adaboost_test.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total running time of this example is : 0.1925187110900879\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAELCAYAAAA2mZrgAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzsnXd0VdUSh7+Tm55AElrovadSQieAUgXpCghSrKio4AMFRYqIoqjwUBQRkSpgeTQFQYTQglIDIkVaINQECIH0Nu+Pndz0AqRB9rfWXrmnz7lJzpzZe89vDBFBo9FoNBoAi8I2QKPRaDRFB+0UNBqNRmNGOwWNRqPRmNFOQaPRaDRmtFPQaDQajRntFDQajUZjJl+dgmEYXQ3DOGkYxmnDMMZnsn24YRghhmEEJLXn8tMejUaj0WSPZX6d2DAMEzAX6ARcBPYZhrFORI6l23WViIzKLzs0Go1Gk3vyM1JoBpwWkbMiEgusBHrl4/U0Go1Gc5/kW6QAVAKCUi1fBJpnsl8/wzB8gX+BMSISlH4HwzBeAF4AcHBwaFK/fv18MFej0WgeXg4cOHBdRMrmtF9+OgUjk3XpNTXWAytEJMYwjJHAYuCRDAeJzAfmAzRt2lT279+f17ZqNBrNQ41hGOdzs19+dh9dBKqkWq4MXE69g4jcEJGYpMVvgCb5aI9Go9FociA/ncI+oI5hGDUMw7AGBgLrUu9gGEaFVIs9geP5aI9Go9FociDfuo9EJN4wjFHAJsAELBSRfwzDeA/YLyLrgNcMw+gJxAM3geH5ZY9Go9FocsZ40KSz9ZiC5n6Jj4/nypUrxMTE5LyzRvOAYWNjQ4UKFbC0TPvObxjGARFpmtPx+TnQrNEUSa5cuYKjoyOVK1fGMDKbD6HRPJiICKGhoVy5coUqVarkfEAmaJkLTbEjJiYGZ2dn7RA0Dx2GYeDi4nJfUbB2CppiiXYImoeV+/3b1k5Bo9FoNGa0U9BoCgGTyYS3tzdubm54eXnx2WefkZiYeE/nmjRpElu2bMly+7x581iyZMm9mmomMTGR1157DXd3dzw8PPDx8eHcuXMAfPDBB/d9/rtl+PDh1KhRA29vb7y9vWnVqlWB2/AwogeaNZpCwM7OjoCAAACCg4N56qmnCAsLY+rUqXd9rvfeey/b7SNHjrwnG9OzatUqLl++zJEjR7CwsODixYs4ODgAyim8/fbbeXKdu2HmzJn0798/y+3x8fFpZuGkX86KhIQETCZTntj4oKEjBY2mkClXrhzz58/niy++QERISEhg3Lhx+Pj44Onpyddff23e9+OPP8bDwwMvLy/Gj1dq9MOHD+enn34CYPz48TRs2BBPT0/Gjh0LwJQpU/jkk08ACAgIoEWLFnh6etKnTx9CQ0MBaN++PW+99RbNmjWjbt267Ny5M4OdV65coUKFClhYqMdG5cqVcXFxYfz48URFReHt7c3gwYMJDAzE3d3dfNwnn3zClClTzNcZM2YMvr6+NGjQgH379tG3b1/q1KnDxIkTAQgMDKR+/foMGzYMT09P+vfvT2RkZK6/zylTpvDCCy/QuXNnhg4dyqJFi3jiiSd4/PHH6dy5MyLCuHHjzBHPqlWrAPDz86NDhw489dRTeHh45Pp6Dxs6UtAUa0aPHm1+Y88rvL29mT179l0dU7NmTRITEwkODmbt2rU4OTmxb98+YmJiaN26NZ07d+bEiROsWbOGv/76C3t7e27evJnmHDdv3mT16tWcOHECwzC4detWhusMHTqUzz//nHbt2jFp0iSmTp1qtjU+Pp69e/eyYcMGpk6dmqFL6sknn6RNmzbs3LmTRx99lCFDhtCoUSNmzJjBF198Yf4eAwMDs71Xa2trduzYwX//+1969erFgQMHKFWqFLVq1WLMmDEAnDx5km+//ZbWrVvzzDPP8OWXX5qdXGrGjRvH+++/D4CbmxvLly8H4MCBA+zatQs7OzsWLVrEnj17OHLkCKVKleLnn38mICCAw4cPc/36dXx8fPD19QVg7969HD16lBo1auT0K3to0ZGCRlNESE4k3bx5M0uWLMHb25vmzZtz48YNTp06xZYtWxgxYgT29vYAlCpVKs3xJUuWxNbWlueee47//e9/5v2SCQsL49atW7Rr1w6AYcOGsWPHDvP2vn37AtCkSZNMH+yVK1fm5MmTfPjhh1hYWPDoo4/yxx9/3PV99uzZEwAPDw/c3NyoUKECNjY21KxZk6AgJZJcpUoVWrduDcCQIUPYtWtXpueaOXMmAQEBBAQEmB1C8jXs7OzMy506dTJ/X7t27WLQoEGYTCZcXV1p164d+/btA6BZs2bF2iGAjhQ0xZy7faPPL86ePYvJZKJcuXKICJ9//jldunRJs89vv/2W7XRDS0tL9u7dyx9//MHKlSv54osv2Lp1a65tsLGxAdQgeHx8fJb7dOvWjW7duuHq6sqaNWt49NFHM9iRetA8Ojo60+tYWFiYPycvJ183/X3e7TTL5LGOzJazU3FIf1xxREcKGk0hExISwsiRIxk1ahSGYdClSxe++uor4uLiAPj333+JiIigc+fOLFy40Ny/nr77KDw8nLCwMB577DFmz56doVvMyckJFxcX83jB0qVLzVFDbjh48CCXLyuh48TERI4cOUK1atUAsLKyMtvr6upKcHAwN27cICYmhl9++eWuv5MLFy6wZ88eAFasWEGbNm3u+hxZ4evry6pVq0hISCAkJIQdO3bQrFmzPDv/g06xiRRWP+eLt80Bzu50IOJ64d72JTs71lWsSFQxnd1Q2Lz55psEBwcXqg1RUVHUqVPHPBumS5cudOrUCX9/fxo2bMiuXbuoX78+IoKzszMzZsygZMmSeHl54ebmhqWlJS1btmTkyJEEBwdz8uRJrK2tGT9+PLGxsYgIL7/8Mv7+/gQFBWFnZ4e/vz9jxoxh5MiRxMTEULFiRd5++238/f25ffs2hw8fJjY2llu3bhETE4O/v38am//880/mz59PbGwsAA0bNqRJkyb4+/vTvXt36tSpQ926dZkyZQpPP/00Xl5eVKhQgbJlyxIUFJThOkePHiU0NNR8neRtzs7OVK9enY8++oijR49SuXJlBg0alMGe4OBgXnvtNSZMmGBet2DBgjT3C3Dq1CmuXLliXnZ1daV06dLUqVMHwzB49tlnOXv2bAZ7iiIlS5ZMM4ifHxQbQbw/R7rTrOUxDJMQd8QgarOJ+OMGmdcCyj8MEUrFxRFqZcXSqlVZX7EicRY6YCtI3nzzTZydnQvbDE0WXLlyhXHjxrFs2bLCNqXIkVuncPr0aWrXrp1mnRbES8dv5Y/Sd8xVvhrzFT18vsLJOwScPaHeaKg+CEy2BWfM3r24TJjAa1u38lp8PLz3HgweDDpyKBAy+4fRFB0CAwOxt7fXyWiFRLF5RR01CgaNKM+T70+l1IgLrAr8lvj4RPjrGVhbDY5MgahrBWNMs2bwxx/w++9QpgwMGwZeXrBmDTxgkZtGk9dUr16do0ePFrYZxZZi4xTKlIFPP4VTp6Dfk7Y89e4zlH76CMuv/E68kw8cnQprq8KfIyD0cMEY1bEj7NsHP/0E8fHQpw+0bAl+fgVzfY1Go0lHsXEKyVStCgsXwpEj0KGDwZCxHak8+BeWh50gocZzcP4H2OgNfzwCF9eD3JseTa4xDOjXD44ehQUL4NIl6NABunSBAwfy99oajUaTjmLnFJJxc1O9Nf7+UK8eDHm5HnWfmstPCUEken0Ed07Bjp6wvh6c/ALiwvPXIEtLePZZFcp8+qlyCE2bwpNPwsmT+XttjUajSaLYOoVkkntrNmyAEiXgicGlaDLkTTbZnEVarQSb0nDgVVhTBQ69CREX8tcgW1t44w04cwbefVcZ5uYGL7wAFy/m77U1Gk2xp9g7BVA9ON26wcGDsGwZhIVB18es6DB8AH86/Qmd/KFCZzjxGayrCbsGQMie/DXKyUnNSjp7Fl55BRYvhtq1Ydw4uHEjf6+tyXceNOnsv//+2yxRXapUKbNkdceOHe/rvJkxZMiQNJLYbdu2zfNraLJBRB6o1qRJE8lvYmJEPv9cpFw5ERDp3Vvk2DERCT8vcnCcyA9OIssR+a25SOBKkYTYfLdJzp0TGTZMxMJCpGRJkWnTRO7cyf/rPoScOnWqsE0QBwcH8+dr167Jo48+KpMmTSpEi3LPsGHD5Mcff8x0W1xc3H2ff/DgwbJ69eps90l/ndxeNy/sexDI7G8c2C+5eMbqSCETrK3VFNYzZ9TL+h9/gLs7PPNqVYLKfAy9L0LTLyD2JuweqKKHYx9DbGj+GVW9OixapEbIH3lEdS3VqgWffw73UY9VU/g8KNLZWbFlyxY6duzIwIEDadSoEQCLFy+mWbNmeHt78/LLL5ujoI0bN9KyZUsaN27MgAEDiIiIyPV1Jk6cyIsvvkinTp0YMWIECxYsYODAgfTo0YNu3bqRmJjIG2+8YZbETv5OMrNPkw258RxFqRVEpJCe4GCRMWNErK1FbGxE/vMfkevXRSQxQeTiepEtj6jIYaW9yN6XRcJO5L9Re/aItG+vQpnq1UWWLBGJj8//6z4EpH+LateuXYY2d+5cERGJiIjIdPt3330nIiIhISEZtuWG1JFCMs7OznL16lX5+uuvZdq0aSIiEh0dLU2aNJGzZ8/Khg0bpGXLlhIRESEiIjdu3BCRlDf3GzduSN26dSUxMVFEREJDQ0VEZPLkyTJz5kwREfHw8BA/Pz8REXn33Xfl9ddfN38Hb7zxhoiI/Prrr/Loo49maXv6SOH3338XBwcHOX/+vIiI/P3339KrVy/zW/nzzz8vy5cvl2vXromvr6/Z/vfff1+mT5+e4fyDBw+W6tWri5eXl3h5ecnTTz8tIiLvvPOO+Pj4SFRUlIiIfPPNN1K1alW5efOmiIisXLlSunTpIvHx8XLlyhWpXLmyXLt2LYN9xQEdKeQzZcvCZ5/Bv//CwIHqc82aMP0DCyKce8Cjf0C3AKg2AM4sgF/qg18PuLol/5LRWrSArVth0yYoVQqGDlUJcOvW6QS4BxQp4tLZ2dGyZUuqVq0KqDfzffv20bRpU7y9vdm+fTtnzpzB39+fY8eO0apVK7y9vVm+fHmW15k1a5ZZEjv1eEivXr2wtU1RH+jcuTMuLi6AksR+6qmnMJlMlC9fnjZt2pAsiZPaPk32FBuZi7ygWjXVgzN2LLzzDkycqHpvJk2C55/3wqrFQvD6EE7Pg1NfwtZO4OyRJKXxVN5LaRgGdO6skuB++kkZ1KuXmlL14YdwFwqYxRm/bJIF7e3ts91epkyZbLfnlgdFOjsr0ktTP/PMM0ybNi3NPqtXr6Zr164sXbr0rs6d1XUyu25uj9NkjY4U7gF3d1i7Fnbvhrp11eSgBg1g5UpItHEFj8nQ6zw0XwgY8NezsKYqHJkMUVfz3iALC5XP8M8/MH8+nD8P7dtD165qSpWmSPOgSGfnlo4dO/LDDz9w/fp1AG7cuMGFCxdo1aoV27dv5+zZswBERERw6tSpPLuur68vK1euJCEhgWvXrrF7926aNs1R/02TDu0U7oNWrWD7dvjlF7C3h0GDVL7Zpk0gFrZQa4TqVnrkDyjTAo5OUzpLe4ZD6JG8N8jKCp5/Hk6fhpkzYe9eaNJE9Xnl4T+f5v5Jrmns5uZGx44d6dy5M5MnTwbgueeeo2HDhjRu3Bh3d3defPFF4uPj6dq1Kz179jR3yyQPHidz584devTogaenJ+3atWPWrFkZrrt48WLGjRuHp6cnAQEBTJo0Kc/vzcPDg8mTJ9OxY0c8PT3p3Lkz165dw9XVlW+//ZYBAwbg5eVFq1at+PfffzM9x5gxY8xTUr29vUlISMjxuv3796d+/fp4eXnRsWNHPvvsM8qVK5fXt/fQU2yks/ObhARYsUJNCgoMVC/qM2ZA8+apdrp9Ck7+F84tgvgIKN8R6r8BFbqAkQ/++dYt+OQTmDVLzVB69lnV11WpUt5f6wFCq6RqHnbuRzpbRwp5hMkEQ4bAiRMwZ47qyWnRAvr2VesAKFkHfL6A3kHgPQPCjoHfY/CrO5xeAPFReWuUszO8/76aW/vSS/DddyoB7s03IV3Xg0aj0YCOFPKNO3fUC/rMmRAZCSNGwJQpULlyqp0SYuHCD3DiUwgNAJuyUOdlqPsy2OZD2HvuHEyerNK2S5ZUwnt3Wfs2zzEMNSA+fLiyqQDQkYLmYed+IgXtFPKZkBCYPh2++ko9/159FcaPh9KlU+0kAsHblYzGpfVgYQM1nob6Y8CpYd4b9fffKiuvKAjtRUWpMRBHRzWtdtQoNWqfj2inoHnY0U7hASAwUL2kL12qZI2mTlU9OlZW6Xa8fRJOzIZziyEhCip0VeMO5TsW/lt9frFvH8ydqwZlYmPh0UeVc3j88XypRqedguZhR48pPABUr6407Q4fVjOUXn8dvL1V8bU0lKwHzb6CXhfA830IPQTbOsNGLzi7CBIeQkkLHx+VAHLxInzwgcoS7NNHZQh+9BEkTW3UaDT5j3YKBYyHB2zerGo5REer3LNevVQPShpsy4D7OyrfocV3at2fI9SU1qPvQ/RD+KAsWxYmTFDKsD//rLSdxo9XAzEjRuicC42mANBOoRAwDOUIjh1T01a3boWGDeGtt9QAdRpMNlBzOHQ7DI/8Di6N4Mi7qnTo3pdUd9PDhqWlmra1dauqSDdiBPzwg8q5aNUKvv9edTM9wGQnnb1//35ee+21fLu2n58f/v7+93Ts66+/TqVKlbKV+a5evbo5cS27fTw8PPD29sbDw4O1a9fekz1ZMXv2bHOSX2r69OmDt7c3tWvXxsnJyZwHcbffx9atW/nzzz8z3bZgwQLKli2bJs/iZFEYv8stuRFIutcGdAVOAqeB8dns1x8QoGlO5ywMQbz85vJlkeHDlbZd+fIi330nkpCQzQGhR0X+fE5khY0S4tvWQ+TqVpEkIbSHktBQkdmzRWrXVl+Uq6vIpEkily7d9amKg3R2dhLRqQXy7oaEhASpUqWKNG/eXLZt25blftWqVZOQkJBsz5V6nxMnTkjVqlXv2p7cnj8ztm3bJt27d7/n87/zzjsya9asTLd98803ZqHBrEj/+0lMTJSEbP/pU4jPhfBlkRTEMwzDBMwFugENgUGGYWSYSmMYRgngNeCv/LKlqFOhgkoh2LtXjT2MGKGS3vZkVcfH2Q2afwO9L4D7ZLjxl6op/VsTOLdMTXV92HB2VgMxJ0/Cxo1qYGbaNCVINWAA7Nr1wAoBppfO9vPzo0ePHgBs377d/LbZqFEj7iSFkplJaLdv3563336bdu3a8d///peQkBD69euHj48PPj4+7N69m8DAQObNm8esWbPw9vZm586dme6XGdu2bcPd3Z2XXnqJFStWmNffuHGDzp0706hRI1588cU0GkS9e/emSZMmuLm5MX/+/EzPe/v2bbOoHcBnn32Gu7s77u7uzJ49O9v1ERERdO/eHS8vL9zd3Vm1ahVz5szh8uXLdOjQgQ4dOuT697Bv3z7atWtHkyZN6NatG9euXQOUOF/Dhg3x8vJiyJAhnDlzhgULFjBz5sy7ijLSS3ifPn0ad3d3Ro4cSePGjbly5QrLli3Dw8MDd3d33n77bQDi4+NxdnZm4sSJNGvWjL179+b6nu6J3HiOe2lAS2BTquUJwIRM9psN9AD8KKaRQmoSEkSWLhWpWFG9EA8eLBIUlMNBcZEip74RWd9ARQ7/qyhy9EOR6BsFYnOhcfq0yBtviDg7qy/Ly0vkm29EkqSZsyLNW9Trr4u0a5e3LYe3RJHspbNTv8X26NFDdu3aJSIid+7ckbi4uCwltNu1aycvvfSS+XyDBg2SnTt3iojI+fPnpX79+iKSMVLIar/0PPvss7JkyRIJCwuTihUrSmysKi716quvytSpU0VE5JdffhHA/JaebFtkZKS4ubnJ9evXRUS9ybu7u4ubm5vY2dnJ+vXrRURk//794u7uLuHh4XLnzh1p2LChHDx4MMv1P/30kzz33HNmG2/dumU+/91ECtHR0dKyZUvzMcuWLZPnn39eRETKly8vMTExIpIiR55TpFCmTBmz9LeXl5fExMRkkPA+deqUGIYhe/fuFRGRoKAgs92xsbHi6+sr69evl7i4OAHk559/zvJ+0lMkIwWgEhCUavli0jozhmE0AqqIyC/ZncgwjBcMw9hvGMb+kJCQvLe0CGFhoTKjT55USqw//QT16qnE5KisEp4t7aD2c9D9KLTfoHIbDk9QdaX3vwp30o9iPyTUqgWffqpmLc2fD4mJSvupcmVVtjRJeO1BQTKJdFq3bs0bb7zBnDlzuHXrFpaWltlKaA8YMMD8ecuWLYwaNQpvb2969uzJ7du3zZFGanKzX2xsLBs2bKB3796ULFmS5s2bs3nzZgB27NjBkCFDAOjevXuat/45c+bg5eVFixYtCAoKSiOAt23bNo4ePcrff//NqFGjCA8PZ9euXfTp0wcHBwccHR3p27cvO3fuzHK9h4cHW7Zs4a233mLnzp04OTndy1fP8ePH+eeff+jYsSPe3t7MmDGDoCD1+HJzc2PIkCEsX74cqwxzyDNn8ODBZunvgIAArK2tgYwS3rVq1cLHxweAv/76i0ceeYQyZcpgZWXFU089ZZY2t7a2pk+fPvd0b3dLfkpnZzap3vxXbxiGBTALGJ7TiURkPjAfVJ5CHtlXpHF0VI7g2WfV8+3dd+Hbb5WUUd++WaQsGBZQsZtqoUdUMtzpr+HfuVC5l8p3KNvm4ct3cHBQzuC552DnTvjiC5VO/umn0KOHynno2FF53PSk6p4oTFJLZx8/fty8fvz48XTv3p0NGzbQokULtmzZgohkKaGdWiI6MTGRPXv2YGdnl+21c7Pfb7/9RlhYGB4eHgBERkZib29P9+7dATK1x8/Pjy1btrBnzx7s7e1p37490dHRGfarVasWrq6uHDt2LEv566zW161blwMHDrBhwwYmTJhA586d70nkT0Tw9PTMtOLcpk2b2L59O2vXruX999/n6NGjd33+ZO5V+tvOzi5b2fS8JD8jhYtAlVTLlYHLqZZLAO6An2EYgUALYJ1hGFrrNhU1aqhoYetWKFEC+vdX1TiP5CSy6uIJLRepKa1ub0PwDtjiC5uaQeBKSIwrCPMLFsMAX181UykwUIVaf/0FXbqoLOnPP4fbtwvbygykl85OzZkzZ/Dw8OCtt96iadOmnDhxIkcJ7WQ6d+7MF198YV5OltIuUaJEmkggq/1Ss2LFChYsWEBgYCCBgYGcO3eOzZs3ExkZia+vL8uXLwdUuc3kEp9hYWG4uLhgb2/PiRMnspytExwczLlz56hWrRq+vr6sWbOGyMhIIiIiWL16NW3bts1y/eXLl7G3t2fIkCGMHTuWg0nTltPfY040bNiQS5cumfvrY2Nj+eeff0hISODixYs88sgjzJw5k5CQECIjI+/6/LmhRYsWbNu2jRs3bhAfH8/KlSvzRdo8R3LTx3QvDRWFnAVqANbAYcAtm/390GMK2RIXJ/LVVyKlS4tYWIiMHCmSwySPVAdHiPz7pci6OmrcYXUVkaPTRcIv5KvNhU50tMiyZSLNm6txB0dHObV1q8jt24VqloWFhXh5eUnDhg3F09NTZs6caZ59krq/e9SoUeLm5iaenp4ycOBAiY6OFhGRDz/8UBo0aCBeXl4yYcIEEVFjCvv27TNfIyQkRJ588knx8PCQBg0ayIsvvigiIidPnhQPDw/x8vKSHTt2ZLlfMhEREeLi4iJhYWFp1vfp00dWrlwp169fl06dOkmjRo1k9OjRUrVqVQkJCZHo6Gjp2rWreHh4SP/+/aVdu3bmWUvJYwrJ38G3335rPu+nn34qbm5u4ubmlqbfPrP1v/32m/lemjZtar7/OXPmSL169aR9+/aZfv+ZzT46cOCAtG7dWjw9Pc02RUdHS6tWrcTDw0Pc3Nzk448/FhGR48ePi7u7u3h7e8vu3bvTnCezMYU///xTfv/9d+nVq5d5v1OnTomXl1eaY5csWWIeaxk/fryIqJlKTk5Omd5HVtzPmEK+ylwYhvEYaiDZBCwUkemGYbyXZNy6dPv6AWNFJFsNiwdV5iIvuXlTyWTMnauihylT4OWXM5HMyAxJhEu/qq6lYD/AgPKPQo1hUKUPWD7EFaqS5DROd+9ObVtbKFNGhWKurg9fl5qmWKO1j4opx47B6NFKKqNBA9U93rnzXZzgzhk4t1TpLEUEgqUjVH1COYhybfOnxkMR4PSJE9Q2mVSFuqgosLODqlXVz8LGZFJqiUmlMTWae0E7hWKMiKr8NmaMKpvw+ONqfLVOnbs5SSIE71TO4cKPEB8ODjWgxlCoORQca+ab/YWB+R8mMRGuXVPjD0VJX8kwwMVFRTDly6tZBxrNXaCdgoaYGFXcZ9o0pak0ejRMnHgPJQriIyBotXIQV/8ABMq2hZrDVBRhVTA1D/KTTFVSo6OVkyhsYmIgOFg5q7Awtc7BQTkHV1flLDKbRaXRpEI7BY2Zq1fh7bdVhrSrqxIdHT78Hp8jEUEQuEyps975F0x2ULmPchCuj4JF3staFwQPjHR2VJRyDlevwo0bymlZW0O5cspJlC2rdKI0mnRop6DJwL59ShVizx6lIzdnjtKSuydElJTG2cVwfiXE3QK7SqoQUI1h4FQ/T23Pbx4Yp5CauDhVsenaNdXi4pSnL1NGeX9X16IxJqIpEminoMkUEVW35s034dIleOopVZ4gTUnQuyUhWlWHO7sYrvwGkgClmynnUG0g2JTK+RyFzAPpFFKTmAihoSqCuHYNIiLUeienlHGIkiX1jKpijC6yo8kUw1CO4ORJlRH9v/8pyYxp07KRzMgJk60aW2j/C/S+CI0+VRXi9r8CqyvAzv5w6ZeHMzkuD7kv6WwLCzVDyc1N1dlu315NP7OwUAWKduyALVtU2dXgYEhISHO4ls6+P+nsuXPnmpP1csPp06exs7NLI6V9N8cXOLlJZihKrTgnr90v586JPPGEyuGqVk1k4UKRyMg8OHFiosiNgyL7Xxf5qaxKjvu5nMj+MSI3A/LgAnnLQy2dHR0tcv68xPn7i/z6q8gSXpFAAAAgAElEQVS6dSIbNojs26eUFWNitHS25CydnZ30+N2SWZJaZqSXxM6tDZntV1QF8TRFjOrVlQKEn5+axPLMM1CxopqplEpu5+4xDCjVCJrMhj6XwHet0lg69QVs9IYN3nBiFkQH59GdPFzkqXT21Km0e/pp/uvvT0jjxvT7+mt8xo3DZ/Bgdi9fTuDSpcz7/HNmffIJ3klaP1o6W1G5cmWmTZtG69atWb16NfPmzcPHxwcvLy+eeOIJopLC64kTJ5qv36ZNG8aPH0+zZs2oV6/eXUUcmUlip7fh4MGDNG/eHE9PT/r160dY0oy0Nm3a8M477+Dr65tGoiRPyI3nKEpNRwp5Q2KiiJ+fyMCBIlZWKnrw9RX5/nv1spknRF8XOfmFyEYfFT18b1IFgc7/JBKfVxe5e1K/RRWScnbBS2cHBkr9unVFjh+XycOGycwRI1QUsXWrDHrsMdn5668iiYnFVjpbRKRSpUry6aefmpeTbRUReeutt+TLL78UkbSy2a1bt5Y333xTRETWrl0rXbp0yXCtU6dOia2tbRrZi927d2cqiZ3ehgYNGph/hxMmTJD//Oc/5uuOGjUqy/u7n0hBz2crphgGtGunWnAwLFoEX3+txiDKlFGFfl54Ae5rPNamNNR9RbWwY2pwOnAZ7PoFrF2g2iA1QF3aRw+Kkr109uDBg+nbty+VK1e+K+nsY8eOmZdvh4dzp1IlFTJaW4O7O1y9ypbduzl2/Lj6HVhaKuns69cpkSrJJVk6e9ZHH1HC1pbmPj5s/vVXuj/2GDu2b+d/q1ZBbCzdO3VSb/2xsRAby5xZs1idNF4QFBTEqWPHKN28OQDbNm2iTJkynDlzhke7daN9q1bs8vOjT8+eOCRptvTt04edO3ciImbpbMAsnd21a1fGjh3LW2+9RY8ePWjbtu19/Q5Sf39Hjhxh0qRJ3Lp1izt37pijt/T07dsXgCZNmhAYGJjpPvXq1csgNBgfH5+pJHayDTdu3CA6Opo2bdoAMGzYMJ5++mnzfgMHDry7m8sl2iloKFdOzVAaO1aNT379NXz2GcycCZ06wciRKlM6l1LymePUEBp9BF4fwNUtKjnu7EI49SXYllPdTWVaq5+lGoHF/Vws9xQR5eyCl862tla6TzVqkGhpyZ7Nm7G7fVu9IcTFZSj799tffxF28yYeDVXxxMiYGOxv3aK7yQR37mDs3AnJtRLi4mDbNvzOn2fLzz+z5733sLexof3bbxO9cyfcuqVmOmzbBiVLUgtwtbHh2HffIcePQ3g4bNqkznX6NISHIzY2KrEvIgKSnCHknXR2Zt/f0KFD2bhxI+7u7ixYsCBLlVebJEkSk8lEfHz8XV0vM0nsZBsye0nIyta8RI8paMxYWCjtpJ9/hgsX1CylEyegXz8lDfTuu2r9/V3EBBW7QOvvoc9VaL4QKnSF0AA49B/Y3Bx+dFblRY9MgiubIa7oyV3nJUVCOnv1amjcGDp3JsDRUc1sStVWHD7MghkzCNyzh8A9ezi3axeb//6byJo18W3bluXHj4ObGxtDQggND4d69QgrXRqXChWwb9yYE7a2/PnvvypKcXNTbxj16oGbG8Gurpy7fp1qbdvi27MnawICiKxZk4gaNVh94ABtvbzwrVRJSWdv3EjE2rWs/v572lapwuXDh7E3me5bOjszIiIiKF++PHFxcXz//ff3da57oUyZMtjZ2ZnHKZYuXVogUto6UtBkSsWKSiZjwgRVEnnePJg+XWVId+umoodu3ZR+2z1j7QS1RqgGEHkZru+G4F0Qsgv+ma50mQwLcPZSUUTZNlC2NdhXyv7cRZyoqCi8vb2Ji4vD0tKSp59+mjfeeCPDfrNnz2bbtm2YTCYaNmxIt27dsLGxISAggKZNm2Jtbc1jjz3GBx98kOHYOXPm8Morr+Dp6Ul8fDy+vr7MmzePxx9/nP79+7N27Vo+//zzLPdLJjIykk27d/P18uVm3RQHoE27dqw/epTJn33GoEGDaNy/P+3atVOVxapXp6uHB/PWrMGzd2/q1atHi5Yt1R9WzZpgaUmHESMwmUzExcUxY+ZMXJs1wxUYfvo0zZK6UJ575RUajRgBiYkMv3OHZu+8AwkJPNetG42srNi0Zg3jvvsOC5MJK1tbvpo+HW7c4IVnn6Vbt25UqFCBbdu23dPv6L333qNZs2ZUrVoVd3f3TAsE5ZaTJ0/i7e1tXn7++ed58cUXczxu6dKlvPTSS0RFRVG7dm2+++67e7Yht+jkNU2uOX8eFixQ7epVqFJFFTx79ln1v57nxN1RmdTJTuL6HkhImnvuUD2Vk2gDTg1yrer6wCevaRTx8Uof6tatlJacm2AYSlfe2TmllShRbHSjdEazpkCJi4P161X08PvvKlro2VNFD1lVvcwTEuMg9LByEMkt+praZu2ixiTKJTmJUk3BlLn8tHYKDzExMWmdRGio+oMF9Yfq5JTWUdjbP5STHLRT0BQap0/DN9/AwoVKfbpmTTVracQINYCdr4hA+Jm0TuL2SbXNwkbNajJHE62U40A7hWKFiIoeUjuJsLAURVxr67ROwtn5oahloZ2CptCJiVEyGl9/Ddu3q3HEfv1U9ODrW4AvY9EhELI7ZWzi5n6QpBkhTm5Qtg2nHZ6hdj13pfr6EL4lanIgMRHu3ElxErduqeVk7O1THISLi4ou7mvwrODRTkFTpDh+XDmHxYvV/1v9+vDiizB0KJQqaL28+Ei4sS8lkrjuz+lq31C7kh1Y2CoBP+ukZlXioa02p8mB5PGJZCeRPG0WzPkbRYKGDdVUwBzQTkFTJImMhB9/VGMPf/4Jtrbw5JMqemjRopBe0hMTOP3vP9SuXAJib0LMTUhMmlVimMDKBaydVVeTtUuW4xKaYkB0tHIOYWEp4xKFTcWKuXqzuh+nUETcn+ZhxN4ehg1T7fBhFT0sWwZLloCnp4oehgy5h+pw94OFCSztwbEGUEP1OSdEKQcRG6pa+Bkg6WXJ5JDkIJx1NFHcsLVVMuTlyxe2JQWK/uvWFAheXvDll3D5Msyfr7poX3lFvfj066ccxtmzhWCYYSgnYV8ZnD2gnC9U6AZlWkHJBsoJxIRA2FEI2QGXf4MQfwg7DlFXISHmni6bLJ3t7u7O448/zq1bt+75FrKTqj506BCGYbApOUM4E6ZMmcInn3yS7TWmTJlCpUqV8Pb2pn79+rz00kvZymffLVnJeX/33XdmQUBra2uz3HayCGBuuXnzZprci9TEx8ebfx/JbebMmfd0Hw8DOlLQFCiOjiq34bnnYP9+lfOwYYMapAY1e6lTJ9UeeUSN8xU4Fial22RTWi2bo4nQVNHEWSDpoWiyT+lusnZRdaxziCbs7OzMGcbDhg1j7ty5vPPOO3l+KytWrKBNmzasWLGCLl263Ne5xowZw9ixY0lMTMTX15ft27fflQppdvj5+eHo6EirdOUBR4wYwYgRKrmxevXqbNu2jTJlytz1+ZOdwsiRIzPdXqJEiQzaROmJj4/HMtXYQvrl3B5X1NGRgqZQMAzw8VERwoULanB6zhylgLB8OfTvr4T5WrRQ8ho7diiNtUIz1tJeZVE7u0O5tkqao0xrKNkQrJwg5kZSNLETLm9UM6DCjkHUFVWtLhtatmzJpUuXzMszZ87Ex8cHT09PJk+ebF6fGxnq1IgIP/30E4sWLWLz5s1pMnKnT59OvXr16NixIydPnjSv/+abb8xy0f369cu0UE1sbCzR0dFmueuAgABatGiBp6cnffr0ITQ0NNv1c+bMoWHDhnh6ejJw4EACAwOZN28es2bNwtvbm507d+Z4bwDh4eEMHz6cZs2a0ahRI9avXw/A33//jY+PD97e3nh6enL27FnGjx9vziq+mygjvZR1esnqc+fO0aFDBzw9PenUqRMXL14EYMiQIfznP/+hQ4cOvP3227m+XpEgN1KqRalp6eyHn9hYkZ07RSZNEmnZUsTCQkl7OziIdO8uMnu2yLFjSv77XkgjK7z/dZHf2+VN29xWZFNLkd1DRK7tFLn4i8jFdapd+V3kxn6RO2dEYm6apbPj4+Olf//+snHjRhER2bRpkzz//POSmJgoCQkJ0r17d9m+fbuIZC9DnZlM9M6dO+WRRx4RESWjnSzRnCxDHRERIWFhYVKrVi1z0Z3UctHvvPOOzJkzR0REJk+eLBUrVhQvLy9xdnaWQYMGmffz8PAQPz8/ERF599135fUk7fCs1leoUEGik/TZQ0NDzefPqfBP+vscN26crFixQkREbt68KXXq1JGoqCgZOXKkrFy5UkREoqOjJSoqKttCN3FxcWJhYZFG2vrHH38UkYxS1uklq7t27SrLli0TEZGvv/5a+vXrJyIigwcPll69eklCQkK295Rf6CI7mocKKyto0wamTgV/f7hxQ3UvDR2qSouOHq1m5lWpopLkvv9eiXsWOoYFWFirbqdybaBi1yQJDjfVrRQbCmH/QMgupX3kXofSpVy4ef0KnR5R8sibN29m8+bNNGrUiMaNG3PixAlOJamPzpkzBy8vL1q0aKFkqJNVSbNgxYoVZnnlgQMHmgvj7Ny5kz59+mBvb0/JkiXp2bOn+ZijR4/Stm1bPDw8WL58Of/8849525gxYwgICCA4OJiIiAhWrlxJWFgYt27dMgu1DRs2jB07dmS5HsDT05PBgwezbNmy++pW2bx5M9OnT8fb25sOHToQHR3NhQsXaNWqFe+//z4ff/wxQUFB2Nra5niu5O6j5Na/f3/zttRy2pBWsvqvv/4yLw8dOjRNlPPEE09g8QDKajw4HV2aYouzM/TpoxpAYKCS1/j9d1i3TtWCADWY3bmzGo9o0wYyU4zOQJN81M42TCnjDMkkREHsLezsbAjYtoiw65foMXwqc2e8wWsvPIFEXmHC6OG8+PwzqsvK0gEsbPHz82PLli3s2bMHe3t72rdvn61AW0JCAj///DPr1q1j+vTpiAg3btwwK4dmJb09fPhw1qxZg5eXF4sWLcLPzy/DPlZWVnTt2pUdO3bQrVu3u/5afv31V3bs2MG6deuYNm1aGsdzN4gIa9asoVatWmnW161bl5YtW/Lrr7/SqVMnFi9eTMX7EOdKL1GdW8nq/JK2zm8ePDemKfZUr64Gq3/4QUUIe/cqBVcXF1UfoXNn9blTJ/j4Yzh0KEXVoNAx2YFdBcACyrbGqX5/5nz+FZ988wtxRkm6+HqxcPEKwi/9Cdf3cOnQKoKPfE/Y+V24OBjYx57jxIE/lLZ/QpRSkc2ELVu24OXlRVBQEIGBgZw/f55+/fqxZs0afH19Wb16NVFRUdy5c8fcFw9w584dKlSoQFxcXJbF5UUEf39/atWqhZOTEy4uLuY35GR556zWJyYmEhQURIcOHfj444+5desW4eHh9yR13aVLF+bMmWNePnToEKBqU9SuXZvXX3+d7t27c+TIkTyR0s6MFi1a8MMPPwCwbNkyfH198/waBY2OFDQPNCaTGrD28YG331Y1WHbsSIkk3npLtbJllVhfp06QVPiraGBY0KiZL17eTVi5+V+efvpNjl+1pmXfqYDgaG/Lsvkf0LVTdeYtXo9nyx7Uq1WRFo3qwM1DcDlBDWTf2A+WlVRkYbJnxfIl9OndK82l+vXrx1dffcXGjRsZMGAA3t7eVKtWLU21smnTptG8eXOqVauGh4dHmgfprFmzWLZsGXFxcXh6evLyyy8DsHjxYkaOHElkZCQ1a9Y0yztntj4hIYEhQ4YQFhaGiDBmzBicnZ0zyHnnpoLa5MmTGT16NB4eHiQmJlK7dm3Wrl3L999/z4oVK7CysqJixYq8//77ODs707RpUzw8POjevTszZsxIc647d+6kkbbu3r0706dPz9GGL774gmeffZYPP/wQV1fXApG2zm90RrPmoebKFVVNbvNm9fPqVVi16jRly9ambFk1w6lMmaKjYpAjIpAYA/ERSS0SElJ9lnSZtyZblYBnmdzsk5btC6y6nabg0RnNGk0WVKgATz+tmggcPaq0z0wmNRX23Dk149TFBVxdlbJriRJFWCfPMJIe9LYpeRTJiCh58YQkB5HsOBIilMR4YrpEOwvrpMjCIWX8InnZwqoIfwma/KRYOYX27dtnWPfkk0/y8ssvExkZyWOPPZZh+/Dhwxk+fDjXr19PMyMhmZdeeokBAwYQFBSUpqh2Mv/5z394/PHHOXnyZKaVliZOnEjHjh0JCAhg9OjRGbZ/8MEHtGrVCn9//0znO8+ePRtvb2+2bNnC+++/n2H7119/Tb169Vi/fj2ffvpphu1Lly6lSpUqrFq1iq+++irD9p9++okyZcqwaNEiFiWP6KZiw4YN2Nvb8+WXX5r7VlOTPFD5ySef8Msvv6TZZmdnx8aNGwHVbfHHH3+k2V66dGl+/vlnACZMmMCedHWDK1euzLJlywAYPXp0huSjunXrmufzv/DCC/z7778AvPnmmzg7B1OlihMVKrgREgLnz0dy86Y9x4+DhUU01tahlCoVh7d3VUwm2L9/P7HpEiXKlClD3bp1ATULJSEhIc12V1dX8yBoZtm6FStWpHr16iQkJPDXX39l2F6lShWqVKlCbGwsmUXH1apVo1KlSkRFRZn701NTs2ZNypevS3h4OEeOHMFkOGBrlYCtZSK2lgmUK+2IHYkkRodgJMak8QHxCQZYOmBp60xUnAUXroQSFWdBdLyJRFE7urm54eTkREhISKYzoTw9PXF0dOTq1auczSRdvVGjRtjZ2XHp0iXOnz+fYXtyZbmgoCCCgoIybG/evDkmk4nAwEAuX76cYXtyItyZM2e4du1amm0mk4nmSf2I//77b4aMcGtra5o2VS/Vx48fN+dYJGNra0vjxo0BNWPr9u20JWMdHBzw8vIC4PDhw0RERKTZXrJkSdzd3QE4ePBghkkDLi4uNGjQAEj7t5c+uS8/KFZOQaNJjWGIufsoKuoEkZGJxMY6ExfnQkxMOS5fNnHtmtoeHV0GC4vrWFgUVgbd/ZMgBhGxlkQk3YK9a03sypblTlgYx48dxSbJWdhaJmJrlUAZJ0uIvYFdQhT1yqpjRCAmwYLoOBO2MachvBTWiQnYmBKJSdDzVh4G9JiCptiRmyI7CQkqPyI4GK5dS6nyWLKk6mZydVVTZYtFD0tifFJXVHjGn8m1KgAwqe4nK0fVBWXlCJaOSdNq9fhFQaLHFDSaPMZkUuML5cop6Y3wcOUcgoNVtblTp1TRruRxiLJlVdLdQ4mFJVg7qZYa86B3OMQlO4twiA2DhCuYlWZBVcJLdhCWqZyFpb1WnS1i5KtTMAyjK/BfwAQsEJEZ6baPBF4BEoBw4AUROZafNmk0d0tyDfgSJaB2baXBFBKinMTVqxAUpPYpXTrFSTg6FrbVBUCaQe90InWSmGqGVHhKdBF9FRJTd8EZqWZGOaZ1GBbWxSQUK1rkm1MwDMMEzAU6AReBfYZhrEv30P9eROYl7d8T+Azoml82aTR5gbU1VKqkWmKiqsNy7Zpq//yjmoNDioMoXRrSqx2YTCY8PDyIj4+nRo0aLF26FGdn53uyp3r16uzfvz9T9dBDhw7RuHFjfvvttyxVUqdMmYKjoyNjx47N8hpTpkzhm2++oWzZskRHR9OhQwfmzp2btYyDYaFkx61KZNyWEJuhK8pv+y6sTfG0alov1Tms+O5HP/674H+ABcdOnqVenZqYLC3p2rkTM2Z8lKtZUkFBQYwdO5ZVq1Zlu19q2rRpQ0hICHZJafH16tW7q+PzDElUTjQhRv20dATL3KTq3zv5GSk0A06LyFkAwzBWAr0As1MQkdRD9g6kiTc1mqKPhYUqhFWqFDRooMYekruZAgNVjQhLS9W9lOwkbGyKuXS2yRpMpVQp1CT8jvyCo4MDrbo/kuIs4sIZMbgPIwZ2g8RoqjcbyraV71KmVFJVpqubAAsw2RIvJiytHVMiF5OtKrdqsqVKpQr39EBftWpVmoS29NyTlLYI8XHRWBoJ6iGfGJP0wI9J+/BPXp8+78TJAxyr3/W93A25cgqGYVQD6ojIFsMw7ABLEckpZ7wSkHoe2UUgQy6pYRivAG8A1sAjWVz/BeAFgKq5qE+q0RQW9vZQo4Zq8fFw/XqKk7hyRe3j4qK648PC1MB1y5YtOXLkiPkcM2fO5IcffiAmJoY+ffowdepUQElnBwUFER0dzeuvv84LL7yQrS2SJJ39+++/07ZtW6Kjo83icNOnT2fJkiVUqVKFsmXL0qRJE0BJZ8+fP5/Y2Fhq167N0qVLsbe3T3PezKSzkzOXa9WqxcKFC3Fxccly/Zw5c5g3bx6WlpY0bNiQGTNmMG/ePEwmE8uWL886o9lkB2XbgosjJEYzcfIHhIQEc/bcBcqXdWbK2CEMf20G4RGRWFhY8OX0kTRvXI/T5y7T/8WPCdi6gAXfb+a3bfu4ExHF2fOX6N/ncT58f2qSI7FRelXZMGTIEFxdXTl48CA+Pj5YW1sTEhzM2bNnKO9ahvmff8TIV8dyMOBvrCxNzJ7+Br4t3Fmw5H9s2b6X8IhIYmJj+X3FtIwnN6yUDRY2YFlCdclZWKtlk01SXkn+90vm6BQMw3ge9UAuBdQCKgPzgEdzOjSTdRkiARGZC8w1DOMpYCIwLJN95gPzQc0+yslmjSa3jP5tNAFXsy+ucrd4l/dmdtfZWFqmVHMUgdu3U7qZEhOVHIeVVQL/+98fDB/+LPHxsHXrZk6dOsXevXsREXr27MmOHTvw9fVl4cKFlCpViqioKHx8fOjXrx+lS5fO0o7du3dTo0YNatWqRfv27dmwYQN9+/blwIEDrFy5kkOHDhEfH0/jxo3NTqFv3748//zzgMqh+fbbb3n11VeBFJmL8+fP061bN/Nb9NChQ/n8889p164dkyZNYurUqcyePTvL9TNmzODcuXPY2Nhw69YtnJ2dGTlyZI5dWIAamE6OMKydOfTPXnbs2IGtrS2RkZH8vr0/ttaWnPjnMMOefYG//NZCSFL2tlVJIJHDR09wcOMsLE0GdX1f4tWBTahYPul7tLBWEUbcHQY80Tup+8iga0dfZkweDbGhnDkexB8rp2Eh8Uz8cD6H9h5kx88fYmtrzUefTMFa7vD35k/452QQjw19h1N/rgDDij0HTxKwazUupculPOwtrFMe+EVkwD03kcIrqK6gvwBE5JRhGOVycdxFoEqq5cpAxgyTFFYCGbOnNJqHAMMAJyfV6taF2Ngoxo/35sKFQGrVaoK9fSc2bYLlyzezfftmPDwaYWEBkZHhnDp1Cl9fX+bMmcPq1asBzNLZ2TmF9NLZS5cupW/fvmmks4EM0tkTJ040C9Wl7nJK7j6Ki4ujf//+rFy5km7dumWQyH7iiScylc5+4okngBTp7N69e9O7d+/7+l579epljn5iYmIYNWoUhw8fxtLSkjNnzqgyq47RKhIo1QRKHKJjl8cpUbc/JMZRv8FcLoSXo6Kzu9KQSoiGxGhAWPXleLwbVsb8Lht+GhKieaJHWywsrMDkCFYl6dWzB7blfcDCml2H5zJu7BtQ4VHcKpqoWGUpp2+XhxI16Ny1By7VWtzX/RYEuXEKMSISmyy1axiGJbnr+98H1DEMowZwCRgIPJV6B8Mw6ohIcipkdyB7gXiNJo+Z3TUfpbOzwc7Ojn/+CSAsLIwePXpw7NhcevV6jfh4oXfvCXTtqrLfbWyUI1myxI8NG7awdeseSpWyp0MHLZ0NaeWpP/30U6pUqWIW7XPMYgqYjY1N0swpa0xWNsSbSoJDum5pq5JQtgVU9EyaLWWhog27lThUbAplW6r9bMvhUKqM+XgxLFU3l0XGR+uDIqWdm3hlu2EYbwN2hmF0An4E1udwDCISD4wCNgHHgR9E5B/DMN5LmmkEMMowjH8MwwhAjStk6DrSaB5mnJyckvrYP6Fu3Tiee64L+/YtxMsrHKWCcInLl4P5998wwAV/f3u+/fYEe/b8yYULapxC1aVLe97iIJ2dnrCwMCpUqIBhGCxevJg8Scw1LJLGG3I3PdbX19f8vR0/fpwrV67kmChZ1MhNpDAeeBb4G3gR2CAi3+Tm5CKyAdiQbt2kVJ9fz72pGs3DSaNGjfDy8mLlypU8/fTTHD9+nO7d1Zuoo6Mjy5Yto0uXrvz55zzGjfOkSpV6NGjQgkuXYP9+iI6GbdugWrWULqply1bQu3efNNd52KSz0zNq1Cj69+/PihUr6Nixo4oI7oMBAwaYp6S6urqyadOmHI959dVXefHFF/Hw8MDKyoolS5ZgbW19X3YUNDnKXBiG8bqI/DendQWFlrnQ3C+5kbl4EEhIUIPXt26pmUxhYUoBNvlf2tpaOQhn5xRnYWen88GKA/ktczEMlZWcmuGZrNNoNAWIyaSmt7qkrvaZ5CjCwlKcxenTGR1FamehHYUmNVk6BcMwBqEGhmsYhrEu1aYSwI38Nkyj0dw9OTmKZGdx5kzmjiLZWWhHUXzJLlLwB64AZYDUQvx3gCOZHqHRPCCISJYzcB42snIUd+6k7XpK7yiSj3FxUY7igalOV8y53wH2LH/NInIeOA+0vK8raDRFDBsbG0JDQ3FxcSk2jiE9JpN60KeWW0p2FGFhEBqqWnJtGsNQ2dcuLkrSw8VFRxNFEREhNDT0vgbZc5PR3AL4HGiAkqIwAREiUvKer6rRFCIVKlTgypUr3Lx5s7BNKbI4OqoWHw8RESktOFhlY4OSCndwUM3RUTmJrPTxNAWHjY0NFSpUuOfjcxMQfoFKPPsRaAoMBR78qRuaYoulpSVVqlTJeUdNBuLjVZ1rf3/Ys0f9TK60aW0NTZpAq1bQsqX6eR/PJk0hkZspqftFpKlhGEdExDNpnb+I5H+x0EzQU1I1mqLFtWspDsLfX+VOxMSobdWrpziIli3B0/MhLkZUxMnLKamRhmFYAwGGYXyMGnx+MPK1NRpNvuPqCr17q0E6OWAAAA6NSURBVAaqCNGhQymOYscOWLFCbbO3h2bNUhxFixaqBram6JCbSKEacA01njAGcAK+FJHT+W9eRnSkoNE8eAQFpe1yOnRIdUWBEghM3eXUsKEem8gPchspZOsUkqqnLRaRIXlp3P2gnYJG8+ATGQkHDqR0Oe3Zo0qcgsqVaN5cOQhv75T6FCUyKeL2MCICN2+qpMPkeuDJn8ePT4nI7pY86T4SkQTDMMoahmEtIrHZ7avRaDS5xd4e2rZVDdSD8MyZtNHEe++lzHQCVdY02UGkb9WqKUXZBwUR5QSTH/bpH/63bqXsaxhQtaqqD14Q4zG56T76GmgMrAMikteLyGf5a1rm6EhBoyke3L4NJ0/CuXMZ2/nzauwiNRUrZu00KldWuRkFiYgahE/9sE/dbqcqRmxhoRxbnTrq4Z/c6tRR9ueFw8vLgebLSc0CJXGh0Wg0+U7JkuDjo1p6EhPh8mVVBzu9w9ixA77/Pm2UYWmp3rZr1FAzotI7DVfXe0vES0xU8uXp3/STW0REyr4mk7pW7drQunXah3/16mpKb1Egx0ihqKEjBY1GkxOxsWpwO7Mo49w5lYSXGju7tM4i/efw8My7es6cgaiolPNYWUHNmhnf9mvXVk6pMKfj5mWkoNFoNA8U1tZQq5ZqmRERkTbKSP15924l9ZHTuevUgc6d0zqAqlULvpsqr9FOQaPRFDscHMDNTbXMCA1N6zAcHVPe+CtVevAf/NmRrVNImpL6mojMKiB7NBqNptBJVodt3LiwLSl4sk0REZEEoFcB2aLRaDSaQiY33Ue7DcP4AlhF2impB/PNKo1Go9EUCrlxCsnCd++lWifAI3lvjkaj0WgKkxydgoh0KAhDNBqNRlP45Cg7ZRiGk2EYnxmGsT+pfWoYhlNBGKfRaDSagiU3WoQLUXWZn0xqt4Hv8tMojUaj0RQOuRlTqCUi/VItTzUMIyC/DNJoNBpN4ZGbSCHKMIw2yQuGYbQGorLZX6PRaDQPKLmJFEYCS1KNI4QCw/LPJI1Go9EUFjllNFsA9UTEyzCMkgAicju7YzQajUbz4JJTRnMiMCrp823tEDQajebhJjdjCr8bhjHWMIwqhmGUSm75bplGo9FoCpzcjCk8k/TzlVTrBKiZ9+ZoNBqNpjDJzZjCEBHZXUD2aDQajaYQyc2YwicFZItGo9FoCpncjClsNgyjn2HcSwVTjUaj0TxI5GZM4Q3AAUgwDCMKMAARkZL5aplGo9FoCpwcIwURKSEiFiJiJSIlk5Zz5RAMw+hqGMZJwzBOG4YxPpPtbxiGccwwjCOGYfxhGEa1e7kJjUaj0eQNuVFJNQzDGGIYxrtJy1UMw2iWi+NMwFygG9AQGGQYRsN0ux0CmoqIJ/AT8PHd3oBGo9Fo8o7cjCl8CbQEnkpaDkc97HOiGXBaRM6KSCywknSlPUVkm4hEJi3+CVTOldUajUajyRdy4xSai8grQDSAiIQC1rk4rhIQlGr5YtK6rHiW/7d39zF71Xcdx98f282NPeFsN7e2kTobJz6smAJ7SDYd/FHUtMSMrUTMMIvExLp10RimBrWJumWLgoEQKmMsikOGM6umgxlgTP2joeVBVgqxAVI62ChxY5tmc6Vf/zinv169e7dcbe/TU3u/X8mV6zxd5/flhLuf6/zOdX4HvjDbiiRXHHiew969e6doWpJ0PKYJhe/3XUEFkGQxsH+Kz832a6WadcPkMmAV8PHZ1lfVpqpaVVWrFi9ePEXTkqTjMU0o/BXwj8Drkvwp8G/An03xuT3Ason5pcDTMzdKciHwB8CaqvreFPuVJA1kmmc035JkO3AB3bf/i6tq5xT7vg9YkWQ58FVgHQevSwCQ5BzgBmB1VT17rMVLkubWNPcpUFWPAo8ey46ral+S9cCdwALgpqrakWQjsK2qNtN1F70S+Gx/b9zuqlpzLO1IkubOVKFwvKpqC7BlxrKrJqYvHLJ9SdKxmeaagiRpnjAUJEmNoSBJagwFSVJjKEiSGkNBktQYCpKkxlCQJDWGgiSpMRQkSY2hIElqDAVJUmMoSJIaQ0GS1BgKkqTGUJAkNYaCJKkxFCRJjaEgSWoMBUlSYyhIkhpDQZLUGAqSpMZQkCQ1hoIkqTEUJEmNoSBJagwFSVJjKEiSGkNBktQYCpKkxlCQJDWGgiSpMRQkSc2goZBkdZLHkuxKcuUs69+Z5P4k+5K8Z8haJEkvbrBQSLIAuA64CDgbuDTJ2TM22w1cDvzdUHVIkqa3cMB9nwfsqqrHAZLcCqwFHjmwQVU92a/bP2AdkqQpDdl9tAR4amJ+T7/smCW5Ism2JNv27t07J8VJkg43ZChklmV1PDuqqk1VtaqqVi1evPgEy5IkHcmQobAHWDYxvxR4esD2JEknaMhQuA9YkWR5kpcC64DNA7YnSTpBg4VCVe0D1gN3AjuB26pqR5KNSdYAJDk3yR7gEuCGJDuGqkeS9OKG/PURVbUF2DJj2VUT0/fRdStJkk4B3tEsSWoMBUlSYyhIkhpDQZLUGAqSpMZQkCQ1hoIkqTEUJEmNoSBJagwFSVJjKEiSGkNBktQYCpKkxlCQJDWGgiSpMRQkSY2hIElqDAVJUmMoSJIaQ0GS1BgKkqTGUJAkNYaCJKkxFCRJjaEgSWoMBUlSYyhIkhpDQZLUGAqSpMZQkCQ1hoIkqTEUJEmNoSBJagwFSVJjKEiSmkFDIcnqJI8l2ZXkylnW/2CSv+/Xb01y1pD1SJKObrBQSLIAuA64CDgbuDTJ2TM2+wDwjar6ceAvgY8NVY8k6cUtHHDf5wG7qupxgCS3AmuBRya2WQv8cT99O3BtklRVzXUxG+7YwINfe3CudytJJ83KH1nJ1auvHrSNIUNhCfDUxPwe4PwjbVNV+5I8D/ww8NzkRkmuAK7oZ7+T5LFBKj55FjHjv3Ge83gc5LE4lMdjwr3cu+garjne4/Gj02w0ZChklmUzzwCm2Yaq2gRsmouiTgVJtlXVqrHrOFV4PA7yWBzK43Gok3E8hrzQvAdYNjG/FHj6SNskWQi8BvivAWuSJB3FkKFwH7AiyfIkLwXWAZtnbLMZeH8//R7g7iGuJ0iSpjNY91F/jWA9cCewALipqnYk2Qhsq6rNwCeBv0myi+4MYd1Q9ZxiTpuusDni8TjIY3Eoj8ehBj8e8Yu5JOkA72iWJDWGgiSpMRROoiTLktyTZGeSHUk+NHZNY0uyIMkDSf557FrGluTMJLcnebT/f+RtY9c0piQf7v9OvpLkM0leNnZNJ0uSm5I8m+QrE8tem+Rfkvxn//5DQ7RtKJxc+4DfqaqfBN4K/NYsQ3/MNx8Cdo5dxCniGuCOqnoz8Bbm8XFJsgT4ILCqqn6a7scq8+WHKAA3A6tnLLsSuKuqVgB39fNzzlA4iarqmaq6v5/+Nt0f/ZJxqxpPkqXALwE3jl3L2JK8Gngn3S/yqKr/rapvjlvV6BYCL+/vYTqDw+9zOm1V1Zc5/J6ttcCn++lPAxcP0bahMJJ+RNhzgK3jVjKqq4HfA/aPXcgp4MeAvcCn+u60G5O8YuyixlJVXwU+AewGngGer6ovjlvV6F5fVc9A9wUTeN0QjRgKI0jySuAfgA1V9a2x6xlDkl8Gnq2q7WPXcopYCPwccH1VnQP8NwN1D/x/0PeXrwWWA28EXpHksnGrmh8MhZMsyUvoAuGWqvrc2PWM6B3AmiRPArcC707yt+OWNKo9wJ6qOnDmeDtdSMxXFwJPVNXeqvo+8Dng7SPXNLavJ3kDQP/+7BCNGAonUZLQ9RnvrKq/GLueMVXVR6pqaVWdRXcB8e6qmrffBKvqa8BTSX6iX3QBhw4zP9/sBt6a5Iz+7+YC5vGF997ksEDvBz4/RCNDjpKqw70D+DXg4SQHHu7w+1W1ZcSadOr4beCWfqywx4FfH7me0VTV1iS3A/fT/WrvAebRkBdJPgP8PLAoyR7gj4CPArcl+QBdaF4ySNsOcyFJOsDuI0lSYyhIkhpDQZLUGAqSpMZQkCQ1hoIkqTEUpCkkWZnkFyfm1ySZk2EokmxIcsZc7Es6Ud6nIE0hyeV0wzivH2DfT/b7fu4YPrOgql6Y61okzxR0WklyVv+Amr/uH9DyxSQvP8K2b0pyR5LtSf41yZv75Zf0D3Z5KMmX+zuMNwLvS/JgkvcluTzJtf32Nye5vn+A0uNJ3tU/JGVnkpsn2rs+yba+rj/pl32QbsC3e5Lc0y+7NMnDfQ0fm/j8d5JsTLIVeFuSjyZ5JMl/JPnEMEdU805V+fJ12ryAs+iGRVjZz98GXHaEbe8CVvTT59ONvwTwMLCknz6zf78cuHbis22e7oEotwKhG9nzW8DP0H3p2j5Ry2v79wXAl4Cf7eefBBb102+kG8JgMd0wNHcDF/frCnjvgX0Bj3HwbP/MsY+9r9Pj5ZmCTkdPVNWBsaW20wXFIfrhy98OfLYfh+oG4A396n8Hbk7yG3T/gE/jn6qq6ALl61X1cFXtB3ZMtP/eJPfTjePzU8BsT907F/hSdaOD7gNuoXv4DsALdCPsQhc83wVuTPIrwP9MWad0VA6Ip9PR9yamXwBm6z76AeCbVbVy5oqq+s0k59M9Fe7BJIdtc5Q2989ofz+wMMly4HeBc6vqG3230mzPHM5R2vhu9dcRqmpfkvPoRg9dB6wH3j1FndJReaageam6hxs9keQS6IY1T/KWfvpNVbW1qq4CngOWAd8GXnUCTb6a7sE5zyd5PXDRxLrJfW8F3pVkUZIFwKXAvTN31p/pvKa6EXY3ANMEl/SiPFPQfParwPVJ/hB4Cd11gYeAjydZQfet/a5+2W7gyr6r6c+PtaGqeijJA3TdSY/TdVEdsAn4QpJnquoXknwEuKdvf0tVzTZu/quAzyd5Wb/dh4+1Jmk2/iRVktTYfSRJauw+0mkvyXV0T72bdE1VfWqMeqRTmd1HkqTG7iNJUmMoSJIaQ0GS1BgKkqTm/wA+1ira4u2gjAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1eef31d6240>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn import datasets\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.metrics import zero_one_loss\n",
    "from sklearn.ensemble import  AdaBoostClassifier\n",
    "import time\n",
    "a=time.time()\n",
    " \n",
    "n_estimators=10\n",
    "learning_rate=1\n",
    "X,y=datasets.make_hastie_10_2(n_samples=12000,random_state=1)\n",
    "X_test,y_test=X[2000:],y[2000:]\n",
    "X_train,y_train=X[:2000],y[:2000]\n",
    " \n",
    "dt_stump=DecisionTreeClassifier(max_depth=1,min_samples_leaf=1)\n",
    "dt_stump.fit(X_train,y_train)\n",
    "dt_stump_err=1.0-dt_stump.score(X_test,y_test)\n",
    " \n",
    "dt=DecisionTreeClassifier(max_depth=9,min_samples_leaf=1)\n",
    "dt.fit(X_train,y_train)\n",
    "dt_err=1.0-dt.score(X_test,y_test)\n",
    " \n",
    "ada_discrete=AdaBoostClassifier(base_estimator=dt_stump,learning_rate=learning_rate,n_estimators=n_estimators,algorithm='SAMME')\n",
    "ada_discrete.fit(X_train,y_train)\n",
    " \n",
    "ada_real=AdaBoostClassifier(base_estimator=dt_stump,learning_rate=learning_rate,n_estimators=n_estimators,algorithm='SAMME.R')\n",
    "ada_real.fit(X_train,y_train)\n",
    " \n",
    "fig=plt.figure(1)\n",
    "ax=fig.add_subplot(111)\n",
    "ax.plot([1,n_estimators],[dt_stump_err]*2,'k-',label='Decision Stump Error')\n",
    "ax.plot([1,n_estimators],[dt_err]*2,'k--',label='Decision Tree Error')\n",
    " \n",
    "ada_discrete_err=np.zeros((n_estimators,))\n",
    "for i,y_pred in enumerate(ada_discrete.staged_predict(X_test)):\n",
    "    ada_discrete_err[i]=zero_one_loss(y_pred,y_test)    ######zero_one_loss\n",
    "ada_discrete_err_train=np.zeros((n_estimators,))\n",
    "for i,y_pred in enumerate(ada_discrete.staged_predict(X_train)):\n",
    "    ada_discrete_err_train[i]=zero_one_loss(y_pred,y_train)\n",
    " \n",
    " \n",
    "ada_real_err=np.zeros((n_estimators,))\n",
    "for i,y_pred in enumerate(ada_real.staged_predict(X_test)):\n",
    "    ada_real_err[i]=zero_one_loss(y_pred,y_test)\n",
    "ada_real_err_train=np.zeros((n_estimators,))\n",
    "for i,y_pred in enumerate(ada_real.staged_predict(X_train)):\n",
    "    ada_discrete_err_train[i]=zero_one_loss(y_pred,y_train)\n",
    " \n",
    "ax.plot(np.arange(n_estimators)+1,ada_discrete_err,label='Discrete AdaBoost Test Error',color='red')\n",
    "ax.plot(np.arange(n_estimators)+1,ada_discrete_err_train,label='Discrete AdaBoost Train Error',color='blue')\n",
    "ax.plot(np.arange(n_estimators)+1,ada_real_err,label='Real AdaBoost Test Error',color='orange')\n",
    "ax.plot(np.arange(n_estimators)+1,ada_real_err_train,label='Real AdaBoost Train Error',color='green')\n",
    " \n",
    "ax.set_ylim((0.0,0.5))\n",
    "ax.set_xlabel('n_estimators')\n",
    "ax.set_ylabel('error rate')\n",
    " \n",
    "leg=ax.legend(loc='upper right',fancybox=True)\n",
    "leg.get_frame().set_alpha(0.7)\n",
    "b=time.time()\n",
    "print('total running time of this example is :',b-a)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-1.93156389, -0.16042612, -0.83221954, -0.20611156,  1.02625492,\n",
       "        0.96996377, -0.22680971, -0.11040249, -0.60073243,  1.13414588])"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-1., -1.,  1.,  1.,  1.,  1., -1.,  1.,  1.,  1.])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x = 1\n",
    "x"
   ]
  },
  {
   "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}