0 1 2 3 4 5] [6 7]\n",
      "2 [0 1 2 3 4 5 6 7] [8 9]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import TimeSeriesSplit\n",
    "\n",
    "\n",
    "\n",
    "tscv = TimeSeriesSplit(n_splits=3) # + появился gap\n",
    "\n",
    "\n",
    "for i, (train_index, test_index) in enumerate(tscv.split(np.arange(10))):\n",
    "    print (i, train_index, test_index)\n",
    "    # if i > 10:\n",
    "    #     print ('...')\n",
    "    #     break"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# оценка работы алгоритмов"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "accuracy = 0.85 ± 0.07\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import ShuffleSplit\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "logreg = LogisticRegression()\n",
    "cv = ShuffleSplit(n_splits=10,\n",
    "                  test_size=0.25,\n",
    "                  random_state=1)\n",
    "scores = cross_val_score(logreg, X, y, cv=cv, scoring='accuracy')\n",
    "print (f'accuracy = {scores.mean().round(3)} \\u00B1 {scores.std().round(3)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'accuracy')"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAm0AAAELCAYAAACLTrFkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nOzdeXzU9Z0/8Nf3O0cmmckkmclBCCBySwDDoaioXSkoHmw9Vq2iZXnYFaTggVt9aL3AfVRWsS51Lba2ta6rWNaf19Z60NWu2vUAjRJCAjnJPTkn1ySZ4/v5/fHJzGRCgBAyV/J6Ph7zIPnON9/5DDnmNe/PpQghBIiIiIgopqnRbgARERERnRxDGxEREVEcYGgjIiIiigMMbURERERxgKGNiIiIKA4wtBERERHFAX20GxBuTU2d0W4CERER0bBkZCQf9z5W2oiIiIjiAEMbERERURxgaCMiIiKKAwxtRERERHGAoY2IiIgoDjC0EREREcUBhjYiIiKiOMDQRkRERBQHxvziukRERDQ6XC75r04XvClKdNs0njC0ERER0Ql1dgLl5SqamxUAAoASCGs6HWAwAEajgMEgYDAgcDMaAb3eH/AEdDpAVeXn/uMMfcPH0EZERERD6uwEKipUNDYqMJkE0tNF/z0icI6mAT6f/LenR0F3t/xcHpOJTAgZzhRFQNOAgaFPVWW48wc+vV7AaETg5q/o6fWhoc9/U8fRQC+GNiIiIgrR1QVUVChwOFQkJMiwdryKmKqeKDiJkx7XtOCtt1cGPX8I1DQFQgRDn/9rhQiGPkUZWOkLhj9/6Buq0hc8Fl+hj6GNiIiIAMiwVlmpoL5eDVTWwt19OZqhr69PVvuC1b/Q0KcogBBDh75g126w0ufv5pVBT8BikZ9HC0MbERHRONfVBRw9KsNaQoJARkb4w9poOd3QJ0Swi9ftBnp7lZAg6A9+Hg+waJGGzMzjXTP8GNqIiIjGqe5uGdbq6lQYjZGprMUaRQl2lQ5NhrSWFhneoimioa2+vh5bt25Ffn4+TCYTvv/97+OBBx6AYYha48svv4xXXnkF9fX1mDZtGh599FHk5eUBAJ566im8+OKL0A34H9br9cjPz4/YcyEiIopX3d1AdbWCmprxG9biUUSH323atAmpqanYu3cvXn31VeTn52Pnzp3HnPfWW2/hF7/4BbZt24Z9+/bh1ltvxfr169HR0QEAaG9vx80334yCgoLAjYGNiIjoxFwuoLhYwRdf6OBwKEhPF0hJ4bIb8SJioa2goACHDh3CfffdB6vVipycHKxfvx579uyBJuf/Bnz00Ue47LLLcO6558JoNOLaa6/FjBkz8MEHHwAAOjo6kJycHKmmExERxTWXCzh8WMHnn8uwZrMJpKYyrMWbiIW2wsJCZGdnw2azBY7l5uaivb0dVVVVx5yvDPpJslqtKCoqAgA4nU7s378fq1evxjnnnINbbrkFBQUF4X0CREREcaanBzhyRIa1hoZgWIunZS4oKGLfNqfTCavVGnIsJSUFANDW1hZy/JJLLsH777+Pffv2we124+OPP8bXX38Np9MJAJg4cSImT56M3/72t/j444+xYMECrFu3Dq2trZF5MkRE41i0B2PTyfX2AiUlMqzV1zOsjRVRnT0q+n/zB1fVrr76atTX1+P+++9HZ2cnVq5ciSuvvBLd3d0AgO3bt4ecf++99+Kdd97Bhx9+iB/+8IeRaTwR0TgihFwdv7FRQW2tCrNZYOJEgbQ0gcTEaLeO/Hp7gZoaBVVVKlQVSEsTDGpjSMRCm81mO6ai1t7eHrhvIEVRsHHjRmzcuDFw7M4778TUqVOHvLZOp0N2djYaGxtHt9FEROOcywU0N8tZhj09Ano9kJws4HYDxcUqhACs1mCAS0qKdovHp74+GdaOHpVhLTVVnGAJC4pXEQtt8+bNg8PhQGNjIzIzMwEABw4cgN1ux+TJk0POraioQFlZGVasWAEAcLvd2LdvH6677jp4vV5s374dN910E6ZPnw4A8Hg8qKqqOuY6RER06txuoK1NQXW1AqdTgU4HWCwCZnPwnMREIDFR9pb09gKHD8sAl5wskJPDABcpfX1Aba2CykoVisKwNtZFrGg6d+5c5OXlYceOHejs7ER1dTV27dqFNWvWQFEUrFq1Cl9++SUAoLGxEVu2bEF+fj7cbje2bduG9PR0XHjhhdDr9aisrMRjjz2GxsZGdHd348knn4TBYMCll14aqadDRDSm+HxAaytw8KCCzz7ToaBAhdcLZGQI2GxyW5/jMZkAu12u9SWEDHCff67DF1+oqKmRG4jT6Orrk3uD/t//6VBVpSI1VX6fGNjGtoiOadu5cye2bt2KFStWICkpCZdffjk2bNgAQFbXXC4XAGDp0qW4++67sXnzZnR1dWHx4sV4/vnnA4vpPvnkk3jiiSdw9dVXw+fzYd68eXjppZdgHvg2kIiITkgIuX2RwyHHqXk8gMkkTmsclMkkrwHIYFFSokLTALNZVuBsttCKHZ0atxuoq1NQUSG/QSkpDGrjiSLE2J4H1NTUGe0mEBHFFJdLbslTU6PA5ZLb91itJ9rG5/T19QFdXcoxAc5iCd9jjiX+sFZZKUMwu0Ejr6VFQW6uhqys8MamjIzjr0PLvUeJiMYBjwdobZVBzelUoKpynFp6emQePyEBSEiQL3ZuN1BWpqKkRI57G1iB42KvoTweoL5eQXk5wxoxtBERjVmaBjid8kXf4ZATBZKS5NizaDIaAZvt2ACXmKhg4kQN6ekMcAxrNBSGNiKiMWTwODWvV1a4YnW9rtAAJ1BZqaKsTMBkUpCTo8Ful12o4yXAeb3BsOb1yrCm5ys19eOPAhHRGNDTI8fcVFcPHKcWX9UZoxEwGoMB7uhRBWVlChISFEyaNLYDnNcrg3ZZmZwQwrBGQ+GPBBFRnPKPU6utVdDWpkBR5DppkRqnFk4ywMmPPR6ByspggPN3oSYnx3+AGxzWUlIEDIZot4piFUMbEVEc0TSgvV3OJIylcWrhZDAAaWnyY49HoKpKQUWFAqNRdqHGY4Dz+WRYKy1V4XbLyhrDGp0MQxsRUYzzj1NrapLbSXk8shsxVsephdPAAOf1BgOcwSArcBkZMsDF6v+Lzyf3by0pkd9Hq1UgJSXaraJ4wdBGRBSjenpk92dVVXCcWnIyxzr56fWhAa62Vu69aTAg5gKczydDd0mJrKwxrNFI8FefiCiGeDxy38/aWgWtrXKcWiTXU4tXej2QmgoAAl4vAgFOrweyszVkZgpYrZEPcD4f0Nwsw1pvrxyzZrVGtg00djC0ERFFmX+cWn29goaG4Dg1u13E1TitWDEwwPl8QEODgurqyAY4TQtW1vr6ZGUt+fgL3RMNC0Mb0Wno7gYSE2Oj+4XiT1eXHN9UUyO7zGJ5PbV4pdOhvxvy2ACXlSW3JBrNLbz8Ya2sTIXLxbBGo4uhjWiEqqsVFBerMBqByZPlu3fuo0gn09sbXE+tu1tWhSwWjm+KhMEBrqlJLkA8GgFO02Q3aFmZApdLQXKyQEbGaD8DGu8Y2ohGoKpKweHDKux2ASFkgKuoUGA2A1OmyG6thIRot5JihccDOJ1y30+OU4sNcvFhYHCA0+mCAS4l5eQBbmBY6+5W+tfJG7vLr1B0MbQRnaLKSjlOxW4PrjYvx8/IKkpRkezbysjQkJPDPQPHK/84tYYGOU5N04DERI5Ti0WDA1xzs4K6umCAy8yUEwgGztoVQlZMS0sVdHX5K2sMaxReDG1Ep2CowDaQyQSYTKJ/XS0F336rQK+Xyw9kZcXfAqB06vzrqVVXB9dTS03lOLV4MTjAtbQoqKtToKpAZqaGCRNkYCsrU9DZqcBiYVijyGFoIxoGIYCKCjm4OD395C/AigKYzYDZLP/w19XJtbaSkoDJk2W1JTExMm2n8DveODWupxbf5Lp4cm08TZNLsTQ0KBACDGsUFfyTQnQSQgDl5QrKy4cX2AbT6YLdp243cOSI7Cqz2wUmTZKzBfniHn+83uB6ai0tHKc21qlqMMARRQtfKohOQAigtFRBZeXIAttgRiMCkxdcLuDAARWqKteOys6WM9fYfRqb/N+zzk4FjY1ASwvHqRFRZDG0ER2HEEBJiVxVfTQC20CDu0/9a3WZTMCUKXID7KSk0Xs8Gpm+PhnSmpuBxkYVHo+AoigwmThOjYgij6GNaAiaBhw5IgeTZ2SEt4oycO0otxsoK1NRUiJnq02eLGCzCRgM4Xt8CvL5gM5OuTyHwyFnBQJyMoHZ7O/GZvcYEUUHQxvRIP7AVlMT/sA2mNEI2GwyFPT0AAcPqlAUYMIE2X2aksLdF0aTEHJXi44OBY2NCtra5CBzVZWVTq63RUSxhKGNaABNA4qL5RpN6enRHaeUmCjHS2maf9kBufuCv/uUuy+MTG9vsMuzqSnY5ZmYKCeFcGwaEcUqhjaifpoGHDqkwOGIfmAbSFWD60Z5vUBlpYrSUrmn4ZQpsvvUaIx2K2OX1yu7PNvaZJenyyW/sQkJAhaLf709VtSIKPYxtBFBjmUqKpIr18fy2kt6PZCWFtp9qqr+3ReGt+3OWKdpwS5Ph0OB0ylDmk7HLk8iim8MbTTu+XxAYaGKpiYlpgPbYP7uUyH8Y7Lkxtc5OXL3BYtl/Cwf0tMjuzybmmSXp88nuzyTkmQlcrz8PxDR2MbQRuOa1ysDW3OzErcVGLmoq1zY1ecDamoUHD0qN6/3775gMkW7laPL45Fdnq2tsju7t1ceN5kErFZ2eRLR2MTQRuOWP7C1tMRvYBtMpwPS0uTHfX3A4cMqhJBdgjk5cm2xeNx9QdPknp5Op5zl2dEhZ3kaDAKJiZyUQUTjQxz++SY6fV4vUFCgoq1Ngd0+NgLbYAkJcrC9f1mL775TodPJzesnTIjtzev9uw90dSlwOOTuA0IEZ3myy5OIxiOGNhp3PB4Z2Nrbx25gG2jw7gsNDXJjc5NJweTJGjIyYmPzerdbjktraZG7D/T1yeOJiQN3Hxj73y8iouNhaKNxxeOR+312diqBRWzHk+DuC4DbLQK7L6SlBTevj9TuCz6f7PJsa5Ndnp2dwd0HEhNlJZCIiIIY2mjccLtlYOvqUgLLZoxnA3dfcLlk9dG/ef2ECXL5kNHsgvR30/pneTY3y/F2Op2s9I2VcYVEROHC0EbjQl+fHNPV0wMGtiEkJQFJSXL3haYmuYVXQgIC3adm88iue6IN19PSuOE6EdGpiGhoq6+vx9atW5Gfnw+TyYTvf//7eOCBB2AYoj/m5ZdfxiuvvIL6+npMmzYNjz76KPLy8gAAQgg8++yzeOedd+B0OjF37lw8/PDDmDlzZiSfDsUJf2Dr7QVSU6Pdmtg2cPcFjweoqJC7L6Smys3r09JOvPuC1xvs8nQ4FHR3c8N1IqLREtH3uZs2bUJqair27t2LV199Ffn5+di5c+cx57311lv4xS9+gW3btmHfvn249dZbsX79enR0dAAAXn31Vbzxxht47rnn8Mknn2DRokVYv349+vwjl4n69fYC334rB7X7x3LR8BgMsvs0I0Nun3XwoIrPPtPh4EEFra1yGQ4hZEirrVWQn6/ik090+PprHY4eVaDTyS7P9HQBqxVxudQIEVEsiVhoKygowKFDh3DffffBarUiJycH69evx549e6BpWsi5H330ES677DKce+65MBqNuPbaazFjxgx88MEHAIDdu3dj7dq1mD17NpKSkvCTn/wEnZ2d+PTTTyP1dCgO9PYC+fkq3G5/9YhGyj/mLC1NwOlUkJ+vw2ef6fDJJzp89ZWKw4dlJdNmE/3ngfuhEhGNsoiFtsLCQmRnZ8NmswWO5ebmor29HVVVVcecrwwaAW21WlFUVITe3l6UlpZi7ty5gfsMBgNmzZqFgoKC8D0Biis9PTKweb0MbKNJVYHkZBngLBaB5GQBux2w2+W+nlw7jYgofCIW2pxOJ6yDXj1T+vur2traQo5fcskleP/997Fv3z643W58/PHH+Prrr+F0OtHe3g4hROBrB15r8HVofHK5gPx8HQNbmOl03JyeiCiSojrKRAg5IHlwVe3qq69GfX097r//fnR2dmLlypW48sor0d3dfdJr0fjmcgHffKMDIBjYiIhoTIlYaLPZbMdUwtrb2wP3DaQoCjZu3IiNGzcGjt15552YOnUqUlNToarqkNeaPXt2mFpP8aC7WwY2VeVelERENPZErHt03rx5cDgcaGxsDBw7cOAA7HY7Jk+eHHJuRUUF/vKXvwQ+d7vd2LdvHxYvXoyEhATMnDkzZPya2+1GcXFxYEkQGn+6uoCvv9ZBp2NgIyKisSlioW3u3LnIy8vDjh070NnZierqauzatQtr1qyBoihYtWoVvvzySwBAY2MjtmzZgvz8fLjdbmzbtg3p6em48MILAQBr1qzByy+/jCNHjsDlcuGZZ55BZmYmli1bFqmnQzGkq0tW2PT6kS8CS0REFOsiOqZt586d2Lp1K1asWIGkpCRcfvnl2LBhAwBZXXO5XACApUuX4u6778bmzZvR1dWFxYsX4/nnn4euf9TzjTfeiJaWFmzcuBHt7e1YsGABfv3rXw+5SC+Nbf4Km9EoZy8SERGNVYoY4yP4m5o6o90ECpPOThnYTCa5dyUREVG4tLQoyM3VkJUV3tiUkZF83Pu48x/FpY4OYP9+BjYiIho/GNoo7rS3y8CWmMjARkRE4wd3A6S44nTKSQdms4DJFO3WEBERRQ5DG8WNtja504HFIpCQEO3WEBERRRa7RykutLbKChsDGxERjVestFHMa25W8O23KlJSBIzGaLeGiIgoOhjaKKY1NSn47jsGNiIiInaPUsxqbGSFjYiIyI+VNopJDoeCAwdUpKUJcKMLIiIihjaKQQ0NCgoKGNiIiIgGYmijmFJfr6CwUIXNJqDnTycREVEAx7RRzKirU3DwoKywMbARERGFGnZoa2lpCWc7aJyrqVFQWKjAbmdgIyIiGsqwQ9vFF1+MH//4x/jv//5v9PT0hLNNNM5UVysoKlJhtwM6XbRbQ0REFJuGHdr27NmDefPmYdeuXVi2bBnuu+8+fPbZZxBChLN9NMYdPaqguFiF3S4Y2IiIiE5AESNIXaWlpfjwww/x3nvvoa2tDatXr8Z1112HGTNmhKONp6WpqTPaTaDjqKxUUFLCwEZERLGvpUVBbq6GrKzwFqsyMpKPe9+IJiLMmDED55xzDpYuXYq+vj588MEHuPnmm7Fp0ya0traOuKE0PggBVFQwsBEREZ2KUwptR44cwY4dO3DJJZdg/fr16O7uxrPPPouPPvoIH3/8MZKTk3H//feHq600BggBlJcrKC1VkZ7OwEZERDRcw+4e/cEPfoCSkhKcc845uOaaa3DppZciKSkp5Jy+vj6ce+65+O6778LS2JFg92js8Ae28nIZ2FQuOENERHEiFrpHh724wmWXXYZdu3Zh4sSJxz0nISEBv//970+tdTQuCAGUliqorGRgIyIiGolhv3SuW7cOv//97/H5558Hjr3++uvYunUrXC5X4NjixYtHt4UU94QASkoUVFSoyMhgYCMiIhqJYb98Pv744zh48CAyMzMDxxYsWICysjL8/Oc/D0vjKP5pGnD4sIKjR1VkZgooSrRbREREFJ+G3T368ccf4/3330dKSkrg2KxZs/Dss8/i8ssvD0vjKL75A1tNjaywMbARERGN3LBDmxACmqYdc7y3txcej2dUG0XxT9OA4mIFdXUMbERERKNh2KHt0ksvxR133IHbbrsNEydOhKZpqKiowO9+9zusXr06nG2kOOPzAUVFChoa5KQDBjYiIqLTN+wlP3p7e/H000/j7bffRkdHBwDAarXi2muvxb333guDwRDWho4Ul/yILJ8POHRIhcOhICODW5wREdHYEAtLfoxoG6u2tjYoioLU1FQActHdWbNmjbyFYcTQFjk+H1BYqKKpSUF6OgMbERGNHbEQ2k558YXm5mb09PTA5XKhrq4O+fn5+OEPf3haDaT45/UCBw8ysBEREYXLsMe0ffvtt7jrrrvQ2Nh4zH3Lli0b1UZRfPF6gYICFW1tDGxEREThMuxK289//nOsXr0ar7/+OvR6Pd588008/vjjuPjii/Hkk0+Gs40Uw/yBzelUYLczsBEREYXLsCttZWVleO2116CqKhRFwZw5czBnzhxMmjQJDz74IJ5//vlwtpNiVEmJgrY2BjYiIqJwG3alLSkpCZ2dclC/2WyGw+EAACxZsgRfffVVeFpHMa2pSS6ca7MxsBEREYXbsEPb8uXLcfPNN8PlcuGcc87Bfffdhz//+c946qmnYLfbh3WN+vp6bNiwAUuXLsX3vvc9bNu27bgL877yyiu47LLLkJeXh5UrV+L555+Hf6LrU089hblz52L+/PmB28KFC4f7VGgU9PUBRUUqUlO5DhsREVEkDDu0/exnP8OVV14Jk8mEn/3sZ/B4PLj//vvx2Wef4fHHHx/WNTZt2oTU1FTs3bsXr776KvLz87Fz585jzvvrX/+Kp556Ctu3b8c333yDZ599Fi+++CJef/11AEB7eztuvvlmFBQUBG75+fnDfSp0moQAjhxRIISA0Rjt1hAREY0Pwx7T1tjYiI0bNwIAsrKy8Oqrr57SAxUUFODQoUN44YUXYLVaYbVasX79ejzyyCPYsmULVDWYHw8cOICZM2cGqmdz5sxBXl4eiouLAQAdHR3IyMg4pcen0dPYKHc7yMxktygREVGkDDu0/f3f/z32798fEq5ORWFhIbKzs2Gz2QLHcnNz0d7ejqqqKkydOjVw/OKLL8aLL76IL774AosWLUJpaSkOHDiAW2+9FQDgdDqxf/9+rF69Gg0NDZg9ezbuv/9+zJ8/f0Rto+Hr6Ql2ixIREcUTTQh4NA1unw9u/78+LeRjj+Y/FnqfcOsxa84kANEbEzTs0LZmzRrs3LkT//RP/wSLxXLKD+R0OmG1WkOOpaSkAJA7LAwMbXl5eXjwwQdx2223wev1QlEUbN68GRdeeCEAYOLEiVBVFTt27IDZbMavfvUrrFu3Dh9++GFIKKTRJQRw+LAKVWW3KBERjQ4hBHxCDAhSAwJT/zFPSIgKvc9/zDNE0PKHsL7++7zayAoOelVBisGEPm82TiE6jbphP/Jf/vIXNDc344UXXoDFYoFOpwu5//PPPz/lB/dPLFAGjWT/4osvsGPHDvz2t7/FokWLUFBQgM2bN+PMM8/EFVdcge3bt4ecf++99+Kdd97Bhx9+yN0Zwqi+XkFTk8Ju0X69Xi/6fD5YjcZjfoZp7NKEQGtvL3q8PmhCyBsEhEDgc4Hgx/KG/nPkxwKD7gtcI/i5PCf0OgOvO9RjBK8r/75qCH2M4DGEfBy4phDQBj5Gf7tDPu9/bFUBVEWBAiX4saJAHfC5/6ZgwP0K+s/x3xD4OsX/dQjeN/C6ysDrIvix/7rKwK8b8BgDrxt6jQGPP+C6ihJ6nZDP+5/PeOXTBPo0DZ6BlSp/cPL54NE09A0VtAad7/Fp6NOCoWokWUoBYNSpMKo6GHUqDP3/Juh0MOl0sBoNgfuC56gw6nRIGHB+4Jz+jw06FQmDzlcVBS0tCiwJGoDovQYOO7Tdfvvtp/VANpsNbW1tIcfa29sD9w20e/duLF++HOeffz4AuazI6tWr8eabb+KKK6445to6nQ7Z2dlD7tZAo6O7Gygu5vIeAODVNPy1pg7vlFfC5fXCYtAjx2JBjsWMSRYzcixm5JjNMOmj926MRkeX24Oari7UdnWjtrsbNV3dqO3qRp/PF5X2KIAMHQgNK4ODjoJBYWWogILQsKJXlWMCijIoPPnvV4DQoDoocGqDw2F/JUXTtBEH2SFDZlS+C3QyelUJCUsyBMnwYzUaYdT575PHjKoKQ/+/CTpdf1AK3uc/3xByvw76/t+B8WTYryrXXHPNaT3QvHnz4HA40NjYiMzMTABywoHdbsfkyZNDztU0DZqmhRzz9f+R9Hq92L59O2666SZMnz4dAODxeFBVVXXMdWh0aJoMbEajwHjOIUIIfNfcgv8qKYPD1YOzbGlYkG5DXZcLtV1d+FtdPfp8wZ/b9EQTJpn7Q5zFgkkWM7KSEqEb4bhQCh+Pz4e6bld/KJMhraarG+1ud+Acs0GPHLMZy7InIMdihsVoGDIIDa7+qIMD1MkC1eDK1YDrjrcXqJMRAyqOg6uEx1YfB1cOj60+DlXhHOq645lOUQaEruNXpSg8hv0SfNddd53w/qGW7hho7ty5yMvLw44dO/Dwww/D6XRi165dWLNmDRRFwapVq7B161YsXboUy5cvx+OPP45/+Id/wMKFC1FUVIQ///nPuOeee6DX61FZWYnHHnsMTz/9NMxmM/7t3/4NBoMBl1566XCfDp2C2lq560FGxvj9Y1Xd2YU9JaUoanViQlIiNp89DwvS7SEvopoQaOnpRU1XV6AiU9PVje+aWwIVAb2iINuchEkhlTkLUhPYxRoJmhBo7ulF7aDvUWOPK9A9o1cVTDSbMdeeFvj+TLKYkcJu8Jjjr/oxJNB4MezQlpSUFPK5z+dDVVUVqqqq8IMf/GBY19i5cye2bt2KFStWICkpCZdffjk2bNgAAKioqIDL5QIgq3odHR14+OGH4XA4kJmZiXXr1uH6668HADz55JN44okncPXVV8Pn82HevHl46aWXYDabh/t0aJi6uoAjR8Zvt2h7nxtvl1fg09p6JOr1+OGsGfi7SROhH6JapioKMpISkZGUiIWZwSVpPD4f6v1VnO5u1HR1oaitDZ83OALnJOn1ga5Vf6DLsZiROJ5Lm6ep0+0eEMxk9ayuuzukGpqRaEKOxYwlWRmBEJ2ZyGooEcUmRYjTq/W+++67yM/Px0MPPTRabRpVTU2d0W5C3PL5gK+/VuH1AuMtD3t8PvyluhbvVhyFR9NwyaSJuGraVFgMhlF7jC63B7Xd3YGqz1DjpeymhEClxx8qspKShgyN45W7v2sz9P+xCx3u4G4rFoMhON6wPxhPNCdx3CERDVtLi4LcXA1ZWeEtYmRkJB/3vtMObT6fD+eddx727dt3OpcJG4a2kausVFBaqo6rblEhBL5ubMLrpZNfSf4AACAASURBVOVo7unFgnQ7bpg5HRPMSSf/4lGgCYGW3t5At51/fFWDK9h9p+vvYh1YlZtkMSMtIWFMd99pQqCppycQbv0hrdHVE+h+Nqhqf/dzsFtzksXMGb5EdNpiIbQN+21mT0/PMcd6e3vx/vvvw8hFu8acjg6gpESF3T5+AltlRwf+eKQMJc525FjM2LJwAebaI7vun6ooyEhMREZiIvIy0gPHPZqGhm5XoJuvpqsbR9ra8WVDcMZ0kl4fqCQNHC+XFIfVpA63GzWdwXFntd3dqOvqhrt/gpICICMxETkWM87NygwE2MykRI5vIqIxa9h/zRcuXDjkO1VVVfHTn/50VBtF0eX1AoWFKiwWgUHL8Y1Jbb19eKOsHJ/XO5BsMODWObNwUU52TL34G1QVk5MtmJwcurB1t8cTUpWr6erGl/UO9AzoYrUlJMhQkxzsZp0QI12sfT4f6gZMCKjtls+hc0DXZrLBgByLGRfnZAeqZxMtZiSMhx9OIqIBht09+tVXXx1zLCEhAZMmTYLdbh/1ho0Wdo+eupISBVVVKtLTx3aVrc/nwwdHq/F+ZRU0IbBiyiRcceYZcVmZGkgIgdbevpCqnL+L1df/665TFExISgpZW26SxQKbKTxdrJoQaHT1HNOmpp5g16ZRVTGxf427gZXClARW8oko+mKhe/SUxrQdPHgQKSkpgfXQDhw4ACEEzj777NNvZZgwtJ0apxPYv18Hu10gBgoxYaEJgS8bHHijtBxtfW4szszAP8yYhoykxGg3Lay8moYGlwu1Xd2o7uxfMLarG619fYFzEnU6TBwQmPyBzjzMCRhCCNm1Oaj6V9/tgmdA12ZmUmIwnPVXADMS2bVJRLErFkLbsEsKb731Fh555BE8++yzgdBWXV2Nhx56CI8++iiuvvrq028pRZXHAxw8qIPFMnYDW4mzHX88UorKjk6ckWzBP82bi1lpqdFuVkToVRWTLBZMsliwdEJW4LjL4wlZ7b+2qxtfORrRU1sfOCfN38U6qCrnGFA981fQujzBrk2r0YAciwV/N2li4OuzzezaJCIaiWFX2i677DL8y7/8C84555yQ49988w0eeOABfPDBB2Fp4OlipW34iosV1NePzTXZmnt68HppOfY7mpCaYMQ106fh/OwsVnaOQwiBtr6+Y9Y5q+8OdrEO5O/anDRoRmsyJykR0RgRV5W2xsZG5OXlHXM8NzcXDodjiK+geNLSoqCmRhlz49h6vF78ubIKe6uqoULB6jPPwKqpU1jpOQlFUWAzmWAzmbAgPThm1atpcLjkQsGtvX3ISpIzONm1SUQUfsMObXPmzMF//Md/YN26dVD7+87cbjd+9atfYfbs2WFrIIVfX5+cLWq1CoyV111NCHxWV4+3yirQ4fbgvAlZuHbGmbCZTNFuWlzTqypyLBbkWCwnP5mIiEbVsLtHCwsLsWHDBvT19SErKwtCCNTW1sJiseCll17CtGnTwt3WEWH36MkVFipoblaQOkaGdhW1tmHPkVJUd3VjeooVN86agWkp1mg3i4iI4lgsdI+e0uzR3t5efPbZZ6iqqoKiKJg8eTIuvvjimF5cl6HtxBwOBQUFcnmPeK+yOVwu/NeRMnzb3AK7KQHXzZiOc7IyuBI+ERGdtlgIbae0IFVpaSlmz56NFStWAJBLfhQVFcX0kh90fL29QHGxipSU+A5s3R4P/rv8KD6uqYVeVXHt9DOxcsokGDhujYiIxpBhL+zw1ltv4eabb0Z5eXngWHV1Nf7xH/8Rb731VlgaR+EjBHD4sApFEYjhQukJeTUNH1XX4MH/+xL/U12DC7In4OcXnIsrzjyDgY2IiMacYVfadu3ahd/97nchS35ceeWVyM7OxgMPPMB12uJMQ4OCxkYFmZnxN1tUCIGCllbsOVKGBpcLc9JSceOsGcds8URERDSWcMmPccjlAoqK4nM9ttqubuw5UorC1jZkJibiJwvmIS/DznFrREQ05nHJj3FG02S3qMEgEE9bbHa63XirrBKf1NYhUa/HDTOnY/nknJjY9JyIiCgSuOTHOFNdreDwYRUZGfFRZfP0j1v7U/lR9Gk+/F1ODlZPO4Mr7RMRUUTF1ezR3Nxc7N27N2TJjylTpuCiiy7CunXr8Morr4xKYyl8urqAkpL46BYVQiC/qRn/VVKGpp5ezLPbcMPM6ZhoMUe7aURERFEx7NDmcrnwm9/8BgcPHoTb7Q4cf/rpp9HR0RGWxtHo0TS5vEdCgkCsT6ys6ujEa0dKccTZjonmJNy9cAHm2W3RbhYREVFUDXtA0NatW7F3715MnToV33zzDc466yxomobExET8/ve/D2cbaRRUVSlob1cQy7sPOfv68GJhMR7/6mvUdbuwZs5MPLp0CQMbERERTmFM2wUXXIA//elPsNlsWLBgAQ4cOAAAeO6552AwGHD77beHtaEjxTFtQGcn8OWXOthssVllc/t8+PBoNd47WgWvJrBiyiRcOXUKkgyGaDeNiIgIQJyNafN6vbDZZMVDr9ejr68PCQkJWLt2LVatWhWzoW288/nkZvBmc+wFNiEEvnI04v+VlKO1rw8LM9LxDzOnISspKdpNIyIiijnD7h6dPXs2nnnmGXg8Hpx55pnYvXs3AKCyshJ9fX1hayCdnspKBd3dCmItB5W1t+OJ/fl44WARLEYD/nnR2fjJ2fMY2IiIiI7jlJb8uOeee/D222/jb3/7G+6++27o9Xq43W6sXbsW999/f7jbOiLjuXvU6QT27dMhPV0gVpYza+ntxf8rKcdXjkakGI24ZsaZuCB7AlQujktERDEsFrpHhx3aBisvL0dRUREmT56MBQsWjLhx4TZeQ5vHIwObqgokJka7NUCv14v3KqvwYVUNAODSKZNw+dQpMMXTCr9ERDRuxUJoG/Er5rRp02J2QV0CyssV9PYCdnt026EJgf+rb8CbpRVod7txblYmrps5DXaTKboNIyIiijMsc4xBra1yiY9o73pwuM2JPx4pRVVnF6ZZk7Hx7FxMT0mJapuIiIjiFUPbGON2AwcP6pCSIhCtYWKNrh78V0kZ8puaYUtIwD/NOwvnZmVyU3ciIqLTwNA2xpSWKtA0ICEh8o/t8nrxbsVR/KWqBnpVwdXTpuLSMybDGGtrjRAREcUhhrYxpLFRQW1t5DeD92kaPqmrx9tllej2eHBB9gRcM+NMpEYjORIREY1RDG1jRG8vUFSkIjU1st2ihS2t+OORUtR1uzArNQU3zpqBM6zHn/lCREREI8PQNgYIARw5ogIQMBoj85geTcOeI6X4uKYOGYkm3LEgF4sy0jlujYiIKEwiuuRqfX09NmzYgKVLl+J73/setm3bBo/HM+S5r7zyCi677DLk5eVh5cqVeP755+FfUk4IgV/+8pdYsWIFlixZgh/96EcoKSmJ5FOJKQ6HAodDQWpqZB6vydWD7fu+wcc1dVg5ZRK2nX8uFmdmMLARERGFUURD26ZNm5Camoq9e/fi1VdfRX5+Pnbu3HnMeX/961/x1FNPYfv27fjmm2/w7LPP4sUXX8Trr78OAHj11Vfxxhtv4LnnnsMnn3yCRYsWYf369eNyO62eHtktmpYWmXFs+Y1N2PbVfjT19OInC+bhxlkzYIiV7RaIiIjGsIi92hYUFODQoUO47777YLVakZOTg/Xr12PPnj3QNC3k3AMHDmDmzJlYuHAhVFXFnDlzkJeXh+LiYgDA7t27sXbtWsyePRtJSUn4yU9+gs7OTnz66aeRejoxQQiguFiFXi9gMIT3sbyahj8eKcVzBwqRlZSEh5cuxsLM9PA+KBEREQVELLQVFhYiOzsbNpstcCw3Nxft7e2oqqoKOffiiy9GaWkpvvjiC7jdbhw6dAgHDhzAJZdcgt7eXpSWlmLu3LmB8w0GA2bNmoWCgoJIPZ2YUFenoKVFgdUa3sdp6e3Fk/vzsbeqBssn5+D+JQuREQt7YxEREY0jEZuI4HQ6YR2ULlL6V8dva2vD1KlTA8fz8vLw4IMP4rbbboPX64WiKNi8eTMuvPBCOBwOCCECXzvwWm1tbWF/HrGiuxs4fDj83aIHmlvwu8Ii+DSBDfPnYklWZlgfj4iIiIYW1dmj/okFgwewf/HFF9ixYwd++9vfYtGiRSgoKMDmzZtx5plnYvHixSe81nigaXIcm9EoEK791n2ahjfLKvD+0WpMtpixYUEuspKSwvNgREREdFIR6x612WzHVMLa29sD9w20e/duLF++HOeffz4SEhKwZMkSrF69Gm+++SZSU1OhquqQ1xp8nbGqpkaB06kgOUzLobX19mHHN9/h/aPVuDgnGw+cs4iBjYiIKMoiFtrmzZsHh8OBxsbGwLEDBw7Abrdj8uTJIedqmnbM5ASfzwcASEhIwMyZM0PGr7ndbhQXFyMvLy+MzyA2dHUBJSUqbLbwVBYLW1qx7cv9qOrsxI9zz8KPzprNbaiIiIhiQMRC29y5c5GXl4cdO3ags7MT1dXV2LVrF9asWQNFUbBq1Sp8+eWXAIDly5dj79692LdvH7xeLwoKCvDnP/8ZK1euBACsWbMGL7/8Mo4cOQKXy4VnnnkGmZmZWLZsWaSeTlT4fEBhoYrERIHRzlGaEHi7rAL/ln8AyUYDHjp3Mc7LzhrdByEiIqIRi+iYtp07d2Lr1q1YsWIFkpKScPnll2PDhg0AgIqKCrhcLgDANddcg46ODjz88MNwOBzIzMzEunXrcP311wMAbrzxRrS0tGDjxo1ob2/HggUL8Otf/xqGcK97EWVHjyro6lKQnj66Vbb2vj68cLAIxW1OXJA9AWvmzEQCq2tEREQxRRFjfAR/U1NntJswKtrbga++0iE9XWA017Itbm3DCweL0OP14uY5M3HhxOzRuzgREdEY0dKiIDdXQ1ZWeGNTRsbxB6xz79E44PUChYU6JCePXmDThMCfK6vwdlkFspIScc+iBZhksYzOxYmIiGjUMbTFgYoKBb29gN0+OtfrdLvx28IiFLa0YemETNw6ZxZM4Vo7hIiIiEYFX6ljXFubHMtmt49OObbE2Y7fFBSi0+PBrXNm4eKcbG70TkREFAcY2mKYxwMcPDg63aKaEPjwaDXeKCuH3WTCg0sWYYo1TAu9ERER0ahjaIthpaUKfD7AZDq963R5PHixsBjfNbdgcWY61s6dgyR2hxIREcUVvnLHqKYmBbW16mkv71He3oFfFxTC2efGTbNmYPnkHHaHEhERxSGGthjU1yf3Fk1JERhpvhJC4H+qa/FfJWVITTDi/iULMS3FOroNJSIioohhaIsxQgBHjigQQsBoHNk1XF4v/lBYjG+amnF2uh3rcufAMsYXHiYiIhrrGNpiTGOjgoYGFZmZI+sWPdrRiecLCtHS24vrZ07HpVMmsTuUiIhoDGBoiyE9PbJbNC3t1AObEAL/W1uH1w6XItloxH2LF2JGakoYWklERETRwNAWI4QADh9WoaoCp9qT2ev14j+KjuArRyPm2W24LXcOkkfat0pEREQxiaEtRtTXK2hqUk65W7Smqwu7DhSi0dWDa6efiVVTp0BldygREdGYw9AWA7q7geJiFTbb8AObEAJ/q2vAK4dLkKTX495FZ2OOLS2MrSQiIqJoYmiLMk2Tgc1oFBjuerd9Ph/+s/gIPq934CxbKn6cOxcpCewOJSIiGssY2qKspkZBW5uCjIzhVdnqurvx/IFC1He7sPrMM7B62lR2hxIREY0DDG1R1NUFlJQMv1v08/oGvFx0BAk6He5euAC5dluYW0hERESxgqEtSnw+4NAhFYmJAjrdic91+3x47UgpPqmtx6zUFNw+fy5SExIi01AiIiKKCQxtUVJdraCj4+Tdog3dLvy6oBDVXd24YuoU/GDaVOhUNUKtJBpfhJDjTH0+wOuV/8qbAiHQv62cgOj/tVUUBcnJI9+9hIjoVDC0RUFHh+wWtdtPHNj2ORrx0qHD0KkK7sybjwXp9gi1kGhsEGKoAAZo2uAQpgT2+TUYAJNJwGyWYSwhAUhIkOsn6vVywpBOJ8NdW5uCmhoFHR2AqiqwWBjgiCh8GNoizOsFCgtVWCzH7xb1aBr2HCnFxzV1mJ5ixfr5c2EzmSLbUKIYNLAK5v/4RCFMUWToMhoFEhNFfwBDYFjCwBCm18vbqczrsVgEJk8W6OoCmpsV1Naq6OiQ1zSbccoLZVNs6+kBenrkz1pKyvBn/BONFv7IRVhFhYKeHuW4VbYmVw+eLyjE0c4uXDplEq6dMQ16dofSGDWw+jW4K9JPCBHYP1enkwHMH8QSEgCTCYElc/R6EQhgBgNOOl50tFgsMsCdcYYPXV1AS4uCmhoV7e2yPcnJ4At8HNK0YFADZFCbNUuDzwdUVKjw+RjeKLL4oxZBTidw9Ojxu0XzG5vw+0PFUKDgJwvmYWFmeoRbSDRy/q7I44WwgePB/CFMr5ddj/7gZTLJrkl/APNXwWTXJBDr718URQa05GQZ4Do7gaYmWYHzeORzSk6OXJikU+fzyQXP3W5ZrU1P1zBjhoDVKn8+/bKzfaivV1BerkLTgNTUk08qIzpdDG0R4vEABw/qkJwsjnnh8Woa/l9pOfZW1WCqNRnr589FRmJidBpKMUuI4O1UPh/euUrg81Nb9i+0K9JgkCHMbA52RSYkhIawgbexvMSgogBWK2C1Cpx5pgxwDoeC+noVXq8MqWYzA1wscLtlUPP5FOj1QFaWhsxMDVbr8SukBgMwZYpAdrYPtbUKKisZ3ij8GNoipKxMgccj34UP1NLbi18XHEJ5eweWT87B9TOnwxDr5YRxzueTIdztDr4bBxDSpQccL5AETwqer5zg/NDrqWrwX/mxCHwcep8IBKnQ83Hc8wce8z9e6E0cc0xVgwFMpxvbIex0qCqQkiK70mbM8KGjA2hoUOBwyAqcDLoMcJHU0wO4XAqEEEhMVDB1qgabTYPFcmoVXYMBmDpVYOJEH+rqZHjzj3nj95NGG0NbBDQ3K6iuVo9Z3uNAcwt+V1gEnyawYf5cLMnKjFILaSjBYCbfgSuKAKD0j1ESsNkELBYRCDonCzj+G3CiYHT8G40NqgqkpsqKzMyZPrS3ywqcwyHHSPkDHN+7jS7/+DSXS4Gqygro7NkaUlPl//fpMhpDw1tFhfwGMrzRaGJoC7O+PrmIbkqKCLzw+jQNb5VX4r3KKky2WLBhwVxkJSVFt6HjlKYFw5nH4+8ilAPfTSbAYpFdJGazgMnkHwAf7VbTWKHTATYbYLP5A5yChgbA4ZDVGrn0CEP7SA0cn6aqgN0ux6elpMjf5XDwh7eB3ab+kMjwRqeLoS3MSktl+d3/B6Kttw+/OXgIJc52XJyTjR/OmgEjf5PDzucLVs28XvkKKISsZlgsAunp/oHGwRmJ/LZQJOn1gN0uYLcDs2bJAFdXBzQ3ywCXmCiQlMQAdzLB8WmAwaAgM/Pk49PCISEBmDZNICcnGN4UhZU3Oj0MbWHkH3Scni67RQtbWvHbg0Vwaz78OPcsnJedFeUWji1ChI410zQEujT1eiAlRUNmZmjVLCGBL4IUewwGID1dID0dcLt9cDoV1NcraGmRP9dms0BiIn92/YLj04CkJIGpU/3DF6LfzewPbxMn+lBTo6CqSu0f43jspDSik2FoC5PeXqC4WHaLCgi8U1aJP1UcRbY5CXcsyEP2aAyiGKf8VTOPR3ZpAsFZj2azQFqarJolJgaXk+A6ShSvjEYgM1MgM1Ogrw9wOhXU1ckABwQD3HgyeHxaSsrojk8LB5MJmDFDYNIkH6qrZXjT6wWs1ugHS4offCkLAyGAw4dVKIpAj+jDC98UobjNiQuyJ2DNnJlIYG18WE42EcBul+EsISFYNeMfPxrLEhKArCyBrCyB3l4Z4GpqFDQ3K4E3LWN185TB49P866eFc3xaOJhMwMyZMrzV1MhJajodwxsND0NbGDQ0KGhsVNCqb8ULXxahx+vFP86djQsnZke7aTGHEwGIRsZkAiZMEJgwQaCnJ7gPqj/AWSzxFWaG4nYDLpfctsxgUDBhgob09MiPTwuHxMRgeKuqkjtosPJGJxPnP/axx+UCCg8p+KKzEv9dUYGspETcs2gBJlks0W5aVHm9oRMB/GNx/BMBMjJCJwIkJHAiANFwJSbKiQoTJwq4XEBrazDA+X/H4uHNjhByaMnA8WlnnBE749PCITERmD1bYPJk2W1aXa3CYBBISeGYRTpWRENbfX09tm7divz8fJhMJnz/+9/HAw88AMOgXZUfeughvP322yHHNE3DokWL8PLLL+Opp57Ciy++CN2AV3W9Xo/8/PyIPI/j0TRgf4EXf6gowuH2NiydkIlb58yCKd7fEg7TiSYCGI2A1aphwoRjq2b8w0Q0epKSZNiZNEmgu1vug1pbq6CjA9DpFJjNsRXgNE2+2e3tlX8IUlMF5szRkJISu+PTwiEpKRjeqqrk1md6PcMbhVKEGLyOe/hcd911mDlzJh588EF0dnZi06ZNWLZsGf75n//5hF8nhMBNN92Ea665BjfeeCMeeughmEwmPPTQQyd9zKamztFq/kl9mN+Bf/20CC6fGzfNmomLc7IDeyzGE59PBjBNC978+0oGj4dOABg4ESA5WZb4ORGAKDYIIceDNTfLbri+PrnshMUiZ6pGmtcr2+PxBMenTZiA/jGqkW9PLOruRiC8GY3yb2ocvpyMKS0tCnJzNWRlhTc2ZWQkH/e+iL2UFhQU4NChQ3jhhRdgtVphtVqxfv16PPLII9iyZQvUE9S9X3/9dXg8Hlx//fUAgI6ODmRkZESq6SelCYHf/60GL3xZCXtiAu5euAhTrMf/Tx9Ng8PVwJAV/PzYfSVl0BL911BC/hj4N+jW6UT/Rt3BjbuNRv+/crFgnU7ez4kARLFLjnGT3aRnnOFDV1cwwLW3y99viyW8b67k+mkKvF4Bo1GOT8vI0JCczDd1QzGbgbPOClbe6upUJCQIJCczvI1nEftVKSwsRHZ2Nmw2W+BYbm4u2tvbUVVVhalTpw75dT09PXjmmWfwy1/+MhDsnE4n9u/fj9WrV6OhoQGzZ8/G/fffj/nz50fiqYTo7PXi0feK8Wl5K862peO2BXOQdIK/QCOpYklDByxFQWAzboPBH6r8YUveDAb5R9m/t6ROF7pfpU4X+i8RjV2KIvdATk4WmDpVbmTf2ChDgdst35BZLKc/pnSo8WlTp2qw2+X1GTyGx2IB5s4VmDLFh6NH5dqfDG/jV8RCm9PphNVqDTmWkpICAGhraztuaNu9ezemT5+OJUuWBI5NnDgRqqpix44dMJvN+NWvfoV169bhww8/DAmFkfBekQOfV7RhVeZMfC87Bz3tCnoAAKJ/Q/Bjq1j+kDUwUA2sYhkMYkCIEoEwNVTI4i8tEY2UogBWq+yWnDZNbmTf2CiDgdcrA9ypbGQ/1Pi0s86S49O4U9/psViA3FxZKa2oUNDQoMJkYngbb6JalPZ3zx1v3JemaXjxxRexbdu2kOPbt28P+fzee+/FO++8gw8//BA//OEPw9PY4/jB/GycNzEDJiWhv4KlnTBkERHFooEb2c+YEdzIvqFBbmTvr8ANrsYPNT5t1iy5flosTXgYKywWYP58gTPP9KG8XEFjo6y8DaqJ0BgVsdBms9nQ1tYWcqy9vT1w31D279+Prq4uLFu27ITX1ul0yM7ORmNj4+g09hQk6FVMyTLC331JRBTvVBVISwPS0vwb2cv1Jx0OFZomJxh5vQp8PhFYPy0jQ66fxjenkWGxAAsWCHR2+lBRoaKpSY4rTo7McGqKkoiFtnnz5sHhcKCxsRGZmZkAgAMHDsBut2Py5MlDfs1f/vIXnHfeeTAOeLvm9Xqxfft23HTTTZg+fToAwOPxoKqq6rjXISKikdHpAJsNsNlEYCN7h0OOUeP4tOhLTgYWLNDQ0QGUl8vwlpgoQx2NPREbdj537lzk5eVhx44d6OzsRHV1NXbt2oU1a9ZAURSsWrUKX375ZcjXHDp0CJMmTQo5ptfrUVlZicceewyNjY3o7u7Gk08+CYPBgEsvvTRST4eIaNzR6wG7XWDuXLkpO8dTxQ6rFcjL03DuuRrMZqC5GejqinaraLRFdK7gzp070dXVhRUrVuBHP/oRvve972HDhg0AgIqKCrhcrpDzm5qahlza48knn8SECRNw9dVXY/ny5SgvL8dLL70E83haiZGIiGgQf3hbskRDUhLQ1CTHHNLYENHFdaMhkovrEhERxRKnU3abtrYqSEoaX7tMjCaPR24Pd/bZ42RxXSIiIoqs1FRg0SINTidQVqaiqYnhbSj+bRj9WzH6fErIbj8mEzBhggazObp1LlbaiIiIxgEhEAhvTqfch3Y8rZ/n8wWDmccTuj+2oqA/zMrJNUlJcpa00Rj53X5OVGljaCMiIhpHhADa2mR4a28fW+HN55OVMn8wk2Qw0+nQH8pkMDOZgsHMaIydSTXsHiUiIiIAMpzYbEBamhYIb42NCpKTBRITo926ExNCLujsr5h5vUr/7kMCiqLAaJR77MrlaERItcxgiHbrTx9DGxER0Tg0MLy1tiooLVXQ1KTAYolueNO0YDA73vgyi0VDcrLs0hwYzMb64s4MbUREROOYosj192w2gdZWBSUl4Q9vwxlfZrXKbkyzWYR0Y0ZyfFmsYWgjIiKikPDW0iLDW2OjAqtVwGQ69esN7MY83viylBTZjekfX+bvxoyV8WWxhqGNiIiIAhQFSE+X4a25WUFZmay8JSeHhreBy2R4PLIbUx4fH+PLooGhjYiIiI6hqkBmpkB6ugxvpaVyb1NVDY4vS0yUgWw8ji+LBoY2IiIiOq5gePOhvV2GMX8wYzdmZDG0ERER0UmpKpCWFu1WjG/jeA4GERERUfxgaCMiIiKKAwxtRERERHGAoY2IiIgoDjC0EREREcUBhjYiIiKiOMDQRkRERBQHsLFj0AAADHBJREFUFCGEiHYjiIiIiOjEWGkjIiIiigMMbURERERxgKGNiIiIKA4wtBERERHFAYa201BbW4vNmzdj6dKlOO+883DXXXfB4XBEu1kR8e233+KWW27BokWLsGzZMmzZsgVNTU3RblZU/PznP8fs2bOj3YyIueCCCzBv3jzMnz8/cHv00Uej3ayI+t3vfoeLL74YeXl5uPnmm1FaWhrtJoXdvn37Qr7n/tvs2bNRW1sb7eaFXVFREdauXYtzzjkH559/Pu68807U1dVFu1kRc/DgQaxduxZLlizBhRdeiOeffz7aTQqrw4cP46qrrsLy5ctDjn/11Ve44YYbsGjRIqxatQq7d++ObMMEjdhVV10l7r33XtHZ2Smam5vFj370I3H77bdHu1lh53Q6xcKFC8Uf/vAH4Xa7RXNzs7jlllvEHXfcEe2mRdyhQ4fEueeeK2bNmhXtpkRMbm6uOHjwYLSbETW7d+8WK1euFIcPHxZdXV3i6aefFvfee2+0mxUVu3fvFjfeeKPQNC3aTQkrj8cjli1bJp566inR19cnOjo6xObNm8VNN90U7aZFhNPpFEuXLhX/+q//KlwulygtLRUrV64Ub775ZrSbFhbvvvuuuPDCC8XGjRvFJZdcEjje2NgoFi5cKF555RXR09Mjvv76a7Fo0SLxv//7vxFrGyttI9TR0YF58+bhpz/9KSwWC+x2O2644Qbs27cv2k0LO7fbjZ/97GdYu3YtDAYD7HY7Vq5cieLi4mg3LaI0TcOjjz6KdevWRbspEdPd3Q2PxwOr1RrtpkTNCy+8gLvuuguzZs2C2WzGli1bsGPHjmg3K+JaW1uxc+dOPProo1AUJdrNCav6+no0NTXhmmuugdFoRHJyMq644goUFRVFu2kRkZ+fj66uLtxzzz1ITEzE9OnTcfvtt+OPf/xjtJsWFt3d3fjjH/+I888/P+T4O++8g5ycHNx8880wmUxYtGgRfvCDH+C1116LWNsY2kbIarXiiSeeQFZWVuBYfX19yOdjVUZGBq677joAgBACZWVlePPNN3HllVdGuWWR9dprr8FkMuGqq66KdlMipr29HQDwi1/8AhdddBEuuugiPPLII+jq6opyyyLD4XCgpqYGLpcLq1evxjnnnIMNGzagoaEh2k2LuOeeew6XXHIJzjrrrGg3JexycnIwZ84cvPbaa+jq6kJbWxvefffdY7rOxpOUlJQx+0b9+uuvx8SJE485XlhYiNzc3JBjc+fORUFBQaSaxtA2WsrLy7Fr1y5s3Lgx2k2JmOLiYsybNw9XXXUV5s+fj7vvvjvaTYqY5uZmPPfcc3jsscei3ZSI8nq9OPvss3H++efjf/7nf/DSSy/hu+++Gzdj2vzh7E9/+hN+85vf4L333oPb7caWLVui3LLIcjgceOONN7Bhw4ZoNyUiVFXFv//7v+Ojjz7C4sWLcd5556G+vn7c/NwvXLgQZrMZzzzzDHp6elBTU4P//M//hMvlgtvtjnbzIsbpdB7Ty5Camoq2traItYGhbRQcPHgQt9xyC9atW4fVq1dHuzkRM2fOHBw8eBB/+tOfUFFRMa5euJ544glcf/31mDZtWrSbElFTpkzBnj17cMMNN8BoNGLatGnYsmUL3n33XfT29ka7eWEn+jeQue2225CdnY309HRs2bIFX3/99biqtr388su46KKLMGXKlGg3JSLcbjfuuOMOXHbZZdi/fz8++eQTZGZm4t5774120yIiJSUFzz33HPbv349ly5bh7rvvxtVXXw0A0Ol0UW5ddAkhIjo8gKHtNH366adYu3YtNm3ahE2bNkW7ORGnKAqmT5+OLVu24P333x8XM0g///xzFBQU4I477oh2U2LCpEmTIIQYF9/79PR0APLdtV9OTg4AoLGxMSptiob33nsPK1asiHYzIubzzz9HZWUl7rnnHiQnJyMrKwt33nknPvnkE7S0tES7eRGxZMkS7NmzB9988w1ef/11pKWlISMjY1yFtrS0tGOqak6nEzabLWJt+P/t3V9IU/0fB/C34hkmC9OCLQlCuogurDatrai1WmAkeBEkuRCKhG60jC48kkg1KYcYW0O0YGCgENi/JdVGLCn6oxDU2JQVLSnGaP1RMlpOl3su4hmsX8+D9uT22/Z+XbnzPed7Pl/Zxvuc7zk7DG3/gcvlwrFjx2A0GqHX65NdTsLcuXMHe/bsiVuWnf3jrZSTk5OMkhLq5s2bCAaD0Gg0UKlUsf+FSqXCrVu3klzdwnK5XGhvb49b5vP5IAgC5HJ5kqpKHLlcjsLCQoyOjsaW+f1+APjlNTDpyOv1wu/3Q6PRJLuUhPn+/XvsLOvfIpFIkqpJvHA4jBs3buDLly+xZQ8fPkRpaWkSq0q8kpISeDyeuGVutxvr1q1LWA0Mbb8pEongxIkTqK+vz6gjTgBQKpV48+YNOjs7MTU1hU+fPsFisUCpVKKgoCDZ5S04URThcDhgs9lgs9lw8eJFAIDNZkv7C5MLCwvR29uLnp4eTE9P4/Xr1zCbzaiqqoIgCMkub8Hl5ORAr9eju7sbPp8Pnz9/hslkglarjZ2FS3cjIyNYvHhx3NnGdKdQKCCVSmEymRAKhTAxMYHu7m4oFAosXbo02eUtOEEQYLFY0NnZiUgkgvv376O/vx8HDhxIdmkJVVlZiQ8fPqCvrw/hcBjDw8MYGBhATU1NwmrIiv58+EBz8vTpU+zfvx8SieR/2ux2e2zKJF25XC6cPXsWo6OjkEqlUKvVaGxszIi7Z3/m9/uh0+nw4sWLZJeSEE+ePMG5c+fw6tUrFBQUYNeuXWhoaPjlZyEdzczMwGg0YmBgAOFwGFqtFidPnsyYEHPhwgVcv34ddrs92aUklMfjgdFohNfrhSAI2LBhA5qamjLiDDPwY/wtLS3w+XyQyWRoaGjA7t27k13WgigvL0cgEMDs7CwikUjsu81ut+Pdu3dob2/Hy5cvUVRUhNra2tj1fYnA0EZERESUAjg9SkRERJQCGNqIiIiIUgBDGxEREVEKYGgjIiIiSgEMbUREREQpgKGNiIiIKAUwtBER/WE1NTUwGo0AgObm5ox5RiURLaz0f+YQEVEStba2xr3u6emBXq/PmB8jJqI/h2faiIgSZHx8HG1tbZiZmUl2KUSUghjaiChtrV69Gg6HA9XV1Vi/fj0qKyvn/LixsbExHDx4EGVlZSgrK8OhQ4cQCAQA/Hj+rCiKaG1tRWlpKVQqFS5duvTLfkRRxJEjRxAMBqHRaBCNRqFWq9Hf34+PHz+irq4OKpUKCoUCer0eXq/3j42fiNILQxsRpTWr1YozZ87g8ePHyM/Ph8VimdN2BoMBy5cvx6NHj/DgwQPI5fLYdWoAcPfuXaxatQpDQ0Noa2uD0WjE8+fP/7E/mUwGq9UKABgaGsLevXthNpvx7ds3OJ1ODA8PQ61Wo7m5+b8NmIjSFkMbEaW1iooKFBcXIy8vDxqNBj6fb07bTU5OQhAESCQS5OXlwWAwwGw2x9qXLVuG6upqCIKA7du3Y+3atXA6nfOq7e995ObmQiKRoL6+HleuXJlXH0SUORjaiCitrVixIvb3okWLEA6H57RdXV0dbt++DZ1Oh5aWFgwPD8e1FxcXx70uKirC+/fv51VbbW0tPB4Ptm3bBlEU4XQ6EY1G59UHEWUOhjYiSmvZ2b/3NafVajE4OIjjx4/j69evOHz4cNz06OzsbNz60WgUWVlZ89pHSUkJ7t27h9OnT0MQBDQ2NuLo0aO/VS8RpT+GNiKiXxgfH4dUKkVFRQU6Ojpw6tQpXL58Odb+9u3buPUDgQBkMtm89jE5OYns7GzodDoYDAZ0dXXB4XBgYmLij4yBiNILQxsR0U+mpqZQXl6Ovr4+TE9PIxwOY2RkBCtXroytEwwGcfXqVczMzGBwcBButxs7d+78135zc3MB/LgzNRQKoaqqKnYzQiQSgdvtxpIlS5Cfn7+g4yOi1MTQRkT0k9zcXFgsFly7dg0bN27E1q1bMTY2ho6Ojtg6W7ZswcjICDZt2gRRFNHU1ISSkpJ/7XfNmjVQKpXYt28fent7YTKZ8OzZM2zevBkqlQpOpxNdXV2/PaVLROktK8qrXomI5kUURYRCIZw/fz7ZpRBRBuHhHBEREVEK4LNHiSjjWK1WmEymf2zfsWNH3G+yERH9P+D0KBEREVEK4PQoERERUQpgaCMiIiJKAQxtRERERCmAoY2IiIgoBTC0EREREaUAhjYiIiKiFPAXqKETv7+SWJIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 720x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "accs = []\n",
    "stds = []\n",
    "pars = [2, 3, 4, 5, 6, 7, 8, 9, 10]\n",
    "for k in pars:\n",
    "    cv = KFold(n_splits=k, shuffle=True, random_state=1)\n",
    "    scores = cross_val_score(logreg, X, y, cv=cv, scoring='accuracy')\n",
    "    accs.append(scores.mean())\n",
    "    stds.append(scores.std())\n",
    "    \n",
    "accs = np.array(accs)    \n",
    "stds = np.array(stds)\n",
    "    \n",
    "plt.figure(figsize=(10, 4))    \n",
    "plt.fill_between(pars, accs - stds, accs + stds, color='b', alpha=0.2)\n",
    "plt.plot(pars, accs)\n",
    "plt.xlabel('n_splits')\n",
    "plt.ylabel('accuracy')\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import validation_curve\n",
    "\n",
    "logreg = LogisticRegression()\n",
    "cv = KFold(n_splits=4, shuffle=True, random_state=1)\n",
    "\n",
    "param_range = [0.01, 0.1, 1.0, 10, 100]\n",
    "train_scores, test_scores = validation_curve(estimator=logreg, X=X, y=y,\n",
    "                                             param_name='C',\n",
    "                                             param_range=param_range,\n",
    "                                             cv=cv, scoring='accuracy')\n",
    "train_scores_means = train_scores.mean(axis=1)\n",
    "train_scores_stds = train_scores.std(axis=1)\n",
    "test_scores_means = test_scores.mean(axis=1)\n",
    "test_scores_stds = test_scores.std(axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10, 4))    \n",
    "plt.fill_between(param_range,\n",
    "                train_scores_means - train_scores_stds,\n",
    "                 train_scores_means + train_scores_stds,\n",
    "                 color='b', alpha=0.2)\n",
    "plt.plot(param_range, train_scores_means)\n",
    "plt.fill_between(param_range,\n",
    "                test_scores_means - test_scores_stds,\n",
    "                 test_scores_means + test_scores_stds,\n",
    "                 color='r', alpha=0.2)\n",
    "plt.plot(param_range, test_scores_means)\n",
    "plt.xlabel('param_range')\n",
    "plt.ylabel('accuracy')\n",
    "plt.xscale('log')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f368f95aad0>"
      ]
     },
     "execution_count": 128,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "logreg = LogisticRegression(C=100)\n",
    "logreg.fit(X, y)\n",
    "plot_decision_regions(X, y, logreg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f368f1f94d0>"
      ]
     },
     "execution_count": 132,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.svm import SVC\n",
    "svm = SVC(C=1.0, kernel='rbf')\n",
    "svm.fit(X, y)\n",
    "plot_decision_regions(X, y, svm)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## cross_val_predict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import cross_val_predict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [],
   "source": [
    "logreg = LogisticRegression(C=100)\n",
    "svm = SVC(C=1.0, kernel='rbf', probability=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [],
   "source": [
    "cv = KFold(n_splits=4, shuffle=True, random_state=1)\n",
    "\n",
    "a_logreg = cross_val_predict(estimator=logreg, X=X, y=y,\n",
    "                       cv = cv, method='predict_proba')\n",
    "\n",
    "a_svm = cross_val_predict(estimator=svm, X=X, y=y,\n",
    "                       cv = cv, method='predict_proba')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x7f368f03bd50>"
      ]
     },
     "execution_count": 142,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(a_logreg[:,1], a_svm[:,1], 20, y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Learning curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'accuracy')"
      ]
     },
     "execution_count": 156,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.model_selection import learning_curve\n",
    "\n",
    "svm = SVC(C=1.0, kernel='rbf', probability=True)\n",
    "cv = KFold(n_splits=4, shuffle=True, random_state=1)\n",
    "\n",
    "train_sizes = np.linspace(0,1,21)[1:]\n",
    "\n",
    "train_sizes, train_scores, test_scores = learning_curve(estimator=svm,\n",
    "                                                        X=X,\n",
    "                                                        y=y,\n",
    "                                                        cv=cv,\n",
    "                                                        train_sizes=train_sizes)\n",
    "\n",
    "train_scores_means = train_scores.mean(axis=1)\n",
    "train_scores_stds = train_scores.std(axis=1)\n",
    "test_scores_means = test_scores.mean(axis=1)\n",
    "test_scores_stds = test_scores.std(axis=1)\n",
    "\n",
    "plt.figure(figsize=(10, 4))    \n",
    "plt.fill_between(train_sizes,\n",
    "                train_scores_means - train_scores_stds,\n",
    "                 train_scores_means + train_scores_stds,\n",
    "                 color='b', alpha=0.2)\n",
    "plt.plot(train_sizes, train_scores_means)\n",
    "plt.fill_between(train_sizes,\n",
    "                test_scores_means - test_scores_stds,\n",
    "                 test_scores_means + test_scores_stds,\n",
    "                 color='r', alpha=0.2)\n",
    "plt.plot(train_sizes, test_scores_means)\n",
    "plt.xlabel('x_size')\n",
    "plt.ylabel('accuracy')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "\n",
    "dtc = DecisionTreeClassifier(max_depth=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'accuracy')"
      ]
     },
     "execution_count": 159,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmUAAAELCAYAAACYmfGQAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nOzdeXhU5d038O89eybJJJlJCCEsYQ8kyCKLCIWqWFFUXNGqrfXSR5FitVprrY/aYF99e2nthdXi09Yu2qrVPnV5S6VFW61aaEERkrCvYQnZZ9/Pud8/TlaWMAmzZfL9XNdcJGfOzPlNjpgv97nP7xZSSgkiIiIiSildqgsgIiIiIoYyIiIiorTAUEZERESUBhjKiIiIiNIAQxkRERFRGmAoIyIiIkoDhlQXcLaamjypLoGIiIgoJkVFuad9jiNlRERERGmAoYyIiIgoDTCUEREREaUBhjIiIiKiNMBQRkRERJQGkhrKdu3ahcsvvxwXXnhhr/utW7cOS5cuxfTp03HllVdi/fr1SaqQiIiIKDWSFsr+8pe/4I477sCoUaN63W/nzp148MEHcc8992Djxo2499578cADD2D37t1JqpSIiIgo+ZLWp8zn8+EPf/gD/v73v2PHjh2n3e+NN97AvHnzsGjRIgDARRddhLlz5+LNN9/EI488kqxye1ClTMlxiYgylU6IVJdAlHaSFsquv/76mParra3F/Pnze2ybPHkyNmzYkIiyzuhQqx9f+90WBCJKSo5PRJSJJhXZcOe8kZg3pgCCAY0IQBp29Hc6nbDZbD225eXloa2tLSX1FOeacePkMrR6GMqIiOIhqqrYUN+Ab79dg7IcG66eMArzxhQgJ0fAbJbIygL0+lRXSZR8aRfKTidV/5KyGPVYsag0JccmIspU/uBI/PKDRry77xB+8nk1/rTLhkVDR2N8bj6E0MFiAWw2Ffn5QFYWYLFIWCwMa5TZ0i6UFRQUnDQq5nQ6YbfbU1QRERHFm9Wiw8pLh+LiPcX40xcN+KjpEF7atxXj8my4ckwZxhkL4HYLNDUJqCoASAghkJUlkZcnYbMBVqsW1CwWQMcGT5QB0i6UVVZWoqampse26upqTJ06NUUVERFRIuh0wKSJArfllmBGdQlqA8fwtyN1eHbLNozPz8OVY8pQXpDfeaVESolIBGhrEzh+vGubEALZ2V1hLStLC2tmM8MaDSxpEcoWL16MqqoqzJkzBzfeeCOuvvpqrF+/HgsXLsQHH3yAzZs347HHHkt1mURElADDhmnzyLK2DsfswhJ87qzHXw4ewo8/34rx+XlYOqYM5fYCCAGYTNojJ6frrviOsNbcLFBfLyAlICUgBJCbK5GbK5GXp4U1m41BjdKXkDI5/R4uueQSHDt2DKqqIhqNwmQyAdAaxV544YV48cUXccEFFwAA3n//fTz//POoq6tDWVkZ7rvvPixYsOCU79vU5ElG+URElGA+H7Btmw7hMJCdq+DjY8fxl4OH4AyFMaFj5MxeEPP7SQmEwx0PLaxZrRLjx0s4HBK86ZNSoago97TPJS2UJQpDGRFR5giHgdpaHVpaBAoLJaKqgn8ercdfDtbBFQ5jYkE+rhxThokF+f16/2AQ8HgEcnK0cGa3M5xRcjGUERHRgKGqwN69AocO6eBwSOj1QERR8NHRerzXHs7K28PZhH6Gs0AA8HoFbDaJceNUFBSA4YySgqGMiIgGnKNHBbZv1yEvT8Js1raFFW3krHs4Wzq2DOPzzy6c5edLjB2rteBgOKNEYigjIqIBqbUV2LpVD7NZIju7a3tYUfDhkWNYd6gO7nAEk+z5uHLMaIzPz+vXcfx+wOcTKCjoCmdEicBQRkREA5bXC1RXazcAnBiWQoqCj44cw3uH6uAJRzDZXqD1OetnOPP5tIDmcABjxqjI69/bEJ0WQxkREQ1oJ94AcOIlxlDHyNnBOngiEVS0h7OxZxHOAoGucHbC6n9E/cZQRkREA56iAHv2CBw+3HUDwIm0cHYU6w4e1sKZoz2c9XPIy+vV7tgsKgJGj1aRe/rfp0QxYSgjIqKMICVw5IjAzp065OdLtLe8PElIUfD3w0fx10OH4Y1EUOmw47KykRiXnwddP2byezxAMChQXKxi9GiJnJyz/CA0aDGUERFRRmlp0W4AsFh63gBwomA0in8cOYa/HqqDNxKF3WzGzOIizCoegjJbbucSTrGQUgtnoZBASYmKUaMYzqjvGMqIiCjjeL3Atm16RCLyjHdLBqNRbGlqxqaGRtS2tEGREoVZFswcUoTZxUMwIjcn5oAmJeB2a6sElJaqGDmy92BI1B1DGRERZaRQCNi+XYfWVhHz0km+SKQzoO1obYMqgWJrFmYVD8HM4iKUZmfHFNCkBFwuIBrtCmdWaxw+FGU0hjIiIspYsdwAcDqecLgzoO1sdUICKMm2YlbxEMwqLkJJDENgqqqNnEWjAsOHa+EsK6v/n4cyG0MZERFltFhvAOiNKxTG541N2NTQiD1OFySA0pzszoBWfIZhsI5wpigCI0aoGDFCwmLp3+ehzMVQRkREg0Jzs8C2bboz3gBwJs5QCJsbtIC2z+UGAIzMzem8xFnUy1CYqgIul4CqAqNGqbDbtdE7nU576PVdDy7pNPgwlBER0aDh9Wp3ZiqKjEtH/tZgsDOgHXBrv3NG23I7A5r9NMNhigK43Vo4EwKQUkJK9JivJgRgNAJ6vYTRCJhMWoAzGACTSfvTaOwKcwaDdsdnXy7RUnphKCMiokElFNJWAGhri/0GgFg0BwLY1B7Q6jxeAMDYPFtnQMvvWDk9RlJqI2uK0vX1id8rimgPddprdDrA4VBRXAzYbJy/NtAwlBER0aCjKMDu3QJHjvT9BoBYNPj9nSNoR7w+CADj8/Mwq3gIzi0ugq0/E9tioKraElCBgJY0rVaJkhKJggKJ3FwttFH6YigjIqJBSUrg8GGBXbv6fwNALOp9vs6AdsznhwBQbs/H9KJC2C0W5BiN2sNkhNVg6NeqAqcTDmtrdSqKgF4PFBZqo2i5ubzRIB0xlBER0aDW1CRQU6Nrn9MFGI0SZrM2Xyvek+2Per2dlzgb/IGTnhdAZ0DrCGu53b7OPuH7HJMRWXp9TL3TFKVrFE0ILZgNHaqNomVncxQtHTCUERHRoBcKAX6/Fljcbu0OSb9fQFUlAAGdLr5hTUqJ1lAInnAY3nAEnkgEvkgU3kgEnnAE3kj7o9vXyml+JeuFQI7RiGE52ah0FKDSYcewGJrcap9ZQFEkDAaBIUNUFBVpc9ESNWo4UEmpPRIdXBnKiIiITkFVgWCwK7y43dodkz6fgJSyM/SYzVqIScTIWgcpJQKK0iOkeXp8HcZ+lxvHfH4AQIHZhMl2OyoL7ZhkL0CO0djr+yuKdpkzHNZG0Ww2bS5aXp42ijZY2nNEItr5DgYF/H7tbl2PRzvnw4ermDAhsbGIoYyIiKgPOsJaMKiNrLlcXb+4kx3WTtQaDKK2pQ01La3Y0doGfzQKAWB0ng0V9gJUFtox2mbrdd6alNpn8/u1z2M0CgwdqqKwULtZ4Az5Lu11nL9wWFtA3u3Wzp/XKxCNAoCElKJbKxJt3/x8iYoKhrJ+YygjIqJk6R7WOkbWXC6BQKArrAmh9RtLRlhTVBUH3B7UtrSitqUVB9weSABWgwGTHQWotNtR4bCjwNJ7q45oVLu02zGKVlCgzUWz2dJ7sXUtdGmjXoEAOsOXdj66fvZGY1f4Ot1duD4fkJfHUHZWGMqIiCjVOtpUdL8M2hXWACG0kRkhAJ1OwmBAj0e8gps3HMH2Vm0UrbalFa5wGABQmp2Niva5aOPz82DspT+IlF03C0gJmM2AxaLVfGLtHY1uOz6DtmqB7Fy9oLdHrBRF+7l2hK+O4OXzCSgK0H3UqyMI9ycMM5TFAUMZERGlK0XRRnOiUW0uUySijehoo21aaAsGOyaZayNtHb+VjUbZGXy0rv99O7aUEke8Pm0UrbUVe9pciEoJk06HiQX5qHDYUemwo9ia1esNA9Go9uhoZnvyn101n6KK9ufESSGpa7kp2S3gyfYbLrQQ5vXq2n8+XZeMjcau8BXP3nMMZXHAUEZERAOZlF2hLRrVLiFGIh2XSNEe3ATC4a6u/iaTdlmxL6EkpCjY1eZETbMW0jradRRaLJ2jaOX2AmQZDAn4lCc7MeB1rGLQsU2nQ2f4SgaGsjhgKCMiosFASm30yOMRaGkBGht1iES0ESSLRVtuqS+XBZv8AdS0tqK2uRU72pwIKQr0QmBsnq1zFG1Ebk5cG92mM4ayOGAoIyKiwUhKbSTN4xFobARaWnRQVW30zGLRuvnHmqeiqoq9ThdqW9tQ09yKw15tXc9ckxFlubnQ6wR0EBBCQCcAAQHdCV8LAe1PdPzZ/rwQPbb32EfglO9r0AkMycpCSbYVNpMppsa5Z4uhLA4YyoiIiLr6kDmdAo2NAm63NtfLYJCwWtGnZrGuUKjzhoFjXj8kJFQpIQHtTwmokJBSQpXoer59e+fX3V8Dbd++yjLoUWK1Ymh2NkqyrRhqtaIk24qiLAv0cez0ylAWBwxlREREJ4tEAI8HaG0VaGzUIRDQRtcsFi2kxXuB9lhpQa57WOsIelpoUyERVhQ0+AM47vPjuN+Pep/26LibFNBWORhizWoPbFpQ6whtln7Mi2MoiwOGMiIiojMLBLRLnc3NQHNz13y0rCztUudAWBfTH43ieHtA08KaD8d9fjQGAj1G4QrMJm1krX1UrSO05fVyKZShLA4YyoiIiPpGSi2EuN0CTU0CLS3apU69XrthwGJJdYV9E1VVNAbaR9baQ1u9X/s6qDUzAwBk6fUY2i2kdYyyFWVlIRTQMZSdLYYyIiKis6Mo2hqQbW3afDSPp6snmNXaNYrW0b4C6GpfceL3Jz538n6i8/uOQauO57oPYun12pJPZ3OZVUoJZyjc7RKoD8fbw1pbqOel0CJLFu6eXo7L5ub0/4Ax6C2UJacZCREREaUtvR7Iy9Mu35WVSYTDXa03mpt1iEa7OvEL0bNjv16vdfEXQgtV2vfdO/z3/F6vlz26+ne8ruuhvVdLi0BdnbZTXp7WYLavhBAosJhRYDFjkr2gx3OB9kuhHYGt0RuCUZeiiXbtOFJGREREaSkUAurrBQ4e1EFRAJtN9uku0r5IhzllHCkjIiKitGQ2A2VlEqWlChobBfbv18HtBnJzJcy9r7E+IDGUERERUVozGoHSUomhQxU0Nwvs36/doJCdrc15yxRJvQG2vr4ey5cvx5w5c7Bw4UKsWrUKkUjklPu+8sorWLx4MaZOnYqrr74aX3zxRTJLJSIiojSj1wPFxRLnnadi+nQVBgPQ1CTQvgDBgJfUULZy5Urk5+dj/fr1ePXVV7FlyxasXr36pP3efvttPPvss1i1ahU2bdqEr33ta7jrrrvgdruTWS4RERGlISEAh0Ni5kwVM2cqyM7Wwpnb3XUn50CUtFBWXV2N7du347vf/S5sNhtKS0tx11134Y033oCqqj32/fvf/45LLrkEs2fPhslkwjXXXINx48bhr3/9a7LKJSIiojQnBJCfD0ybpuK88xQ4HBItLQJOp9a+Y6BJWiirra1FSUkJ7HZ757aKigq4XC7U1dWdtP+JHXdtNht27NiR8DqJiIho4MnJASZPlpg7V0FJiURbm0Brq0C33rFpL2mhzOl0wmaz9diWl5cHAGhra+ux/YILLsC6deuwadMmhMNh/OMf/8Bnn30Gp9OZrHKJiIhoALJagQkTJObNUzB6tAq3W6C5WeCkKexSQhcOwuBpg6nlOLKP7IaxuT4lNXdI6d2XHS3SThwVu+qqq1BfX4+HHnoIHo8HF198MZYsWQKfz5eKMomIiGiA6WynURJB4+Ew6naHEfIFkC9csIadMAS8PZYSCPmi0FtKAQxNWc1JC2V2u/2kETGXy9X5XHdCCKxYsQIrVqzo3Patb30LZWVlCa+TiIiIBiBVBYJBiFAQIhAA3G7oXE4YfT6USYmRJgFXQOBokxlOxQxLbgHMWV0XDKNRL1Lbzz+JoayyshINDQ1obGzEkCFDAADbtm2Dw+HAiBEjeux74MAB7Nu3D4sWLQIAhMNhbNq0Cddee22yyiUiIqJ0pChd4cvv7wxfIhDoWlBTADCaIM0mSLsdEAICQH4RYBujLcR+5KhAW5uAxaItwp4OkjanbPLkyZg2bRqeeeYZeDweHD58GGvWrMHNN98MIQQWL16Mf//73wCAxsZG3H///diyZQvC4TBWrVqFwsJCzJ8/P1nlEhERUSpFo4DXC9HSAl3dIehrqmH41ycwfvR3GDf+C4Ytn0O/Zxf0bS2AwQBpt0MWFkI6HJB2B2RuLmAy91zlHNqam/n5EhWTVUyapMBiAZxOIBgSpykkeZI6p2z16tWoqqrCokWLYLVacemll2L58uUAtNExv98PAJgzZw7uu+8+3HPPPfB6vTj33HPx4osvQn82S8UTERFR+lMU6Hfvgq7+qPa9BKDXQZpMgMkMmZ0dl8MIAdhsgM2mwucDGg+o0Kd46SYuSE5ERETpIRSCobYaoq0N0uE4aZQroXxeqHkFUCsqE3oYLkhORERE6c3rhWHbFxDRKGRhYaqrSYmkLrNEREREdCLR1grjJm1euWzvYToYcaSMiIiIUkZ37Cj0tTVaGDOneFJXijGUERERUfKpKnT790F/YL/WtsKQwkgiVYhwKyDzU1cDGMqIiIgo2SIR6Hdsh67huDZ/TJfk2VRRPwyeXTB4dsLg2Qm9exd0ig+eyY8giCnJraUbhjIiIiJKnkAAhuqtED4fZHsz+YSSErrA0R4BTO8/BAEJCQHVOhKRovmImkYjXDg38fX0gqGMiIiIkkJ43NBv2QIhpHbJMhGUAAye3VoIc++E3rMLuqjWPkvVZ0OxTUSw8HxEc8sRzZ0AGNr7nvm8gC61sYihjIiIiBJONDXBUP0FpNUKmWWNz5tKCV2wHgZ3+yiYZyf0vkMQUAEAinUEIo7ztABmK4eaNRwQ6dt4gqGMiIiIEkdK6A7XwbB7J9S8fMBk6v97KUEYPHs6A5jBvRO6qFs7jN6KaO4EBEcsQ9Q2CUruBEhDTpw+RHIwlBEREVFiKAp0e3ZDf7gOqsMB9GW5RCmhCzW0X4JsvxTpO9A1CpZVioh9FqK2SYjmlkO1DgfEwF6OkaGMiIiI4i8chr62BrqWFsiiopiXTBKhFlj3/wIGdy10EScAQOqzEM2ZgOCI6xHNLYeSOxHSePrligYqhjIiIiKKL59PWzIpHIIs6sOSSVIie+/zMLi2IVw4vz2AlUPJHjngR8FiwVBGREREcSNcThi+2AIY9JD5BX16rbH5YxjbNsM/+g6ESpcmqML0xVBGREREcSGO18NQUw2ZmwtYLH17bcQN6/6fI5ozHqFhlyeowvSWvveFEhER0cAgJXQHD8BQvQ0yP7/PgQwAsg68BBH1wj/+nkFxqfJUOFJGRERE/ReNQr97F3RHj/R7ySSD8wuYG/+OwIhlULJHJ6DIgYGhjIiIiPonFIKhZhvgcvXpDsselCCse56HklWK4Igb4l9jb6JRCL8PCEcAAaA0tX3NGMqIiIio77xeGLZugVAUSIej32+TVfcq9KEGeKY8CejOorFsLKQEggEIfwCQEtJogjK0BNJRCGmzAUZjYo9/BjGHspaWFjjO4odOREREmUG0tsDwxRZIiwUyL6/f76P37oX56DsIDb0E0bwpcaywm0hEGw2LRAEByAI7lBGjtNUFsrP7N7qXIDGHsgULFmDu3LlYunQpFi1ahKysrETWRURERGlId+wo9LU1kPl5gMnc/zdSo7Du+SmkKR+Bsm/ErT5ICQQCEAG/NhpmzoJSUgrpcEDmpn40rDdCSilj2bG2thbr16/H3/72Nxw/fhyLFi3ClVdeiXnz5kGkMGU2NXlSdmwiIqJBQ1Wh27cX+oMHIB12QH92M6DMR/4X1oO/gbf8YUQKzz+72iIRCJ8PUBQAgHQ4oA4phmrLA6zWtBoNKyo6/UoEMYey7vbu3Yu//e1veO+999DW1oYrrrgC1157LcaNG3dWhfYHQxkREVGCRSLQ79gOXWODNn+sH3dYdqcLHINtyz2IFJwL36Tv9/0NVLV9NCwAQEJasqAOLYG027XRMEP6TpmPeygDgE2bNuGvf/0r3nnnHeTm5sLr9WL27NlYtWoV7HZ7v4vtK4YyIiKiBAoEYKjeCuH3QRbE4fe7lMip+W8YvHvhmvEzSHOM89XDYW00TFUBIaA6HJDdR8MGiLiFst27d+Pdd9/F2rVr4XK5cMkll2Dp0qU477zz4PP58MMf/hDNzc34xS9+EZfCY8FQRkRElAA+H3T1x6CvOwQYjdrdiXFgangf2XtWwzfumwgPXXz6HVUVCPi1OyUFIK1WqEOHQRYUaKNh+oHZYDYuoWzp0qXYs2cPZs2ahauvvhpf+cpXYD0hmYZCIcyePRtbt249u4r7gKGMiIgoTqSEcLZBV3cIuqYmwGiIawAS4TbYPl8BxToK3ilPAuLUl0GF0wmoKlRHIdTiYkhbHpAhNxj2Fspivuh6ySWXYM2aNRg2bNhp9zGbzfjVr37Vt+qIiIgotaJRiJZm6A/sh/B6gSyL1p0/zhPkrft/DqEE4R+38rSBDKoKKFFEzv9Sv5ZrGshinql322234Ve/+hU2bNjQue2Pf/wjqqqq4Pf7O7ede+658a2QiIiIEiMQgO7gARg//ScMNdWAEJBFRZA5uXEPZMaWf8PU/AmCI2+Eah1+2v2E2w2ldMSgC2RAH0LZE088gZqaGgwZMqRz2znnnIN9+/bhySefTEhxREREFH/C7YJ+ey2M//oY+oP7IXNytZGxRAWhqB/WfWsQtZYhWHpN7/tGIpClpYmpI83FfPnyH//4B9atW4e8bp17J0yYgJ/+9Ke49NJLE1IcERERxYmiQLS2Qn9wP4TbBZhMkI74X6I8laxDL0OEW+Gf9DCg66V5q88LWViojdQNQjGHMiklVFU9aXswGEQkEolrUURERBQnoRB0jQ3QHTgAEQlDZlshC4uSdni9ezvM9X9BaNgVUHIn9rqv8AcQLZ+cpMrST8yh7Ctf+Qruvvtu3H777Rg2bBhUVcWBAwfw0ksv4YorrkhkjURERNRHwuuBOHoU+qNHIAHAZoM0xqetRczUCLL3PA/VXIjAqFt63zcUgrRa49MLbYCKuSVGMBjEj3/8Y7zzzjtwu90AAJvNhmuuuQYPPPAAjClaS4otMYiIiNqpKoSzDfpDByFaWrr6i51lB/7+shx6FVmHX4Nn8uOI2mf2uq9oaoZSWQl1aEmSqkuNuHf0b2trgxAC+fn5ALSmshMmTOh/hWeBoYyIiAa9SAS6pkboDh6A8PshrVYgOzulJen8dbBtuRfhwnnwT/xO7zsrUQi3B5H5C9J6iaR4iEufsg7Nzc0Ih8MAAL/fj4aGBtx+++34/PPP+18hERER9Z3fD93xeugPHQRUFdKWC1mUvPlipyVVZO/5KaQ+C4Ex/3XG3YXLBaVsTMYHsjOJ+dN/8cUXuPfee9HY2HjSc/PmzYtrUURERHQaUkK4nNDV1UHX1ADo9ZB5NkCfPoHGXP8eDJ6d8E34NqQxr/edpdS692f4ZctYxHwGn3zySVxxxRW49NJLceONN+LNN99ETU0N1q9fj6eeeiqRNRIREVEgoPUX6+i6b7EkraVFX4hQE7IO/RaR/OkIF11w5v09HqhDh2XMMkpnI+ZQtm/fPrz++uvQ6XQQQqC8vBzl5eUYPnw4vv/97+PFF18843vU19ejqqoKW7ZsgcViwUUXXYSHH374lDcJ/P73v8fLL7+MhoYGFBUV4dprr8Vdd90FkWb/8REREcWdqkL4fRBeL0RLi/aIhAAIyJyc9LhEeSpSwrp3DSBV+MetiC0whkNQR4xIfG0DQMyhzGq1wuPxIC8vD9nZ2WhoaEBxcTFmzpyJFStWxPQeK1euxPjx47F+/Xp4PB6sXLkSq1evxne+03MC4Icffoinn34av/71rzF16lTs3r0bt956KxwOB66//vq+fUIiIqJ0F4lA+LwQbjd0Lc3tC3IrWqgxm7XeYoYkt7PoB2PzJzC1bYJ/9O1QLUPP/IKAH9KWpy04TrGHsgsvvBA33XQT3nzzTcyaNQvf/e53ccMNN2Dr1q1wOBxnfH11dTW2b9+OX/ziF7DZbLDZbLjrrrvw2GOP4f7774eu2+2627Ztw/jx4zF9+nQAQHl5OaZNm4adO3f24yMSERGlmUAAwueDcLZpIcznBSQAnYDMyoLMz09ZG4v+EhEPrPt/jmjOOISGxda/VHh9iE6bkeDKBo6Yz/gjjzyCJUuWwGKx4JFHHkEkEsFDDz2ETz75BE888cQZX19bW4uSkhLY7V1N4SoqKuByuVBXV9dj3wULFmDv3r3YuHEjwuEwtm/fjm3btuGCC858bZqIiCitqCqE16PdJVlbA8M/P4LxXx/DsHUL9MeOaEHMUagtL2R3AFnWARfIACDrwK8gIm74x30LEPozvyAchjRbIO2Dt1nsiWIeKWtsbOy8TFlcXIxXX321TwdyOp2w2XoOvXaso9nW1oaysrLO7dOmTcP3v/993H777YhGoxBC4J577sH8+fP7dEwiIqKkkhIIBiH8fgiPB7rWZgiXa0BeiuwLg3MrzI3vIzD8Oig5o2N7kccDdcLEARlAEyXmUHbllVdi8+bNPS4znq2OvrUnTt7fuHEjnnnmGfzyl7/EjBkzUF1djXvuuQejR4/GZZddFrfjExERnZVgECLQHsCcrUBrG4SiaM/p9ZAW84C8FNknShDWvS9AsZQgOOLGGF+jQAgBdUhxYmsbYGIOZTfffDNWr16N//qv/0JOTk6fD2S329HW1tZjm8vl6nyuu9deew0XXngh5s6dCwCYOXMmrrjiCrz11lsMZURElBrhsBbAvF6I1laItjaIaBiAAAQgLVna+pL6GC7dZZCsutehD9bDU/l/AL05ptcIjwfK8BGAyZTg6gaWmEPZ+++/j+bmZvziF79ATk4O9Cf8R7dhw4ZeX19ZWYmGhgY0NjZiyJAhALQJ/Q6HAyNOuEBe3YYAACAASURBVBVWVVWoqtpjm9LxLw8iIqJEi0a1lhR+P9DaCp2zDSLgB4ROC2BmM5BjhdRn1mXIvtJ798F89C2Eir+CaP45sb1ISiAagTqsNLHFDUAxh7I777zzrA40efJkTJs2Dc888wweffRROJ1OrFmzBjfffDOEEFi8eDGqqqowZ84cXHjhhXjiiSdw3XXXYfr06dixYwf+8pe/4Nvf/vZZ1UBERHQSRdECl8+v3Q3Z2qJ9L9E+D8ykTUhP8VqSaUcqsO75KaTRhsDo22J+mfB5oRYWpXxtznQUcyi7+uqrz/pgq1evRlVVFRYtWgSr1YpLL70Uy5cvBwAcOHAAfr+/81hutxuPPvooGhoaMGTIENx2223sUUZERGdHVQG/HyIQ0JYqamuF8Li15yQAkwmyo1M+9cp89B0YfPvgLf8epKEP05oCQaiTKxNX2AAmZMds+zO49957e31+9erVcSmor5qaPAk/hmhuhv7gvoQfh4gofQmoxUO1S04Dbc5UNArR1gb9sSMQra3anZAQgNGgzQMzmdJuqaJ0pwvUw7ZlJSL50+Gb9EjsP79gEJAS0TlzB+3PvKgo97TP9amjf3eKoqCurg51dXVYunRp/6sbAITfB+FyQeYO7rkDRDSISRX63bugO1wHZeIkrbdUOv9SlRLC64Gor4f+2FFAUbSmrAUF6V33QCAlrHtfAIQB/rF39+nnKbxeRCun8BycRsyh7HSLjq9duxZbtmyJW0FpS28AzLHdVUJElImkxQIEgzBs2Qy1cAiUceOBftyNn1ChEHQtzdDVHdK65BuMkDbbwBvdS2Omxr/D6NoK39gVkOYzr+jTKRqFNBh4abgXMV++PB1FUXDeeedh06ZN8aqpT5Jx+VJXdwj6/fu0f2ERERGE2w2EQlBGlUEdVZba1gaqqk3QP3oUuqYGQAIyN5f/kE4AEW6D7fMVUKwj4Z3ylHY3aqyvbW2FMnos1G7N4gejuFy+DAQCJ20LBoNYt24dTOwzQkQ0qEibDVBV6I8ehv7YUSjjJ0AdWpLcJqk+H3RNTdDVHYSIRLRu+XYHL40lkHX/LyGUIPzjVvYpkEFVAVWBOjSGRcoHsZhD2fTp00/qvA8AOp0ODz74YFyLIiKiAUCngyywA5EI9NtroTt0EEr5JG1bokSjEC0t0B+pg2hrAwx6yJxcSKMxccckAIChdRNMzf9EYOTNUK0jzvyCboTHDaV0OGCxJKi6zBBzKHv55ZdP2mY2mzF8+HA4HH24pkxERJnFaIQsKgICfhg2b4JaXAxl7Pj49aGSEsLj7pq0r6ra+pFFRfF5fzqzqB/Z+34GxToSweHX9v31kQhk6fD415VhYg5ls2fPRk1NDfLy8jo78G/btg1HjhxhKCMiIiDLCpllhc7lhG7jp1DKxkAdOQro7yhWMNg+af8ghM8PmIyQeXmctJ8CWYdegQi1wHvOQ4Cuj+fT54MssLODQQxiviD89ttv46abbsL+/fs7tx0+fBjf+MY38PbbbyekOCIiGnikLQ/S7oD+0EEYN3wCXf0xbU5RLFQVorUF+m1fwPivj6HbtRMwGCCLiiDz8hnIUkDv3glz/VqESpZAsZX3+fXC74cyanQCKss8MY+UrVmzBi+99BJmzZrVuW3JkiUoKSnBww8/jKuuuiohBRIR0QCk00E6HEA4DH1tDXR1h6BMLIfMP81d7F4vdE2N0B2ugwiHgSwLJ+2nAzWC7L3PQZodCIz6Wt9fHw5BWrLYvSBGMYeyxsZGTJs27aTtFRUVaGhoiGtRRESUIUwmbe6X3w/Dpv9AHVoCZew4wGoFIhGI1lboDx+EcLkAvR4y16ZdoqS0kHXod9D7D8Mz+THAYD3zC04g3B4okyYn967cASzmUFZeXo6XX34Zt912G3TtP9xwOIyf/exnmDhxYsIKJCKiDGC1QmZlQdfWAt3GBqiFRdA1NwNSapP2CzlpP92Yj62F5eifEBp6KaL2WWd+wYmUKKDTQS0aEv/iMlTMzWNra2uxfPlyhEIhFBcXQ0qJo0ePIicnB7/97W8xZsyYRNd6SmweS0Q0wCgKRMAPac3mCEqaMjZ/iuydP0LEPhu+SQ8Dou9z+URbK5SRZVDHjE1AhQNXXJrHVlRUYP369fjkk09QV1cHIQRGjBiBBQsWsHksERHFTq/1FqP0ZHDVInvXj6HkToRv4nf6FcggJaAoWkNhilnMoQwA9u7di4kTJ2LRokUAtJYYO3bswNSpUxNSHBERESWPzncI2dufgGophnfyo4C+f81ehdcLtahYmztIMWNLDCIiIoIINSG39geA3gxvRRWk8Sz6igWDUEeNilttg0XMoayjJcbChQs7ty1ZsgQvvfQS1qxZk5DiiIiIKPFE1Ivc2schFD+8k38A1XIWk/MDAUibDdLGu2j7KuZQxpYYREREGUgNI3v7D6EL1MM76REoOWfX6FV4vVDKxrDHXD/EHMo6WmKo3boysyUGERHRACYVZO96BkZ3LXwT7kc0/5yze79IBNJk1hoHU5+xJUYM2BKDiIgyjpTI2v8/sNSvhX/0HQiVLj3rtxQtLVDGjdfWPKVTSlhLjJEjR+JLX/oSbrvtNvz+97+PS7FERESUeJYjb8JSvxbB0mviEsg61jdVi4ee/XsNUjGHMr/fj5///OeoqalBOBzu3P7jH/8Ybrc7IcURERFR/Jka3kfWoVcQKvoyAmW3xuU9hdsNpXQ4YDbH5f0Go5jnlFVVVWH9+vUoKyvD559/jkmTJkFVVWRlZeFXv/pVImskIiKiODG0boZ1z08RyZ8G//hvASJOqypEIlBLh8fnvQapmEfKPv74Y/z5z3+G3W7HG2+8gYcffhgA8MILL+DDDz/EhAkTElYkERERnT29Zzdydv5fKNll8JY/DOiM8XljnxeysBDIyYnP+w1SMcfjaDQKu90OADAYDAiFQgCAW2+9FS+//HJiqiMiIqK40AWOIae2CqoxH96KHwCG+HXbF/4AFE7uP2sxh7KJEyfiJz/5CSKRCEaPHo3XXnsNAHDw4MHOgEZERETpR4TbkFPzGADAW7kK0hTHbgKhEGS2FbLAHr/3HKRiDmXf+9738N577yEajeLuu+/GM888g2nTpmHZsmW47rrrElkjERER9VfUj5zaKugiTngrHoeaNSyuby88Hqijx7JZbBzE3KfsRPv378eOHTswYsQInHPOWTabOwvsU0ZERHQaagQ521fB4NwG7+RHEbXPjO/7K1EIjweReQsAQ8zT1Ae1uPQpO9GYMWNS1jCWiIiIzkCqsO55DkbnF/CNvzf+gQyAcLmgjB7LQBYncboPloiIiNJJ1sGXYW76EIFRtyBcvCj+B5ASUFU2i40jhjIiIqIMYz76LixH/xfBkssQHL4sIccQHjfUklIgKysh7z8YMZQRERFlEGPTx8g68EuEHXMRGHNn4ibgh8NQh7NZbDwxlBEREWUIg3Mbsnc/i6htEnwTHgCEPjEH8vsh8/IhbXmJef9BiqGMiIgoA+h9B5Cz4/9AzSqBb/KjgD5xa1AKnw9KGW/2izeGMiIiogFOF2xETu0PIPVZ8FRUQRoSuNxROAxptkDa2Sw23hjKiIiIBjARcSOn9nFACWmBzFyU2OO53VDHjAF0jBDxxsYiRERE8SIldKEGGNw7oPfshD7YiEj+NIQL50GaC+N/PCWInO1PQBdsgLdyFdTsBK8/qSiAEFCLhiT2OINUUkNZfX09qqqqsGXLFlgsFlx00UV4+OGHYTT2XKX+v//7v/HOO+/02KaqKmbMmIFXXnklmSUTERGdnhKEwbsXes9OGNw7YfDsgi7iBABIfRZUkx3WA5thPfBLRG2TEHbMbw9ojrM/tlSQvesZ6D274Ct/CNG8yrN/zzMQHre28LjJlPBjDUb9XmapP6699lqMHz8e3//+9+HxeLBy5UrMmzcP3/nOd3p9nZQSX/3qV3H11Vfjhhtu6PEcl1kiIqKkkBK6UGO3ALYTet8BCKkAABTLMERt5YjmlkOxlUOxjgSEHrrAUZiaP4Gx6RMY/AcBABHbZEQK5yPsOL9/AU1KWPe9APPxv8I/5i6Ehl0exw96+mOK5mZE5s4DsrMTf7wM1dsyS0kLZdXV1Vi2bBk+/fRT2NsnB65btw6PPfYYNm7cCF0v16bffPNNvP7663jzzTdP2o+hjIiIEkINQ+/d2xnADO6d0EXaAABSZ0E0dwKiuROh2MoRzZ0IaTxzewid/whMzZ/C2KwFNAmBaEdAKzwf0hTb5HlL3WvIqnsVgeHXI1j29bP6mLESXg/UnFwoU6cn5XiZKiFrX/ZVbW0tSkpKOgMZAFRUVMDlcqGurg5lZWWnfF0gEMBPfvITPPfcc70GNyIiorMhQk09Apjetx9CRgEAiqUEkfyp7QGsHEp2Wb96gKnW4QiOvAHBkTdA5z8MU/OnMDV/Auv+/0HW/p8jaqvoFtBOPRBgOv5XZNW9itCQCxEc9bWz+ch9EwhCrZiSvOMNQkkLZU6nEzabrce2vDztXxVtbW2nDWWvvfYaxo4di5kz47+QKhERDVJqBHrvvs4AZvDshC7cAgCQOhOiOeMRKl2KaO4kbRTMlB//EqwjEBx5I4Ijb4TOX9ctoL2IrP3/g2heJcKF8xFxzO0MaMaW/8C692eIFMyAf9w9ievWf6JgEDInBzIv/j8H6pLSuy87rpyK0/xHpaoqfv3rX2PVqlXJLOskevc+6P17IPWnH3IkIspoOguU7FGAGJhXLESopUcA03v3do2CmYsRyauEkluOqK0cirUM0CX316NqHYngyJEIjvwqdL5DnQEte98ayH1aQIvmnwPL4Teg5IyBt/x7Sa1ReDyITpmavBA4SCXtjNrtdrS1tfXY5nK5Op87lc2bN8Pr9WLevHkJr+909E21yP8oOdfriYjSmWouRNgxD+HC+VByJ6bvL2g1Ar3vQHsA26GNgoWaAXSMgo1DaNiViLaHsNNdJkwVNXsUgtmjtIDmr4Op+ROYmj9G1qHfQbGUwDv5cUCfxEXAo1FIoxGyMAEtPaiHpIWyyspKNDQ0oLGxEUOGaP1Ntm3bBofDgREjRpzyNe+//z7OO+88mFJ4661SOBmu+f8D/eG9kDkJ7JBMRJTGdOE2GFv+BXP9WliOvQPFXIRIYXtAy5mQ0oAmwq1dc8E8O6H37oNQwwAAxVykXYIc1j4Klj0a0BnP8I5pQohuAe0m6AJHII22mG4oiGsZbjeU0WMBfYLW0aROSQtlkydPxrRp0/DMM8/g0UcfhdPpxJo1a3DzzTdDCIHFixejqqoKc+bM6XzN9u3bMWnSpGSVeGpCIFpQCdmWxbsviWhQCxdfBER9MLX+G8amT2A+9mdYjr4NxTykW0Abn9iApka1UbCOAObeCX2oEQAghQFKzjiEhl7W2ZoiLv3A0oEQUK2nHsBIKCUKSAm1pCT5xx6EknrRfPXq1aiqqsKiRYtgtVpx6aWXYvny5QCAAwcOwO/399i/qakJCxYsSGaJRETUG0M2wkMuRHjIhRBRL4wt/4ap+ROYj/0/WI6+1R7Q5rcHtHFnHdBE2NkjgBm8ezpHwVSTA9HccoSGXaHdEZkzduCMgg0Qos0JZex4wJy4xc2pS1KbxyYC+5QREaWeFtA2wtT8CQzOLyCkok2gL5yPcNF8KNljzxzQpNI+F2yX1qDVsxP64HHtKWGAkj2mcwQsaitP+BqPg144DASCiJ4/DzBwVcZ4SYvmsYnCUEZElF5ExANj679hav4YBudWLaBZhnZd4mwPaCLi6hHADJ7dEGoIAKAaC3oEMCVnHKDj0j7JJJqaEJ1yDmTx0FSXklEYys4SQxkRUf+IiKfnCBpUKJYSAAL64DEAgBR6KNmjuwJYbjlU85D0vbtzMAj4AQhEZ58HsHF7XKVFR38iIhp8pDEX4aEXIzz0YoiISwtoLRshhR6hoRdrvcFyxgF6S6pLpW6E14fouTMZyJKMoYyIiJJCGvMQHnoJwkMvSXUp1AvhdkMtLIIsiG0dToofRmAiIiLSqCoQDEIdNy7VlQxKDGVEREQEABBuF5SRoyBzuKxgKjCUERERkdYoVlGhjipLdSWDFkMZERERaY1ix4wDLLzpIlUYyoiIiAa7cAjSaII6fHiqKxnUGMqIiIgGOeFyQ5k4kZ37U4yhjIiIaDAL+CFzcyGLhqS6kkGPoYyIiGgQE14flAkT2Sg2DfAMEBERDVJsFJteGMqIiIgGI1UFQiE2ik0jDGVERESDkHA5oYwYyUaxaYShjIiIaLBRooAEG8WmGYYyIiKiQUa0OaGMZaPYdMNQRkRENJh0NIodVprqSugEDGVERESDCBvFpi+GMiIiosHCz0ax6YyhjIiIaJAQPjaKTWc8K0RERIOAcLuhDhnCRrFpjKGMiIgo07U3ilXGsFFsOmMoIyIiynAdjWKRk5PqUqgXDGVERESZjI1iBwyGMiIiogzGRrEDB0MZERFRpgqHIE1mNoodIBjKiIiIMpRwubUWGGwUOyAwlBEREWWizkaxRamuhGLEUEZERJSB2Ch24OGZIiIiyjBsFDswMZQREREliqIAUib3mGwUO2Bx5h8REVG8RCIQgQAQDgOQkAYjRDgM6XAAen1SShAuJ5RRZWwUOwAxlBEREfVXOKSFsIgCQIW0ZEEZUgxpd0BmZwNZWdAdOQz9rh2Q+QWA0ZjYejoaxY4cldjjUEIwlBEREcVCSiDUHsJUFZAqZHYOlNIRkPkFWgg7RYNWdcRISJMZhuqtkHk2wGROWImizQll/ATAnLhjUOIwlBEREZ2KlEAwABEMAqoEICFtNihlYyDzbJDZOYDJFNtbFRcjajwXhi8+h7QqQJY1/vWyUeyAl9RQVl9fj6qqKmzZsgUWiwUXXXQRHn74YRhPMZzb3NyMVatW4eOPP4bJZMLll1+Ohx56CKYY/wIQERH1iaoCgfYQ1k4WFGgjYbb2EHYWTVil3YHorDkwbPkMUBTInNx4VN1JOF2InjONjWIHsKTefbly5Urk5+dj/fr1ePXVV7FlyxasXr36pP2klJ37fvTRR/jjH/+InTt34sMPP0xmuURElMmUKITXA9HSrD1cLsicHCjlkxCdNRuRhRcgOv1c7fJjXn5cwo7MtSFy7mxIoYNwu+LwIdr5/ZD5+ZBDhsTvPSnphJTJuVe3uroay5Ytw6effgq7Xeubsm7dOjz22GPYuHEjdN2a223atAnf/OY38c9//hOWMyyg2tTkSWjdAKCrOwT9nt2QeXkJPxYRUdrS6ZJ2B2HcKQqgKNooWDgMAJBGI6SjEGphoTYKZrUmr9FqKARD9RcQXm9ceomJpiZEZ87SbiagtFZUdPoR0qSNcdbW1qKkpKQzkAFARUUFXC4X6urqUFZW1rl98+bNmDBhAl544QX87//+L0wmE6677jqsWLGiR3hLFpllhbRagUgk6ccmIkobkShENNxtg9D+kFIbRTLoIfUG7Wu9Pr4BR1W1YKWqgKr9KRQFUFRtG2RXPR01CdG5XRoMgNEIpWgIpKOw885IbZ8UMJsRnToD+u010DU3ay0z+lmLcLu0RrEMZANe0kKZ0+mEzWbrsS2vfeSpra2tRyg7fvw4qqurcf755+ODDz7Atm3b8M1vfhPFxcW4/vrrk1VyJ1lUhCjXDiMi0gJQNKr144pGtKAWCQPBIBAIQBcKal97vBBSac9EHWFDdo62Sb0eoiNQqe0NVjv2FaI9aKEr9+m1UAWjEciyQjWa2r83AWaT9rxeD+h1gMEAqdO3f9/+SFX46o3RCKXyHGDXTuiOHoEsLOx7kFVVIBxmo9gMkdLZgB1XTsUJf1mklMjJycGKFSsAAHPmzMHSpUuxdu3alIQyIiJqp9NpdxyaTOiY+9J9Dozafd/u4S0a1QJcOAQEAtqVB4NBaw9hMnYbYWsPbR2hqmPULR1DVTzo9VAmTYY0maA/sL/PTWaFsw3KyDI2is0QSQtldrsdbW1tPba5XK7O57orKirqHEXrUFpaig0bNiS2SCIiih+DFrQksjo3JXnBoYFBCKjjxgMmE/S7dkLa7bHdVBCNAhBsFJtBkjZBq7KyEg0NDWhsbOzctm3bNjgcDowYMaLHvpMnT8aRI0fg8XRN4j9y5AiGDRuWrHKJiIiSSh05CtEp50C0tgLh0Bn3F04nlHHj2Sg2gyQtlE2ePBnTpk3DM888A4/Hg8OHD2PNmjW4+eabIYTA4sWL8e9//xsA8OUvfxnFxcV48skn4fV6sWXLFrzzzju47rrrklUuERFR0smhJYieOxPC49Uu855OR6PYEg5WZJKk3sq4evVqeL1eLFq0CF//+texcOFCLF++HABw4MAB+P1+AIBer8eLL76Iw4cP4/zzz8c999yD++67D4sXL05muUREREkn7Q5EZ87S5t/5vKfcR7jcUCZMZKPYDJO0PmWJkow+ZUREREnn88GwdQuEEoW0dZtn7fcDBgOiM2dn7g0QGay3PmXJb/pFREREZ5adjeiMmZBmC0Rbq7ZNSgivV1t0nIEs4zCUERERpSuLRVvqKa8AoqlJaxRbPJSNYjMUQxkREVE6MxqhnDMV6rBSIBqFMpaNYjMV55QRERENBFJCeD2QubYz70tpi3PKiIiIBjohGMgyHEMZERERURpgKCMiIiJKAwxlRERERGmAoYyIiIgoDTCUEREREaUBhjIiIiKiNMBQRkRERJQGBnzzWCIiIqJMwJEyIiIiojTAUEZERESUBhjKiIiIiNIAQxkRERFRGsi4UFZfX4/ly5djzpw5WLhwIVatWoVIJJLqsjLerl27cPnll+PCCy/ssf0///kPli1bhhkzZmDx4sV47bXXUlRh5jt69CjuuecezJkzB+eddx7uvfdeNDQ0AOB5SJYvvvgCt9xyC2bMmIF58+bh/vvvR1NTEwCeg1R58sknMXHixM7veR6S4/zzz0dlZSWmTJnS+Xj88ccB8Bz0SmaYa665Rj700EPS5XLJI0eOyKuuuko+/fTTqS4ro61du1bOnz9frlixQl5wwQWd2xsbG+X06dPl73//exkIBORnn30mZ8yYIT/66KMUVpu5Lr/8cvnAAw9Ij8cjm5ub5de//nV555138jwkidPplNOnT5e/+c1vZDgcls3NzfKWW26Rd999N89Bimzfvl3Onj1bTpgwQUrJ/yclU0VFhaypqTlpO89B7zJqpKy6uhrbt2/Hd7/7XdhsNpSWluKuu+7CG2+8AVVVU11exvL5fPjDH/6AuXPn9tj+7rvvorS0FDfddBMsFgtmzJiBpUuX4vXXX09RpZnL7XajsrISDz74IHJycuBwOLBs2TJs2rSJ5yFJwuEwHnnkEdx6660wGo1wOBy4+OKLsXPnTp6DFFBVFY8//jhuu+22zm08D8nh8/kQiURgs9lOeo7noHcZFcpqa2tRUlICu93eua2iogIulwt1dXUprCyzXX/99Rg2bNhJ22tra1FRUdFj2+TJk1FdXZ2s0gYNm82Gp556CsXFxZ3b6uvrUVxczPOQJEVFRbj22msBAFJK7Nu3D2+99RaWLFnCc5ACr7/+OiwWCy6//PLObTwPyeFyuQAAzz77LL70pS/hS1/6Eh577DF4vV6egzPIqFDmdDpPSuZ5eXkAgLa2tlSUNKid6nzk5+fzXCTB/v37sWbNGqxYsYLnIcl27tyJyspKXH755ZgyZQruu+8+noMka25uxgsvvIAf/OAHPbbzPCRHNBrF1KlTMXfuXHzwwQf47W9/i61bt+Lxxx/nOTiDjAplpyLbFywQQqS4EgK088FzkVg1NTW45ZZbcNttt+GKK6445T48D4lTXl6Ompoa/PnPf8aBAwdw//33n3I/noPEeeqpp3D99ddjzJgxZ9yX5yH+Ro4ciTfeeAPLli2DyWTCmDFjcP/992Pt2rWdv5O74znoklGhzG63n5S2O4ZRu1/SpOQoKCg46Xw4nU6eiwT6+OOPceutt2LlypVYuXIlAJ6HVBBCYOzYsbj//vuxbt06KIrCc5AkGzZsQHV1Ne6+++6TnuPfhdQZPnw4pJSn/D3Nc9Alo0JZZWUlGhoa0NjY2Llt27ZtcDgcGDFiRAorG5ymTJmCmpqaHtuqq6sxderUFFWU2bZu3Ypvf/vb+NGPfoSbbrqpczvPQ3K89957uOaaa3ps0+m0/8UuXLiQ5yBJ3n33XTQ0NGDBggWYM2dO5zmZM2cOJkyYwPOQBFu3bsXTTz/dY9u+fftgNBoxadIknoPepOy+zwS54YYb5IMPPijdbresq6uTl112mXz++edTXdag8Morr/RoidHS0iLPPfdc+bvf/U4Gg0G5ceNGOW3aNPmf//wnhVVmpkgkIpcsWSJ/85vfnPQcz0NyHD9+XM6YMUM+//zzMhAIyObmZnn77bfLG2+8kecgiZxOp6yvr+98bNmyRU6YMEHW19fLI0eO8DwkQV1dnTznnHPkr3/9axkKheS+ffvkZZddJquqqvh34QyElKe4wDuANTQ0oKqqCp999hmsVisuvfRSPPDAA9Dr9akuLWNdcsklOHbsGFRVRTQahclkAgCsW7cOx48fx9NPP43du3dj2LBhuOOOO3DVVVeluOLMs3nzZtx8882dP/vueB6SZ+vWrXjqqaewfft25OTk4LzzzsNDDz2E4uJifPbZZzwHKXDkyBFcdNFF2LVrFwDwPCTJhg0b8Oyzz2Lv3r0oKCjA4sWLcd9998FkMvEc9CLjQhkRERHRQJRRc8qIiIiIBiqGMiIiIqI0wFBGRERElAYYyoiIiIjSAEMZERERURpgKCMiIiJKAwxlRERERGmAoYyI6DTefvttLFiwINVlENEgweaxRERERGmAI2VEREREaYChjIgywh/+8AcsXLgQfr8fAOB2uzF37ly88847vb4uEAjg4Ycfxty5czF9+nRcc8012LBhAwDgT3/66kWDGQAAA15JREFUE+bMmQMA+NnPfoYpU6Z0PiorKzFx4kQcPXoUALBr1y584xvfwKxZszBnzhw89dRTiEQiCfzERJRpGMqIKCMsW7YMo0aNwpo1awAAzz33HCoqKrB06dJeX/fb3/4WtbW1WLt2LTZv3oyvfvWrePDBBxGNRnvst2LFClRXV3c+Lr74YixYsADDhg1DIBDAHXfcgVmzZuHTTz/FW2+9hU2bNuG5555L2OcloszDUEZEGUEIgR/+8Id4/fXX8be//Q3vvPMOqqqqzvg6t9sNg8GArKws6PV6XH/99fj4449hMBhO+5o33ngDn332GX70ox9BCIEPP/wQkUgE3/zmN2EymTBs2DAsX74cb731Vjw/IhFluNP/X4eIaIAZOXIk7rzzTnzrW9/C9773PZSWlp7xNTfddBM++OADLFiwAPPmzcOXv/xlLFmyBEaj8ZT779mzB0899RR+/vOfw263AwAOHz4Mp9OJKVOm9NhXVVWEw2GYTKaz/3BElPH+fzt3z5pIFIVx/CEoYqOF2ARBrPIB7MQm1tY2+gUEBRHsBi0cUiSdoFXaYMTGQhQbSZNCbNTKMiBpM/iCooGwVQQJyS5sdhkm/19551yY0wwPc+4MoQyAoywWC3m9Xj09Pf1RfSgUUq/X02g00nA41PX1tZrNpu7u7j7U7nY7FQqF46jyncfjUSQSUb/f/642APxAjC8BOMZ4PNZgMND9/f3xjNjvbLdbvb6+KhaLyTAMtdttTSYTzefzD7WmaSoYDCqbzZ6sh8NhPT8/a7PZHNeWy6XW6/XfNwXgxyCUAXCE/X4vwzBULBZ1cXGhXC4nwzB0OBy+3JfP51WpVLRarfT29qbpdCq3263z8/OTum63q4eHB93c3Ojs7PTRGY/HFQwGdXV1pfV6rZeXF5VKJVWr1W/vE4BzEcoAOEK9Xpff71cqlZIkZTIZeb1e1ev1L/eZpinLsnR5ealoNKrb21vVarXjebF3rVZLlmUpkUic/Bqj0+nI5XKp0WhosVgoHo8rmUwqEAioXC7/s34BOA9/9AcAALAB3pQBAADYAF9fAnCs2WymdDr96XWfz6fHx8f/eEcA8DnGlwAAADbA+BIAAMAGCGUAAAA2QCgDAACwAUIZAACADRDKAAAAbOAXN02D2eRtfuEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 720x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "train_sizes = np.linspace(0,1,21)[1:]\n",
    "\n",
    "train_sizes, train_scores, test_scores = learning_curve(estimator=dtc,\n",
    "                                                        X=X,\n",
    "                                                        y=y,\n",
    "                                                        cv=cv,\n",
    "                                                        train_sizes=train_sizes)\n",
    "\n",
    "train_scores_means = train_scores.mean(axis=1)\n",
    "train_scores_stds = train_scores.std(axis=1)\n",
    "test_scores_means = test_scores.mean(axis=1)\n",
    "test_scores_stds = test_scores.std(axis=1)\n",
    "\n",
    "plt.figure(figsize=(10, 4))    \n",
    "plt.fill_between(train_sizes,\n",
    "                train_scores_means - train_scores_stds,\n",
    "                 train_scores_means + train_scores_stds,\n",
    "                 color='b', alpha=0.2)\n",
    "plt.plot(train_sizes, train_scores_means)\n",
    "plt.fill_between(train_sizes,\n",
    "                test_scores_means - test_scores_stds,\n",
    "                 test_scores_means + test_scores_stds,\n",
    "                 color='r', alpha=0.2)\n",
    "plt.plot(train_sizes, test_scores_means)\n",
    "plt.xlabel('x_size')\n",
    "plt.ylabel('accuracy')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## GridSearch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {},
   "outputs": [],
   "source": [
    "svm = SVC(C=1.0, kernel='rbf', probability=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.69444444, 0.69444444, 0.69444444, 0.69444444, 0.73611111,\n",
       "       0.68055556, 0.69444444, 0.73611111, 0.68055556, 0.69444444,\n",
       "       0.69444444, 0.69444444, 0.69444444, 0.80555556, 0.68055556,\n",
       "       0.69444444, 0.80555556, 0.68055556, 0.69444444, 0.69444444,\n",
       "       0.69444444, 0.69444444, 0.84722222, 0.68055556, 0.69444444,\n",
       "       0.84722222, 0.68055556, 0.80555556, 0.73611111, 0.73611111,\n",
       "       0.97222222, 0.86111111, 0.76388889, 0.97222222, 0.86111111,\n",
       "       0.76388889, 0.80555556, 0.69444444, 0.73611111, 0.97222222,\n",
       "       0.81944444, 0.76388889, 0.97222222, 0.81944444, 0.76388889,\n",
       "       0.80555556, 0.69444444, 0.73611111, 0.97222222, 0.83333333,\n",
       "       0.76388889, 0.97222222, 0.83333333, 0.76388889, 0.80555556,\n",
       "       0.73611111, 0.73611111, 0.97222222, 0.86111111, 0.76388889,\n",
       "       0.97222222, 0.86111111, 0.76388889, 0.80555556, 0.69444444,\n",
       "       0.73611111, 0.97222222, 0.81944444, 0.76388889, 0.97222222,\n",
       "       0.81944444, 0.76388889, 0.80555556, 0.69444444, 0.73611111,\n",
       "       0.97222222, 0.83333333, 0.76388889, 0.97222222, 0.83333333,\n",
       "       0.76388889])"
      ]
     },
     "execution_count": 164,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import GridSearchCV\n",
    "parameters = {'C':(0.1, 1, 1.0),\n",
    "              'kernel':('rbf', 'poly', 'sigmoid'),\n",
    "              'degree':(1,2,3),\n",
    "              'gamma':(0.1,1,1.0)} # , scoring='roc_auc'\n",
    "clf = GridSearchCV(svm, parameters, cv=cv, )\n",
    "clf.fit(X, y)\n",
    "clf.cv_results_['mean_test_score']\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean_fit_time</th>\n",
       "      <th>std_fit_time</th>\n",
       "      <th>mean_score_time</th>\n",
       "      <th>std_score_time</th>\n",
       "      <th>param_C</th>\n",
       "      <th>param_degree</th>\n",
       "      <th>param_gamma</th>\n",
       "      <th>param_kernel</th>\n",
       "      <th>params</th>\n",
       "      <th>split0_test_score</th>\n",
       "      <th>split1_test_score</th>\n",
       "      <th>split2_test_score</th>\n",
       "      <th>split3_test_score</th>\n",
       "      <th>mean_test_score</th>\n",
       "      <th>std_test_score</th>\n",
       "      <th>rank_test_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.000776</td>\n",
       "      <td>0.000205</td>\n",
       "      <td>0.000530</td>\n",
       "      <td>0.000234</td>\n",
       "      <td>0.1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.1</td>\n",
       "      <td>rbf</td>\n",
       "      <td>{'C': 0.1, 'degree': 1, 'gamma': 0.1, 'kernel'...</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.555556</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.694444</td>\n",
       "      <td>0.092128</td>\n",
       "      <td>57</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.000473</td>\n",
       "      <td>0.000096</td>\n",
       "      <td>0.000311</td>\n",
       "      <td>0.000006</td>\n",
       "      <td>0.1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.1</td>\n",
       "      <td>poly</td>\n",
       "      <td>{'C': 0.1, 'degree': 1, 'gamma': 0.1, 'kernel'...</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.555556</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.694444</td>\n",
       "      <td>0.092128</td>\n",
       "      <td>57</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.000452</td>\n",
       "      <td>0.000011</td>\n",
       "      <td>0.000363</td>\n",
       "      <td>0.000039</td>\n",
       "      <td>0.1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.1</td>\n",
       "      <td>sigmoid</td>\n",
       "      <td>{'C': 0.1, 'degree': 1, 'gamma': 0.1, 'kernel'...</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.555556</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.694444</td>\n",
       "      <td>0.092128</td>\n",
       "      <td>57</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.000470</td>\n",
       "      <td>0.000015</td>\n",
       "      <td>0.000333</td>\n",
       "      <td>0.000011</td>\n",
       "      <td>0.1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>rbf</td>\n",
       "      <td>{'C': 0.1, 'degree': 1, 'gamma': 1, 'kernel': ...</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.555556</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.694444</td>\n",
       "      <td>0.092128</td>\n",
       "      <td>57</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.000449</td>\n",
       "      <td>0.000051</td>\n",
       "      <td>0.000326</td>\n",
       "      <td>0.000009</td>\n",
       "      <td>0.1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>poly</td>\n",
       "      <td>{'C': 0.1, 'degree': 1, 'gamma': 1, 'kernel': ...</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.555556</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.736111</td>\n",
       "      <td>0.106683</td>\n",
       "      <td>47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>76</th>\n",
       "      <td>0.000434</td>\n",
       "      <td>0.000016</td>\n",
       "      <td>0.000317</td>\n",
       "      <td>0.000015</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>poly</td>\n",
       "      <td>{'C': 1.0, 'degree': 3, 'gamma': 1, 'kernel': ...</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.068041</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>0.000428</td>\n",
       "      <td>0.000009</td>\n",
       "      <td>0.000326</td>\n",
       "      <td>0.000016</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>sigmoid</td>\n",
       "      <td>{'C': 1.0, 'degree': 3, 'gamma': 1, 'kernel': ...</td>\n",
       "      <td>0.944444</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.763889</td>\n",
       "      <td>0.106683</td>\n",
       "      <td>35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>0.000438</td>\n",
       "      <td>0.000014</td>\n",
       "      <td>0.000319</td>\n",
       "      <td>0.000008</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>rbf</td>\n",
       "      <td>{'C': 1.0, 'degree': 3, 'gamma': 1.0, 'kernel'...</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.944444</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.944444</td>\n",
       "      <td>0.972222</td>\n",
       "      <td>0.027778</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>79</th>\n",
       "      <td>0.000434</td>\n",
       "      <td>0.000019</td>\n",
       "      <td>0.000321</td>\n",
       "      <td>0.000020</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>poly</td>\n",
       "      <td>{'C': 1.0, 'degree': 3, 'gamma': 1.0, 'kernel'...</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.068041</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80</th>\n",
       "      <td>0.000432</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000319</td>\n",
       "      <td>0.000010</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>sigmoid</td>\n",
       "      <td>{'C': 1.0, 'degree': 3, 'gamma': 1.0, 'kernel'...</td>\n",
       "      <td>0.944444</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.763889</td>\n",
       "      <td>0.106683</td>\n",
       "      <td>35</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>81 rows × 16 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    mean_fit_time  std_fit_time  mean_score_time  std_score_time param_C  \\\n",
       "0        0.000776      0.000205         0.000530        0.000234     0.1   \n",
       "1        0.000473      0.000096         0.000311        0.000006     0.1   \n",
       "2        0.000452      0.000011         0.000363        0.000039     0.1   \n",
       "3        0.000470      0.000015         0.000333        0.000011     0.1   \n",
       "4        0.000449      0.000051         0.000326        0.000009     0.1   \n",
       "..            ...           ...              ...             ...     ...   \n",
       "76       0.000434      0.000016         0.000317        0.000015       1   \n",
       "77       0.000428      0.000009         0.000326        0.000016       1   \n",
       "78       0.000438      0.000014         0.000319        0.000008       1   \n",
       "79       0.000434      0.000019         0.000321        0.000020       1   \n",
       "80       0.000432      0.000003         0.000319        0.000010       1   \n",
       "\n",
       "   param_degree param_gamma param_kernel  \\\n",
       "0             1         0.1          rbf   \n",
       "1             1         0.1         poly   \n",
       "2             1         0.1      sigmoid   \n",
       "3             1           1          rbf   \n",
       "4             1           1         poly   \n",
       "..          ...         ...          ...   \n",
       "76            3           1         poly   \n",
       "77            3           1      sigmoid   \n",
       "78            3           1          rbf   \n",
       "79            3           1         poly   \n",
       "80            3           1      sigmoid   \n",
       "\n",
       "                                               params  split0_test_score  \\\n",
       "0   {'C': 0.1, 'degree': 1, 'gamma': 0.1, 'kernel'...           0.666667   \n",
       "1   {'C': 0.1, 'degree': 1, 'gamma': 0.1, 'kernel'...           0.666667   \n",
       "2   {'C': 0.1, 'degree': 1, 'gamma': 0.1, 'kernel'...           0.666667   \n",
       "3   {'C': 0.1, 'degree': 1, 'gamma': 1, 'kernel': ...           0.666667   \n",
       "4   {'C': 0.1, 'degree': 1, 'gamma': 1, 'kernel': ...           0.777778   \n",
       "..                                                ...                ...   \n",
       "76  {'C': 1.0, 'degree': 3, 'gamma': 1, 'kernel': ...           0.722222   \n",
       "77  {'C': 1.0, 'degree': 3, 'gamma': 1, 'kernel': ...           0.944444   \n",
       "78  {'C': 1.0, 'degree': 3, 'gamma': 1.0, 'kernel'...           1.000000   \n",
       "79  {'C': 1.0, 'degree': 3, 'gamma': 1.0, 'kernel'...           0.722222   \n",
       "80  {'C': 1.0, 'degree': 3, 'gamma': 1.0, 'kernel'...           0.944444   \n",
       "\n",
       "    split1_test_score  split2_test_score  split3_test_score  mean_test_score  \\\n",
       "0            0.555556           0.777778           0.777778         0.694444   \n",
       "1            0.555556           0.777778           0.777778         0.694444   \n",
       "2            0.555556           0.777778           0.777778         0.694444   \n",
       "3            0.555556           0.777778           0.777778         0.694444   \n",
       "4            0.555556           0.777778           0.833333         0.736111   \n",
       "..                ...                ...                ...              ...   \n",
       "76           0.833333           0.888889           0.888889         0.833333   \n",
       "77           0.722222           0.722222           0.666667         0.763889   \n",
       "78           0.944444           1.000000           0.944444         0.972222   \n",
       "79           0.833333           0.888889           0.888889         0.833333   \n",
       "80           0.722222           0.722222           0.666667         0.763889   \n",
       "\n",
       "    std_test_score  rank_test_score  \n",
       "0         0.092128               57  \n",
       "1         0.092128               57  \n",
       "2         0.092128               57  \n",
       "3         0.092128               57  \n",
       "4         0.106683               47  \n",
       "..             ...              ...  \n",
       "76        0.068041               19  \n",
       "77        0.106683               35  \n",
       "78        0.027778                1  \n",
       "79        0.068041               19  \n",
       "80        0.106683               35  \n",
       "\n",
       "[81 rows x 16 columns]"
      ]
     },
     "execution_count": 167,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "pd.DataFrame(clf.cv_results_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "({'C': 1, 'degree': 1, 'gamma': 1, 'kernel': 'rbf'}, 0.9722222222222223)"
      ]
     },
     "execution_count": 168,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.best_params_, clf.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=1.0, break_ties=False, cache_size=200, class_weight=None, coef0=0.0,\n",
       "    decision_function_shape='ovr', degree=3, gamma='scale', kernel='rbf',\n",
       "    max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
       "    tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 172,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "svm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "from scipy.stats import uniform\n",
    "from sklearn.utils.fixes import loguniform\n",
    "\n",
    "distributions = dict(gamma=uniform(0, 10),\n",
    "                     C=loguniform(0.01,10),\n",
    "                     kernel=['rbf', 'poly', 'sigmoid'],\n",
    "                     degree=[1,2,3])\n",
    "\n",
    "clf = RandomizedSearchCV(svm, distributions, random_state=1,\n",
    "                         scoring='accuracy', cv=cv)\n",
    "search = clf.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=3.4588581370567506, break_ties=False, cache_size=200, class_weight=None,\n",
       "    coef0=0.0, decision_function_shape='ovr', degree=2, gamma=6.852195003967595,\n",
       "    kernel='rbf', max_iter=-1, probability=False, random_state=None,\n",
       "    shrinking=True, tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 185,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "search.best_estimator_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'C': 3.4588581370567506,\n",
       " 'degree': 2,\n",
       " 'gamma': 6.852195003967595,\n",
       " 'kernel': 'rbf'}"
      ]
     },
     "execution_count": 186,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "search.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0"
      ]
     },
     "execution_count": 188,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "search.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f368d5383d0>"
      ]
     },
     "execution_count": 190,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "svm = search.best_estimator_\n",
    "svm.fit(X, y)\n",
    "plot_decision_regions(X, y, svm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}