Random Forest Solution.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np  # to build the algorithm\n",
    "import matplotlib.pyplot as plt  # to visualize\n",
    "from sklearn.datasets import make_circles  # to generate a dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x277d0dd0ef0>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1800x1200 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Generate a dataset\n",
    "X, y = make_circles(n_samples=100, noise=0.1, factor=0.5, random_state=0)\n",
    "plt.figure(dpi=300)\n",
    "plt.scatter(X[:, 0][y == 0], X[:, 1][y == 0], label=0)\n",
    "plt.scatter(X[:, 0][y == 1], X[:, 1][y == 1], label=1)\n",
    "plt.xlabel('feature 0')\n",
    "plt.ylabel('feature 1')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy:  0.96\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "clf = RandomForestClassifier(n_estimators=30)\n",
    "clf.fit(X, y)\n",
    "\n",
    "X_test, y_test = make_circles(n_samples=100, noise=0.1, factor=0.5, random_state=42)\n",
    "y_pred = clf.predict(X_test)\n",
    "acc = sum(y_pred == y_test)/len(y_test)\n",
    "print('Accuracy: ', acc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.48652932, 0.51347068])"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.feature_importances_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "code_folding": [
     0
    ]
   },
   "outputs": [],
   "source": [
    "# Challenge 1\n",
    "def gini_calculator(y):\n",
    "    '''\n",
    "    Calculates the gini impurity of a set\n",
    "    Arguments\n",
    "        y: np.array() containing the labels\n",
    "    Returns\n",
    "        gini: 1 - p0^2 - p1^2\n",
    "    '''\n",
    "    probs = []\n",
    "    for label in np.unique(y):\n",
    "        prob = np.mean(y == label)\n",
    "        probs.append(prob)\n",
    "    gini = 1-np.sum(np.square(np.array(probs)))\n",
    "    return gini"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gini_calculator(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Gini impurity')"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1800x1200 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Sanity check\n",
    "num_datapoints = 20\n",
    "plt.figure(dpi=300)\n",
    "for i in range(num_datapoints+1):\n",
    "    num_ones = i\n",
    "    num_zeros = num_datapoints-i\n",
    "    p1 = num_ones/num_datapoints\n",
    "    combined_set = np.concatenate((np.ones(num_ones), np.zeros(num_zeros)))\n",
    "    gini = gini_calculator(combined_set)\n",
    "    plt.scatter(p1, gini, color='k')\n",
    "plt.xlabel('Ratio of 1')\n",
    "plt.ylabel('Gini impurity')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def gini_of_a_split(y1, y2):\n",
    "    '''\n",
    "    Weighted average gini of two sets\n",
    "    Arguments\n",
    "        y1: np.array() containing the labels of set 1\n",
    "        y2: np.array() containing the labels of set 2\n",
    "    Returns\n",
    "        avg_gini: Weighted average gini of y1 and y2\n",
    "    '''\n",
    "    g1 = gini_calculator(y1)\n",
    "    w1 = len(y1)/(len(y1)+len(y2))\n",
    "    g2 = gini_calculator(y2)\n",
    "    w2 = len(y2)/(len(y1)+len(y2))\n",
    "    avg_gini = g1*w1 + g2*w2\n",
    "    return avg_gini"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "code_folding": [
     4
    ]
   },
   "outputs": [],
   "source": [
    "# Challenge 2\n",
    "from operator import itemgetter\n",
    "\n",
    "\n",
    "def split_finder(X, y):\n",
    "    '''\n",
    "    Finds the best split that minimizes the average gini.\n",
    "    Best split is defined by a feature index and its value.\n",
    "\n",
    "    Arguments\n",
    "        m = num_of_datapoints\n",
    "        n = num_of_features\n",
    "        X: np.array() shape (m, n)\n",
    "        y: np.array() shape (m, 1)\n",
    "\n",
    "    Returns\n",
    "        best_split_feature: integer, best feature index\n",
    "        best_split_value: float, best value\n",
    "\n",
    "    '''\n",
    "    #Finding the best split point in a decision tree \n",
    "    #involves evaluating the cost of each value in the training dataset for each input variable.\n",
    "    # Calculate avg gini for all splits\n",
    "    splits = []\n",
    "    for j in range(X.shape[1]):  # j is the split feature\n",
    "        for i in range(X.shape[0]):\n",
    "            Xij = X[i, j]  # Xij is the split value\n",
    "            y1 = y[X[:, j] < Xij]\n",
    "            y2 = y[X[:, j] >= Xij]\n",
    "            # Splitting using >= implies\n",
    "            # Xij is the upper bound\n",
    "            avg_gini = gini_of_a_split(y1, y2)\n",
    "            splits.append([j, Xij, avg_gini])\n",
    "\n",
    "    # Find the j and upper bound with min_avg_gini\n",
    "    best_split = min(splits, key=itemgetter(2))\n",
    "    best_split_feature, upper_bound, min_avg_gini = best_split\n",
    "\n",
    "    # Find the lower bound\n",
    "    j_values = X[:, best_split_feature]\n",
    "    lower_bound = max(j_values[j_values < upper_bound])\n",
    "\n",
    "    # Find the decision boundary\n",
    "    best_split_value = np.mean((lower_bound, upper_bound))\n",
    "\n",
    "    return best_split_feature, best_split_value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize the first split\n",
    "best_split = split_finder(X, y)\n",
    "plt.figure(dpi=200)\n",
    "plt.scatter(X[:, 0][y == 0], X[:, 1][y == 0], label=0, s=5)\n",
    "plt.scatter(X[:, 0][y == 1], X[:, 1][y == 1], label=1, s=5)\n",
    "plt.xlabel('feature 0')\n",
    "plt.ylabel('feature 1')\n",
    "plt.legend()\n",
    "\n",
    "split_feature = best_split[0]\n",
    "split_value = best_split[1]\n",
    "boundary_limits = [min(X[:, 1]), max(X[:, 1])]\n",
    "\n",
    "if split_feature == 0:\n",
    "    plt.plot([split_value, split_value], boundary_limits, lw=.5)\n",
    "if split_feature == 1:\n",
    "    plt.plot(boundary_limits, [split_value, split_value], lw=.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def splitter(X, y):\n",
    "    '''\n",
    "    This is one node split i.e. building block of a tree.\n",
    "\n",
    "    Given X and y,\n",
    "    - finds the split\n",
    "    - splits the datasets into 2 subsets\n",
    "    - returns the subsets and the split\n",
    "\n",
    "    Arguments\n",
    "        m = num_of_datapoints\n",
    "        n = num_of_features\n",
    "        X: np.array() shape (m, n)\n",
    "        y: np.array() shape (m, 1)\n",
    "\n",
    "    Returns\n",
    "        subset1: list of np arrays, [X1, y1]\n",
    "            X1: np.array() shape (m1, n)\n",
    "            y1: np.array() shape (m1, 1)\n",
    "        subset2: list of np arrays, [X2, y2]\n",
    "            X2: np.array() shape (m2, n)\n",
    "            y2: np.array() shape (m2, 1)\n",
    "        where m = m1 + m2\n",
    "\n",
    "        split: a tuple, (split_feature, split_value)\n",
    "            split_feature: integer, best feature index\n",
    "            split_value: float, best value\n",
    "    '''\n",
    "    split = split_finder(X, y)\n",
    "\n",
    "    X1 = X[X[:, split[0]] < split[1]]\n",
    "    y1 = y[X[:, split[0]] < split[1]]\n",
    "\n",
    "    X2 = X[X[:, split[0]] > split[1]]\n",
    "    y2 = y[X[:, split[0]] > split[1]]\n",
    "\n",
    "    return (X1, y1), (X2, y2), split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABCQAAAK2CAYAAACW6E17AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAAewgAAHsIBbtB1PgAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAIABJREFUeJzs3X98k/W9P/zXp0kpkCa0kcKkwR/cyuySoIQf9ftQ06z36jbBo9doNzkeWzxOlG5nzvvrd/6coO52+97qmIfRuZ3dSjvnj1kIKu5Mezxtg8psJWh/BEUdeJvUjSLglQRWaHrdf7TJWvoraa/8fj0fjzxImivv653CRZv35/N5f4SiKCAiIiIiIiIiSqScZCdARERERERERNmHBQkiIiIiIiIiSjgWJIiIiIiIiIgo4ViQICIiIiIiIqKEY0GCiIiIiIiIiBKOBQkiIiIiIiIiSjgWJIiIiIiIiIgo4ViQICIiIiIiIqKEY0GCiIiIiIiIiBKOBQkiIiIiIiIiSjgWJIiIiIiIiIgo4ViQICIiIiIiIqKEY0GCiIiIiIiIiBKOBQkiIiIiIiIiSjgWJIiIiIiIiIgo4ViQICIiIiIiIqKEY0GCiIiIiIiIiBKOBQkiIiIiIiIiSjgWJIiIiIiIiIgo4ViQICIiIiIiIqKE0yY7AZqcECIPgHXoYS+AUBLTISIiIiIiosynAVA0dL9TUZQ+tU/AgkR6sAJoT3YSRERERERElJVWAHhH7aBcskFERERERERECccZEumhN3ynra0NZ599djJzISIiIiIiogz32WefYeXKleGHvRMdO1UsSKSHSM+Is88+GyaTKZm5EBERERERUXaJSx9DLtkgIiIiIiIiooRjQYKIiIiIiIiIEo4FCSIiIiIiIiJKOBYkiIiIiIiIiCjhWJAgIiIiIiIiooRjQYKIiIiIiIiIEo4FCSIiIiIiIiJKOBYkiIiIiIiIiCjhWJAgIiIiIiIiooTTJjsBIiIiIiKigYEBBAIByLKMU6dOIRQKJTslorSl0WgwY8YMGAwG5OfnIycnNecisCBBRERERERJ5ff74fP5oChKslMhygj9/f3o6+uD3++HEALFxcXQ6/XJTmsUFiSIiIiIiChpxipGCCGg0WiSmBVReguFQpFrSlEU+Hy+lCxKsCBBRERERERJMTAwMKIYkZ+fD6PRiNmzZ0MIkeTsiNKXoig4ceIEjh49ikAgEClKLF68OKWWb6ROJkRERERElFXCH5SAwWKEyWSCTqdjMYJomoQQ0Ol0MJlMyM/PBzBYpAgEAknObCQWJIiIiIiIKClkWY7cNxqNLEQQqUwIAaPRGHk8/JpLBVyyQURERBQnsizD6/UiEAhERn8NBkOy0yJKGadOnQIw+KFp9uzZSc6GKDOFl0ApihK55lIFZ0gQERERqUhRFDQ3N6OyshJGoxFmsxmlpaUwm80wGo2oqqpCc3MzdxMgAiJbe2o0Gs6OIIqT4U1iU207XRYkiIiIiFTidrthtVpRXl6O7du3j/rFLxQKobGxEeXl5bBarXC73UnKlIiIKPlYkCAiIkohsizD4/Ggra0NHo8n5dZ60viamppgt9vR3d0d1fHd3d2w2+1oamqKc2ZERESpiQUJIiKiJOMU//TndrshSRKCwWBMrwsGg5AkiTMliIgoK7EgQURElESc4p/+FEVBdXV1zMWIsGAwiJqaGhaciIgo67AgQURElCSc4p8ZWlpaov47HE9XVxdaW1tVyoiIiCg9sCBBRESUBJzinznq6upSKg4RUaratm0bhBAQQuDQoUPJTodSAAsSRERECcYp/plDlmU4nU5VYu3YsYNNTImIstDhw4exa9cu3H///fjmN7+JuXPnRgo369atS3Z6caVNdgJERETZRs0p/g6HQ52kaEq8Xq9qe7qHQiH4fD4YDAZV4hFNx8lTIXzcG8D/UZSPWTM0yU6HKGUcOnQI559/PgDgqaeeUqVgMH/+/GnHSFecIUFERJRgnOKfOQKBgKrx/H6/qvGIpurj3gBWb3kDH/eq+2+ciCZ2zjnn4Morr0x2GgnDGRJEREQJFI8p/hxRT578/HxV4+n1elXjEdHUyLIMr9eLQCCA/Px8mEwm/l9LcXP//fdjxYoVWLFiBebPnz9iFkamY0GCiIgogTjFP7OYTCZoNBpV/k61Wi2Ki4tVyIqIpkJRFLS0tGDr1q3YuXPniOtao9FAkiTU1tbC4XBACJHETCnTPPDAA8lOIWm4ZIOIiCiBOMU/sxgMBkiSpEosSZJYXCJKErfbDavVivLycmzfvn1UkTEUCqGxsRHl5eWwWq1pvdNRT08P7rrrLthsNsyZMwe5ubmYP38+rFYr1q5di23bto1qsHvo0KFIk8Vt27ZNGP+8886LuhljX18fHn300UguBoMBpaWlqKurm7TQu3fvXtx0001YvHgxdDodZs6ciYULF2LZsmX43ve+h5deemnC5s8fffQRbr/9dlitVsyZMwezZs3CokWLsG7dOrzzzjtjvkYIMWLmwo033hj5voRvmzZtmvR90z9whgQREVECcYp/5tmwYQMaGxunHae2tlaFbIgoVk1NTTFtw9zd3Q273Q6n04mKioo4Z6eu3bt3Y/Xq1aMKDocPH8bhw4fR1dWF5557DnPnzsXq1avjmsuxY8dQWVmJvXv3jvh6W1sb2tra8Pzzz+OVV14Z8+fm5s2bcccdd2BgYGDE171eL7xeL9xuN+rq6uD3+8d8/aOPPop77rkHp0+fHvH1gwcP4uDBg2hoaMB9992HBx98UIV3ShNhQYKIiCiBOMU/s7jdbvzgBz+YdhyLxYKysjIVMiKiWLjd7piKEWHBYBCSJMHlcsFms8UpO3X19fXhuuuugyzL0Ov12LBhA7761a9i3rx5OHXqFA4ePIi33npLtT5Hk7nllluwd+9efOc730FNTQ3mzZuHAwcOYPPmzWhvb4fL5cINN9wwKp+Ojo5IMeL888/H97//fVxyySUwGo3w+/344IMP0NzcjBdffHHM8z7yyCP40Y9+BABYsmQJNmzYgAsvvBAFBQX44IMP8Mtf/hJ79uzBQw89hLlz5474P76zsxM9PT34+te/DgD4yU9+gmuuuWZE/Hnz5qn5bcp4LEgQERElUHiKvxoj6pzin1yxjqqOR6fTob6+nmvSiRJMURRUV1dP+RoOBoOoqalBR0dHWly/b775Jnp6egAAzzzzzKgZEJdeeinWrl2LzZs348SJE3HPp729HQ8//DDuvvvuyNeWLVuGqqoqrF69Gq+++ip27tyJP/7xj7jqqqsixzQ2NmJgYAA6nQ579uwZtWXmFVdcge9+97v44osvMHv27BHPeTwe3HvvvQCAjRs3YuPGjSP+7pYtW4brrrsONTU1ePrpp3HvvffihhtuQGFhIYDB4vHwGRfFxcWwWCzqfVOyEHtIEBFlKVmW4fF40NbWBo/HM2r6JsWPWlPzOcU/eaY6qnomnU4Hp9OZNiOsRJmkpaUF3d3d04rR1dWF1tZWlTKKr7/+9a+R+3a7fdzjtFptQordS5YswV133TXm+X/7298iNzcXwOgtrsPvY/HixaOKEcPNmTMHOTkjP+4+9thjOH36NJYvXz6qGBGWk5ODLVu2IC8vD4FAQJUBBBofCxJERFlEURQ0NzejsrISRqMRZrMZpaWlMJvNMBqNqKqqQnNz84RNoGj6HA4HzGbztGJwin/yTHdUNcxsNsPlcqXdGnSiTHHmB91kx4m3s88+O3L/qaeeSmImg2pqasadWWIymXDllVcCGCwcDV/mGH4f4UGVWLz88ssAgDVr1kw4q6WgoABWqxUAsGfPnpjOQbFhQYKIKEtkUwfxVCeEQENDA3Q63ZRezyn+yaXGqCoAbNmyhTMjiJJElmXVeiXs2LEjLWYZXn755Vi0aBEA4Ic//CFWrlyJn/70p3jzzTdx6tSphOezYsWKCZ9fuXIlgMGlMX/5y18iX1+7di1yc3PR19eHyy67DFdffTWeeOIJdHV1TTig8sknn6C3txcAcPfdd4/aHePMW3injeEzS0h9LEgQEWWBpqYm2O32qD9EhTuINzU1xTmz7GWz2eB0OmMuSnCKf/KpNRr6q1/9SpU4RBQ7r9erSnNhYLCg7/P5VIkVT7m5uXj55ZdRUlICYLCHwz333IPLL78cBQUF+MY3voFnnnlGte/LZCZr/jh8OcbRo0cj9y+66CI8++yzKCwsRH9/P3bt2oUNGzbAarVi3rx5uOGGG7B79+5R8Q4fPjylPBPRTyObsSBBRJThpttBnDMl4qeiogIulyvq5RsWi4VT/JMsG0dViTJRIBBQNZ7f71c1Xrx85StfQWdnJ5xOJ/71X/8VF1xwAQDg5MmTePXVV3H99dejtLR0yh/eYzGdWX5r1qzBwYMH8etf/xrf+ta3UFRUBAA4cuQInn76adjtdqxbt27EtqDDCy33338/Ojs7o7qlwvKWTMZdNoiIMli2dRBPRzabDZ2dnWhtbcXWrVvhdDpH/NKk1WohSRJqa2tRVlbGv4cki8eoKndKIUq84TslqEGv16saL540Gg2uvfZaXHvttQCAzz77DH/605+wdetW7N27F3v37sUtt9wyovg6vDnk8A/5Y4n2d46//e1vWLx48YTPhxmNxlHPz5kzB+vXr8f69esBAPv378eLL76ILVu2oKenB/X19Vi6dCluu+02AMBZZ50VeW1ubi53x0gRnCFBRJTBsq2DeLoSQsDhcOCFF17A0aNH4fF48Pbbb8Pj8eDzzz/HH/7wBzgcDhYjUkC2jqoSZRqTyQSNRqNKLK1Wi+LiYlViJcPZZ5+NG2+8EXv27IksB9y1axdOnjwZOWZ4weXYsWPjxjp69Cg+//zzqM7b3t4e1fOzZ8+O9L6YSElJCe666y78+c9/jiyH/MMf/hB5ftGiRZgzZw6AwS1Qp4o/i9XFggQRUQbLtg7imcBgMKCkpAQrV65ESUkJR89TTDaPqhJlEoPBAEmSVIklSVJG/F+dm5sb2b2pv78fx48fjzxXWFiIgoICAIg0exzLc889F/VOXb/73e/GPdbn8+G1114DMLgzVSzFo4ULF0ZmXhw5ciTydY1Gg6uuugoA8Nprr2H//v1Rxxxu5syZkft9fX1TikH/wIIEEVGG4lp3IvVxVJUoc9TW1qZUnHjbvXs3Pvroo3GfP3XqVGRGZH5+fqQvQ5jdbgcAvPjii/j4449Hvf6DDz7Aj3/846jzeffdd/HII4+M+np/fz9uvvnmyM4fGzZsGPH8zp07RxRLzvTpp5/i/fffBwCcf/75I567++67odFoMDAwgMrKSni93nHjhEIh/P73vx91zFlnnYUZM2YAwJjfB4oNe0gQEWUornUnUl94VLWxsXHasTJlVJUoXTkcDpjN5mktbbRYLJFZBanu9ddfx0MPPYQrrrgCq1atwpIlS1BUVISTJ0/iwIEDeOKJJyKNrG+66SZotSM/KtbW1uKll17CyZMn4XA4sGnTJixduhSBQACvv/46Hn/8cRQVFUGj0US215zI8uXLceedd+Ldd99FdXU15s2bhw8//BA///nP0dbWBgC4+uqrsXr16hGv+8UvfoHrr78eq1atQnl5OUpKSjBnzhwcO3YM77zzDrZs2RJZbnLrrbeOeK3VasWjjz6K22+/HR6PBxaLBevXr0d5eTnmz5+Pv//97zh06BD27NmDxsZGfPbZZ+js7ITJZIrE0Gq1WLFiBd588008+eSTWLp0KS655BLk5uYCGOx3MVbPi4m88cYbI4pFw2d2fPTRR9i2bduI49etWxdT/JSmKApvKX4DYAKgAFA+/fRThYgoGm+//bYS/r9Djdvbb7+d7LdElBL++7//W5Vrqrm5OdlvhWhcnd7jyrl37lI6vcfjep4DBw4oHo9HOXDgQFzPM569e/cqOp1uStewTqdT9u7dm5S8p2Ljxo1Rva9rrrlGOXHixJgxfvCDH4z7unPOOUfxeDzKueeeqwBQampqRr3+qaeeihzvdruVpUuXjhvvsssuU2RZHhWjrKxs0veQk5OjPPTQQ+N+L37zm98os2fPnjTOjBkzlA8//HDU63ft2qUIIcZ8zcaNG6P+OwmrqamJ6d/eVEzlWvv000+Hn9ekxOGzLpdsEBFlKK51J4qP8KjqdKTTqCpRJrPZbHA6nZEmiNHS6XRwOp2RJpDp4I477sD27duxYcMGXHrppTjnnHMwc+ZMzJw5E+eddx6+/e1vY9euXdi5cydmzZo1ZozHH38czzzzDOx2OwwGA2bNmoUvf/nLuOuuu+B2u1FSUhJ1PoWFhXjrrbfw05/+FJdccgn0ej3y8/OxYsUKbNmyBa2trWP+7vHss8/iN7/5Df75n/8Zl1xyCb70pS9Bq9UiPz8fZrMZGzZswL59+3DfffeNe+6bb74Zf/nLX/DAAw/gsssuw9y5c6HVaqHT6bB48WKsWbMGTzzxBHw+X2Rr1OFWrVqF119/Hddccw0WLFgQmR1BsROKEl3TEUoeIYQJwKfA4Jqo4VOGiIjGI8syjEajKss2tFotPv/8c04vJxridrtht9untKWuTqeDy+VKqw8ylH26fF9g9ZY3sOvfLoeleE7czvPhhx+iv78fWq0WF154YdzOMxm3243q6uqolm9YLBbU19fzGqa0MpVrzev1YuHCheGHCxVFGb/pxhRxhgQRUYZiB3Gi+MmmUVWibGCz2dDZ2Ynm5mZUVlaOal6r1WpRVVWF5uZmdHR08BomUgmbWhIRZbDa2lpVmu+lSwdxokSqqKiAy+XiqCpRhhBCwOFwwOFwQJZl+Hw++P1+6PV6FBcXszBPFAecIUFElMG41p0ovjiqSpSZDAYDSkpKsHLlSpSUlLAYQRQnnCFBRJTBhBBoaGiY1lr3+vp6CCHikB1RZsj0UVVZluH1ehEIBJCfnw+TyZT274mIiFJD2s6QEELME0KsFkI8KIT4TyHEESGEMnTbFqdzrhVCvCaE+KsQ4u9CiE+EEE8LIf5HPM5HRKQGrnUnSpxMGVVVFCUy68NoNMJsNqO0tBRmsxlGozEy64PN0YmIaDrStiAB4G8AXgbwYwDfAHBWvE4khJglhHgFwDMAKgDMB5AH4BwA1wN4QwixMV7nJyKarvBa92iXb1gsFrhcLlRUVMQ5MyJKNW63G1arFeXl5di+ffuonXpCoRAaGxtRXl4Oq9UKt9udpEyJiCjdpXNBYrj/D8BrcYz/JICrhu43A7gWwEoANwH4GIPfx01CiPVxzIGIaFq41p2IJtPU1AS73R5Vk04A6O7uht1uR1NTU5wzIyKiTJTOPSQeBNAOoF1RlL8JIc4DcFDtkwghygFcN/TwZQCSoijhoYJ2IcRLAPZicLbE/xZCvKAoyjG18yAiUkOmr3Unoqlzu92QJCnmfjPBYBCSJMHlcrGQSUREMUnbgoSiKIlaInHH0J/9AGqHFSPCeRwRQtwJ4FkABQC+C+CRBOVGRDRlBoOBBQgiAjDYM6K6unpKzW+BwaJETU0NOjo62ASXiIiililLNuJCCKEH8H8OPfwvRVG84xy6A4A8dF+Ke2JEREREKmppaYl6mcZ4urq60NraqlJGRESUDViQmNgKADOG7o/7E1ZRlFMA/hx+jRAiN96JEREREamlrq4upeIQEVF2SNslGwnylWH335/k2PcBXInB7+mFADzRnkQIYZrkkC9FG4uIiIgoFrIsw+l0qhJrx44dkGWZy8GIiCgqLEhMbHihYLzlGmGfDru/EDEUJM54LREREVHCeL3eUVt7TlUoFILP52NBgoiIosIlGxPTD7sfmOTY4V2g8uOQCxEREZHqAoHJfsWJjd/vVzUeERFlLs6QmNjMYfdPTXJs37D7s2I8z8JJnv8SBrc4JSIiIlJVfr664yh6vX7yg4iIiMCCxGT+Puz+jHGPGpQ37P7JWE4ywe4dAMDts4iIiChuTCYTNBqNKss2tFotiouLVciKiIiyAZdsTGz4nMPJhg90w+6rO/eRiIiIKE4MBgMkSZ1dyyVJYv8IIiKKGgsSExs+c2GynTCGL7tgk0oiIiJKG7W1tSkVh4iIsgMLEhMbvlPGRZMcG36+H8CH8UmHiIiISH0OhwNms3laMSwWC8rKylTKiIiIsgELEhNrxz+aWY77E1YIMQPApeHXKIpyOt6JEREREalFCIGGhgbodLrJDx6DTqdDfX09+14R0YS2bdsGIQSEEDh06FCy06EUwILEBBRF8QN4fejh14QQ4y3b+BaA8IJJZ9wTIyIiIlKZzWaD0+mMuSih0+ngdDphs9nilBlNRpZleDwetLW1wePxQJblZKdERDF455138OCDD+LKK6+EyWRCXl4e8vPzsXjxYtx444144403kp1i3GR1QUIIsU4IoQzdNo1z2KNDf2oBbBVCaM6IMRfA/x56eBzAb+OSLBEREVGcVVRUwOVyRb18w2KxwOVyoaKiIs6Z0ZkURUFzczMqKythNBphNptRWloKs9kMo9GIqqoqNDc3Q1GUZKdKlFEOHToUmeWxbdu2acez2+1YsWIFNm7ciKamJvh8Ppw6dQrBYBAffvghtm3bhiuuuAI1NTU4derU5AHTTNpu+ymEuBzABcO+NHfY/QuEEOuGH68oyrapnEdRlP8WQjwH4DoA/wSgSQjxCwA9AKwA7gVwztDhdyqKcmwq5yEiIlKDLMvwer0IBALIz8+HyWTirgcUE5vNhs7OTrS2tmLr1q1wOp0jtgTVarWQJAm1tbUoKyvjMo0kcLvdqK6uRnd395jPh0IhNDY2orGxEWazGQ0NDZzBQpSienp6AAALFixAVVUVrrjiCpxzzjkIhULYs2cPHnvsMfh8PjQ0NOD06dN45plnkpyxutK2IAHguwBqxnnusqHbcNumca5/xeCSjKsAfHXoNtwAgIcURfnNNM5BREQ0JYqioKWlBVu3bsXOnTtHfHjUaDSRD48Oh4MfHikqQgg4HA44HA7Isgyfzwe/3w+9Xo/i4mIWuZKoqakJkiQhGAxGdXx3dzfsdjucTidnssRA7u+Ht68PgVAI+RoNTHl5MGjT+aMTpaqLLroIDz/8MNasWQONZsRkfFx66aW44YYbcNlll+HAgQN49tlnceutt8JutycpW/XxqoqCoignAawSQvwzgHUALgZQAOBvAHYD+KWiKHuSlyEREWUrjpRSvBkMBhYgUoTb7Y6pGBEWDAYhSRJcLhev/wkoioKW48ex1efDziNHEBr2nAaAVFSE2gUL4CgoYHGXVLNr164Jn587dy4ee+wxXH311QCAxsbGjCpIpG0PCUVR1imKIqK9jRNj27BjNkVxzmcURblSUZT5iqLkKYpyjqIo17MYQUREydDU1AS73T5uMeJM4ZHSpqamOGdGRGpTFAXV1dUxFyPCgsEgampq2FNiHG6/H9b2dpS/9x62n1GMAIAQgMbeXpS/9x6s7e1w+/3JSFMVPT09uOuuu2Cz2TBnzhzk5uZi/vz5sFqtWLt2LbZt2zaqMWosfRPOO+88CCGwbt26SXPp6+vDo48+GsnFYDCgtLQUdXV1I2b7jWXv3r246aabsHjxYuh0OsycORMLFy7EsmXL8L3vfQ8vvfTShP/eP/roI9x+++2wWq2YM2cOZs2ahUWLFmHdunV45513xnyNEALnn39+5PGNN94Y+b6Eb5s2bZr0fcfqq1/9xwT9jz/+WPX4ycQZEkRERGmII6VE2aWlpSXq4uN4urq60NraCofDoU5SGaLp6FFIXV0IDgxEdXz3iROw79sHp8WCCqMxztmpa/fu3Vi9evWogsPhw4dx+PBhdHV14bnnnsPcuXOxevXquOZy7NgxVFZWYu/evSO+3tbWhra2Njz//PN45ZVXkJ+fP+q1mzdvxh133IGBM/7OvF4vvF4v3G436urq4Pf7x3z9o48+invuuQenT58e8fWDBw/i4MGDaGhowH333YcHH3xQhXc6fX19fZH7Zy7rSHcsSBAREaUZtUZKOzo6OO2YKE3U1dWpFocFiX9w+/0xFSPCggMDkLq64Fq6FDa9Pk7Zqauvrw/XXXcdZFmGXq/Hhg0b8NWvfhXz5s3DqVOncPDgQbz11ltwOp0JyeeWW27B3r178Z3vfAc1NTWYN28eDhw4gM2bN6O9vR0ulws33HDDqHw6OjoixYjzzz8f3//+93HJJZfAaDTC7/fjgw8+QHNzM1588cUxz/vII4/gRz/6EQBgyZIl2LBhAy688EIUFBTggw8+wC9/+Uvs2bMHDz30EObOnYsf/OAHkdd2dnaip6cHX//61wEAP/nJT3DNNdeMiD9v3jw1v00AgNbW1sj9kpIS1eMnEwsSREREaYYjpUTZRZZl1T4k7tixA7Issy8Ihoq7+/fHXIwICw4MoOb999GxfHlaFHfffPPNyI4OzzzzzKgZEJdeeinWrl2LzZs348SJE3HPp729HQ8//DDuvvvuyNeWLVuGqqoqrF69Gq+++ip27tyJP/7xj7jqqqsixzQ2NmJgYAA6nQ579uzB/PnzR8S94oor8N3vfhdffPEFZs+ePeI5j8eDe++9FwCwceNGbNy4ccTf3bJly3DdddehpqYGTz/9NO69917ccMMNKCwsBDC41fHwGRfFxcWwWCzqfVPGMDAwgJ/97GeRx9/+9rfjer5ES9seEkRERNlKzZFSIkp9Xq930vX00QqFQvD5fKrESnctx4+je5ofvLuCQbQeP65SRvH117/+NXJ/oqaIWq02IQWrJUuW4K677hrz/L/97W+Rm5sLYPTPqvD7WLx48ahixHBz5sxBTs7Ij7uPPfYYTp8+jeXLl48qRoTl5ORgy5YtyMvLQyAQQGNjY8zvTU2bN29GW1sbAOBb3/oWli1bltR81MaCBBERURqJx0gpEaW2QCCgajx/GjdkVFPd0GyBVIkTb2effXbk/lNPPZXETAbV1NSMO7PEZDLhyiuvBDA4K3B4QS78PjweT+SDerRefvllAMCaNWsmnNVSUFAAq9UKANizJ3n7F7S2tkaKNvPmzcOvfvWrpOUSLyxIEBERpRGOlBJln7Ga8k2HPk16HsST3N8PZ2+vKrF29PZC7u9XJVYE1HmwAAAgAElEQVQ8XX755Vi0aBEA4Ic//CFWrlyJn/70p3jzzTdx6tSphOezYsWKCZ9fuXIlgMG+R3/5y18iX1+7di1yc3PR19eHyy67DFdffTWeeOIJdHV1TbirxieffILeob/zu+++e9TuGGfewjttDJ9Zkkjd3d2QJAn9/f2YOXMmXnjhhbj0p0g2FiSIiIjSCEdKibKPyWRSrbO+VqtFcXGxKrHSmbevb9TWnlMVAuAbtgtCqsrNzcXLL78caYrY3t6Oe+65B5dffjkKCgrwjW98A88884xqRe/JTPbhevhyjKNHj0buX3TRRXj22WdRWFiI/v5+7Nq1Cxs2bIDVasW8efNwww03YPfu3aPiHT58eEp5JqKfxpkOHjyIK6+8EseOHYNGo8Fzzz034TKbdMamlkRERGmEI6VE2cdgMECSJFXWskuSFPf+AHJ/P7x9fQiEQsjXaGDKy4NBm1ofOwIqf+j2J+hD/HR95StfQWdnJ15++WW8/PLLcLlc+Oijj3Dy5Em8+uqrePXVV/Hzn/8cf/zjH+M+Gj+dRqBr1qzB1772NTz//PN49dVXsXv3bvT29uLIkSN4+umn8fTTT6OmpgZPPvlkpI/E8ELL/fffj6qqqqjOpdPpppznVPT09OBrX/saenp6IITAk08+OWonj0ySWv8zEBER0YTCI6VqjGBxpHQ0WZbh9XoRCASQn58Pk8nE3QgoJdTW1qpSkKitrVUhm9EURUHL8ePY6vNh55EjI2YfaABIRUWoXbAAjoKClNiRIl+lGSdhepXjxZNGo8G1116La6+9FgDw2Wef4U9/+hO2bt2KvXv3Yu/evbjllltG9Csa3hxyYJJdSaLdkvpvf/sbFi9ePOHzYUajcdTzc+bMwfr167F+/XoAwP79+/Hiiy9iy5Yt6OnpQX19PZYuXYrbbrsNAHDWWWdFXpubmxv33TGm4siRI6ioqIgsUdmyZQuqq6uTnFV8cckGERFRGgmPlKohESOl6UBRFDQ3N6OyshJGoxFmsxmlpaUwm80wGo2oqqpCc3PzhGuTieLN4XDAbDZPK4bFYkFZWZlKGf2D2++Htb0d5e+9h+1nFCOAwSUNjb29KH/vPVjb2+FOgaViprw8qFVC0AqB4rw8laIl3tlnn40bb7wRe/bsgc1mAwDs2rULJ0+ejBwzfDbdsWPHxo119OhRfP7551Gdt729ParnZ8+eHel9MZGSkhLcdddd+POf/xyZ1fCHP/wh8vyiRYswZ84cAINboE5VvApqX3zxBb7+9a/D4/EAAH72s5/he9/7XlzOlUpYkCAiIkozao1wxmukNJ243W5YrVaUl5dj+/bto2aehEIhNDY2ory8HFarFW63O0mZUrYTQqChoWHK08d1Oh3q6+tV/zDVdPQo7Pv2Rb19ZveJE7Dv24emYT0BksGg1UIqKlIlljR3bsotSZmK3NzcSMGqv78fx4dtZ1pYWIiCggIAiDR7HMtzzz0XdfH2d7/73bjH+nw+vPbaawAGi3Gx9FBZuHBhZObFkSNHIl/XaDS46qqrAACvvfYa9u/fH3XM4WbOnBm536dS75ATJ05g1apVkZ8x9957L+68805VYqc6FiSIiIjSTCqPlKaTpqYm2O12dHd3R3V8d3c37HY7mpqa4pwZ0dhsNhucTmfMRQmdTgen0xkZ/VaL2++H1NWF4CRT+M8UHBiA1NWV9JkStQsWpFSceNu9ezc++uijcZ8/deoUWltbAQz2Kyo6o2ATbqr44osv4uOPPx71+g8++AA//vGPo87n3XffxSOPPDLq6/39/bj55psjO39s2LBhxPM7d+4cUSw506effor3338fAHD++eePeO7uu++GRqPBwMAAKisr4fV6x40TCoXw+9//ftQxZ511FmbMmAEAY34fYnXq1ClIkhSZtXHbbbfhJz/5ybTjpov0L+URERFlmfBIqd1uj3qt7nDxGilNJ263G5Ikxfz9CwaDkCQJLpdL9Q93RNGoqKiAy+VCdXV1VMU0i8WC+vp61f+9KoqC6v37Yy5GhAUHBlDz/vvYMTTqngyOggKYZ8+OenbHWCw6HcqS+B5i8frrr+Ohhx7CFVdcgVWrVmHJkiUoKirCyZMnceDAATzxxBOREfqbbroJ2jNmfdTW1uKll17CyZMn4XA4sGnTJixduhSBQACvv/46Hn/8cRQVFUGj0US215zI8uXLceedd+Ldd99FdXU15s2bhw8//BA///nP0dbWBgC4+uqrsXr16hGv+8UvfoHrr78eq1atQnl5OUpKSjBnzhwcO3YM77zzDrZs2RJZbnLrrbeOeK3VasWjjz6K22+/HR6PBxaLBevXr0d5eTnmz5+Pv//97zh06BD27NmDxsZGfPbZZ+js7ITJZIrE0Gq1WLFiBd588008+eSTWLp0KS655BLk5uYCGOx3MVbPi/GsXbs2MhukvLwcN910E7q6usY9fsaMGRP23kg3gushU58QwgTgU2Cw4jf8giAiouzV1NQU84fq8EhpRUVFHDNLbYqiwGq1Rj0zYiwWiwUdHR1ZXdSh5FIUBa2trdi6dSucTueI5UZarRaSJKG2thZlZWVT+nfa5fsCq7e8gV3/djksxXNGPd987BjK33tvWu8BAP48bx4MQkCr1eLCCy+cdrxYuf1+2Pftm1JhRZeTA9fSpbClyW5FmzZtwgMPPDDpcddccw2effZZzJo1a9Rzt912G/793/99zNedc845+NOf/oRvfvOb+OSTT1BTU4Nt27aNOGbbtm248cYbAQwWhm+66Sbs27dvzHiXXXYZ/vM//3PUblAOhyMyk2M8OTk5eOCBB3DfffeN+fx//Md/4Ic//OGkW3rOmDED3d3duOCCC0Z8/ZVXXsHVV1895pKTjRs3YtOmTRPGHS7W6/Pcc8/FoUOHYnoNAHz44Yfo7++P6Vrzer1YuHBh+OFCRVHGn1IyRZwhQURElKZSZaQ03bS0tEyrGAEAXV1daG1thcPhUCcpohgJIeBwOOBwOCDLMnw+H/x+P/R6PYqLi+PesLaup0eVOJ/09cE6bE1+otn0ejgtlpiXnuhycuC0WNKmGAEAd9xxB5YsWYL/+q//wr59+9DT04PDhw8DAL70pS9h5cqVqK6uxqpVq8aN8fjjj+PSSy/FE088gXfffRenT5/GOeecA0mScMcdd4zYyWIyhYWFeOutt/CLX/wCzz//PD7++GMoioKSkhJUV1djw4YNY/aOePbZZ7Fr1y60tLTA4/Hgr3/9K44cOYKZM2fi3HPPhd1ux6233oolS5aMe+6bb74Z//RP/4Rf//rXeO211/DBBx/g+PHjyMvLQ3FxMaxWKyoqKrBmzRrMnTt31OtXrVoVmRXS3t6O3t5enD59Our3Tv/AGRJpgDMkiIhoIvEeKc00VVVVqmyfWFVVNaKDO1EmmWiGhNzfD+Mbb4zaTWMq/n3WLJTr9ZiRm5uUGRJhbr8f1fv3R7V8w6LTof6ii9KqGEHEGRJEREQUF8keKU0nsizD6XSqEmvHjh2QZZnfX8o63r4+VYoRAKAACKXAAKlNr0fnihVoPX4cW3t64OztHfEetUJAmjsXtQsWoKyggMVdIpWwIEFERJRBDAYDPyBPwOv1jtrac6pCoRB8Ph+/35R1AipdQ2FTa4upPiEEHIWFcBQWQu7vh6+vD/5QCHqNBsV5eRmxtSdRquFVRURERFkjEAioGs+f5G0LiZIhf4x1/dORo2o0dRi0WhYgiBKAVxkRERFljfz8fFXjndn9nWgqZFmG1+tFIBBAfn4+TCZTSs+8MeXlQQOosmxDANBw+QNR1krFgiQRERFRXJhMpjG7tk+FVqtFcXGxKrEo+yiKgubmZlRWVsJoNMJsNqO0tBRmsxlGoxFVVVVobm4ec1vBZDNotZCKilSJNX/GDOSwIEGUtViQIKKEkmUZHo8HbW1t8Hg8kGU52SkRURYxGAyQJEmVWJIkpfQoNqUut9sNq9WK8vJybN++fVRfk1AohMbGRpSXl8NqtcLtdicp0/HVLligSpxz8/JUiUNE6YkFCSKKu3QeBSKizFNbW5tScSi7NDU1wW63o7u7O6rju7u7Ybfb0dTUFOfMYuMoKIB59uxpxbDodDDm5qqUERGlIxYkiCiuMmEUiIgyi8PhgNlsnlYMi8WCsrIylTKibOF2uyFJEoLBYEyvCwaDkCQppX5GCiHQUFICXc7UPk7ocnJQf9FFKmdFROmGBQkiiptMGQUioswihEBDQwN0Ot2UXq/T6VBfXw/Bde8UA0VRUF1dHXMxIiwYDKKmpialZhPa9Ho4LZaYixK6nBw4LRbY2BSWKOuxIEFEcZFJo0BElHlsNhucTmfMRQmdTgen0wmbzRanzChTtbS0RF2gH09XVxdaW1tVykgdFUYjXEuXRr18w6LTwbV0KSqMxjhnRkTpgAUJIlJdJo4CEVHmqaiogMvlinr5hsVigcvlQkVFRZwzo0xUV1eXUnHUZNPr0bliBZovvhiVRUU4cx8brRCoKipC88UXo2P5cs6MIKIIbbITIKLMo+YokMPhUCcpIqIx2Gw2dHZ2orW1FVu3boXT6RzR60ar1UKSJNTW1qKsrIzLNGhKZFmG0+lUJdaOHTsgy3LK7fAihICjsBCOwkLI/f3w9fXBHwpBr9GgOC8PBu3YHzs0Gg36+/sRCoWgKAqvMaI4UBQl8rNNra2v1cKCBBGpTs1RIBYkiCjehBBwOBxwOByQZRk+nw9+vx96vR7FxcUp98GP0o/X6x3V1HmqQqEQfD5fSv+7NGi14xYgzjRjxgz09fVBURScOHFiyr1diGh8J06ciMw8njFjRpKzGYkFCSJSVTaMAhFR5jIYDPw/h1QXCARUjef3+1WNl0wGgyHyfo4ePYrZs2dzlgSRihRFwdGjRyOPU+1nHHtIEJGq4jEKRERElM7y8/NVjafPoB4M+fn5kQJEIBCA1+tFMBhkHymiaVIUBcFgEF6vN1IUFUKo/v/RdHGGBBGpiqNAREREI5lMJmg0GlUK9lqtFsXFxSpklRpycnJQXFwMn88HRVEQCAQQCAQghEi5te5E6STclyVMCIHi4mLkxLhNb7yxIEFEquIoEBER0UgGgwGSJKGxsXHasSRJSrkp19MV7tcSLkoAg6O7/f39Sc6MKDOEixGp+Hs1CxJEpCqOAhEREY1WW1urSkGitrZWhWxSj16vx+LFixEIBCDLMk6dOqXaElCibKTRaDBjxgwYDAbk5+en3MyIMBYkiEhVHAUiIiIazeFwwGw2T2tbbIvFgrKyMhWzSi05OTlsLEuUZVKzTEJEaU2t0ZtMHQUiIqLsI4RAQ0PDlLe11Ol0qK+v5w4URJRRWJAgItWFR4GmI9NHgTKVLMvweDxoa2uDx+OBLMvJTomIKGXYbDY4nc6YixI6nQ5OpxM2my1OmRERJQcLEkSkOo4CZRdFUdDc3IzKykoYjUaYzWaUlpbCbDbDaDSiqqoKzc3N3MKNiAhARUUFXC5X1IV7i8UCl8uFioqKOGdGRJR4LEgQUVxwFCg7uN1uWK1WlJeXY/v27aMakIVCITQ2NqK8vBxWqxVutztJmRIRpQ6bzYbOzs5IMffM7S21Wm2kmNvR0cGfiUSUsdjUkojiJjwKVF1dHVUTL4vFgvr6ev7ilSaampogSRKCwWBUx3d3d8Nut8PpdHKkj4iynhACDocDDocDsizD5/PB7/dHtsBkY0ciygacIUFEccVRoMzkdrtjKkaEBYNBSJLEmRJERMMYDAaUlJRg5cqVKCkpYTGCiLIGZ0gQUdxxFCizKIqC6urqmIsRYcFgEDU1Nejo6GCfECIiIqIsxoIEESUU9xdPfy0tLVEtwZlIV1cXWltb4XA41EmKiIiIiNIOl2wQEVFM6urqUioOEREREaUnFiSIiChqsizD6XSqEmvHjh2QZVmVWERERESUfliQICKiqHm93lFbe05VKBSCz+dTJRYRERERpR8WJIiIKGqBQEDVeH6/X9V4RERERJQ+WJAgIqKo5efnqxpPr9erGo+IiIiI0gcLEkREFDWTyQSNRqNKLK1Wi+LiYlViEREREVH6YUGCiIiiZjAYIEmSKrEkSeIWsERERERZjAUJIiKKSW1tbUrFISIiIqL0xIIEERHFxOFwwGw2TyuGxWJBWVmZShkREVEyyf398ASDaJNleIJByP39yU6JiNKENtkJEBFRehFCoKGhAXa7HcFgMObX63Q61NfXQwgRh+yIiCgRFEVBy/Hj2OrzYeeRIxi+IbQGgFRUhNoFC+AoKOD/90Q0LhYkiGhaZFmG1+tFIBBAfn4+TCYT+wJkAZvNBqfTCUmSYipK6HQ6OJ1O2Gy2OGZHRETx5Pb7Ub1/P7pPnBjz+RCAxt5eNPb2wjx7NhpKSmDjrkpENAYu2SCimCmKgubmZlRWVsJoNMJsNqO0tBRmsxlGoxFVVVVobm6GoijJTpXiqKKiAi6XK+rlGxaLBS6XCxUVFXHOjIiI4qXp6FHY9+0btxhxpu4TJ2Dftw9NR4/GOTMiSkcsSBBRTNxuN6xWK8rLy7F9+3aEQqERz4dCITQ2NqK8vBxWqxVutztJmVIi2Gw2dHZ2RgpUZ24JqtVqIwWqjo4OzowgIkpjbr8fUlcXggMDMb0uODAAqasLbr8/TpkRUbrikg0iGteZyzEOHDiAf/mXf4l6in53dzfsdjucTidHxTOYEAIOhwMOhwOyLMPn88Hv90Ov16O4uDirl/BwSRMRZQpFUVC9f3/MxYiw4MAAat5/Hx3Ll7OnBBFFsCBBRCMoioKWlhZs3boVO3fuHDUDIlbBYBCSJMHlcnF0PAsYDIas/8A90TWk0WggSRJqa2vhcDj4SzkRpY2W48ejXqYxnq5gEK3Hj8NRWKhSVkSU7rhkg4giJluOMVXBYBA1NTXsKUEZj0uaiChT1fX0pFQcIsoMLEgQEQCgqakJdrsd3d3dcYnf1dWF1tbWuMQmSgWxXkPhJU1NTU1xzoyIaHrk/n44e3tVibWjtxdyf78qsYgo/bEgQURwu90xb984FXV1dXGNT5QsU72GwkuaOFOCiFKZt68P6syZHNwS1NfXp1I0Ikp3LEgQZTlFUVBdXR33YgQA7NixA7Isx/08RNGSZRkejwdtbW3weDxT+vc53WuIS5qIKNUFVFrCGeZXOR4RpS8WJIiyXEtLS9yWaZwpFArB5/Ml5FxE41EUJbJNqdFohNlsRmlpKcxmM4xGY2Sb0mgLBGpcQ1zSRESpLP+MLZ2nS69yPCJKXyxIEGW5RC+j8HMPckqieDSdVOsa4pImIkpVprw8qFVC0AqB4rw8laIRUbpjQYIoi8myDKfTmdBz6vX6hJ6PKCweTSfVvIa4pImIUpVBq4VUVKRKLGnuXBi0WlViEVH6Y0GCKIt5vV7VtvaMhlarRXFxccLORxQWr6aTal5DXNJERKmsdsGClIpDRJmBBQmiLBYIBBJ6PkmSYDAYEnpOong2nVT7GuKSJiJKVY6CAphnz55WDItOh7KCApUyIqJMwIIEURbLz89P6Plqa2sTej4iIL5NJ9W+hrikiYhSlRACDSUl0OVM7eODLicH9RddBCGEypkRUTpjQYIoi5lMJmgS1OnaYrGgrKwsIeciGi6eTSfVvIa4pImIUp1Nr4fTYom5KKHLyYHTYoGNRVciOgMLEkRZzGAwQJKkuJ9Hp9Ohvr6eoyKUcPFuOqnmNcQlTUSUDiqMRriWLo16+YZFp4Nr6VJUGI1xzoyI0hELEkRZLt7LKHQ6HZxOJ2w2W1zPQzSWRDSdVOsa4pImIkoXNr0enStWoPnii1FZVDRqS1CtEKgqKkLzxRejY/lyzowgonFxzx2iLOdwOGA2m6e9xn4sFosF9fX1LEZQ0iSi6aQa1xCXNBFRuhFCwFFYCEdhIeT+fvj6+uAPhaDXaFCcl8etPYkoKpwhQZTlhBBoaGiATqdTJZ5Wq0VVVRWam5vR0dHBYgQlVSKaTk73GuKSJiJKdwatFiU6HVYaDCjR6ViMIKKo8X8LIoLNZoPT6YQkSTFtjajT6fD000/jy1/+Mvx+P/R6PYqLi7kOnlJGuOmkGss2Jmo6OZ1riEuaiIiIKFtxhgQRAQAqKirgcrlgNpujOt5iscDlcuHaa69FSUkJVq5ciZKSEhYjKKUksunkVK+hiooKVfIjIiIiSjcsSBBRhM1mQ2dnJ5qbm1FZWTlqO0Mux6B0lMimk7yGiIgoWnJ/PzzBINpkGZ5gEHJ/f7JTIko4LtkgohGEEHA4HHA4HJBlGT6fj8sxKK0luukkryEiIhqPoihoOX4cW30+7DxyBMMXFGoASEVFqF2wAI6CAvYWoqzAggQRjctgMPDDE6W9cNNJu90eU3+HsOk0neQ1REREYW6/H9X796P7xIkxnw8BaOztRWNvL8yzZ6OhpIRbplLG45INIiLKeOGmk7HuhMGmk0REpIamo0dh37dv3GLEmbpPnIB93z40HT0a58yIkosFCSIiygqlpaXYtm0bFi1aFNXxbDpJRERqcPv9kLq6EBwYiOl1wYEBSF1dcPv9ccqMKPm4ZIOIiDKWoihoaWnB1q1bsXPnzlHbfwohoChK5LFWq4UkSaitrUVZWRnX7xIRZQG5vx/evj4EQiHkazQw5eXBoFXnY5KiKKjevz/mYkRYcGAANe+/j47ly/kziTISCxJERJSR3G43qqurJ2xmGS5GnH/++XjkkUdQUVHBng9ERFkgUc0lW44fj3qZxni6gkG0Hj8OR2HhtOIQpSIu2SAioozT1NQEu90e9c4aBw8eRE1NDd5+++04Z0ZERMnm9vthbW9H+XvvYfsZxQjgH80ly997D9b29mktmajr6ZlWrtOJw21FKR1whgQREWUUt9sNSZJi3lEjGAxCkiS4XC42sSQiylBNR4/G1M8h3FzSabGgwmiM6Vxyfz+cvb1TSXOUHb29kPv7J11Kwm1FKd1whgQREWUMRVFQXV09pe09gcGiRE1NzYi+EkRElBkS3VzS29c3avbFVIUA+Pr6JjwmkTM/iNTCggQREWWMlpaWqJdpjKerqwutra0qZURERKlAreaSsRSsAyG1yhGD/BPE47ailK5YkCAiooxRV1eXUnGIiCg1qNlcMlr5Gs20zncm/TjxuK0opTMWJIiIKCPIsgyn06lKrB07dkCWZVViERFR8iWjuaQpLw9qlSS0QqA4L2/U15Mx84NITRlRkBBCnCuEeEwI8b4QIiiEOCqEaBdC/C8hxOxpxl4nhFCivK1T6S0REVGMvF4vQipNjw2FQvD5fKrEoswjyzI8Hg/a2trg8XhYvCJKcfFoLhkNg1YLqahIlfNKc+eO2dAyGTM/iNSU9gUJIcTVADoA/F8AvgxgNoBCAMsB/D8A9gkhLkhehkRElAiBQEDVeH5OYaVhFEVBc3MzKisrYTQaYTabUVpaCrPZDKPRiKqqKjQ3N3OUkSgFJbq55HC1Cxaoct7x4iRzW1EiNaT1tp9CiKUAngcwC0AAwE8BNA89vg7AzQAWA3hFCLFcUZTp/nb5dQATXa3eacYnIqIpys/PVzWeXq9XNR6lL7fbjerq6nEbpoZCITQ2NqKxsRFmsxkNDQ3cOpYohSSyueSZHAUFMM+ePa1ZDBadDmUFBaO+noxtRYnUlu4zJB7HYPGhH8CViqI8rCjKHkVR/ltRlPUAfjR03GIA/1OF8x1QFKVrghvnOhERJYnJZIJGpQZiWq0WxcXFqsSi9NbU1AS73R717i3d3d2w2+1oamqKc2ZEFK1ENZccixACDSUl0OVM7WOXLicH9RddBCHEqOeSOfODSC1pW5AQQqwEcMXQw/9XUZQ9Yxz2GID9Q/dvE0LkJiQ5IiJKOIPBAEmSVIklSRIMBoMqsSh9ud1uSJKEYDAY0+uCwSAkSYLb7Y5TZkQUi0Q0l5yITa+H02KJuSihy8mB02KBbZwZe8mc+UGklrQtSAC4dtj9p8Y6QFGUAQANQw8LAHw13kkREVHy1NbWplQcSl+KoqC6ujrmYkRYMBhETU0Ne0oQpYBENJecTIXRCNfSpTDPjq7fvkWng2vpUlQYjeMek8yZH0RqSeeCxOVDfwYB7J3guNZh9y+LXzpERJRsDocDZrN5WjEsFgvKyspUyojSVUtLS9TLNMbT1dWF1tbWyQ8koriLd3PJaNj0enSuWIHmiy9GZVHRqFkbWiFQVVSE5osvRsfy5ePOjAhL9swPIjWkc9eSkqE/P1IUZaK9d94f4zVT9ZQQ4ssA5gKQAXwE4L8A/EpRlCnvDyeEME1yyJemGpuIKJsIIdDQ0AC73T6lkW2dTof6+vox1+pSdqmrq1MtjsPhUCUWEU1dPJtLxkIIAUdhIRyFhZD7++Hr64M/FIJeo0FxXl5Msy/CMz8aVWhsOdWZH0TTlZYzJIQQMzFYFAAm2dlCUZRjGJxFAQALp3lqB4CzAeQCOAtAKYB7AXwkhLhlGnE/neTWPo3YRERZxWazwel0QqfTxfQ6nU4Hp9PJ3REIsizD6XSqEmvHjh2QZVmVWIkkyzI8Hg/a2trg8XjS8j0QDRfP5pJTZdBqUaLTYaXBgBKdbkoFgVSY+UE0HWlZkAAwfP5SNBvPhwsSU90T7i8AHgWwBsDKodt1AF4AoACYCeAJIcT6KcYnIiIVVVRUwOVyRb18w2KxwOVyoaKiIs6ZUTrwer0IqdTcLRQKweeb8iTKhFIUBc3NzaisrITRaITZbEZpaSnMZjOMRiOqqqrQ3NzMvhiUtuLVXDKZwjM/pkONmR9EU5WuBYmZw+6fiuL48B42s6ZwLieACxRF+V+KouxQFKV96Pa8oijfBvBPAE4PHbtZCDGV5RULJ7mtmEJMSlMclSJSh81mQ2dnZ+QD1plbgmq12sgHrI6ODs6MoIhAIJqxjuj5/X5V48WD2+2G1WpFeXk5tm/fPqogE9pQS0AAACAASURBVAqF0NjYiPLyclitVu4gQmkrHs0lkykVZ34QxSJdFwr9fdj9GVEcH+7QcjLWEymK8sUkz+8SQjwI4CEAswHcBOD/jvEcEy474X8QmU9RFLS0tGDr1q3YuXPniF8ENRoNJElCbW0tHA4H/z0QxUAIAYfDAYfDAVmW4fP54Pf7odfrUVxczK09aUz5+VOdUDk2fQqOqg7X1NQU0/am3d3dsNvtcDqdnFVEaSncXLL1+HFs7emBs7cXw0twWiEgzZ2L2gULUFZQkPK/e4VnfkhdXQgODET9ulSe+UHZQ6TjtLuhHhLh4sIriqKsnuT4AAAdgD8rivI/4pDPPAB/BSAANCmKcqXK8U0Y7CWBTz/9FCbTZD0wKZ243W5UV1dH1c3dbDajoaGBI7lERHEkyzKMRqMqyza0Wi0+//zzlC1+ud3uaTWBdblc/JmUobp8X2D1ljew698uh6V4TrLTiavpNpdMFW6/H9X790fVuNOi06H+ootYjKAJeb1eLFwYacO4cLKB9KlIyyUbiqL8HcDnQw8n/HQuhCjEYDECGPpQH4d8Dg/Lpzge56DM1NTUBLvdHvXWcuFRqaampjhnRkSUvQwGAyRJUiWWJEkpW4xQFAXV1dVTKkYAQDAYRE1NDXtKUNpTo7lkKlB7W1GiREjPq22QB8AVAC4QQmgn2PrzomH398cxH/40ppi43e6YpsiGBYNBSJLEUSkiojiqra1FY2PjtOOsXLlShWzio6WlJeqC+Hi6urrQ2trKrU2JUoSa24oSJUJazpAY8sbQnzoAyyY4rmzY/TfjkYgQogj/2Ia0Jx7noMzCUSkiotTmcDii3qVlIps2bUrZBpB1dXUpFYeI1JUpMz8os6VzQWLnsPs3jnWAECIHQPXQw+MAmuOUy3oM9o8AgNY4nYMyiJqjUkREpD4hBBoaGjB7mtvppWoBWZZlOJ1OVWLt2LGDO0IREdGUpG1BQlGUNgC7hx7eJIQYq1nl/wRQMnT/cUVRTg9/UgjhEEIoQ7dtZ75YCHGeEGLpRHkIIVYDuH/o4UkAT8XwNihLqTkqxW1CiYjiw2azYdOmTdOOk4oFZK/Xq0rTTmBwS1Cfz6dKLCIiyi7pPm/nNgwuw5gF4DUhxMMYnAUxC8B1GJy5AAAHADw2hfjnAWgWQuwB8DKA9wAcHnpuEYDKoVt4dsQdiqLwJzJNSM1RqRdeeAE7duzgNqGUlmRZhtfrRSAQQH5+PkwmU8o2/6Ps1dbWpkqcurq6lOqzEAgEVI3n9/tVjUdERNkhrQsSiqLsE0J8B8D/z979B7ed5/d9f31FxPQJBE5CljeOyO2dfXeOeACyFborxfEF/Jo3bKa13QQ+aTJtPVRSZy4ZTi9pnMSZ5jLpNBM7U3ccJ72R2nOaJmTSjqOFhE1zcWcCz5FENk1WcqAsRWI3zt2eLyK2vXCXp/2C0B2vBD/9Q4SWIkUSPz4Avl98n48ZjkgReH8/EgWR3/fn/Xm//4GkuKRfes7DflvSTxpjuvlO+WP7b8d5LOnPGmN+rYtrICRs7kpJOhKr0Wgon88rn88zJhS+Y4zR8vKybty4oddee41kGnytF8ca/JJ0GxsbsxovRrd+AEAHAntko8kY848l/T5Jv6onyYfHetIv4rck/UVJl4wxX+8w/L+S9LOSbkh6Q9K/24//fUnflvQ1SV+S9MMkI9Aq27tSJ2FMKA4b5BGfcrmsdDqtmZkZ3b59+9hk2szMjNLptG8bASI8hvVYgzFG77zzjrV4kUhEExNMPQcAtC/QFRJNxphvSfr5/bd2nresD49bPO/zNUn/+/4bYIXtXanTMCYUfqhKKBaLbY25bSbTCoWCZmdne7Im4DTDeKyhXC5rbm6u68bKB+VyOd9UfgAAgiXwFRJA0ExOTmpkZKSv1/Rrl3f0nh+qEsrlclvJiKZmMo1KCQzKsB1rKBaLymazVpMRkjQ/P//c36fpMgDgNCQkgD6Lx+PK5XJ9v64fu7yjt9q9+ejFER9jjObm5tpORjSRTMMg2UwgD/pYQ6eJwdOkUilNT08//dgYo6WlJV29elWJRELJZFJXrlxRMplUIpHQtWvXtLS0xGsaACCJhAQwEMftJvWarXGj8D+/VCUsLy93vRtLMg2d6naH3mYCeZDHGrpNDB4nGo1qYWHh6VEvP1RkAQCChYQEMACu6yqZTPb9us0u7xhufqpKsJUEI5mGVtneobeVQB5UIlqykxg8LBqNqlAoPO1N5IeKrE5wrAQABouEBDAAjuNocXFR0Wi0r9f1U5d39I5fqhJ6MTIROEkvduhtJJAPH2voN9sJvVQqpVKp9LThrF8qslrFsRIA8A8SEsCAZDIZFQqFvicl/NDlHb3ll6qEYR2ZCH/q1Q59twnkw8ca+s1mYtBxHH31q1/V6urq08oIP1VktYJjJQDgLyQkgAGanZ1VqVTq6/GNQXd5R2/5qSphGEcmwp96vUPfaQL58LGGQbCZGDTG6Ed+5EeeSa74pSKrFUE9VgIAw4yEBDBgmUxGDx48eFo+erij+8jIiLWdtUF3eUfv+akqYdhGJsKf+rVD324C+fCxhkHpdWLQLxVZpwnasRIACAsSEoAPOI4j13X16quvamtrS5VKRW+88YYqlYq2trb0+c9/3sp1BtnlHf3hp6qEYRqZCP/q5w79aQnkSCTytP/AwWMNg9TLxKCfKrJOErRjJQAQJpFBLwAYJp7naWNjQ9vb2xobG9Pk5GTbCYB4PH7kOfPz88rn812vb5Bd3tEffqpKaI5MtPFvl2QajmNzh9513VMf10wgu64rz/NUrVZVq9UUi8U0MTHhu3+nzcSgjcqpw4nBXlRk9eLvz2bSqpV/IwCA1lEhAXSpH926h6HLO/rDb1UJwzAyEf416B36eDyuqakpXb58WVNTU75LRkgfJgZtOJwY9FNF1kmCcqwEAMKIhATQhX516w56l3f0Ty9vPjpBMg295KeeKX7Wq8SgnyqyjjPopBUA4GQkJIAO9btbd5C7vKO//FSVQDINvRSUHfpB61Vi0G8VWc9D0goA/I2EBNCBQXXrDmqXd/SX36oSSKahV4KwQ+8HvUoM+q0i63lIWgGAv5GQANo06G7dQezyjv7yY1UCyTT0QhB26P2iV4lBP1VkPQ9JKwDwNxISQJv6OWLuOCeNCX3//fd169Ytua5LmXuI+bEq4WAy7fOf/7zOnHn+t6AzZ87o4sWL+uCDDxizhxMFYYfeT3qRGPRbRdZhJK0AwN9ISABt8lu37iB0ecdg+LEqwXEcxeNxvf3229rb23vuY/b29qw0g0U4+H2H3m9sV9n5sSLrIJJWAOBvDrtP/uc4zqSkh5L08OFDTU5ODnhF4eV5nhKJhJUGWSMjI9ra2uKHG/ScMUYrKyu6ceOGCoXCM/9+I5GIcrmc5ufnNT093fOqmmKx2Hb/lWbVBsc38DzGGKXT6a4q11KplFZXV0NZVeZ5nqrVqmq1mmKxmCYmJjr6vuTn1/bS0pJmZmasxHFdt/sFBcRa9QP91Jdf11e/+FmlJj466OUAGICNjQ29+OKLzQ9fNMZs2L5GxHZAYJj1ols3CQn0WvOIj+u61m4+OtFtM9hSqURPFBzR3KHPZrMd9fYJ+ySXeDxu5f+AZkXW3NxcS8mhVCqlhYWFvrymm8dKuk1aMX4YAOzjyAbQBrp1I+gGdcRn0M1gMdz82DMljPzadNnvx0oAIMxISABtoFs30Bk/NIPFcPNjz5Qw8mvTZZJWAOBPJCSANtCtG+iM35rBYjj5dYc+rPzWdJmkFQD4Dz0kgDY0u3Xn8/muY9GtG2HheZ4KhYKVWHfu3JHnebx2cCy/9EyBPzWTVn5p9AsAYUdCAmjT/Py8lYREWEbMATSDxaDYatiI4ULSCgD8g4QE0Ca6dQPtoRksgH7wPE8bGxva3t7W2NiYJicnT00uHJe06iQWAKB99JAA2kS3bqA9NIMF0CvGmKc9QxKJhJLJpK5cuaJkMqlEIvG0Z0grE3psxgIAtIaEBNABunUDraMZLIBeKJfLSqfTmpmZ0e3bt48cDWs0Gsrn85qZmVE6nVa5XO5LLABA60hIAB2iWzfQmmYzWBtoBgtAkorForLZbMvHJ9fX15XNZlUsFnsaCwDQHhISQBcYMQe0xlYTV5rBAiiXy8rlcqrX6209r16vK5fLPVPdYDMWAKB9NLUEukS3buB0NIMFYIMxRnNzc20nEJrq9bquX7+u1dVVSbIWi75QANAZEhKARYyYQ5D0s4t8sxlsNpvt6Id/msECkKTl5eWuEpuStLa2ppWVFRljrMVyXberOAAQVhzZAIAQGWQXeZrBAujWzZs3rcWxGQsA0BkSEgAQEn7oIk8zWACd8jxPhULBSqw7d+7ozp071mJ5nmclFgCEDQkJAAgBP3WRpxksgE5sbGwcSaR2qtFoaG9vz1qsarVqJRYAhA09JABgyHXbRb5UKllPCtAMFkC7tre3B72EY9VqtUEvAQACiYQEAAwxmx3pe9VQkmawAA46ruHu2NjYoJd2rFgsNuglAEAgcWQDAIaYzY70ANArrTTcfeedd3TmjJ0fXSORiNVYExMTVmIBQNiQkACAIWar+/sv/uIvWokDAIe12nD3p3/6p61VSeRyOf3Mz/yMtVhUeQFAZziyAQBDymZH+t/8zd/UBx98oI9+9KNW4gGA9KThbjs9bmxNs5ifn5cxRvl83kosAEBnqJAAgCFlsyO9JL322mvWYgFApw13u5VKpTQ9PS3XdVseQXxaLABAZ0hIAMCQst2R/u///b9vNR6A8Oq24W6notGoFhYW5DiOHMfR4uKiotFo17EAAJ0hIQEAQ8p2R/qlpSVr5dIAws1Gw912nTlzRr/8y7/8zBjjTCajQqHQdlIiGo2qUChYH4kMAGFDQgIAhtTk5KRGRkasxdvb21O1WrUWD0B42Wq42469vT39wi/8gorF4jO/Pzs7q1Kp1PLxjVQqpVKppNnZ2V4sEwBChYQEAAypeDxu/WxzrVazGg9A+NhsuNuuer2uXC6ncrn8zO9nMhk9ePDg6ejRw8ncSCSia9euaWlpSaurq1RGAIAlTNkAAB/yPE8bGxva3t7W2NiYJicnOxor97M/+7P62te+Zm1dsVjMWiwA4WS74W676vW6rl+/rtXV1Wf6PziOI9d15bquPM9TtVpVrVZTLBbTxMQEoz0BoAeokAAAnzDGPN2dSyQSSiaTunLlipLJpBKJxNPdOWNMyzF/5md+xtr6IpGIJiYmrMUDEE62G+52Ym1tTSsrK8d+Ph6Pa2pqSpcvX9bU1BTJCADoERISAOAD5XJZ6XRaMzMzun379pHdw0ajoXw+r5mZGaXT6SPlxsf56Ec/qs997nNW1pjL5fihHEDXbDfc7ZStPhae56lSqeju3buqVCo0/wWANpCQAIABKxaLymazLXecX19fVzabPdKY7Thf+tKXulneU/Pz81biAAg32w13O3Xnzp2Okwe9qGgDgDAiIQEAA1Qul5XL5VSv19t63nGN2Z7Hdd2Wu8cfJ5VKWW+QCSCc4vG4crncoJehRqPR0eSgXlW0AX7g7e6qUq/rruepUq/L290d9JIw5EhIAMCAGGM0NzfXdjKiqdmY7bQdOMdxtLi4qGg02tF1otGoFhYWnmn+BgDd8EvFVbuTg3pd0QYMgjFGS9/5jq6urSnx+utK3runK+WykvfuKfH667q2vq6l73yHih/0BAkJABiQ5eXlln+oPc5pjdmaMpmMCoVC20mJaDSqQqHAiDsAVtmo3LKhnclB/ahoA/qtXKspfe+eZt58U7ffe0+H5980JOU3NzXz5ptK37unMuO/YRkJCQAYEFsN1VqNMzs7q1Kp1PJNQCqVUqlU0uzsbDfLA4Ajuq3csqGdyUH9qmgD+qm4taXs/ftaf/y4pcevP36s7P37Km5t9XhlCBMSEgAwAJ7nqVAoWInVTmO2TCajBw8ePG3GdrixXCQSedqMbXV1lcoIAD3TaeWWrYaY7UwO6mdFG9AP5VpNubU11ff22npefW9PubU1KiVgDQkJABiAjY2NI43QOtVuYzbHceS6rl599VVtbW2pUqnojTfeUKVS0fvvv69bt27JdV16RgDouU4qt27cuGHl2u30seh3RRvQS8YYzb31VtvJiKb63p6uv/02FT+wgoQEAAzA9va21XjtNmZrisfjmpqa0uXLlzU1NdXybiEA2NJu5dYXvvCFvk4OGlRFG9Ary48etXxM4zhr9bpWHj2ytCKEWWTQCwCAMBobG7Mar53GbADgN83KLdd15XmeqtWqarWaYrGYJiYmjiRLFxcXlc1mO+rp0O7koF5UtJH8xSB4u7va2NnRX/vWt6zEu/nuu3LPn7cSC+FFQgIABmByclIjIyNWfshtpzEbAPhdPB4/9Ya92X+i3akXnUwO8ktFG9AJY4yWHz3SjWpVrz1nikY37mxuytvdVTzCLSU6x5ENABiAeDyuXC5nJVY7jdkAYFj0a3IQFW0IqtNGenarIam6s2M5KsKGhAQADEg7DdX6EQcAgqYfk4OaFW02UNGGfml3pGenapaOMyG8qK8BgAFxXVfJZLKrUXLtNGYDgGHUbv+JdjUr2vL5fNdrpaIN/dDpSM9OxCwl6xBeVEgAwIA4jqPFxUVFo9GOnt9uYzYAGHa9mhxERRuCotuRnu2IOI4mRkd7fh0MNxISADBAzcZs7SYlOmnMBgDoTLOirRtUtKEfbIz0bFXuhRdoaImukZAAgAHrV2M2AEBnqGhDUNx8992+XWv+woW+XQvDi4QEAPhAPxqzAQA6R0Ub/M7b3VVhc7Mv10pFo5o+d64v18Jwo8YGAHyi143ZAADdaVa0zc3NtdSQOJVKaWFhwVoywvM8bWxsaHt7W2NjY5qcnOR7A57a2NmxPtrzeaJnzmjh4kUqfmAFFRIA4EO9aswGAOhOvyvajDFPr5VIJJRMJnXlyhUlk0klEomn1zLGdHUdBN92H0ZwRs+cUSGVUiYW6/m1EA5USAAAAABt6FdFW7lcPrEao9FoKJ/PK5/PK5lManFxkaMhITbW4xGcqWhUCxcvkoyAVSQkAAAAgA7F4/GeVLEVi0XlcjnV6/WWHr++vq5sNqtCoUDT45CaHB3ViGT12EbEcZR74QXNX7ig6XPnOKYB60hIAAAAAD5SLpfbSkY01et15XI5lUolKiVCKB6JKDc+rryFxpafO3dOX/70pzUxOspoT0u83V1t7Oxou9HQ2MiIJvm7lURCAgFDMycAADDMjDGam5trOxnRVK/Xdf36da2urrKbHULzFy5YSUj85Y9/XFMdjrnFh4wxWn70SDeqVb323nvPVK+MSMqNj2v+wgW5Ia4+oaklfI9mTgAAICyWl5dbmuBxkrW1Na2srFhaEYLEPXdOybNnu4rBSE87yrWa0vfuaebNN3X7UDJCenK0Jr+5qZk331T63j2Va7VBLHPgSEjA18rlstLptGZmZnT79m01DnUPbjZzmpmZUTqdVrlcHtBKAQAAunfz5k1fxUGwOI6jxakpRc90dpvHSE87iltbyt6/r/XHj1t6/Prjx8rev6/i1laPV+Y/JCTgW8ViUdlstuVdgmYzp2Kx2OOVAQAA2Od5ngqFgpVYd+7cked5VmIhWDKxmAqpVNtJCUZ62lGu1ZRbW1N9b6+t59X39pRbWwtdpQQJCfhSt82c/F4p4XmeKpWK7t69q0qlwg8MgHhdAMDGxsaRatBONRoNVatVK7EQPLOJhEqXLrV8fCMVjap06ZJmE4ker2y4GWM099ZbbScjmup7e7r+9tuhOopOQgK+Y6uZk99eyPTCAI7idQEAH9re3rYarxaynVY8KxOL6cErr2jppZd0dXxcI4c+H3EcXRsf19JLL2n15ZepjLBg+dGjlo9pHGetXtfKo0eWVuR/TNmA79hs5uS6rp1FdalcLmtubu7YP1ezF0Y+n1cymdTi4iLjujD0eF0AwLPGxsasxotxgxl6juPIPX9e7vnz8nZ3Vd3ZUa3RUGxkhJGePXDz3XetxXHPn7cSy++okIDvDFszJ3phAEfxugCAoyYnJzUycngfuzORSEQTExNWYmE4xCMRTUWjuhyPayoaJRlhmbe7q4KFkauSdGdzU97urpVYfkdCAr4ybM2chr0XBtAJXhcA8HzxeFy5XM5KrFwup3g8biUWgNNt7OwcGe3ZqYak6s6OpWj+RkICvjJMzZyGtRcG0A1eFwBwsvn5eV/FAdCabUv3ME01y/H8ioQEfGWYmjnZ7IUB2DaoiRa8LgDgZK7rKplMdhUjlUppenra0ooAtGLM0nGrppjleH5FQgK+MkzNnIatFwaCzw8TLXhdAMDJHMfR4uKiotFoR8+PRqNaWFiQ4ziWVwbgJJOjo0cmmXQq4jiaGB21FM3fSEjAV4almdOw9cJA8JXLZaXTac3MzOj27dtHjkY1J1rMzMwonU73pE8DrwsAaE0mk1GhUGg7KRGNRlUoFJhIBAxAPBJRbnzcSqzcCy+EpukoCQn4yrA0cxqmXhgIPr9MtOB1AQCtm52dValUavn4RiqVUqlU0uzsbI9XBuA48xcu+CpOEJCQgO8MQzOnYeqFgWDz00QLXhcA0J5MJqMHDx48PW53uIo0Eok8PW63urpKZQQwYO65c0qePdtVjFQ0qulz5yytyP/CUQeCQGk2c+qm8d2gmzkNUy8MBNfe3p4+//nPdz3RYnV11cpZZF4XANA+x3Hkuq5c15XneapWq6rVaorFYpqYmGC0J+AjjuNocWpK2fv3Vd/ba/v50TNntHDxYqh6wFAhAd8ZhmZOw9ILA8FVLpf1yU9+Ur/zO7/TVRybEy14XQBAd+LxuKampnT58mVNTU2RjAB8KBOLqZBKKXqmvVvt6JkzKqRSyoRsw4WEBHwp6M2chqUXBoKp2TOi22REk62JFrwuAABAGMwmEipdutTy8Y1UNKrSpUuaTSR6vDL/ISEB3wp6M6dh6IWB4Om0Z8RJbE604HUBAADCIBOL6cErr2jppZd0dXz8yEjQiOPo2vi4ll56Sasvvxy6yoimoegh4TjOxyX9aUk/KelFSTuSviHplqQbxpjHlq7zn0j6gqRXJI1L2pR0T9KvGWP+LxvXwLOazZxWVlZ048YNFQqFZ7r0RyIR5XI5zc/Pa3p62lfnrYahFwaCxRijubk5q8kI6cOJFjYqEnhdAACAsHAcR+7583LPn5e3u6vqzo5qjYZiIyOaGB0NzWjPkwT+b8BxnJ+W9A8kHfxJ+aykl/ff/oTjOD9pjPl6F9c4I+nXJP3coU9N7L/9Ecdx/ldJf9IY0373EpwoqM2cmr0wstlsRzeIfuiFgWBZXl7u6kb/JLYmWvC6AAAAYRSPREhAPEegj2w4jnNJ0j/Uk2TEtqQvSfoDkj4n6W/vP+xHJf0Tx3G6qYH5RX2YjLgv6T+XdHn/1/v7v/8nJP21Lq6BFgStmVPQe2HADs/zVKlUdPfuXVUqFWvHHw6z1evheWxOtOB1AQAAACngCQlJf0vSRyTtSvqPjTG/ZIz5F8aYrxljviDpF/Yf96OS/lwnF3Ac50cl/fn9D39L0o8bY37dGHPPGPPrkj67//uS9Bccx/lUp38YDKeg98JAZ4wxT+fGJxIJJZNJXblyRclkUolE4unceGOMlet5nqdCoWAl1mG9mGjB6wIAAACBTUg4jnNZ0h/c//DvGGP+xXMe9iuS3tp//884jvO7OrjUf6MPj7Z80Rjz3YOf3O9P8cX9DyOS/mwH18CQa/bCaN6gHh59GIlEnt6grq6usgMccOVyWel0WjMzM7p9+/YzfU+kJz0Z8vm8ZmZmlE6nVS6Xu77mxsbGkevY0quJFrwuAAAAwi3Ih1j+yIH3/+7zHmCM2XMcZ1HSX5d0TtJPSPqnrV7AeXJI+Q/vf/i2MeZfHnOdf+k4zr+R9Hsl/WHHcf5rY2vbE0MjqL0w0J5isdjWlIv19XVls1kVCoWudv+3t7c7fu5pejnRgtcFAABAeAU5IfHZ/V/rkv7VCY9bOfD+j6uNhISkH5Z04TlxjrvO79WTJpefkPTNNq6DkInH49xoDaFOR27W63XlcjmVSqWOqwDGxsY6et5p+jnRgtcFAABAuAQ5ITG1/+vXjTG7Jzzu7ec8p1WfOSZOK9dpOSHhOM7kKQ/5oVZjARiMbkdu1ut1Xb9+Xaurqx1NkJicnNTIyIjVYxtMtAAAAEAvBbKHhOM4Pyjphf0PN056rDHmO3pSRSFJL7Z5qYOJghOvI+nhgffbvc7DU97utRkPQJ/ZGLm5tramlZXTirGeLx6PK5fLdXX9g5hoAQAAgF4LZEJC0sH5c60cnG4mJNqtaW7nOge3RXtTOw3At2yN3Owmjq1eD5/4xCeYaAEAAICeC+qRjR888P73W3j8zv6vH+nhdXYOvN/udU6rqPghUSUB+JbNkZt37tyR53kd9VJwXVfJZLKrSo0f/uEf1te//nWdORPUfDUAAACCIqgJie8deP8HWnj86P6v3z3xUd1dZ/TA+21dxxhz4nEQzm8D/mZz5Gaj0VC1Wu0oIeE4jhYXF5XNZjvqZRGNRpXP50lGAAAAoC+C+lNn7cD7rRyPiO7/2u5cvHauEz3wfu/m7wHwHdsjN2u12ukPOkYmk1GhUFA0Gj39wQfQMwIAAAD9FsiEhDHme5Le3//wxAkVjuOc14fJgocnPfY5DlYunDYJ4+Cxi3avAyDAbI/cjMVipz/oBLOzsyqVSkomky09PpVK0TMCAAAAfRfIhMS+yv6vn3Ic56SjJxcPvP9Wh9c4HMf2dQAEWHPkpg2RSEQTExNdx8lkMnrw4IGWlpZ09erVI+uLRCK6du2alpaWtLq6SmUEAAAA+i6oPSQk6XVJf1BPqh/+I0lvHPO46QPv//M2r/FNSe9KunAozvNk93+tlMD10QAAIABJREFUSvqdNq8DIMCaIzfz+XzXsXK5XEf9I57HcRy5rivXdeV5nqrVqmq1mmKxmCYmJqxdBwAAAOhEkCskXjvw/h9/3gMcxzkjaW7/w0eSltq5gDHGSPpH+x9edBzn9x9znd+vDysk/tH+8wCEiK2Rm7biHBaPxzU1NaXLly9ramqKZAQAAAAGLrAJCWPMXUn/bP/Dn3Mc58ee87A/J2lq//2/ZYz5/w5+0nEc13Ecs//294651N+U1Gyf/2XHcZ4Z6bn/8Zf3P9zdfzyAkGmO3OxGKpXS9PRpxVgAAADAcAhsQmLfn9GTEZsRSf/UcZz/1nGc3+84zk84jvMVSb+8/7jflvQrnVzAGPPbkv7H/Q9flvTPHcf5o47jvOw4zh/Vk2MgL+9//n80xvzbTv8wAIKrOXKz3ekWTdFoVAsLC4z5BQAAQGgEuYeEjDH395MC/0BSXNIvPedhvy3pJ40xnc/Rk74k6WOS/itJlyT9+nMe83ck/eUurgEg4JojN3O5nOr1esvPY+QmAAAAwijoFRIyxvxjSb9P0q/qSfLhsZ70i/gtSX9R0iVjzNe7vMaeMebnJP2knvSUeFfS9/d//UeS/lNjzJ8wxux1cx0AwcfITQAAAKA1ga6QaDLGfEvSz++/tfO8ZUkt10cbY35D0m+0tTgAodMcubmysqIbN26oUCio0Wg8/XwkElEul9P8/Lymp6c5pgEAAIBQGoqEBAD4DSM3AQAAuuft7mpjZ0fbjYbGRkY0OTqqeITb2GHBVxIAeiwej5OAAAAAaJExRsuPHulGtarX3ntPjQOfG5GUGx/X/IULcs+do9I04EhIAAAAACHheZ42Nja0vb2tsbExTU5OkjSHr5RrNc299ZbWHz9+7ucbkvKbm8pvbip59qwWp6aUicX6u0hYE/imlgAAAEBYeZ6nSqWiu3fvqlKpyPO8I48xxmhpaUlXr15VIpFQMpnUlStXlEwmlUgkdO3aNS0tLckYM4A/AfCh4taWsvfvH5uMOGz98WNl799XcWurxytDr5CQAAAAAAKknQRDuVxWOp3WzMyMbt++/UyTZUlqNBrK5/OamZlROp1WuVwe0J8KYVeu1ZRbW1N9r73BhfW9PeXW1lSu1Xq0MvQSRzYAAMeitBcA/KVcLmtubk7r6+vP/XwzwZDP5/Xxj39c3/72t/W9732vpdjr6+vKZrMqFAr6PZ+5bHPZwImMMZp76622kxFN9b09XX/7ba2+/DI9JQKGCgmgDa2URQJBR2kvAPhTsVhUNps9Nhlx2Le+9a2WkxFN9XpduVxOb739didLBDqy/OhRy8c0jrNWr2vl0SNLK0K/kJAATsHNGcKE0l4A8KdyuaxcLqd6vd7za9Xrdf13f+Wv9Pw6QNPNd9/1VRz0DwkJ4ATcnCFM2t15a5b2FovFHq8MAMLNGKO5ubm+JCOavvGNb/TtWgg3b3dXhc1NK7HubG7K2921Egv9QUICOAY3ZwiTTnfemqW9JOMAoHeWl5db/nkECJqNnR01Tn9YSxqSqjs7lqKhH0hIAM/BzRnCpNudt3q9ruvXr3NsCUBPhbmP082bNwd27e0+VmUgnLYbttIRT9Qsx0NvkZAADuHmDGFjY+dtbW1NKysrllYEAE/Qx+lJIqZQKAzs+pv//t8P7NoIh7GREavxYpbjobdISACHcHOGsLG18zbIHTwAw4c+Tk9sbGwc+bP3Uz/7ViCcJkdHZSuFEHEcTYyOWoqGfiAhARzCzRnCxObO2507d0JVQg2gd+jj9KHt7e2BXj8ajQ70+hh+8UhEufFxK7FyL7ygeCRiJRb6g4QEcAA3ZwgbmztvjUZD1WrVSiwA4UUfp2eNjY0N9PrjH/vYQK+PcJi/cMFXcdA/JCSAA7g5Q9jY3nmr1WpW4wEIF/o4HTU5OamRAZ6JH6NCAn3gnjun5NmzXcVIRaOaPnfO0orQLyQkgAO4OUPY2N55i8ViVuMBCBf6OB0Vj8eVy+UGvQygpxzH0eLUlKJnOrs9jZ45o4WLF+U4juWVoddISAAHcHOGsLG58xaJRDQxMWElFoBwoo/T883Pz/f9mp/85Cf7fk2EWyYWUyGVajspET1zRoVUShl+7g4kEhLAAdycIWxs7rzlcjnF43ErsQCED32cjue6rpLJZN+uF41G9d//1b/at+sBTbOJhEqXLrV8fCMVjap06ZJmE4kerwy9QkICOICbM4SRrZ23QezgARge9HE6nuM4Wlxc7MvEi2g0qkKhoKmLF3t+LeB5MrGYHrzyipZeeklXx8ePjASNOI6ujY9r6aWXtPryy1RGBBwzUYBD5ufnlc/nrcQBgqC589bNue1UKqXp6WmLqwIQNvRxOlkmk1GhUGh7AslHPvIRfexjH9O3vvWtUx+bSqW0sLCgTCajteoH3SwX6IrjOHLPn5d7/ry83V1Vd3ZUazQUGxnRxOgooz2HCBUSwCE2yiK5OUOQdLvzFo1GtbCwQCOpPvI8T5VKRXfv3lWlUhmq0nSEF32cTjc7O6tSqdTyzympVEqvv/66vvnNb2ppaUlXr149cjQ1Eono2rVrWlpa0urqqjKZTC+WDnQsHoloKhrV5XhcU9EoyYghQ0ICOISbM4RRc+et3X/3zdJefoDtPWPM0xuKRCKhZDKpK1euKJlMKpFIPL2hGKZxhwgX+ji1JpPJ6MGDB20lGBzHkeu6evXVV7W1taVKpaI33nhDlUpF77//vm7duiXXdfnZBUDfOfzg4n+O40xKeihJDx8+1OTk5IBXFA7FYrHtssjmzdns7GwPVwb0Trlc1tzcXEvHNw6W9qK32vm6JJNJLS4u8nVBIF27ds3Ksclr167p1q1b8jxPGxsb2t7e1tjYmCYnJ4euv5PneapWq6rVaorFYpqYmLDyZ1yrfqCf+vLr+uoXP6vUxEctrBRA0GxsbOjFF19sfviiMWbD9jWokACO0UlZZKlUIhmBQOtk5w29VSwWlc1mW+7xsb6+rmw2q2Kx2OOVAfbZ6r/0Yz/2Y6GpJorH45qamtLly5c1NTU1dAkXAMONCokAoEJisIwxWllZ0Y0bN1QoFJ7pAB6JRJTL5TQ/P6/p6WlKHTF0erXzhtaUy2Vls9m2KrWaotGoSqUSSSMEijFG6XS6qya7o6Oj2tnZOfVxVBOdjAoJAP2okCAhEQAkJPyDmzMA/WLjxiyVSml1dZVkKQKlm0RcuzhqeTwSEgA4sgH4DGWRAPpleXm5q2SEJK2trWllZcXSioD+6LTJbifq9bpyuZzK5XLPrwUAOIqEBAAAPnTz5k1fxQH6qd0+TqOjox1fq16v6/r160PTUwIAgoSEBAAAPuN5ngqFgpVYd+7cked5VmIB/dRqk92/8Tf+Rks9I05CNREADEZk0AsAAADP2tjYeKaBbjcajYaq1SpHzBBIjuPIdV25rntsH6dr165ZudbNmzfluq6VWH4ThvGnAIKJhASwj2/WAPxie3vbarxarWY1HjAI8Xj8yPflXlQTDcv3fmOMlpeXdePGDb322mvPJDlHRkaeTglzXZfGtwAGhiMbCDVjzNNS0DDMKgcQDGNjY1bjxWIxq/EAv+hFNdEwKJfLSqfTmpmZ0e3bt4/8HTUaDeXzec3MzCidTtPUE8DAkJBAaPHNGoBfTU5OHjkv36lIJKKJiQkrsYB2eZ6nSqWiu3fvqlKpWO9nQjXRUcViUdlstuUpPevr68pmsyoWiz1eGQAcRUICocQ3awB+Fo/HlcvlrMTK5XJDU4KOYOik+rDTxAXVRM8ql8vK5XKq1+ttPY/xpwAGhYQEQodv1gCCYH5+3ldxgFa0W334la98patjk1QTfcgYo7m5ubZ/vmli/CmAQSAhgVDhmzWAoHBdV8lksqsYqVRK09PTllYEnKyT6sM/9af+VFfHJoNWTdTLIyzLy8st/90fh/GnAPqNhARChW/WAILCcRwtLi4qGo129PxoNKqFhQW656MvOq0+bMVpxyb9Xk3UrwbaN2/etLJeW3FwlLe7q0q9rruep0q9Lm93d9BLAgbOYafX/xzHmZT0UJIePnyoycnJAa8ouK5du6Z8Pm8lzq1btyysCBhO7YzRZeTuyYrFYts3etFoVIVCQbOzsz1cGfCEMUbpdLrrhP9potGoSqWSMpmM9eunUimtrq5aT+CVy2XNzc21tLZkMqnFxcUjf75WeJ6nRCJhZeLIyMiItra29O9qRj/15df11S9+VqmJj3YdN6yMMVp+9Eg3qlW99t57OvgVGpGUGx/X/IULcs+dI4EM39nY2NCLL77Y/PBFY8yG7WtQIYHQ6MWscgAfamcXkJG7rZudnVWpVGr5+EYqlVKpVCIZgb6xUX3YiuOOTfq1mqifDbQZf+pP5VpN6Xv3NPPmm7p9KBkhSQ1J+c1Nzbz5ptL37qk8BFNegHaRkEBo8M0a6J12Gtl9+tOf1qc+9SlG7rYhk8nowYMHT5M4h5v4RSKRp0mc1dXVjnZYgU71s8T/uGOTmUxGhUKh7aREs5rI9mum3w20GX/qP8WtLWXv39f648ctPX798WNl799XcWurxysD/IWEBEKDb9ZAb7S7C/iNb3xD77zzTkuPZeTuhxzHkeu6evXVV7W1taVKpaI33nhDlUpF77//vm7duiXXdSn5RV/ZrD5s1XEJEL9UEw2igTbjT/2lXKspt7am+t5eW8+r7+0pt7ZGpQRChYQEQoNv1oB9vWxk18TI3aPi8bimpqZ0+fJlTU1N0XMDA2Oz+rBVJx2b9EM10SAaaDP+1D+MMZp76622kxFN9b09XX/7bY4sIjRISCA0+GYN2NXtLmA7GLkL+JPt6sNWnHZsctDVRIOYdhG08afDbPnRo5aPaRxnrV7XyqNHllYE+BsJCYQG36wBu/rVyK6JkbuA/9iuPmxVq8cm+11NNMgG2n4ffxoWN99911dxAL8jIYFQ4Zs1YM8gZtUP4poAjmez+rAdfj02OcgG2q7rttw/4zipVErT09NdxQgzb3dXhc1NK7HubG7K2921EgvwMxISCBW+WQN2DKKRncTIXcBvbFYftsrPxyYH2UDbr+NPw2RjZ+fIaM9ONSRVd3YsRQP8i4QEQoVv1oAdg2hkJzFyF/CjflcN+vnY5KAbaPtt/GnYbFv+vlgbwPdZoN9ISCB0+GYNdG8QjeyaGLkL+IuN6sN2+PnYpB8aaPtl/GkYjVk+vhQbwHEooN9ISCCU+GYNdGdQjewk/54dB8Kq2+rDdvj92KRfGmj7YfxpGE2OjspWCiHiOJoYHbUUDfAvEhIILb5ZA50bVCM7P58dB8Ks0+rDdgTl2KRfGmgPevxpGMUjEeXGx63Eyr3wguKRiJVYgJ/xrxyh1vxm7bquPM9TtVpVrVZTLBbTxMSEb8+oAoPW3AXM5/N9va6fz44DYdesPpybm7M+EjhIxyabR1i6+TuwXQkSj8f5v7NP5i9cUN7CpI35CxcsrAbwPyokgH39nlUOBN0gznH7+ew4gNarD7/yla/oM5/5TEsxg3Zskgba4eaeO6fk2bNdxUhFo5o+d87SigB/IyEBAOhIvxvZ+f3sOIAnWjkq8IUvfEFra2tDe2ySBtrh5TiOFqemFD3T2W1W9MwZLVy8SEIKocGRDQBAR5q7gNlsVvV6vafXYscQCKaTjgoM+7HJdo+wpFIpLSwskIwYAplYTIVUSrm1NdX39lp+XvTMGRVSKWVo3owQoUICACR5nqdKpaK7d++qUqnI87xBLykQ+tXIjh1DYLgN67FJGmiH12wiodKlSy0f30hFoypduqTZRKLHKwP8hQoJAKFljNHy8rJu3Lih1157TY1G4+nnRkZGlMvlND8/TxfyU7S7C/jJT35Sxhi98847pz6WHUMAQTfslSA4XiYW04NXXtHKo0e68e67KmxuqnHg8xHHUe6FFzR/4YKmz53jZw2EEgkJAKFULpdPvIFuNBrK5/PK5/NKJpNaXFzkpvgEzV3AlZUV3bhxQ4VC4ZkETyQSeZrgafaBaPWx/IAGYFgw7SJ8HMeRe/683PPn5e3uqrqzo1qjodjIiCZGRxntidDjFQAgdIrFonK5XMt9D9bX15XNZlUoFALT5X0Q2t0FZMcQABAm8UiEBARwCK8IAKFSLpfbSkY01et15XI5lUolKiVa0M4uIDuGAAAA4URTSwChYYzR3NxcxxMh6vW6rl+/LmOM5ZUBAAD0h7e7q0q9rruep0q9Lm93d9BLQohRIQEgNJaXl1tquniStbU1raysyHVdO4sCAADoMWOMlh890o1qVa+9994zzTVHJOXGxzV/4YJcmmuiz6iQABAaN2/e9FUcAACAXivXakrfu6eZN9/U7UPJCElqSMpvbmrmzTeVvndP5VptEMtESJGQABAKnuepUChYiXXnzh15nmclFgAAQK8Ut7aUvX9f648ft/T49cePlb1/X8WtrR6vDHiChASAUNjY2HhmtGQ3Go2GqtWqlVgAAAC9UK7VlFtbU31vr63n1ff2lFtbo1ICfUFCAkAobG9vW41X45s0AADwKWOM5t56q+1kRFN9b0/X3347EI28adIZbDS1BBAKY2NjVuPFYjGr8QAAAGxZfvSo5WMax1mr17Xy6JHc8+ctrcoemnQODxISAEJhcnJSIyMjVo5tRCIRTUxMWFgVAACAfTfffddaHL8lJMq1mubeeuvYhEuzSWd+c1PJs2e1ODWlDBtJvsWRDQChEI/HlcvlrMTK5XKKx+NWYgEAANjk7e6qsLlpJdadzU1fHYGgSefwISEBIDTm5+d9FQcAAMC2jZ2dI6M9O9WQVN3ZsRStOzTpHE4kJACEhuu6SiaTXcVIpVKanp62tCIAAAC7ti1NFWuqWY7XiTA16QwbEhIAQsNxHC0uLioajXb0/Gg0qoWFBZojAUPC8zxVKhXdvXtXlUpFnucNekkA0LWxkRGr8WKW43XCZpNO+AsJCQChkslkVCgU2k5KRKNRFQoFZTKZHq0MQD8YY7S0tKSrV68qkUgomUzqypUrSiaTSiQSunbtmpaWlthFAxBYk6OjspVCiDiOJkZHLUXrnM0mnfAXEhIAQmd2dlalUqnl4xupVEqlUkmzs7M9XhmAXiqXy0qn05qZmdHt27ePTN1pNBrK5/OamZlROp1WuVwe0EoBoHPxSES58XErsXIvvKB4ZLCDGYe5SSdISAAIqUwmowcPHjzdKR05VI4YiUSe7pSurq5SGQEEXLFYVDab1fr6ekuPX19fVzabVbFY7PHKAMC++QsXfBWnG8PapBNPDDbdBQAD5DiOXNeV67ryPE/ValW1Wk2xWEwTExOM9gSGRLlcVi6XU71eb+t59XpduVxOpVKJpCSAQHHPnVPy7Nmu+i6kolFNnztncVWdGcYmnfgQCQkAkBSPx0lAAEPIGKO5ubm2kxFN9Xpd169f1+rqKg1tAQSG4zhanJpS9v79jiZTRM+c0cLFi774f28Ym3TiQxzZAAAAQ2t5ebnlYxrHWVtb08rKiqUVAUB/ZGIxFVIpRc+0d8sXPXNGhVRKmVisRytrzzA26cSHSEgAAIChdfPmTV/FAYB+mk0kVLp0ScmzZ1t6fCoaVenSJc0mEj1eWeuGrUknnkVCAgAADCXP81QoFKzEunPnjjzPsxILAPopE4vpwSuvaOmll3R1fPxItUHEcXRtfFxLL72k1Zdf9k1lxEHD1KQTzyI9BAAAhtLGxsaR0Z6dajQaqlarfe8143meNjY2tL29rbGxMU1OTtLvBkDbHMeRe/683PPn5e3uqrqzo1qjodjIiCZGR31fNTBMTTrxLCokAADAUNre3rYar1arWY13HGPM05HEiURCyWRSV65cUTKZVCKReDqS2BjTl/UAGC7xSERT0agux+OaikZ9n4yQPmzS2W4/jCY/NenEs0hIAACAoTQ2NmY1XqwPZczlclnpdFozMzO6ffv2kQqPRqOhfD6vmZkZpdNplcvlnq8JAPxgWJp04lkkJAAAwFCanJzUiMXxbhd6fPa4WCwqm822PBVkfX1d2WxWxWKxp+sCAL8YhiadeBYJCQAAMJTi8bhyuZy1ePfv37cW67ByuaxcLqd6vd7W8+r1unK5HJUSAEJjGJp04kP+PzAEAADQofn5eeXzeSuxbt68Kdd1rcQ6yBijubm5tpMRTfV6XdevX9fq6irnowGEQtCbdOJDfKUAYAjQiR94vkwmYy1Wc/Sn7dfW8vJyy8c0jrO2tqaVlZWeJEwAwM/ikQgJiADjyAas8jxPlUpFd+/eVaVSYWY70EN04gdOV61WrcVqjv607ebNm76KAwBAvwQ+IeE4zlnHcX7BcZx7juNsOY5TdxznbcdxfsVxnI9biP8Jx3FMi29/z8IfKXC4KQL6j078HyIRipP4ffSn53kqFApWYjUrOAAACIpAJyQcx/mUpH8t6X+Q9LKk85LOSvq9kn5e0qrjOD81uBUOP26KgP6jEz+JULTO76M/NzY2jnzv7FSvKjgAAOiVwCYkHMeJSfonkj69/1t/W9LnJP0BSV+StC0pLukfOo7zH1q67F+WlD7h7UuWrhMI3BQB/UcnfhKhaI/N0Z+RSEQTExNWYjX5vYIDAIBeCmxCQtJfkPSj++//gjHmC8aYrxlj/oUx5pck/SFJu3pSMfE3LV2zaoxZO+EtNNsS3BQB/WerE3+QqwZIhEJq75iOzdGfuVzOekNLv1dwAADQS4FMSDiO87sk/en9D9+S9CuHH2OM+b8l/Z39D6cdx3mlT8sbetwUAYNhsxN/EJEIDbdujunMz89bWYOtOAf5vYIDAIBeCmRCQtJPSPro/vsLxpi9Yx739w68b2d7BKG/KQo6GgAGV5g78ZMIDbduj+m4rqtkMtnVGlKplKanp7uK8Tx+r+AAAKCXgpqQ+OyB90+6q/0tSY/33//x3i0nXMJ8UxRUNAAMvrB34icRGl42juk4jqPFxUVFo9GO1hCNRrWwsCDHcTp6/mn8XMEBAEAvBTUh8ZkD77993IOMMbuSvr7/4ZSF637RcZyvO47zPcdxPnAcZ91xnP/FcZxMN0Edx5k86U3SD1lYuxVhvykKIhoADoewd+InERpONo/pZDIZFQqFtpMS0WhUhUJBmUxX3+pP5OcKDgAAeimoCYnJ/V/rxphHpzz24f6v447jjHZ53YykT0oa1ZMJHp+R9Ccl/av9xESn8R+e8navy3VbE/aboqChAeDwCHMnfhKh4dSLYzqzs7MqlUot3/ynUimVSiXNzs52tIZW+b2CAwCAXglqQqLZQrqVn9AP/iTTaSvrR5L+rqTrejJWNCPpJyX9rQNr+JOS/rcO4wdGmG+KgoYGgMMlzJ34SYSGU6+O6WQyGT148ODpMbbDDSUjkcjTY2yrq6s9rYw4vC6/VnAAANArkUEvoEM/uP/r91t47M6B9z/SwbXelTRhjHl86PfvS/oNx3FuSPpNSf+BpP/CcZx/aIz5P9u8xounfP6H5JMqiTDfFAWJrZ3F1dVVdtx8otmJ38aNedA68ZMIDSebx3Rc133m9xzHkeu6cl1XnuepWq2qVqspFotpYmJiYI0hmxUcc3NzLSVjUqmUFhYWApOM8DxPGxsb2t7e1tjYmCYnJ2nCCQAh19MKCcdxjIW3P/ac0N/b//UHWljGwWMU3233z2CM+f5zkhEHP/9vJf3sgd/6YgfX2DjpTdL/227MXmE8WTDQAHD4hLkTP4nQ8OnnMZ14PK6pqSldvnxZU1NTA39t+LWCo1M0VQYAnCSoRzaa21ut/JR6sPbR7jbbPmPMP5NU2f/ws47jBPXv9VRhvikKEhoADqewduIfxkQo43dPFvZjOs0KjldffVVbW1uqVCp64403VKlU9P777+vWrVtyXdf3FWw0VQYAnKbXRzZsTLb4f57zexuSrkiKOo5z7pTGls3jEJvGmJ0THtetip40ufxBSb9b0mYPrzVQ8/PzyufzVuLAvl7sLJI48odmJ/5uql+C2Im/mQi18f/OIBOhxhgtLy/rxo0beu211565ORsZGVEul9P8/HwgbjR7jWM6H4rH44H8P7hYLLbVx6jZVLlQKPS8iSgAwD96upNvjHnbwtsHzwldOfD+xeOu7zhORE+mYkjSWzb/bM8RmlpDxpP5W9h3FodZmDvxB706hJ3i9nBMJ9hoqgwAaFVQjxa8fuD9k+5qX9aHRzb+ee+WI+lJdYT0pInm+z2+1kCF+aYoCNhZHG5h7cQf5EQo43fbN4zHdMKiF+NaAQDDK6gJiWVJzcqJ687xd7Z/7MD7dmrYn8NxnB+X1PxJ+XVjzF6vruUXYb0pCgJ2FodfsxN/qzfoqVRKpVIp0GXQ3SZCP/KRj+jGjRt9T4SyU/xEuz0z6FcUXDRVBgC0I5AJCWPM9yX9T/sfTkn684cf4zjOj0n6uf0PV4wxzx2beWCax+8c8/k/ckLCQ47jfErS/3Hgt0LTBTCMN0VBwM5iOAxbJ/5WdJoIlaTvfve7mpmZ6WtH/7DvFHc7XSHox3TCiqbKAIB2OEH9QcdxnJik35L0o/u/9WuSfl1PRnv+hKS/pCdTOL4r6Q8YY/71MXGafwHfMsZ84pjPf13SHUl39aSh5o6k3yPpD+lJ0qO5JX3LGPNHu/2zPWcNk5IeStLDhw81OTlp+xJdMcZoZWVFN27cUKFQeOZsdCQSedqobXp6mmMafXLt2jUrDQCvXbumW7duWVgRes3zPFWrVdVqNcViMU1MTAztrnC5XNbc3FxXu7DJZFKLi4s9TdQsLS1pZmbGShzXdbtfUB+18zU67mthjFE6ne66ievq6irfe/rE8zwlEgkrfYxGRka0tbU1tP+PBcFa9QP91Jdf11e/+FmlJj466OUAGICNjQ29+GJzRoReNMZs2L5GYBMS0tPqhN+Q9OljHuJJ+i+NMV89IUYrCYlW/M+S/mwvJnn4PSFxUJhuivwszDdCCIfjosNuAAAgAElEQVSTEqGtah4h61XVVlgTg+1OV5CO/1qUy2Vls9mOqkyi0ahKpdJQVAcFRaVS6brXy+F4U1M2BrahEyQkAPQjIRHIIxtNxpivS7ok6S/qSbXEI0mPJf0bSb8q6fedlIxo0X8m6a9L+pqkb+hJ74pdSVv71/xVSWljzHyPx4oGQjwe19TUlC5fvqypqSmSEQMS5AaAQCscx5Hrunr11Vf1ta99TR/5yEfajtHLPg29GL8bBLZ7ZtCvKFhoqgwAaFegExKSZIypG2N+2RjzijHmvDEmaoy5aIz5eWPMt1p4vrP/9oljPv+PjTF/yRjzOWPMp4wx54wxv8sY87v3r/nzxpg1638woAuDnoTSbgM7oFPGGM3Pz+u73/1uR8/vVZ+GMI7f7VXPDPoVBQdNlQEA7Qp8QgLA8/V7Z7HbBnZAJ/za0T+MO8W9/FqEsYlrENFUGQDQLhISwBDr185iuVxWOp3WzMyMbt++fWRnuNFoKJ/Pa2ZmRul0emhGGWLw/NrRP4w7xb3+Whw8prO1taVKpaI33nhDlUpF77//vm7duiXXdWlgOUCMawUAtIuEBDDker2zWCwWlc1mW94ZXV9fVzabVbFYbOs6wGF+7tMQtp3ifn8t6FfkX4xrBQC0IzLoBQDovebOouu6ViehdNvAjg746EYv+jTYurFt7hTbmLIRhJ1iP38t0F/NpsrdjmulqTIAhAMVEkDI2NpZ7FUDO6BVfu/TEKadYr9/LdA/g26qDAAIFhISADri12aCCA+/92kI0/hdv38t0F+MawUAtIqEBICO+LWZIMLD730awrRT7PevBfqPca0AgFaQkADQNj83E0R4BKGjf1h2ioPwtUD/Ma4VAHAaEhIA2taLBnZAJ4LQpyEsO8VB+Fqg/xjXCgA4CQkJAG2jgR38Iih9GsKwUxyUrwUGh3GtAIDDGPsJoG00sINfNPs0ZLPZjia+9LNPQ6/G7/pFkL4WAADAH6iQANA2GtjBT4LYp2FYd4qD+LUAAACDQ0ICQNtsN7CTpEqlort376pSqdDkEm0LS5+GIOBrAQAAWkVCAkBHbDWe+/a3v61EIqFkMqkrV64omUwqkUg8PU9vjLFyHQy/MPRpCAq+FgAAoBUOP+z7n+M4k5IeStLDhw81OTk54BUBkjFG6XRa6+vrPb1OMpnU4uIiNyxo2zD2aQgqvhZA8KxVP9BPffl1ffWLn1Vq4qODXg6AAdjY2NCLL77Y/PBFY8yG7WvQ1BJAR7ptYNeq9fV1ZbNZFQoFSrrRlng8zk2vT/C1AAAAz8ORDQAd67SBXbvq9bpyuZzK5XJPrwMAAMLF291VpV7XXc9TpV6Xt7s76CUBoUKFBICuNBvYzc3N9fT4Rr1e1/Xr17W6uspYQAAA0DFjjJYfPdKNalWvvfeeGgc+NyIpNz6u+QsX5J47x88cQI9RIQGga600sJuenu76Omtra1pZWek6DgAACKdyrab0vXuaefNN3T6UjJCkhqT85qZm3nxT6Xv3VK7VBrFMIDRISACwwnEcua6rV199VVtbW6pUKnrjjTdUqVT0/vvva3x83Mp1bt68aSUOAAAIl+LWlrL372v98eOWHr/++LGy9++ruLXV45UB4cWRDQDWHW5g53meCoWCldh37tyR53k0yAOADnmep42NDW1vb2tsbEyTk5P8n4qhV67VlFtbU31vr63n1ff2lFtbU+nSJWVisR6tDggvKiQA9NzGxoYajcNFkZ1pNBqqVqtWYgFAWBhjnh6rSyQSSiaTunLlipLJpBKJhK5du6alpSUxDh7DyBijubfeajsZ0VTf29P1t9/m9QH0AAkJAD23vb1tNV6N85wA0LJyuax0Oq2ZmRndvn37SIK40Wgon89rZmZG6XSaiUYYOsuPHrV8TOM4a/W6Vh49srQiAE0kJAD03NjYmNV4MUomAaAlxWJR2Wy25SlI6+vrymazKhaLPV4Z0D83333XV3EAfIiEBICem5ycPDJ5o1ORSEQTExNWYgHAMCuXy8rlcqrX6209r16vK5fLUSmBoeDt7qqwuWkl1p3NTXm7u1ZiAXiChASAnovH48rlclZi5XI5mq8BwCmMMZqbm2s7GdFUr9d1/fp1zswj8DZ2do6M9uxUQ1J1Z8dStMHxdndVqdd11/NUqddJsmCgmLIBoC/m5+eVz+etxAEAnGx5ebnlYxrHWVtb08rKilzXtbMoYAC2LTXVbqpZjtcvxhgtP3qkG9WqXnvvvWeSNCOScuPjmr9wQe65c3IcZ1DLRAhRIQGgL1zXVTKZ7CpGKpXS9PS0pRUBwPC6efOmr+IAgzJm6choU8xyvH4o12pK37unmTff1O1DyQjpSeVHfnNTM2++qfS9eyrTPBx9REICQF84jqPFxUVFo9GOnh+NRrWwsEDWHgBO4XmeCoWClVh37tyR53lWYmG4BKXsf3J0VLZSCBHH0cToqKVo/VHc2lL2/v2Wp4ysP36s7P37Km5t9XhlwBMkJAD0TSaTUaFQaDspEY1GVSgUlMlkerQyABgeGxsbR0Z7dqrRaKharVqJheAzxmjpO9/R1bU1JV5/Xcl793SlXFby3j0lXn9d19bXtfSd7/iq90g8ElFufNxKrNwLLygeCc6J93Ktptzamup7e209r763p9zaGpUS6AsSEgD6anZ2VqVSqeXjG6lUSqVSSbOzsz1eGQAMh+3tbavxatyUQMEu+5+/cMFXcfrBGKO5t95qOxnRVN/b0/W33/ZVcgnDiYQEgL7LZDJ68OCBlpaWdPXq1SMjQSORiK5du6alpSWtrq5SGQEAbRgbG7MaLxaLWY2H4Al62b977pySZ892FSMVjWr63DlLK+q95UePWv56HWetXtfKo0eWVgQ8X3BqjgAMFcdx5LquXNeV53mqVquq1WqKxWKamJgI/WhPz/O0sbGh7e1tjY2NaXJyMvR/JwBaMzk5qZGRESvHNiKRiCYmJiysCkH1Vr2uP/6Nr3dc9l+6dEmZASe1HMfR4tSUsvfvd1QxED1zRgsXLwaqj9XNd9+1Fsc9f95KLOB5qJAAMHDxeFxTU1O6fPmypqamQnvjbYx5WjWSSCSUTCZ15coVJZNJJRKJp1UjlE8COEk8Hlcul7MSK5fLhfb/ZDzxV775zaEo+8/EYiqkUoqeae/2J3rmjAqp1MCTKu3wdndV2Ny0EuvO5qZvG5ZiOJCQAAAfKJfLSqfTmpmZ0e3bt4/sbDYaDeXzec3MzCidTqtcLg9opQCCYH5+3ldxEFzvfO97XT3fT2X/s4mESpcutXx8IxWNqnTpkmYTiR6vzK6NnZ0jPT461ZBU3dmxFA04ioQEAAxYsVhUNpvV+vp6S49fX19XNptVsVjs8coABJXrui03Dz5OKpXS9PS0pRUhzGwdH7AhE4vpwSuvaOmll3R1fPzISNCI4+ja+LiWXnpJqy+/HKjKiKZtS1N2mmqW4wEH0UMCAAaoXC4rl8upXq+39bx6va5cLqdSqUTTTwBHOI6jxcVFZbPZtv9/kZ6MW15YWAjUmXnYZfOmtln275eRmY7jyD1/Xu758/J2d1Xd2VGt0VBsZEQTo6O+WWenxkYOp1m6E7McDziICgkAGBBjjObm5jq6WZCeJCWuX7/ui7O5APwnk8moUCgoGo229bxoNKpCoUCyM+T+/fe/by2Wn8v+45GIpqJRXY7HNRWNBj4ZIUmTo6NHKj86FXEcTYyOWooGHEVCAgAGZHl5ueVjGsdZW1vTysqKpRUBGDazs7MqlUotH99IpVIqlUqanZ3t8crgd487bGR5HMr++yceiSg3Pm4lVu6FF4YiSQP/IiEBAANy8+ZNX8UBMJwymYwePHjwdIrPyKHy60gk8nSKz+rqKpURkCSdbXMaxWko+++v+QsXfBUHOA7pLgAYAM/zVCgUrMS6c+eOPM9jNB+AYzmOI9d15bquPM9TtVpVrVZTLBbTxMQE/3/giI/9wA9Yi0XZf/+5584pefas1h8/7jhGKhrV9LlzFlcFHEVCAgAGYGNj48hoz041Gg1Vq1VuKAC0JB6P8/8FTmWzMSJl//3nOI4Wp6aUvX9f9Q6O30TPnNHCxYs0tkXPcWQDAAZge3vbarxarWY1HgAAtlD2PxiZWEyFVErRNo/fRM+cUSGVCuTIUwQPCQkAGICxsTGr8WL80AAA6IEf+cEf7Or5lP0P1mwiodKlS0qePdvS4///9u4+OLKrPvP481s1NqalBmQ7ENQGA4ZitqVl6dgzGBypLRBUMBA3kSprSGnMS0GVUt6EWsBbS1heNkmFeKnNLplZcMBYKgoCq5k2gZCAYFvqkJiModloRoIlBkwsgXkz0FLjl3Vz9o97e6atUUvq7ntvv30/VSrdO310zhEcXd9++txzRuNxFZ73PE0ND4fcM8BDIAEAbZBMJs9bWK5ZsVhMIyMjgdQFAECt9zz96Q1/wl7FtP/OkB4a0umrrlL+uc/V9KWXnrclaMxMM5deqvxzn6vVK69kZgQixcNcANAGiURC2WxWi4uLLdeVzWZ5HhwAEIpD8bhyo6PKnjnT0FoETPvvLGamzBOfqMwTn6jSI49o86GHtFWpaGhgQCMXXsgaH2gbZkgAQJvMzc11VD0AAOyGaf+9JRGL6VA8rsOJhA7F44QRaCsCCQBok0wmo1Qq1VIdo6OjmpiYCKhHAADsjmn/AMJAHAYAbWJmWlhY0Pj4uMrlcsM/H4/HNT8/z7O5AIBIMO0fQNCYIQEAbZROp5XL5RSPxxv6uXg8rlwup3Q6HVLPAACoj2n/AIJAIAEAbTY1NaVCoXDgxzdGR0dVKBQ0NTUVcs8AAACA8BBIAEAHSKfTOn36tPL5vKanp8/bEjQWi2lmZkb5fF6rq6vMjAAAAEDXY24VAHQIM1Mmk1Emk1GpVNLm5qa2trY0NDSkkZERtvYEAABATyGQAIAOlEgkCCAAAADQ03hkAwAAAAAARI5AAgAAAAAARI5AAgAAAAAARI5AAgAAAAAARI5AAgAAAAAARI5dNgAAuyqVStrY2ND29rYGBweVTCbZ+QMAAACBYYYEAOAs55zy+bymp6c1PDysVCqlI0eOKJVKaXh4WDMzM8rn83LOtburAAAA6HLMkAAASJKKxaJmZ2e1tra26+uVSkWLi4taXFxUKpXSwsKC0ul0xL1EP2GWDgAAvY0ZEgAALS0taXx8vG4YsdPa2prGx8e1tLQUcs/Qb5ilAwBA/yCQAIA+VywWlc1mVS6XG/q5crmsbDarYrEYUs/Qb4rFosbGxjQ5OakTJ06oUqk86vXqLJ3JyUmNjY0x9tDxSqWS1tfXderUKa2vr6tUKrW7SwDQUQgkAKCPOec0OzvbcBhRVS6XdfToUT6tRsuYpYNewSwfADg4AgkA6GPLy8sHfgNYz5kzZ7SyshJQj9CPmKWDXsEsHwBoDIEEAPSx48ePd1Q96D/M0kGvYJYPADSOQAIA+lSpVFIulwukrpMnT/JsNJrCLB30Amb5AEBzCCQAoE9tbGycN524WZVKRZubm4HUhf7CLB10O2b5AEDzCCQAoE9tb28HWt/W1lag9aH3MUsHvYBZPgDQPAIJAOhTg4ODgdY3NDQUaH3ofczSQS9glg8ANI9AAgD6VDKZ1MDAQCB1xWIxjYyMBFIX+gezdNDtmOUDAK0hkACAPpVIJJTNZgOpK5vNKpFIBFIX+gezdNDtmOUDAK0hkACAPjY3N9dR9aC/MEsH3Y5ZPgDQGgIJAOhjmUxGqVSqpTpGR0c1MTERUI/QT5ilg27HLB8AaA2BBAD0MTPTwsKC4vF4Uz8fj8c1Pz8vMwu4Z+gXzNJBN2OWDwC0hkACAPpcOp1WLpdrOJSIx+PK5XJKp9Mh9Qz9gFk66GbM8gGA1hBIAAA0NTWlQqFw4DeGo6OjKhQKmpqaOu+1Uqmk9fV1nTp1Suvr66wajz0xSwfdjlk+ANA8AgkAgCRvpsTp06eVz+c1PT193jTkWCymmZkZ5fN5ra6uPmpmhHPu7M8NDw8rlUrpyJEjSqVSGh4ePvtzzrmofy10AWbpoJsxywcAmmfcHHY+M0tKuleS7r33XiWTyTb3CEA/KJVK2tzc1NbWloaGhjQyMrLrdOJisajZ2Vmtra3tW2cqldLCwgJvILGrRsbS6Oio5ufnGUvoCMViUePj4yqXyw3/bDweV6FQ6LixfGbz53r5+7+kz9x0jUZHHt/u7gBog42NDV122WXV08uccxtBt8EMCQDArhKJhA4dOqTDhw/r0KFDu4YRS0tLGh8fP9AbSElaW1vT+Pi4lpaWgu4uekArs3SAdmKWDwA0J9buDgAAulOxWFQ2m234E8FyuaxsNtuRnwii/cxMmUxGmUzmwLN0gE5QXYuHWT4AcHAEEgCAhjnnNDs729T0ZMkLJY4eParV1VUWI0RdiUSCAAJdpTrLZ2VlRceOHVMul1OlUjn7eiwWUzab1dzcnCYmJkK5/pVKJW1sbGh7e1uDg4NKJpP8HQHoWAQSAICGLS8vH/gxjXrOnDmjlZUVZTKZYDoFAB2gHbN8nHNaXl7WsWPHdMcddzwqBBkYGDgbgmQyGUJgAB2FQAIA0LDjx48HVg+BBIBeFcUsn/0Wg61UKlpcXNTi4iILCwPoOCxqCQBoSKlUUi6XC6SukydPqlQqBVIXAPQbFhYG0O0IJAAADdnY2HjUdOBWVCoVbW5uBlIXAPSTVhcWLhaLIfUMAA6OQAIA0JDt7e1A69va2gq0PgDodUEtLOycC7hnANCYrg0kzGzQzMbN7C1m9kkz+46ZOf/rnpDafIGZfdTMvmtmD5rZfWb2OTO7IYz2AKATDQ4OBlrf0NBQoPUBQK8LcmFhAGinbl7U8tOSMlE1ZmbvkvQOPTrEeZKkl0h6iZm9RtK0c+7BqPoEAO2QTCY1MDAQyGMbsVhMIyMjAfQKAPoHCwsD6BVdO0NCUu2eRfdL+rykYOcRVxsye5Okd8r73+tbkl4v6bCk6yXl/WLXSbotjPYBoJMkEglls9lA6spms6GvQA8AvYSFhQH0km4OJD4m6dWSnuWcu9g591JJPwm6ETMblvRe//RfJD3fOXebc+4u59ynJE3Jm60hSTeYWSboPgBAp5mbm+uoegCgX7CwMIBe0rWPbDjnbo2oqTdIerx/fLNz7sc7+lExszlJL5M0IOmtkpYj6hsAtEUmk1EqlWrpGebR0VFNTEwE2Kv2K5VK2tjY0Pb2tgYHB5VMJpkBAiBQLCwMoJd08wyJqFzvfy9JOrlbAefchqQv+KcvMjNWaAPQ08xMCwsLisfjTf18PB7X/Py8zGz/wh3OOad8Pq/p6WkNDw8rlUrpyJEjSqVSGh4e1szMjPL5PKvZAyEplUpaX1/XqVOntL6+3vOPILCwMIBeQiCxBzO7QN5aEZJ0p3Pu4T2KV5cpvlDSlaF2DAA6QDqdVi6XaziUiMfjyuVySqfTIfUsOsViUWNjY5qcnNSJEyfOm0ZdqVS0uLioyclJjY2NqVgstqmnQG/p5yCwurBwEFhYGEC7EUjs7dnyHsOQpG/sU7b29UPhdAcAOsvU1JQKhYJSqdSByo+OjqpQKGhqairknoVvaWlJ4+PjB35sZW1tTePj41paWgq5Z0Bv6/cgkIWFAfQSAom9JWuON/Ype2/N8WWNNGJmyb2+JD25kfoAIErpdFqnT58++2nlzk/uYrHY2U8rV1dXe2ZmRDabVblcbujnyuWystlsz71BAqJCEOhhYWEAvaJrF7WMSO1DdfutIFR7V9row3337l8EADqXmSmTySiTyahUKmlzc1NbW1saGhrSyMhIT30C55zT7Oxsw2FEVblc1tGjR7W6utoTa2gAUWk1CCwUCj0RiEosLAygdzBDYm+PrTnea/0ISXqo5viiEPoCAF0hkUjo0KFDOnz4sA4dOtRTYYQkLS8vt/QmQJLOnDmjlZWV/QsCkBRcENgra0qwsDCAXhFqIGFmLoCvG8Ps4z4erDm+YJ+yF9YcP9BgO5ft83VVg/UBAEJy/PjxjqoH6AcEgedjYWEAvYAZEnur3Zh5v8cwav9r0NAG0c65jb2+JN3XSH0AgHCUSiXlcrlA6jp58mTPb08IBIUgcHf9vLAwgN4Q9hoSQew28f0A6mhW7UKWybqlPLULWbImBAD0oI2NjfNW9G9WpVLR5uZmzz3SAgQtjCCwl/7uqgsLr6ys6NixY8rlco+6TsViMWWzWc3NzWliYoLHNAB0lFADCefcfltldrpvSqrI2/rzOfuUrX3966H1CADQNtvbDU2A29fW1tb+hYA+RxC4v35aWBhAb2GXjT045x42s1OSrpZ0tZld4Jyrt7hldZnihyR9JZIOAgAiNTjY6CZKexsaGtq/ENDnCAIbk0gkCCAAdA3WkNjfHf73hKRX7VbAzJKSXuyfftE519v/pQOAPpVMJjUwMBBIXbFYTCMjI4HUBfQygkAA6F19HUiY2eU1u3ks1yn2IUk/94//xMwu3lHHgKTj8h7rkKRbQuksAKDtEomEstlsIHVls1k+xQQOgCAQAHpX1wYSZnaFmd1Y+6VzO2EM7nzNzJ7cTDvOufsl3eyfPk3SP5rZa83sSjN7paQlSa/wX/+4c265+d8KANDp5ubmOqoeoNcRBAJA7+rmNSSukfSROq9dvMtr16rJ7TOdcx80s6dIeoekZ0q6bZdin5X0umbqBwB0j0wmo1QqpbW1tabrGB0d1cTExP4FAUjyArzFxcVA6gEAdI6unSERNefcO+WFIB+Tt63nw5J+KG+GxKudc9c55x5sYxcBABEwMy0sLCgejzf18/F4XPPz82y9BzSgGgS2giAQADpP1wYSzrnbnXPWwNfyLnXcU/N65gBt/oNz7jXOuac65y50zj3JOfcS59zHw/gdAQCdKZ1OK5fLNRxKxONx5XI5pdPpkHoG9CaCQADoTV0bSAAA0E5TU1MqFAoH/tR2dHRUhUJBU1NTIfcM6E0EgQDQewgkAABoUjqd1unTp5XP5zU9PX3eTgCxWEwzMzPK5/NaXV3lDRHQIoJAAOgt3byoJQAAbWdmymQyymQyKpVK2tzc1NbWloaGhjQyMsKK/kDAqkHgysqKjh07plwup0qlcvb1WCymbDarubk5TUxM8JgGAHQwAgkAAAKSSCQIIIAIEAQCQG8gkAAAAEDXIggEgO7FGhIAAAAAACByzJAAAKALlEolbWxsaHt7W4ODg0omk3wqDAAAuhozJAAA6FDOubM7eAwPDyuVSunIkSNKpVIaHh4+u4OHc67dXQUAAGgYgQQAAB2oWCxqbGxMk5OTOnHixKN2EZCkSqWixcVFTU5OamxsTMVisU09BQAAaA6BBAAAHWZpaUnj4+NaW1s7UPm1tTWNj49raWkp5J4BAAAEh0ACAIAOUiwWlc1mVS6XG/q5crmsbDbLTAkAANA1CCQAAOgQzjnNzs42HEZUlctlHT16lDUlAABAVyCQAACgQywvLx/4MY16zpw5o5WVlYB6BAAAEB4CCQAAOsTx48c7qp4olUolra+v69SpU1pfX1epVGp3lwAAQMgIJAAA6AClUkm5XC6Quk6ePNkVb+jZ1hQAgP5GIAEAQAfY2Ng4b2vPZlUqFW1ubgZSV1jY1hQAABBIAADQAba3twOtb2trK9D6gsS2pgAAQCKQAACgIwwODgZa39DQUKD1BYVtTQEAQBWBBAAAHSCZTGpgYCCQumKxmEZGRgKpK0hsawoAAGoRSAAA0AESiYSy2WwgdWWzWSUSiUDqChLbmgIAgFoEEgAAdIi5ubmOqido/bytKQAAOB+BBAAAHSKTySiVSrVUx+joqCYmJgLqUXD6cVtTAACwNwIJAAA6hJlpYWFB8Xi8qZ+Px+Oan5+XmQXcs9b127amAABgfwQSAAB0kHQ6rVwu13AoEY/HlcvllE6nQ+pZa/ppW1MAAHAwBBIAAHSYqakpFQqFAz++MTo6qkKhoKmpqZB71rx+2dYUAAAcHIEEAAAdKJ1O6/Tp08rn85qenj5vS9BYLKaZmRnl83mtrq527MyIqn7Y1hQAADQm1u4OAACA3ZmZMpmMMpmMSqWSNjc3tbW1paGhIY2MjHTk1p71VLc1XVxcbLmuTt3WFAAANIZAAgCALpBIJLr+Tfjc3FwggUSnbmsKAAAawyMbAAAgEr28rSkAAGgcgQQAAIhEL29rCgAAGkcgAQAAItOr25oCAIDGEUgAAIBI9eK2pgAAoHEEEgAAIHK9tq0pAABoHLtsAACAtuilbU0BAEDjCCQAAEDb9cK2pgAAoDE8sgEAAAAAACJHIAEAAAAAACJHIAEAAAAAACJHIAEAAAAAACJHIAEAAAAAACJHIAEAAAAAACJHIAEAAAAAACJHIAEAAAAAACJHIAEAAAAAACJHIAEAAAAAACJHIAEAAAAAACJHIAEAAAAAACJHIAEAAAAAACJHIAEAAAAAACJHIAEAAAAAACJHIAEAAAAAACIXa3cHAAAA+lmpVNLGxoa2t7c1ODioZDKpRCLR7m4BABA6ZkgAAABEzDmnfD6v6elpDQ8PK5VK6ciRI0qlUhoeHtbMzIzy+bycc+3uKgAAoSGQAAAAiFCxWNTY2JgmJyd14sQJVSqVR71eqVS0uLioyclJjY2NqVgstqmnAACEi0ACAAAgIktLSxofH9fa2tqByq+trWl8fFxLS0sh9wwAgOgRSAAAAESgWCwqm82qXC439HPlclnZbJaZEgCAnkMgAQAAEDLnnGZnZxsOI6rK5bKOHj3KmhIAgJ5CIAEAABCy5eXlAz+mUc+ZM2e0srISUI8AAGg/AgkAAICQHT9+vKPqAQCgExBIAAAAhKhUKimXywVS18mTJ1UqlQKpCwCAdiOQAAAACNHGxsZ5W3s2q1KpaHNzM5C6AABoNwIJAACAEG1vbwda39bWVqD1AQDQLgQSAAAAIRocHK9PPjsAABAhSURBVAy0vqGhoUDrA3bzzEsH9ZmbrtEzLw12/AJArVi7OwAAANDLksmkBgYGAnlsIxaLaWRkJIBeAXu76IIBjY48vt3dANDjmCEBAAAQokQioWw2G0hd2WxWiUQikLoAAGg3AgkAAICQzc3NdVQ9AAB0AgIJAACAkGUyGaVSqZbqGB0d1cTEREA9AgCg/QgkAAAAQmZmWlhYUDweb+rn4/G45ufnZWYB9wwAgPYhkAAAAIhAOp1WLpdrOJSIx+PK5XJKp9Mh9QwAgPYgkAAAAIjI1NSUCoXCgR/fGB0dVaFQ0NTUVMg9AwAgegQSAAAAEUqn0zp9+rTy+bymp6c1MDDwqNdjsZhmZmaUz+e1urrKzAgAQM+KtbsDAAAA/cbMlMlklMlkVCqVtLm5qa2tLQ0NDWlkZIStPQEAfYFAAgAAoI0SiQQBBACgL/HIBgAAAAAAiByBBAAAAAAAiByBBAAAAAAAiByBBAAAAAAAiByBBAAAAAAAiByBBAAAAAAAiByBBAAAAAAAiByBBAAAAAAAiByBBAAAAAAAiByBBAAAAAAAiByBBAAAAAAAiByBBAAAAAAAiByBBAAAAAAAiByBBAAAAAAAiByBBAAAAAAAiFys3R3AgQxUD77//e+3sx8AAAAAgD6w473nQL1yrTDnXBj1IkBmdqWku9rdDwAAAABAX7rKOfeVoCvlkQ0AAAAAABA5Zkh0ATO7UNKYf/ojSZU2dqcbPVnnZphcJem+NvYFCAvjHP2AcY5+wDhHP2Ccd4cBSZf6x6edcw8F3QBrSHQB///4wKfH9Aszqz29zzm30a6+AGFhnKMfMM7RDxjn6AeM867y3TAr55ENAAAAAAAQOQIJAAAAAAAQOQIJAAAAAAAQOQIJAAAAAAAQOQIJAAAAAAAQOQIJAAAAAAAQOQIJAAAAAAAQOXPOtbsPAAAAAACgzzBDAgAAAAAARI5AAgAAAAAARI5AAgAAAAAARI5AAgAAAAAARI5AAgAAAAAARI5AAgAAAAAARI5AAgAAAAAARI5AAgAAAAAARI5AAgAAAAAARI5AAgAAAAAARI5AAj3DzAbNbNzM3mJmnzSz75iZ87/uCanNF5jZR83su2b2oJndZ2afM7MbwmgPqDKzx5nZ28zsLjO738zKZvYNM3ufmT0tgPovr/n72e/r9gB+JfQJM3uaP06/4Y/b+/1x/FYze1yA7fyGmeXMbMPMHvK/58zsN4JqA6gnzHFuZjc2cH2+MaBfCTjLzH7FzF5uZu8xs78xsx+HfU9gZjeY2ef9e+0H/Xvvj5rZ1WG0h+iYc67dfQACYWZ5SZk6L3/XOXd5wO29S9I7VD/Y+2tJ0865B4NsFzCzKyR9VtKz6hQpSXqNc+4zLbRxuaTvHLD4vHPuxmbbQv8ws1dI+qikRJ0i35R0nXPu7hba+FeSbpX0+j2KfUjSm5xzv2y2HaCesMe5HzJ85IDFX+ucu72ZdoB6zGyvN5CB3hOY2UWSFiW9rE6RX0p6j3Pu3UG1iWgxQwK9xGqO75f0eUnboTRk9iZJ75T3N/QteTe+hyVdLynvF7tO0m1htI/+ZWZD8sKuahjxF5JeJOkFkt4ub8wnJH3CzP5tQM3+gaSxPb7eHlA76GFm9jxJn5A3PrfljZsXyBu/f+EXe7akv/bHebP+SOfCiK9JukHe9fkG/1yS3iDpD1toA9hVhOO86qXa+/p8RwBtAHv5F3n33GG5TefCiLy8e+3D8q7z35J3L/4uM3tjiH1AiJghgZ7hX4i2JN1V/dTBf1TjaQpwhoSZDUv6tqTHy7sI/5pz7sc1rw9Iykl6hf9P1zrnloNoGzCz98ibmSNJb3PO3bLj9RdIWpEUk7TinMs02c7lOjdDgk/Y0DIzK0j6dUmPSBp3zt254/W3SvpT//Tdzrl3NdHGsyWtyRv/X/HbeaDm9cfJ+/u40u/HoVZmYwA7RTTOb9S5GRJPd87d02x/gWaY2bsl3SXvnvsHO+4ZApshYWaTkr7on35aUtY5V6l5/RJJX5X0VEk/k/QM59xPg2gb0WGGBHqGc+5W59zHI7i5fIO8MEKSbq4NI/x+VCTNSapeMN8acn/QJ8zsMZL+vX/6dUnv21nGOfcPkj7sn06Y2VURdQ+oy8wOy3uTJkkf3vkmzfc+eeNakn7PH++N+n15YYQk3VQbRkiSc+4Xkm7yT2OS3txEG8CuIhznQFs5597pnPuMc+4HITf1Fv/7I5LmasMIvx8/lnSzf/oEeffo6DIEEkDjrve/lySd3K2Ac25D0hf80xcFNC0TuFbnwrD5PZ5/v73mOBtqj4CDub7meNdn3/3xvOCfPkHeeD8wMzNJv+mffsM59+U67XxZ0v/1T3/T/zkgCKGPc6Bf+PfOL/JPv+DfW+/mpLx7col7nq5EIAE0wMwukPfcmiTd6Zx7eI/iK/73C+VNDwZadU3N8UrdUt5U9V/4xy8MrzvAgVXHblne9Np6asd1o2P36ZKesks9e7UzIunyBtsB6olinAP94ipJF/jHda/p/r14NYC+illH3YdAAmjMsyUN+Mff2Kds7euHwukO+sy/rjmuO/6cc49Iqj66FMTYu8nM7va32fq5ma2Z2QfMLB1A3egP1XF4tz8+62nlunmgv48A2gHqiWKc7/QRM/uemT3sb734ZTP7QzMbabFeoN2auabHVH8HMnQoAgmgMcma43pTx6rurTm+LIS+oP9Ux1/ZOfezfcpWx9+lZnZhi+2mJT1T3myfhLybhDdJ+qofTLRaP3qYmT1W0iX+6Z7XTX8xsrJ/2uh1k+sz2ibCcb5TRtKvSnqMpIslHZG3s8fd/o5gQLfimt4nYvsXAVCjdi2I/bYULdccD4bQF/Sf6vg7yHa2O8ffQ0209zN5O8YsS/pnSQ/Ku/F9ibzttgblBRNDkl7TRP3oD41cNyVv7MbV+HWT6zPaKapxXvVtec/O36lzb8aeIem3JE1LeqykD5iZc87d2mQbQDtxTe8TBBJAYx5bc7zX+hHSo98AXhRCX9B/quNvv7EntT7+vidpxN+VoNbXJH3WzI7JW7j1qZJebWafcM79VRPtoPc1ct2Uzo3dRsct12e0U1TjXPKC4nnnnNvx73dJ+oSZvVxeWPEYSf/NzP7KOXdfE+0A7cQ1vU/wyAYiZWYugK8b2/grPFhzfEHdUp7aaewP1C2FnhPiOK+Ov/3GntTi+HPOPbxLGFH7+j9L+p2af7qpXln0vUaum9K5sdvouOX6jHaKapzLOffzXcKI2tc/I+k9/unj5M1oA7oN1/Q+QSABNGar5ni/KWHxmuODTN8E9lMdfweZjhj6+HPO/Z2kdf/0GjPjvynYTSPXTenc2G103HJ9RjtFNc4P6lZJ1dBiIqQ2gDBxTe8TPLKBqAWxmvn3A6ijWbWL6iTrlvLULqpzb91S6EVhjfMNeQuWxc3sCfssbFkdfz9yzjWzfsRBrctb5PKx8hZU+1GIbaELOeceNLOfyBsfe143zeyJOndj2eh1k+sz2ibCcX7Q/vzQ788l8ra3BbrNzmv6V/YoyzW9ixFIIFLOuf227el035RUkbf153P2KVv7+tdD6xE6TojjfF3egmWSN76+vFshM4vJ2xVDCn/s1Z02DNRYl/Trkq4ws9geWyK2ct1crznm+ox2iGKcN4LrM7pZM9f0R+Qtwo0uwvRaoAHOuYclnfJPrzazvZ5pq06RfEh7p7rAQX2p5nivKbhX6tynb38fXnckndsn/CFJPwm5LXSv6tiNS/q1PcrVjutGx+535C3GurOe3Yz73zcl3dNgO0A9UYzzAzGzS3VuG9Lv7VUW6FB36dxilnWv6f69+POrP+Oc+39hdwzBIpAAGneH/z0h6VW7FTCzpKQX+6dfdM5t7VYOaNCypJ/7x0fNzOqUu7HmOBdWZ8zshZJS/umXnHO/DKstdL07ao5fu1sBfw2SWf/0Z5LyjTTgL/L3Kf/0OWb2/N3K+f9e/TTtU3stDgg0KPRx3oA3Sqr+N2IlpDaA0Pj3zl/0T1/s31vv5lXy7smlEO95EB4CCaCGmV1es8vBcp1iH9K5N4V/YmYX76hjQNJxeY91SNItoXQWfcefofM//NNDkt6ys4yZXa1zK6qvOOfu2q2umnF+T53Xr98j8JCZXSHpYzX/dHz/3wD9yjl3StLf+aev98fpTv9B59Zf+e87P+Uys0zNuL29TlN/Ju+xOkl6v5k9avs3//z9/ukjfnkgEFGMc/8+5Xl79cPf9vM/+6cPSPpIA78GEAkzu7FmrL+rTrH/6n+PSTrm32PX1nGJpPf6pz+Td4+OLsMaEugZ/huka3b8c3VV3sFdtlH822b25XbO3W9mN0v6gKSnSfpHM/sjSaclPUXS70u61i/+cefccqNtAHu4RdJvS3q2pD/1x/1fyrvpvFbSf5J3bX9A3lhsVk7S3WZ2Ut5jShvyHsv4VUkvlRd6VP++PumcO9lCW+gPvydvevpFkj5vZn8s79PhiyT9O3mf6EreWj3va6YB59w3zewWSf9R3qNLf29m75X0LXnrqtwsqfpm7hZ/+1ogSGGP88sl5c3sTkmflvRPkn7ov/YMSdP+VzVQfotzbrOJdoC6zOwaSVfU/NMlNcdX7Lznds7d3kw7zrn/bWZ/Ke9v55WSlszsz+Q9hjQm6e2SnuoXv9k599Nm2kF7EUigl1yj+p8CXLzLa9dKajiQkCTn3AfN7CmS3iHvJve2XYp9VtLrmqkfqMc5t2Vm18kbX8+Sd3P7xh3FSpJe45z7Py02d4Wkt+1T5n9KenOL7aAPOOe+Zma/Lemj8qbX/vEuxb4p6boWH3N7u6RfkXf9fZ68wG6nD0v6gxbaAHYV4Ti/2v+q5xeS3uycu7WFNoB63iDpaJ3XXuh/1bq9hbZeJ+9v6WXy7t2v3fH6LyX9F8Z69yKQAJrknHunmX1O0u/KW1X7SfKmi/2TpI845z7ezv6hdznn7van7P6upBl5wcEF8ra6+qy8acDfbbGZV8q72T0ibybQJfIWaitJ+ra8acm3OefOtNgO+ohz7tNm9m/kfYp8nbyt3B6WdLek/yXpz51zv2ixjV/Kmy5/Ql5Yd5W88ftjeYukfdA59zettAHsJeRx/lVJvyPv+nylvFlrl8i7p/+ppDV5z91/yDn3w3qVAN3COfeApOvM7NXy1sh6rqQnSPqBvHuRP3fO3dm+HqJVxlpOAAAAAAAgaixqCQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIkcgAQAAAAAAIvf/AdUJGqPLbKyYAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize the split and the subsets\n",
    "subset1, subset2, split = splitter(X, y)\n",
    "X1 = subset1[0]\n",
    "X2 = subset2[0]\n",
    "\n",
    "plt.figure(dpi=200)\n",
    "plt.scatter(X1[:, 0], X1[:, 1], color='k', label='subset 1')\n",
    "plt.scatter(X2[:, 0], X2[:, 1], color='c', label='subset 2')\n",
    "plt.legend()\n",
    "\n",
    "split_feature = split[0]\n",
    "split_value = split[1]\n",
    "boundary_limits = [min(X[:, 1]), max(X[:, 1])]\n",
    "\n",
    "if split_feature == 0:\n",
    "    plt.plot([split_value, split_value], boundary_limits, lw=.5)\n",
    "if split_feature == 1:\n",
    "    plt.plot(boundary_limits, [split_value, split_value], lw=.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "code_folding": [
     1
    ]
   },
   "outputs": [],
   "source": [
    "# Challenge 3\n",
    "def fit_tree(X, y):\n",
    "    '''\n",
    "    Repeat the splitter to fit a tree to X and y.\n",
    "    return the tree i.e. the trained model\n",
    "\n",
    "    Arguments\n",
    "        m = num_of_datapoints\n",
    "        n = num_of_features\n",
    "        X: np.array() shape (m, n)\n",
    "        y: np.array() shape (m, 1)\n",
    "\n",
    "    Returns\n",
    "        tree: a dictionary of nodes\n",
    "            key: node name e.g. 'root', 'rootRL'\n",
    "            value: node dictionary\n",
    "                key:\n",
    "                    'depth' int\n",
    "                    'data' (subset_X, subset_y)\n",
    "                    'class' majority class\n",
    "                    'leaf' binary, leaf(1) or not(0)\n",
    "                    'split_feature' 0 or 1\n",
    "                    'split_value' float\n",
    "    '''\n",
    "\n",
    "    tree = {}\n",
    "\n",
    "    depth = 0\n",
    "    max_depth = 10\n",
    "\n",
    "    # Init the root node\n",
    "    node_name = 'root'\n",
    "    tree[node_name] = {}\n",
    "    tree[node_name]['depth'] = depth\n",
    "    tree[node_name]['data'] = (X, y)\n",
    "    tree[node_name]['class'] = np.bincount(y).argmax()\n",
    "    tree[node_name]['leaf'] = 0\n",
    "\n",
    "    while depth < max_depth:\n",
    "\n",
    "        all_nodes = list(tree.keys())\n",
    "        nodes_at_depth_i = []\n",
    "        for node_name in all_nodes:\n",
    "            if tree[node_name]['depth'] == depth:\n",
    "                nodes_at_depth_i.append(node_name)\n",
    "\n",
    "        for node_name in nodes_at_depth_i:\n",
    "\n",
    "            if tree[node_name]['leaf'] == 1:continue\n",
    "            X_parent, y_parent = tree[node_name]['data']\n",
    "            Xy_child1, Xy_child2, split = splitter(X_parent, y_parent)\n",
    "\n",
    "            tree[node_name]['split_feature'] = split[0]\n",
    "            tree[node_name]['split_value'] = split[1]\n",
    "\n",
    "            # Init child 1\n",
    "            child1_name = node_name + 'L'\n",
    "            tree[child1_name] = {}\n",
    "            tree[child1_name]['depth'] = depth + 1\n",
    "            tree[child1_name]['data'] = Xy_child1\n",
    "            tree[child1_name]['class'] = np.bincount(Xy_child1[1]).argmax()\n",
    "            # Termination condition\n",
    "            if gini_calculator(Xy_child1[1]) != 0 and len(Xy_child1[1]) > 3:\n",
    "                tree[child1_name]['leaf'] = 0\n",
    "            else:\n",
    "                tree[child1_name]['leaf'] = 1\n",
    "\n",
    "            # Init child 2\n",
    "            child2_name = node_name + 'R'\n",
    "            tree[child2_name] = {}\n",
    "            tree[child2_name]['depth'] = depth + 1\n",
    "            tree[child2_name]['data'] = Xy_child2\n",
    "            tree[child2_name]['class'] = np.bincount(Xy_child2[1]).argmax()\n",
    "            # Termination condition\n",
    "            if gini_calculator(Xy_child2[1]) != 0 and len(Xy_child2[1]) > 3:\n",
    "                tree[child2_name]['leaf'] = 0\n",
    "            else:\n",
    "                tree[child2_name]['leaf'] = 1\n",
    "                \n",
    "        depth += 1\n",
    "        \n",
    "    return tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "tree = fit_tree(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "code_folding": [
     1
    ]
   },
   "outputs": [],
   "source": [
    "# Challenge 4\n",
    "def predict_tree(X, tree):\n",
    "    '''\n",
    "    Given X and the model (i.e. tree), return predictions\n",
    "\n",
    "    Arguments\n",
    "        X: np.array() shape (m, n)\n",
    "        tree: a dictionary of nodes\n",
    "\n",
    "    Returns\n",
    "        y_pred: a list of predictions for each row of X,\n",
    "                class labels 0 or 1.\n",
    "\n",
    "\n",
    "    '''\n",
    "    y_pred = []\n",
    "    for i in range(X.shape[0]):\n",
    "        node_name = 'root'\n",
    "        data_point = X[i]\n",
    "        x0 = data_point[0]\n",
    "        x1 = data_point[1]\n",
    "\n",
    "        while True:\n",
    "            split_feature = tree[node_name]['split_feature']\n",
    "            split_value = tree[node_name]['split_value']\n",
    "\n",
    "            if split_feature == 0:\n",
    "                feature_value = x0\n",
    "            else:\n",
    "                feature_value = x1\n",
    "\n",
    "            if feature_value <= split_value:\n",
    "                node_name = node_name + 'L'\n",
    "                if tree[node_name]['leaf'] == 1:\n",
    "                    y_pred.append(tree[node_name]['class'])\n",
    "                    break\n",
    "            else:\n",
    "                node_name = node_name + 'R'\n",
    "                if tree[node_name]['leaf'] == 1:\n",
    "                    y_pred.append(tree[node_name]['class'])\n",
    "                    break\n",
    "\n",
    "    return y_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = predict_tree(X, tree)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def accuracy(y_pred, y):\n",
    "    return sum(y_pred == y)/len(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy(y_pred, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training acc: 1.0\n",
      "Testing acc: 0.94\n"
     ]
    }
   ],
   "source": [
    "# Putting all together\n",
    "# with Train/Test\n",
    "X_train, y_train = make_circles(n_samples=100, noise=0.1, factor=0.5, random_state=0)\n",
    "X_test, y_test = make_circles(n_samples=100, noise=0.1, factor=0.5, random_state=42)\n",
    "tree = fit_tree(X_train, y_train)\n",
    "y_pred_train = predict_tree(X_train, tree)\n",
    "y_pred_test = predict_tree(X_test, tree)\n",
    "\n",
    "print('Training acc:', accuracy(y_pred_train, y_train))\n",
    "print('Testing acc:', accuracy(y_pred_test, y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "code_folding": [
     1,
     10
    ]
   },
   "outputs": [],
   "source": [
    "# Challenge 5\n",
    "def fit_forest(X, y):\n",
    "    '''\n",
    "    Fit 30 trees\n",
    "    by randomly sampling from\n",
    "    X and y\n",
    "    return 30 trees\n",
    "    '''\n",
    "    num_trees = 30\n",
    "    forest = []\n",
    "    for i in range(num_trees):\n",
    "        idx = np.random.choice(np.arange(len(y)), int(len(y)*0.8))\n",
    "        X_sample = X[idx]\n",
    "        y_sample = y[idx]\n",
    "        tree = fit_tree(X_sample, y_sample)\n",
    "        forest.append(tree)\n",
    "    return forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "code_folding": [
     1
    ]
   },
   "outputs": [],
   "source": [
    "# Challenge 6\n",
    "def predict_forest(X, forest):\n",
    "    '''\n",
    "    Predict the labels for X\n",
    "    for all 30 trees\n",
    "    calculate the average of 30 trees\n",
    "    return avg. predictions\n",
    "    '''\n",
    "    y_preds = []\n",
    "    for tree in forest:\n",
    "        y_pred = predict_tree(X, tree)\n",
    "        y_preds.append(y_pred)\n",
    "    y_preds = np.array(y_preds)\n",
    "\n",
    "    y_pred = np.mean(y_preds, axis=0)\n",
    "    y_pred[y_pred > 0.5] = 1\n",
    "    y_pred[y_pred <= 0.5] = 0\n",
    "    return y_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training acc: 1.0\n",
      "Testing acc: 0.96\n"
     ]
    }
   ],
   "source": [
    "# Putting all together\n",
    "# with Train/Test\n",
    "X_train, y_train = make_circles(n_samples=100, noise=0.1, factor=0.5, random_state=0)\n",
    "X_test, y_test = make_circles(n_samples=100, noise=0.1, factor=0.5, random_state=42)\n",
    "forest = fit_forest(X_train, y_train)\n",
    "y_pred_train = predict_forest(X_train, forest)\n",
    "y_pred_test = predict_forest(X_test, forest)\n",
    "\n",
    "print('Training acc:', accuracy(y_pred_train, y_train))\n",
    "print('Testing acc:', accuracy(y_pred_test, y_test))"
   ]
  },
  {
   "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": "3.7.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}