hands-on/05_support_vector_machines.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Chapter 5 – Support Vector Machines**\n",
    "\n",
    "_This notebook contains all the sample code and solutions to the exercises in chapter 5._"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Setup"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First, let's make sure this notebook works well in both python 2 and 3, import a few common modules, ensure MatplotLib plots figures inline and prepare a function to save the figures:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# To support both python 2 and python 3\n",
    "from __future__ import division, print_function, unicode_literals\n",
    "\n",
    "# Common imports\n",
    "import numpy as np\n",
    "import os\n",
    "\n",
    "# to make this notebook's output stable across runs\n",
    "np.random.seed(42)\n",
    "\n",
    "# To plot pretty figures\n",
    "%matplotlib inline\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "plt.rcParams['axes.labelsize'] = 14\n",
    "plt.rcParams['xtick.labelsize'] = 12\n",
    "plt.rcParams['ytick.labelsize'] = 12\n",
    "\n",
    "# 한글출력\n",
    "plt.rcParams['font.family'] = 'NanumBarunGothic'\n",
    "plt.rcParams['axes.unicode_minus'] = False\n",
    "\n",
    "# Where to save the figures\n",
    "PROJECT_ROOT_DIR = \".\"\n",
    "CHAPTER_ID = \"svm\"\n",
    "\n",
    "def save_fig(fig_id, tight_layout=True):\n",
    "    path = os.path.join(PROJECT_ROOT_DIR, \"images\", CHAPTER_ID, fig_id + \".png\")\n",
    "    print(\"Saving figure\", fig_id)\n",
    "    if tight_layout:\n",
    "        plt.tight_layout()\n",
    "    plt.savefig(path, format='png', dpi=300)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Large margin classification"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The next few code cells generate the first figures in chapter 5. The first actual code sample comes after:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=inf, cache_size=200, class_weight=None, coef0=0.0,\n",
       "  decision_function_shape='ovr', degree=3, gamma='auto', kernel='linear',\n",
       "  max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
       "  tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.svm import SVC\n",
    "from sklearn import datasets\n",
    "\n",
    "iris = datasets.load_iris()\n",
    "X = iris[\"data\"][:, (2, 3)]  # petal length, petal width\n",
    "y = iris[\"target\"]\n",
    "\n",
    "setosa_or_versicolor = (y == 0) | (y == 1)\n",
    "X = X[setosa_or_versicolor]\n",
    "y = y[setosa_or_versicolor]\n",
    "\n",
    "# SVM Classifier model\n",
    "svm_clf = SVC(kernel=\"linear\", C=float(\"inf\"))\n",
    "svm_clf.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saving figure large_margin_classification_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA1gAAAC7CAYAAAB4pz8ZAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzs3Xd4VGX2wPHvO+mNAAkkJNTQS+hgQQQpKioqrLoW/AErNTTXtiqKLAKKNHVXFlGxgwiCXXdFwYK0AAEJVbqEhEAgIaTPvL8/JokJpAyZmdyZyfk8zzxk7sy9cybAvHPue+55ldYaIYQQQgghhBD2MxkdgBBCCCGEEEJ4CkmwhBBCCCGEEMJBJMESQgghhBBCCAeRBEsIIYQQQgghHEQSLCGEEEIIIYRwEEmwhBBCCCGEEMJBJMESQgghhBBCCAep1gRLKTVWKbVTKRWvlNqllIor53kTlFL7lFK7lVJfKqUiqjNOIYQQnkfGICGEENWh2hIspZQX0BLopbXuDgwC5iuloi95Xl/gH0BvrXUHIB54q7riFEII4XlkDBJCCFFdlNbamBdWyh9IAWK11sdLbP8PcE5r/XTh/TpAKhCmtU43JFghhBAeRcYgIYQQzmLkNVgvAytKDmyFYoBDRXe01ueAdKBp9YUmhBDCw8kYJIQQwim8jXhRpdRMIBr4S1kPA+ZLthVQRjKolBoDjAEICgrq1qZNGwdHKoRwJ5ZcC9kHsrHkWfAK8CKgVQDKWxkdlnCQbdu2ndFa17P3OI4agwqPVSPHofT0dH7//XeUUrRr1w5/f3+jQ7JbTk4Oe/bsQWtNixYtCA0NNToku1ksFhITE8nLyyMqKooGDRoYHZIQbs3WcajaEyyl1DygOfAXrXVeGU/5A2hc4vmBQFjh9lK01kuAJQDdu3fX8fHxTolZCOH6Lu65yM4BO8nLyyPkqhA6ftMRnzo+RoclHEgpdcwBx3DYGAQ1exwaPnw47733HuPGjWPixIlGh+MQCxYs4NFHH+XixYts2bKFOnXqGB2S3dauXcvAgQNp2rQpP/30E15eXkaHJITbsnUcqs4mFyal1GKgEXB30cCmlPJSSn2vlOpd+NT3gQeUUkWnjiYAG7TWqdUVqxDCvVzYcYGEPgnkncqjdt/adPquk8snV6v2rGL/mf0YdR1sTSNjkOO9/PLL/PTTTx6TXAFMmTKFXr16cerUKaZMmWJ0OA4xYMAAvvrqK9avXy/JlRDVpDqvwboFGIu1vv0XpdQmpdQmrJ2cmgF1AbTW64DXgB+VUluB64B7qzFOIYQbSd+YTsINCeSfyafuoLrEfh2Ld4gh1c82O5N1hrtX3k3XJV0x60ur0YSTyBjkYHXq1KF3796VP9GNeHl5sXTpUvz9/Xn//ff54osvjA7JIW655RZ8fFz7pJMQnqTaEiyt9Zdaa6W17qG1vrrE7UutdYzW+rMSz31Va9258Ll3aK1PVVecQgj3cW79OXYO3Ik53Uz40HA6rOmAV4Drn6HdenIrAF0bdMXb5NrJoKeQMci5fvjhB+6//37MZvc/YdCqVStmz54NwJgxY0hLSzM4IsdJSUnh3nvv5bfffjM6FCE8mpFdBIUQosrOfn2W3wb9huWihYhhEbRb0Q6Tn3t8pG1NsiZYPaN6GhyJEPbLycnh//7v/1i+fDmvvPKK0eE4xOTJk+nVqxfJyck8/PDDRofjMC+99BIrVqxgxIgR5OfnGx2OEB7LPb6NCCFECamfpLL7zt1Yciw0GNuANu+2weTtPh9nW05uAaBHdA+DIxHCfv7+/rz++usATJ06lf379xsckf08tVRw+vTpNGnShO3btzNnzhyjwxHCYxm20LCjVdS9yWKxcObMGc6fP+8R5QvCufz9/WnYsKHUq7uo5PeT2TdiH1ig4SMNaT6vOUq5Tyt2rTWR8yM5ffE0hyYfIqZOjNEhuQ2l1DatdXej4yhPTesieKkRI0bw7rvvcs011/Dzzz97REOFhQsX8sgjjxAZGUliYiJ169Y1OiS7ff/99wwYMAAfHx/i4+Pp2LGj0SEJ4TZsHYdqRIJ1/PhxlFJERETg4+PjVl/GRPXSWnP27FkuXLhAs2bNjA5HXOLk4pMcHH8QgCbTmtB0elO3+/987Pwxmr7SlLCAMFIfT3W7+I0kCZZrO3fuHB06dCApKYl58+bx6KOPGh2S3cxmM3369GHDhg0MGzaM999/3+iQHGL8+PEsXryYLl26sHnzZjmhKISNbB2H3Kemxg4XL14kOjoaX19f+TIjKqSUIiwsjJycHKNDEZc4seBEcXIVMyeGZv9s5pb/nw+mHcTXy5ce0T3cMn4hylOnTh2WLFkCwDPPPOMxpYJvv/02/v7+fPDBB3z++edGh+QQL730Ek2aNGHHjh1SKiiEE9SIBAvAZKoxb1XYSb70uhatNUdnHOXQo4cAaPlaSxo/0biSvVzXgJgBXHjqAm/f8Xa1vF5kJCh1+S0yslpeXtQwt956K8OHDyc0NJRTpzyj+WLLli2LuwqOHTvWI7oKhoSE8NZbb2EymcjMzDQ6HOHhauI4JP2BhRAuS2vN4X8c5sTcE2CCNkvbEDnc/T+Rfb18iQyunveRknJl24Ww18svv4zZbCYsLMzoUBxm8uTJfPLJJ2zYsIEpU6Z4RKlg//79OXjwIDExch2ocK6aOA7JtI4L69u3L08++WSZj506dYqwsDC7OxuNHj2agQMHlvnYlClTuPXWW+06flm2b99O7dq12b59u93Hquh3JNybtmgOTjjIibknUN6Kdh+1c/vkSmuNRVuMDkMIp6pdu3ap5MoTmksVlQoGBAR4VKlgyeTKE/6ehHAVkmC5KX9/f7p06WL3GcJhw4axbt06Ui45jWCxWFi5ciUPPPCAXccvS0hICF27dqVWrVoOP7bwDJYCC/tG7iPpP0koP0X7Ne2pf3d9o8OyW2JqInXm1OHBNQ8aHYoQTpebm8szzzzDDTfc4BFf3j2xVLDI//73P1q3bs2uXbuMDkUIjyAJlo1crX60Tp06rF27lmuvvdau41x//fVER0ezcuXKUtvXr19PRkYGd955p13HL0vLli354YcfaNGihcOPLdyfJc/C3vv3kvJeCqZAEx2/6kj4beFGh+UQW09uJSM3g5wCaaIiPF92djZvv/02P//8MwsXLjQ6HIeYPHky1113HcnJyUyZMsXocBzm008/5dChQ7IAsRAOIgmWjYysH50+fTpdunThscceIzIykquvvpqCggKUUqxfvx6ADRs20LNnT6KiooiOjmbkyJGcP3++0mMrpbj//vtZvnx5qe3Lly9nyJAhBAYGAnDmzBkeeughWrVqRXR0ND169CgukVi/fj1KKRYtWkSzZs0IDAwkMzOTN998k1atWhEZGUnz5s2ZOXMmAL///jtKKY4ePQpAfn4+8+bNIzY2loiICGJiYpg3bx4AZ8+e5aGHHqJevXrUrVuXq666ig0bNpT7fj799FNiY2MJCwujQYMGPPbYY2RnZ1cYp3Ad5mwzu4fuJnVlKl61vOj0v07U6V/H6LAcpmiB4Z5RPQ2ORAjnq127Nm+88QZg7Sq4b98+gyOyn8lkYunSpR5XKvjSSy/RtGlTduzYwYsvvmh0OEK4PUmw3ERCQgIhISGcPHmSdevWXfb45MmT6d+/P0lJSRw4cIBu3bqRnp5u07EffPBBNm7cyPHjxwHIy8vjk08+YdiwYYA1ARo8eDC33347+/bt4+TJkyxatIj777+/VBveb775hoSEBNLS0jCbzYwZM4Z//etfJCcn88svv2CxlH3tycMPP8w777zDhx9+SEpKCgkJCdSvXx+tNbfeeisWi4Vjx45x9uxZnn32WQYOHMiePXsuO84333zDyJEjWbJkCWfPniUxMZHExERGjRp12fOK4gwKCrLpdyScryCzgN9u+420r9LwDvOm87rOhPYKNTosh9qSVJhgRVdfghURcWXbi7jarL1wT7fccgsjRowgNzeXkSNHSqmgiwoODuatt94C4Pnnn5dSQeFQNXIc0lp7xK1bt266PHv27Cn3MVtB+Tdn6dOnj/7HP/6hn3vuOR0dHa0tFkvxY/n5+RrQ69at01prPX78eN2zZ0+9devWy47TvXt3HRERcdmtpM6dO+s5c+ZorbX+/PPPdUREhC4oKNBaa718+XLt4+Nz2f4hISH6gw8+0OvWrdOAPnDgQPHxzGazbt++vf6///s/feLEiVKvdfDgQQ3oI0eO6KSkJK2U0uvXr78s7k2bNmlAnz9/vtT2u+66S0+cOLHU70hrrW+++Wb997//vdRzd+/erQGdmppaZpzlccS/GWG7vHN5ets12/Q61ukNkRt05u5Mo0NyuOz8bO09w1ur6Upn5GQYHU6ljPjMqzwm4rULjDfl3Soah2qyc+fO6ejoaA3ouXPnGh2OQ5jNZn3ddddpQD/wwANGh+Mw48eP14Du0qWLzsvLMzocUcO58zgkM1huomHDhhWuz/Taa68RFxfHuHHj6NatGytWrCh+bOvWrSQnJ192K2nYsGF89NFHgLU88N5778XLywuAw4cP06JFi8v2z8jIKNUEo1GjRsU/m0wmNm7cSPPmzenduzeDBw9m27Ztl8V95MgRtNZ07NjxsseOHTtGaGgooaGlZzGaNGnC4cOHy3x+48aNL3tu0euUFacwXt6ZPHb220nGxgz8GvvR+efOBLX3vJnFhOQECiwFtKvXjhC/EKPDEaLaeGqpYFFXwQ8//JDPPvvM6JAcQkoFhXAMSbA8hFKK4cOHEx8fz6JFi4iLi+PTTz+1ef/777+fXbt2sWPHDj7//PPi8kCA6OhoDh8+zNmzZ68oppCQEKZNm8ahQ4cYMmQIvXv3vmzhyejoaIBSpYZFGjduTHp6+mXXkh09epRmzZqV+fxjx45d9lyApk2bXlHsonrknsoloU8CmTsyCWgRQJefuhDYItDosJxi68mtQPWWBwrhKgYNGsTIkSO54YYbCA4ONjoch2jRogUvvPAC4HmlgtHR0XTt2tXocIRwW5Jg2aiq9aPV4fz58zz33HOcOHECsCYTvr6+Nl+DBdCgQQNuuOEGRo0aRcOGDenevXvxY3fddReNGzdm5MiRpKamApCSksL06dPLva5qw4YNvPHGG2RkZGAymWjRogU5OTnk5uaWel6TJk24//77iYuLK76uKjMzk9dff52rrrqKq6++mr///e9kZWWhtebLL7/kq6++Ii4u7rLXLLqWa9OmTQCcO3eOxx9/nPvuu4969erZ/LsQ1SPnWA47eu8ga08Wge0D6fxTZ/yb+BsdltPc2upWFt+6mAc7Sot2UTMtWrSIr7/+moYNGxodisNMmjSJ6667jpSUFCZPnmx0OA7Rr18/fv/9d6esgylETSEJlo2Sk8uuAr2k0s4QAQEBZGdn069fPyIiIujRowcjRowoNQtliwcffJDt27dftvZVUFAQa9euJSQkhC5duhAREcF1112H2Wwut2wxOjqaH3/8kXbt2tGgQQPGjRvHe++9V+ZM0htvvMEdd9zB0KFDqV+/Ph06dCA9PR2lFF9//TVgTcTCwsKYMWMGa9eupV27dpcd5+abb+add95h1KhRhIWF0bZtW9q1a1d84a5wHVkHs9jRewc5h3II7hpM5/Wd8WvgZ3RYThVTJ4ax3cdyQ7MbjA5FCEP4+/sXjxlms7n4hJ0789RSQX//P092XbpOphCicsp6vZb76969u46Pjy/zsb1799K2bdtqjki4M/k34zyZuzPZOWAn+Sn51Lq2Fh2/7oh3qLfRYYlLREaWvQxFRIRxJ5aUUtu01t0rf6YxKhqHxJ+OHz/Offfdh8Vi4Zdffim+3tedvfLKKzz88MNERESQmJhIWFiY0SHZTWvN1KlTWbhwIZs2baJTp05GhyRqGHceh2QGSwhRbS5su0BCnwTyU/Kp3a82Hf9bM5KrbUnbeOK7J1h35PIlFi5Vlba0zmhlW9GsvVu3zhWGq1WrFseOHWPTpk0sWLDA6HAcYtKkSfTu3dujSgWVUmRkZJCTkyMLENcgVf18l3GoNEmwhBDVIn1DOgn9EihIK6DurXWJ/SoW72DPT64A/nfof8z9dS5r9q2p9LlVWdS8uhdCN3LhdeH+SnYVfPbZZ9m7d6/BEdmv5ALEy5Ytu6ImU67sxRdfpFmzZiQkJBQ39BCeraqf7zIOlSYJlhDC6dLWprHzxp2YM8zUu7seHVZ3wMvf/cuCbGXEAsNCuLKiroKetABxya6C48aNu+LOu64oODiYpUuXAtYFiHfu3GlwREK4B0mwhBBOdeaLM/x2229YsixEjoik3fJ2mHxr1kfPlpOSYAlxqQULFhAdHc3mzZuZP3++0eE4hCeWCvbt25cJEyZQUFAgpYJC2KhmfcsRQlSr0x+fJnFoIjpXExUXReu3WqO8yl8w2xOdzDhJ0oUkQv1CaVG3hdHhCOEySpYKTps2jYMHDxockf1qQqng3LlzjQ5HCJcnCZYQwilOvX2KPfftQRdoGj3RiJb/boky1azkCmBrknWB4R7RPTAp+cgVoqRBgwYxYcIEZs6cSUxMjNHhOESLFi148cUXAc8rFbz99tsZMWKE0eEI4fKqdbRXSvkopR5TSuUrpe4t5zndlFJpSqlNJW6PVmecQgj7nHztJPv/th8s0HRGU2JejCl3zTR3ZksXo+LywCj7ywPLe73ymCr5hPfyKvt4lXXNduWF1ysiY5Br+ve//81jjz3mEe3ai0ycONEjSwU/++wzoqKijA5FlFCd3fQqeq3yxpuaOg5V9+nU0YAGNlXwnHDgE6311SVunlGcLUQNcHzOcQ5OtJb6NJ/fnKbPNvXI5Aps62JUP6g+nSI6cXXDq532euWxWKr2eGX7ufLC65WQMcjFJSUlcejQIaPDsJunlgoWMZvN/Prrr0aHIajebnoVvVZVxxNPHYeqNcHSWi8qHKgqahcUDtyolNpceJullAqpphCFEFWktebIs0c4/ORhUNDq9VY0eqSR0WEZ7uGrHyZhXAKDWw82OpQaT8Yg1/bLL7/Qvn17HnjgAY/pKuhppYIAeXl59O7dmz59+pCQkGB0OEK4JFe8IGA10FRrfRUwCGgGvFfWE5VSY5RS8Uqp+NTU1OqMUQhRgtaaQ48e4tjMY2CCNu+1IWqMlJEIt2TzGAQyDjlShw4dCAoK8qiughMnTuT6668nJSWFSZMmGR2OQ/j6+tK1a9firoJ5eXlGhySEy3G5BEtrna211oU/pwGPAoOVUgFlPHeJ1rq71rp7vXr1qjtUp+vbty9PPvlkmY+dOnWKsLAwvvjiC7tf58yZM4wePZomTZoQGRlJTEwMkyZNIjs726b9ExISmDdvnt1xCPekzZoD4w7wx8I/UD6K9ivbEznMRZZSN9ix88c4cu4IhR9pwg1cyRhU+ByPHoeq06VdBT1tAeLly5ezZk3li427g6Kugjt37pQFiIUog8slWGXwAnKAXKMDcSX+/v506dKFsLAwu4911113kZOTQ2JiIsnJyWzbto0mTZrYfFZKEqyay1JgYe/wvZxacgqTv4kOn3Wg3lD5kllk/sb5xLwaw/yNnnE2voaSMagaDRo0iL/97W/k5uYyYsQICgoKjA7Jbs2bNy8uFRw/frxHlAqWXIB45syZUiooxCUMT7CUUmFKqQ1KqZaF9+9VStUu/NkbeAF4X2tdyeVuzpeS8iEbNzZl/XoTGzc2JSXlQ8NiqVOnDmvXruXaa6+1+1hbtmzh1ltvJTg4uPjYjz32GKGhoXYfW3guS66FPffs4fSHp/EK9iL2m1jCBtmf8LuTyroYFXUQ7BLZxamvV16Xpsq6KVW165MncacxqKZYsGABDRs2ZMuWLVIq6MIuXYBYSgWNUZ3d9Cp6rarG4anjkCuEHwg0AYq+zQcA3yultgK/AsnAIwbFViwl5UP27x9Dbu4xQJObe4z9+8dUS5I1ffp0unTpwmOPPUZkZCRXX301BQUFKKVYv349ABs2bKBnz55ERUURHR3NyJEjOX/+vE3H79u3L8899xy//PJLuc9ZvXo1PXv2pHHjxjRt2pQxY8Zw/vx5HnzwQaZMmUJqaiqRkZF07doVgIKCAmbPnk3jxo0JCwujRYsWvPnmm8XH27t3LzfccANRUVFERkZy1113ceLEieLX6tOnD9HR0URGRjJw4MDix4RrMGeZ2X3nbs6sOYN3bW86fteROn3rGB1WhZzRyraiLkZ55jx2JO8AoHtU91L7VdSWtqzjFd3KU163pdTUit+32Vz263hAf4Er4RZjUE0SGhpaPF7MmDHDI2Z8ikoFAwMDPbZU8N133zU6HJdX3eNQeRw9BoF1vClve00chwxJsLTWfbXWHxX+fEJr3VBrHV94/22tdTetdQ+tdU+t9eNaa9suBnKiw4enYrFkldpmsWRx+PDUann9hIQEQkJCOHnyJOvWrbvs8cmTJ9O/f3+SkpI4cOAA3bp1Iz093aZjv/vuu7Rr147rr7+eTp068e6775bq4LRy5UreeecdvvjiC44fP86ePXvIyspi8uTJvP/++7zyyivUq1eP5ORktm/fDsDTTz/N6tWr+fXXXzl79ixffPEFs2bNKq6vf+aZZ2jQoAEnT57k2LFjDBkyhOTCT4Pdu3cze/Zsjh8/zh9//EHdunWZOXOmvb9C4SAFFwrYdcsu0r5Nwyfch07rOhF6tevPdlZnK1uAXSm7yDPn0TqsNaH+pX8/VW1LW5U27dX9vt2BO45BNc1NN93E888/z7p16xxSCu8KSpYKjhs3jjNnzhgckf2Cg4N5++23WbhwIX/729+MDsflucrnsaPHoMratLvK+65O3kYH4C5yc49f0XZHi46OZtq0aSilCAgIuKwu/aqrruKHH34gPj6e7t27M3HixOLHevToUeYMUFFCU69ePdasWcPevXv597//zdixY3nzzTf59ttvCQoK4tlnn+X06dN06tSpeN+CggIiypn31Vrz2muvsXz5cho2bAhA27Zteeyxx3j11VcZPXo0V111FW+++Sbr1q2jX79+PPDAA8X7T5s2jTNnzvD9999z7NgxMjIyPGIg8gT55/LZNWgXFzZfwDfKl05rOxHUNsjosFzS1pNbAegZbf8Cw0LURM8884zRITjchAkTWLVqFT/99BOTJ09m2bJlRodktz59+tCnTx+jwxDCpbhCiaBb8PNrfEXbHa1hw4YVLtb62muvERcXx7hx4+jWrRsrVqwofmzr1q0kJydfdrtU27Ztee211/j111+Jj4/n1VdfBeDw4cMsXry41L5nzpwhMTGxzFhSU1PJysqicePSv5smTZpw+PBhAJ544gnmzZvHjBkzaNOmDYsXL8ZisZCfn8/999/P9ddfz9q1a8nLy6N58+YesSaKu8s7nUfCDQlc2HwB/6b+dPm5iyRXFdiSZL3+ShIsIez35ZdfelRXQU8rFSzyxx9/sHz5cqPDEMJwMoNlo5iYWezfP6ZUmaDJFEhMzCwDo/qTUorhw4czfPhwNm/ezC233IKfnx933nlnhftprdFaYypxNWHXrl1p165d8axXdHQ0W7du5Z577rEplvDwcAIDAzl27BidO3cu3n706FGaNWtWfP/222/n9ttvZ9++fdx2223k5ORQu3ZtfvzxRw4dOoS/vz9grfHevXu3zb8L4Xi5J3PZOWAnWfuyCGgVQKe1nfBv5G90WC5tW9I2AHpE9TA4EiHc29tvv83f/vY3evbsyYYNG/D2du+vLkWlgpMnT2bcuHH07t2b8PBwo8Oy29mzZ4mNjSUzM5O2bduWGv+FqGlkBstGEREP0Lr1Evz8mgAKP78mtG69hIiIByrd19nOnz/Pc889V5wQNW3aFF9fX5uuwUpPT2fQoEHs2bOneNu3337Lnj17GDJkCGC9nuq1117j008/xWKxYDabWbt2bfGZt8DAQLKyssjMzCQpKYns7GwmTZrEjBkz+OOPPwBrU4v58+czefJkAJ566ikOHDgAWBO4oKAg0tPTyc3NJT8/n7S0NADWrVvHggULHPSbElWRfSSbHb13kLUvi6DYILr81EWSKxtsHrWZX//2K50j5UuGEPYYOnSox3UVnDBhAn369OH06dMe01UwLCyMBx98kIKCAoYPHy5dBUWN5t6ngapZRMQDLpFQXSogIIDs7Gz69etHRkYGfn5+jBgxgmHDhlW6b3BwMNdccw1Dhw4lIyMDpRRNmzZl+fLlDBw4EIDRo0fj7e3Nc889x6hRo/D19SU2NpZZs6yzdwMGDKB9+/aEh4fToEEDfvrpJ2bNmkWtWrW45ppryMrKonbt2jz99NOMHj26+HWHDBlCamoqPj4+3HbbbTz++OMopfjxxx+JjY3Fx8eH3r17ExcXV9wtUVSvrP1ZJPRPIO9kHiE9Quj4bUd86vpUy2tHRpZ9AWxERMXdkcoTEVH+8SqovsVkKvvi3aJJ3/IfC8Biuabc/cpT0fsuL5bymExQr175xxPCHRR1Fbz55puZNm0agwcPpl27dkaHZZeiUsHY2Fg++ugj7r77boYOHWp0WHZ74YUX+Oqrr9i1axezZ89m+vTpRodUZY4eg4r2rd5x6MrGDKjaGFRZHDVxHFK6sr6LbqJ79+46Pj6+zMf27t1L27Ztqzki4c7k3wxk7spk58Cd5J/OJ7R3KLFfxuJdq/rOyVQ02Dj6Y6ui13IXHvJRXiGl1DatdffKn2mMisYhYb9Ro0bx1ltv0aNHD3799Ve3LxUE+Pe//82kSZOoX78+iYmJHlEq+OOPP9K3b1+8vb3ZunWr25YKVucYVNnruYOaMAaB7eOQlAgKIS6TsSWDhL4J5J/Op86Ndej4bcdqTa6EEOJS8+fPp2HDhmzdupV58+YZHY5DxMXFeVypYJ8+fZg0aZKUCooaTRIsIUQp5386z84BOyk4V0DYHWHEfh6LV6CX0WEJIWq4kgsQL1261CO+uJfsKvjRRx+xevVqo0NyiBdeeIGYmBgSExP55ZdfjA5HiGp3RQmWUsqklJrsrGCEEMZK+28au27ehfmCmfr31af9yvaY/OQ8jBDCNdx00028//77xMfH4+vra3Q4DhHhz+GAAAAgAElEQVQTE8OcOXMAGD9+vEes+xgUFMSHH37I5s2b6devn9HhCFHtrvSbkw+w0BmBCCGMlfppKr/d/huWbAuRD0XS9v22mHwkuRJCuJZhw4ZRq1Yto8NwqLi4OPr27cvp06eZOHGi0eE4xNVXX023bt2MDkMIQ5T77Ukp9bVSyqvw54xLHlNKqUdK3MY5O1AhhPOkLEsh8a5EdJ4menI0rZe0RnkZe8Vted2FqrvrUHld/0ymijoCmkGV3brJZKr4vVX1MSFqmqysLB599NFyF713JyaTibfeeovAwEBWrFjBJ598YnRIDvXpp5+ycKF7nZ93pc/bqoxDFY9RZZMxyHEq+tXfDLRRSjUGgi95zAeYBwwuvN3knPCEEM6W9GYSe4ftBTM0froxLV5ugTIZ384oOdnalejSW1Xb40ZGWrs0XXqLjKx40DCby47DbLa2ni1TUCqrdq8pd7+qvu/yHoPy35sQnmrWrFksWLCAESNGUFBQYHQ4ditZKhgXF+cRpYIA+/btY+jQoTz++ONs377d6HBs5ugxCKp3HKpXr/z9KkqIqjIGJSdX/N5qospy213AUaCo+aIq+afW+obC2xDnhOc4ntKOXjhfTfq38scrf3Bg9AHQ0GxWM2JmxaDcvVdsOcpag6Noe0WPVeWYXIykR3SPKsVSFY4+nhDu4B//+AeNGjUiPj7eo7oKelqpYJs2bZg0aRJms5mRI0d6RHOSqqrOcaii/ZwxZsg4VFplCVY9IKzEfbf85unj40N2drbRYQg3kZ+f7xHrq1Tm2Oxj/P7w7wC0eKUFTZ5uYnBEnqVRrUZGhyCER6tVq1ZxV8HnnnvOo0oFg4KCPKpUcPbs2TRv3pxdu3Yxc+ZMo8MRwukqS7AytdbnsF52lQbYMTFqnPr163Py5EmysrJq1OyEuHIWi4WUlBRCQ0ONDsVptNYcfvowR6YeAQWt32xNw8kNjQ7L43jqTKAQruTGG29k9OjR5OXleWSp4Pjx40lNTTU4IvsFBQWxdOlSwJpsbd+yBY4cgQ0bYN06659HjoAH/P0JAaDKSziUUhbAT2udX/jzYMAXWAUEAtlaa5dpMda9e3cdHx9f7uMZGRmcPn2a/Pz8aoxKuKOgoCAaNmyI6UqvDnUD2qL5/eHfOfmvk+AFbd9vS8R9NePq1KrmOxWdk6nomM7Yr7qO56qUUtu01t2NjqM8lY1DwjkyMjKIjY3l+PHjzJ49m6eeesrokOxmsVjo378/69ev569//SsfffSR0SE5xJTJk3n1X/+iY5MmbJ0zB9+S46xX4XqLLVtChw5V/9B2Ya4yDjljzJBxqLTK6qC2KqUKAK21/kop5e+Y8KpfrVq1PK6tqxBXQps1+8fsJ3lpMspX0f7j9oTfEW50WEIIYZeiUsEbb7yR3bt3o7V2+xnkolLBjh07smLFCu6++27+8pe/GB2WfbRm9uDBfLVyJRlZWRxLSaFlgwZ/Pl7UAejgQUhPh169PDLJEjVDZafolwFvlrivy/lZCOHCLPkW9g7bS/LSZEwBJmK/iK1xyZUzWs+W93j9+mW3aLcllqqQ1rmiphs4cCBbtmzhgw8+cPvkqojHlQru3k3QhQt88cQT/DZ/funkqiSzGU6fht27qze+alCd41BF+zljzJBxqLTKSgT9tdZ5SimL1tqklPIDsoAA4KLW2qcaY62QlGYIUTZzjpk9f93D2c/P4hXixV0+vTiddvm5lYgI+9rPuoLIyLI7FhV9wJf3mCPed0JyAl1e70LzOs35ffLv9h9QXEZKBIWtPGEWC0qXCt5zzz2sWLHC6JCqpqAAPv+8eJYqcvRgUtIvL4qKCM0h+Y0vrHe8vOD228HNmk4ZOQ4J57N1HKpoBstHa13US/OaksfWWue5UnIlhCib+aKZ3bfv5uznZ/Gu402n7zuVmVyBZ7RSdUYLXFttPbkVgJ7RPR1zQCFElfz222/06tXLY7oKLl26lKCgID7++GNWrVpldEhVc+JEqbtlJVdlbr9kP3dg5DgkXEe5CZbW2lzi582FP1qAH50dlBDCfgUZBey6eRfnvjuHT30fOq/vTK0ech2is2w5uQWQBEsIoy1atIiNGzd6TFfBZs2a8dJLLwHWdbLcslQwKanyVdYvZTZb9xPCDV1RmzStdb7W+gZnBSOEcIz8s/ns7L+T9F/S8WvoR5efuhDcMdjosDzaliRrgtUjqvwFhoUQzjdnzhwaN25MfHw8c+fONTochxg3bhx9+/YlNTXVPRcgruriwtL5Wbgpz+tDLUQNl5ucS0LfBC7EX8A/xp/OP3cmsHWg0WF5tIt5F0k8nYiX8qJLgy5Gh+OZLl40OgLhJkouQDx9+nR2e0CzBLcvFfT1rdp+PnI1inBPkmAJ4UFyTuSQcH0CF3dfJLBNIF1+6kJA0wCjw/J4SReSaBXWik6RnQj0kWTWZmaz9eKDhAT49lt4+2144QWYMgXuuQeuv966Jk5ICATLDKyw3cCBA4sXIB45cqSUChotKurPda5s5eVl3U8IN1StrVmUUj7AFOAF4EGt9WUr5ylr258ZwD2AGdgOjNVay+lLISqQfSibhP4J5B7LJbhzMB3/1xHfepefNYyIqLjDUVV4eYGljM7kJlPFZfdV3a+8Lk0mU9nHq6x7k71dn1qGtWTPhD3km6WcBbDONiUnw6lT1j9L/lzyz9Ony/4LK4u/P+Tk2BWWjEE1y7x58/jvf/9LfHw8L730Ek8//bTRIdlt3LhxrFq1inXr1jFhwgQ+/vhjo0OyTaNGsGNH8d2I0JwyG114mU6TV1CAb1HnwEaNbH6Jqownjh6DitqtV2U8ke6DnqXMBEspFVfBPm8Ds7TWjxQ+1wSs11pfb8Prjca6ftamCp4zHLgF6Ky1zlZKvQ3MAdyw6FiI6nFxz0V2DthJ3qk8Qq4KoeM3HfGpU3ZphTM+jMv7jlzZd+eq7ldexyWLpWorxpfXzbmizk5lPebj5cHlLBYLpKaWnzCV/PnCBduPGx5u/WbRoEHpPy/dVquW9VuPfWQMqkGKSgWHDh1KeLhnrPtXtABxbGwsK1euZOXKldx9991Gh1U5b2/rbPTBg2A2/9mKvdDFnBw6Pf443Zs3JztvDL5+ftbnX0GL9qqMJ44eg1JSqjYGgePGIeEaylwHSym1sZzna+A2IBVoi/XsHsABrbXNc79KqfXA4nLOHn4DrNFaLym83xn4XmsdVtExZf0RUVNd2HGBXTfuIv9MPrX71qbD5x3wDqnedUMqWm6mosGmuveryvEqUvRayZnJRAZHVu0gRsvKqjxhKpptsrULmJ9f5QlTZCTUr39F12Y4ah0sZ4xBIOOQq0pLS6Nu3bpGh+FQixYtYsKECdSrV4/ExETq1atndEiV0xo2bCj3syQtM5O6wcHWaaX69aFXryv6cK7KuOAqY1Blx6xIVV9PVI2t41CZ38K01teUtb3wwF6AAvYW/gnWxMtRYoBDJe4fAuoqpUK11umXxDIGGAPQuHFjB4YghHtI35jOrkG7MKebqTuoLu0/aY9XwBXWuQu7nMk6Q4P5DWhRtwUHJh5wjcVNLRY4c8a2Mr0rmW0KC6s4YSr6OTS06t8WjGfzGAQyDrmDksnVxYsXCQoKMjAax3DLUkGlrEnT7t3WmSwolWjVrV3bmi20bEleq1aQn49vVZtjCGGwqp7m1kBI4f7+wCmHRWRN2kqe2ii6MvWy2pDCM4xLwHrm0IExCOHyzq07x2+Df8Ny0UL40HDaLWuHyU/61lS3+CTrjEVkcKTzk6vs7MoTpuRka92IrbNNvr6XJ0plJVEREVXvBOZebB6DQMYhd7J48WKmTp3Kjz/+SIcOHYwOxy5uWyqoFMTGQtu21kWEk5Ksrdh9fKwNLRo1YteePfzfVVdx2223MXPmTKMjFqJKyk2wlFKjgAcK736ktX695ONa66zC5zn6wt8/gJKnARsDmcB5B7+OEG7r7NdnSfxLIpYcCxHDImj9dmtM3pJcGaF4geGoKi4wbLHA2bMVJ0xFP2dk2H7cunVtK9OrXdudZ5ucQcYgD7Vjxw7S0tIYMWIEGzduxMfNW4A3a9aMuXPnEhcXR1xcHH379nWPUkGwXlvVrJn1domMjAx27drF7t27GTJkCN26dTMgQCHsU16Ti0nAk8A8rLNVU5VSIVrreY4OQCkVBnwOjNBaHwTeB0YppT7SWucBk4DVuqyLxYSogVI/SWXPfXvQ+ZoGYxvQalErlMnYL8jlde+rrCdBVfdzdCfEqnZ9ghIJVvQlCVZ2tm0NIVJSwNYW0r6+5SdKJZOoiAjrdVCiUjIG1Rxz587l22+/Zdu2bbz00ktMnTrV6JDsNnbsWFauXOlepYKVuO6665gyZQovv/wyI0aMID4+Hj8bPs+qMp64yhhkyzEd/XrCucprcrEXGKe1/rHw/tXAMq11TOE1WPnAUsAPa4ng0Ko2uVBKNQI2AndqreMLjz8DuBVracYeYKLWusJTt3JxsagJkt9PZt+IfWCBho80pPm85q5xzU9NYbFAWlpxcqSTknj+4wkEp11kTNRtBKdl/pk4pV92uU756tSxbbapTp0aPdvkjCYXjhqDQMYhd7B27VoGDhyIj48P27dvd/tSQYAjR44QGxvLxYsX+fjjj92jVLASWVlZdOrUid9//52pU6dKqaBwGbaOQ+UlWBeAxlrrc4X3A4EzWuvAEgnWv4A8rAPQE1eSYDmDDGzC051cfJKD460XBjeZ1oSm05tKcuUoOTmVN4Qoutk62+TjU3nC1KCBzDZdAUclWM4i45B7GDduHK+//jrdunXziFJBgP/85z/ExcURHh5OYmIi9evXNzoku/3yyy9cf/31mEwmNm/eLKWCwiXYm2BtAv6ltf6w8P5QYJrWunPhule5Wmufwse8gTyttaEXgMjAJjzZifknOPSYtbFZzEsxNH5cupVVSutSs00Vlumdv4LLa+rUKU6YjgXmsyrtF0KbtmbULc+UTqbq1q3Rs03OIAmWcIQLFy4QGxvLsWPHmDlzpkeUClosFgYOHMgPP/zAXXfdxcqVK40OySH+/ve/8/LLL9OhQwe2bdsmXQWF4exq0w5MA9Yope7E2k3pNuBeAK21BSg+3aO1LlBKXX6VohDCblprjs04xtHpRwFo+VpLouOijQ3KaDk51mL0ytZtSkmxdqeyhbe3bYvdRkSAv3/xbuF5F+meFI+3yRsa93LSGxZCOFJISAhvvfUWTzzxBIMHDzY6HIco6irYoUMHVq1a5T5dBSsxa9YsfvrpJ8aOHesRM42i5ihzBgtAKdUVuAtra9pVWmuXPi0nZw6Fp9Fac/gfhzkx9wSYoM3SNkQOd9PFbCtTNNtkS5neuXO2H7d2bdvK9OrUqfyqZmE4mcESjmSxWDB52P97TywV9MS/J+G+7J3BQmu9Hdhe4oC3a60/d1B8QogKaIvm4MSDJP0nCeWtaLusLfXvdsOBMje39PVLFc06XelsU2Xd9CIiICDAue9PCOG2ir60a63Zv38/bdq0MTgi+40dO5ZVq1bxww8/MGHCBI8oFSyZXB0+fJjo6GibugoKYaQrWWh4DVBmIwulVNfChEwIYSdLgYX9D+0n5b0UlJ+i/ar2hN8WbnRYf9LaOotUWcJ06tSVzTaFhtpWple3rkvMNu0/s58n1j7Bzc1vZnyP8UaHI4SoArPZzL333stnn33Gtm3biI2NNToku5RcgHjVqlV8/PHH3HPPPUaH5RDLli1j9OjRPPzww8yaNcvocISoUEULDd8PpAOpwFGsq9uXfNxXa52nlIoGPgMaOTFOIWoES56FvcP2kroyFVOgidjPY6nTv071vHhe3uVJUnlJVF6ebcf08ro8SSoviXKz2aZfT/zK5/s/x8/LTxIsIdyUl5cX4eHh5OfnM2LECDZt2uT21/o0bdqUuXPnMn78eCZMmEDfvn09olSwSZMmZGdnM2fOHIYMGUL37i5bLSxEhTNYHwD7gHAgB+uCwyUdUUr1w7rK/U7nhCdEzWHONpN4dyJpX6XhVcuLjl93JLRXqH0H1draIa+yhOnUKes1ULaqVavyxW4bNICwMJeYbXKGogWGe0T1MDgSIYQ9XnrpJb755hu2b9/OnDlzeOaZZ4wOyW5FCxB7Uqlgr169ePjhh1m4cCEjRoxg27ZtUiooXFZFTS4sQLDWOkspFQyka629lFLfAsOAZOA34L+Fj71QXUGXRS4uFu6sILOA3Xfs5vwP5/EO86bT/zoR0jWk/B3y8qxd8ior00tOtl4HZQsvL+t1S5WV6UVGQmCgY964G+u2pBvbT21n/fD19Gnax+hwagRpciGc5fvvv2fAgAH4+PgQHx9Px44djQ7JbkePHiU2NpbMzExWrFjhEaWCWVlZdO7cmYMHD/L0009LqaCodnatg1V4ADPQFChqwb6uMMHKAloAvwMrsbZv76q1TnRE4FUlA5twV/nn8/ntlt/I2JhOQP08Yt8IJzAko+JZp7NnbX+BkJDKu+hFRlpnm7wMXS/cbeQU5BDyQggWbSH9yXSCfYONDqlGkARLONP48eNZvHgxXbp0YfPmzW5fKgiwePFixo8f71FdBTds2EDv3r0xmUxs2rRJSgVFtXJEgmUB/gosK9xkKkywsrEmWIeBLsAmrXUtx4RddTKwCZeVn1963aYSiZLl2Emy1h/BO+s0vpzDhI3XNplM5c82lUymIiMhKMi5768G2vTHJq556xra12vP7rjdRodTY0iCJZyp5ALEL7/8MlOmTDE6JLtprRk4cCDff/89f/nLX1i5ciXKAxZAf+SRR1i4cCFdu3YlPj7eI96TcA92t2kHJgOrtdY+hQc0F24vysiU1nqPUmqvUqqb1nqbfSEL4Ua0hvT08meYSiZTZ86UexgTUGruIzi48uuaIiMhPFxmmwxUdP1Vz+ieBkcihHCUkJAQli5dyhdffMHo0aONDschlFLFCxB/8sknfPzxx/z1r381Oiy7zZw5k1OnTvHMM89IciVcUkUJ1qvAy0qpc8COEtsv/Zf8FXAnIAmWcH/5+XD6dMUJU9HPOTm2HdNkgvr1SyVKBYH1OLGigKwztVBNG9B82XX4xTa2JljC5bWo24J72t/Djc1vNDoUIYQD9evXj379+hkdhkM1adKEefPmMW7cOCZMmMANN9zg9qWCgYGBLF++3OgwhChXZetgdQd8gXpA/8JtDwIlLwBZB0x3eGRCOIrWkJFRefvxU6cqnG26THBw5dc1RUZCvXqlZpuyDmaxs/9Ocs/kEtw1mI7/7YhvuK8T3rhwllta3sItLW8xOgwhhBNlZGTwyy+/cMst7v9/fcyYMaxcuZLvv/+euLg4jykVBGsZ5GeffcagQYOkq6BwGZUlWAe01llA8X9ErfWqwvsnC5+TCHRwVoBClKug4M9OepWV6WVn23ZMpUrPNlVUpleF2abM3ZnsHLCT/JR8al1bi45fd8Q79ErW+xZCCOFsGRkZdOzYkaSkJI/oKuippYIAjz76KAsXLuSpp55i9uzZRocjBFBxk4upwItaa3PhfbPWusyLPpRSU7XWhvbKlIuLPYTWcOFC5QlT0WxTOf9+LxMYaE2MKuumFx4O3s5JeC5su8DOG3dSkFZA7f61if0sFq8guY7K3RxKO8TBtIP0jO5J3YC6RodTo0iTC1GdPLGr4Ouvv864ceMICwsjMTGRiIgIo0OyW1FXQaUUmzZtokcPWZtQOI/dXQTLOOARrXWzyp9pDBnYXFxBgfXapsquazp16spnm8pLmEr+HFLBmlLVIH1DOrtu2YU5w0zdW+vSflV7vPwluXJHs3+ezdQfpjK552ReGfSK0eHUKJJgiepUsqvgjBkzePbZZ40OyW5aa2688UbWrl3L0KFDWbVqlUeUCj766KMsWLCAdu3asX37dikVFE7jiC6CpbhyciUMojVkZlaeMCUnQ2qq7bNNAQGlZ5vKm3WqV89ps02OlLY2jd137MaSZaHe3fVo+0FbTL4mo8MSVVTUQbBHtJwlFcKTFXUV7N+/P88//zx33HGHR5QKvvnmm3To0IHVq1d7TKngzJkz+fLLL9mzZw///Oc/pVRQGM7mGSxXJ2cOHaigwJoQ2VKml5Vl2zGVsiZEhclRyjVZHO65k1z/DPws4cR4jSMi6sE/Z5s84IwawJkvzpB4dyI6VxM5IpLWb7ZGeXnGe6upohdEk3Qhif0T99MqrJXR4dQoMoMlHCkl5UMOH55Kbu5x/PwaExMzi4iIBy57XlxcHP/5z3+kVNDFSamgqA4OLxF0dTKw2aCs2aayZp1SU8Fise2YRbNNlZXp1asHhYNSSsqH7N8/Bovlz+TMZAqkdeslZQ5u7ur0itPsHbYXXaCJmhBFy1dbokySXLmzkxknabiwIaF+oaT9Iw2TkpnI6iQJlnCUKxmHMjMziY2N5ejRo6xevZohQ4ZUd7gO5+mlgnfeeSdr1qwxOhzhgRxeIihclNlsTYhsKdO7eNH245aYbaqwTK8Ks02HD08tNagBWCxZHD481WMSrFNvn2L/qP1ggUZPNCLmxRiPGLxquq1JWwFreaAkV0K4rysZh4KDg3n33XdJT09n8ODB1Rmm0xSVCsbGxrJ69WpWrFjBvffea3RYdps5cybh4eE8/PDDRociajhJsFxVZmblCVNysrVxhK2zTf7+lSdMkZHWxhFOLIHIzT1+RdvdzcnXTnJw4kEAms5oSpNnmkhy5SGKrr/qGdXT4EiEEPa40nHo+uuvd2Y4hihagHjs2LFMnDiRG264we1LBQMCAnjqqaeMDkMISbCqVdFsU0UJU9HPmZm2Hzc83LYyvVq1XOLaJj+/xuTmHitzu7s7Puc4h588DEDz+c1p9EgjgyMSjnQi4wQgDS6EcHf2jENbt25l586djBo1yhmhVavRo0ezcuVK1q5dy/jx4/nkk0885oRgeno68+bNY+rUqfj7+xsdjqhhJMFyhIsXbWs/fiWzTX5+ti12GxHh1NkmZ4iJmVVm7XtMjKFLqdlFa83RaUc5NvMYKGi1uBVRY6KMDks42PtD3uflm14mwCfA6FCEEHao6jh09OhRrr32WpRS9OzZ02O6CsbGxrJmzRqPKRUEuOuuu1i7di0FBQW88MILRocjahhpclEes9m6kK0tZXoXLth+3LCwyhe7jYyE0FCXmG1yFlu7N7kDrTWHHj3EHwv/ABO0ebcNkcMijQ5LCI8jTS6EI1V1HJowYQKLFi3yqK6CS5YsYezYsR7VVXDjxo306tULpRQbN26kZ08p7Rb2c8kugkqpPsBCrDNnecBErfWmS57TDfgOOFBi80qt9fyKjm3zwJaVVXnCVDTbZDbb9sb8/CpPmIpmm3x9bTtmDeZOyZc2aw7EHeDUklMoH0W7j9pRb2g9o8MSTlBgKcDbJJP+RrI3wXLmGASSYHmSisahkl0F//nPfzJt2jSDo7Wf1pqbbrqJ7777jiFDhnhMqeBjjz3G/PnzadeuHdu2bZNSQWE3l0uwlFK1gUPAbVrrjUqpvsAKoJnWOqvE824C7tJaj76S43fv3FnHv/de5WV6VzLbVLeubWV6tWt79GxTdXKnFu6WAgv7Ruzj9IenMfmbaL+6PWGDwowOSzjJpK8n8eXBL1lw4wKGtHX/Ns3uyJ4Ey9ljEEiC5SlsGYfWrVtHv3798Pb2Jj4+nk6dOhkVrsMcP36cDh06cOHCBZYtW8Z9991ndEh2y87OpkuXLuzfv58nn3xSSgWF3VwxwforMEVrfW2JbQnAs1rrL0psewCYDSQXbloLvKi1rjAz6q6UtmlY8/WtvItegwbWTnp+flf6NoWdNm5sWs6Fx0245pqj1R9QOSy5Fvbct4cza87gFexFhy86UKdvHaPDEk501ZtXseXkFr578DsGxAwwOpwayc4Ey6ljEEiC5SlsHYeKSgU7d+7Mli1bPKJU8I033mDMmDHUrVuXPXv2SKmgEJdwxXWwYrCePSzpUOH2klYDy7TWWilVF/g38B5w2SljpdQYYAxAV5MJ2rSpfLapTh2ZbXJh7tDC3ZxlJvEviaR9m4Z3bW9iv4kl9OpQo8MSTpRnziMhOQGA7lEuewmQqJjDxyAoPQ41buz+nVCF7ePQnDlz+Oabbzhy5AiJO3fSOSwMkpIgL896MjcqCho1Am/3KS0eNWoUK1eu5LvvvvOYroLXXHMNjzzyCPPnz+fbb7+VBEtUi+r8X6+ASy9qKgBKrdaptc4u8XOaUupR4IRSKqDkY4WPLwGWgPXMIXLm0O25egv3ggsF/Db4N9J/TMcn3IeO33UkpHOI0WEJJ9uVsos8cx6tw1pT27+20eGIqnH4GFT4nNLjkHB7to5DwcHBrFq5kohz54g+fhyOHy997XZKCuzYAS1bQocObnFyt6irYIcOHVizZg0fffSRR5QKPv/889x8880MGCDVB6J6mCp/isP8AVz6Lblx4faKeAE5QK4zghKuJSZmFiZTYKltrtLCPf9cPjsH7iT9x3R8o3zp/FNnSa5qiK0ntwLQM1rOfLoxGYOETWweh7Sma3Y20RkZ1sTq0sZYRdsOHoQNG8BNujY3btyY+fOtPV0mTpxIcnJyJXu4voCAAEmuRLWqzgTrM6CjUioWQCnVE2gD/KCU2qCUalm4/d7Ci5FRSnkDLwDva61tXEBKuLOIiAdo3XoJfn5NAIWfXxOXaHCRdzqPhBsSuLD5Av5N/enycxeC2gYZGpOoPluStgDQI0oWGHZjMgYJm9g8Du3eXdxx2GyxMPfzz3nx008vP6DZbH3e7t3VEr8jjBo1ioEDB5KWlsb48ePxlCV9ADZt2sSQIUPIyckxOhThwaotwdJapwN3A0uVUluAl4FbgECgCVB0EUsA8L1SaivwK9YLjR+prjjF5VJSPmTjxqasX29i48ampKR8aNN+CQkDWL9eFd8SEv48eyNQEoEAABdUSURBVFTVYzojzsrknswloU8CF3deJKBVAJ1/6kxAjCw0W5NsOWlNsGQGy33JGOTeXG4cKiiwzkwVzlrFHzrEf099QLPbl7E+/B421o0jxe/nP59fNJNVUGD7axioqFQwJCSETz/9lI8++sjokBzCYrEwatQoPv30U6ZPn250OMKDyULDokJVbZuekDCA8+e/v2x77dr9adBgZLnHBKr0es5q7559JJud/XeScySHoNggOn3XCd8IWcuspll7eC2b/tjEY9c+hr+3rKNiFFlouGZyyXHoyBHr9VWFCVaK38/8FvAa3j5/TnSatC+tL4wlIre3dYOXF3TpAs2aXfkvwSAluwomJiYSGRlpdEh227RpE7169QLg119/5aqrrjI4IuFOXK5Nu7PJwOYcVW2bvn59+Rfz+vk1KfeYQJVezxnt3bP2Z5HQP4G8k3mE9Aih47cd8anr/m14hXBXkmDVTC45Dm3YYO0YWBRj3Thyvc5cvo85nGvSFv25ISoKCr/cuwOtNTfffDP/+9//uPPOO1m9erXbdxUEeOKJJ5g7dy5t2rRhx44dsgCxsJmt41B1XoMl3JAz2qZXdMyqvp6j48zclcmO63eQdzKP0N6hdFrbSZIrIYQwgEuOQ3l5pR8znS17n0u35+fbHqQLUErxxhtvFJcKLl++3OiQHOKf//wnrVu3Zt++fVIqKJxCEixRofLao9vTNr2iY1b19RwZZ8aWDBL6JpB/Op86N9ah47cd8a7lPuuYCMea/+t8XtrwEqcunDI6FCFqJJcch3xLl4r7WcLK3sd8yXY3XIy4cePGLFiwAIBJkyZ5TFfBd955B5PJxNy5c9m8ebPRIQkPIwmWqFBV26bXrt2/3O0VHbOqr+eo9u7nfzrPzv47KThXQNgdYcR+HotXoNcVHUN4lle3vMo/1v6DcznnjA5FiBrJJcehqCjrNVVFMV68D5MunXTl5SrCUu/4c4OXl3U/N/TQQw9x4403elRXwauvvppHH32U0NBQTp8+bXQ4wsNIgiUqFBHxALVqXVNqW61a1xRf6Fteh6bOndcSENCu1H4BAe3o3HlthS1wIyIeIDJyONalZwC8iIwcXmmjCke0d0/7bxq7bt6FOdNM/fvq035le0x+8l+kJkvJTOF4+nGCfYNpHdba6HCEqJFcchxq1Kh0jLm9aX1hLH7mcNAK77y6xOZMoJX3TaXfzCX7uQtPLhVMTExk8ODBRociPIw0uRAVOnAgjqSk/1y2PSpqPFlZB6rUocmIboCVSf00lT1/3YPO00Q+FEnr11ujvNz/Ql5hny8PfMng5YPp27Qv64avMzqcGk+aXNRMLjsO/fZbqVbt5dFaY1EKrzZtIDa2wue6ujfffJPRo0d7VFfBkgoKCvD2lksCRPmkyYVwiKSkJeVuL2tQAzh//nsOH55aanACsFiyOHx4aoWvV9X97JGyLIXEuxLReZroydG0XiLJlbAqXv8qSta/EsIoLjsOdegA9euXKhW81B9nz3Lriy/y/DffWJ/v5h566CFuuukm0tLSGDdunEeUCoI1sZozZw49evSQBYiFQ0iCJSpR3pm5is/YuUo3wMokvZnE3mF7wQyNn25Mi5dboEySXAmrogSrR3QPgyMRoiZz0XFIKWvL9ZYtrUnWpYmWtze/nz7NNzt2MOv990nYubPC13UHJUsFP/vsM48pFSwoKOCdd94hISGB5557zuhwhAeQBEtUorwzcxU3fnCFboCV+eOVPzgw+gBoaDarGTGzYjxifQ/hGFprtiZtBaBntMxgCWEcFx6HlLKW/d1+u3UR4agoqFfP+mfnzvR9+mkmTpxIQUEBw4cPJ++S9u7uqFGjRh7XVdDf35+3334bk8nEvHnz2LRpk9EhCTcnCZaoUFTUmHK3V7VDU0Uc1Q2wMsdmH+P3h38HoMUrLWjydBOHHl+4v8y8TPo06UOXyC40quWeF6YL4QncYhzy9oZmzawzWn37Wv9s1gy8vXnxxReJiYlh165dzJrl2LHMKJ5YKljUVdBisTBy5EgpFRR2kQRLVKhVq0VERY2nZDelqKjxtGq1iM6d1142uNWu3b/SDk0VcUQ3wIporTn89GGOTD0CClq/2ZqGkxs65NjCs4T4hbD6r6vZPna7zGwKYSB3H4eCgoJ46623AJg9ezY7duyo0nFcSVGpYK1atfjss89YtmyZ0SE5xIwZM2jTpg379u1j2rRpRocj3JgkWB4oJeVDNm5syvr1JjZubEpKyoc27XfgQBzr13sXtrr15sCBOADOnfuRP2vdzYX3rc6f/6nUMUre379/Irm5xwBNbu4x9u+fWPzYhg3RpdrqbtgQXaX3WlJl71tbNL9P+Z3jLxwHL2j7YVsaPNTA7tcVQgjxJ0ePQeAe41BF77tv377FpYIjRozwyFLBU6fcfzH2kqWC8+fPl1JBUWXSpt3DVLXNeXltcE2m2lgs5y/bHhDQjuzsg0B+GUfzwWQKKnM/k6k2Xl6B5OcnXb6XTxQtWrzklLa62qzZP2Y/yUuTUb6K9h+3J/yO8HKPJ8T/t3f/wVHXdx7Hn+/8Rn5TqBQ5oVpUIiSICnQoAQdqAavXa0tbT9sOTitnq9z16virwIh2OLmpet7ZXo9zRC32tNM6/Lip+LOod4KeJxSIFRV/HAXlZwolJiQh7/tjN7rZzSabzTf7/e7m9ZjJ6H6z3+++9zNJXnx2P/v+vLLvFcYNG8fgisFhlyJxatMefUFn0KhR11BX9xwNDa+lfC9KOZTJ866vr2fy5MksWLCAZcuWUVZW1uG18om7M2/ePJ544gkuu+wy1q5dWxDv+N9www1s2bKF1atXc+aZZ4ZdjkRIpjmkCVaB2bx5bPzVuvbKy8fw2c++m/a8TZtK6KojUy6Ul4/Jqv7OnvfUC97m9W+9zoFHDlDUr4gJaycw7OJhQZYtBaaltYVB/zCIxpZGjtx4hCEVQ8IuSdAEKx8En0HFaY73nmxyKNPn3djYSEVFRVClRsKePXuYMGECx44dY82aNVxxRe/tWZkrTU1NlJSUUFSkhV7SnvbB6qOyb3Me/uQKeqetbu1XaznwyAGKBxZT9USVJlfSpdoDtTS0NHDG0DM0uRLphuAzKPfZlM1zyPScxMnVkSNHtFQwosrKyj6aXLk7+/fvD7kiyTeaYBWY7Nucd97uNleCbqtbdHQkh9cfpmRoCdXPVDNkhv6xLF37aINhtWcX6ZbgMyj32ZTNc+juOU8++SSVlZUF01XwqquuYu7cudTV1RVMV0GAgwcPMnfuXGbOnElDQ0PY5Uge0QSrwGTbljZdG9yioo4nJP36VQKlaa5Wmva8oqIhlJaO6vis0lGBttWlqYLWf7mK0k+WMmnTJAZdOKjTa4i00QRLJDtBZ9CoUVfH8yZVlHKou+eUl5ezf//+guwquH79eh5+OLPGJlE3cOBA9uzZw65du7QBsXSLJlgFpqv2sum6HKVrg1tTU5cSbv36VTJ1ai2zZjWR+iNUxKxZTdTU1HVYX01NHdOn700Jt9LSUUyfvjewtrpWNxL+8YeU77qE8144jwFVA7ocO5E2L++LTbAuHHVhyJWI5JegM+iss37G1Km1kc+h7p4zc+ZMrrvuuoLqKjh69GjuvvtuABYvXlwQSwUrKip44IEH1FVQuk1NLvqQbLs7ZXO9d99dkbbr09Sptdk9gQyc+OAE2z+/nfqd9VScUUH1M9X0G9uv1x5PCk99Uz2D74h1Djx28zFOKT2lizMkV9TkIr8FnUFdXTOsHMpUfX091dXV7N69m2XLlrF8+fKwS+oxd2f+/Pls3LixoLoK3nTTTaxcuZKzzz6brVu30q+f/l3RV6nJhaR4++0ftQshgNbWD3n77R8Ffr2OQg1IezwIjXsa2Vazjfqd9Zxyzimc9/x5mlxJt+04sIOTfpKJp07U5EokQEFnUFfXDCOHuqN///7cf//9QGFuQFxISwVvvfVWxo8fr6WCkjFNsPqQ7Ls75eZ6PdGwu4GtM7bS8GYDAyYNYNLzkyg/rTzndUj+mzZ6Gu//8H0e/NKDYZciUlB6IzOilEPZqKmpYfHixbS0tLBw4UJOnoxGR9+e6AtLBfVOtXRFE6w+JPvuTrm5XrbqX6tn64ytnHjvBIOmDaL62WrKRuT/Bo4SnpEDRlJ1alXYZYgUlN7IjKjkUE+sWLGC2bNnc+edd1JcHI2Ovj21cOFC5s2bR11dHYsWLSqIroJTpkzhlltuYcmSJVRVKR+kc5pg9SHZdnfK5nqdd30Kzp+3/pltM7fR9H4TQ2YNoerJKkqHpusqJSIiYQk6g7q6Zq5yqKf69+/P008/zezZs8MuJTBmxqpVqxg0aBAbNmwomKWCt99+O8uXL6esTC/iSuc0wepDsu3Ql831Ouv6FJSjm4+y7aJtNB9qZti8YUz87URKBpYEdn3pew5/eJiz7z2bRRsWhV2KSMEJOoO6umYucqg3bN68WV0F88ChQ4eorY32z5KEJ6ddBM1sJnA3UAI0Ade6+5ak+xhwG/A1Ylu4vwoscvf6zq6t7k19S93v6thx6Q5a61sZ/uXhVP6ykqJyvV4gPbPxrY3Me3genzv9c7yw8IWwy5EkPe0i2JsZBMoh6bk77riDm2++maVLl3LbbbeFXU6PuTuXXHIJjz/+OJdeeinr1q0riK6C27dvZ86cOQwdOpRt27apq2AfErkugmY2BHgM+L67VwHXA+vMLLlN17eB+cAkd68EmoGVuapTou/wbw+zY35scnXqladS+agmVxKMjzYYHqUNhguNMkjywfTp0zEzVqxYwauvvhp2OT3WtlRw8ODBbNiwgTVr1oRdUiDOOussRowYwRtvvMHSpUvDLkciKJf/Kv0CsMvdNwO4+ybgfSB50fHXgX9z94b47XuAy3NVpETbwd8cZOeXdtLa2MqnFn2Kcx48h6ISTa4kGG0TrAtP0wbDBUgZJJE3Y8YMFi9ezMmTJwt2A+J9+/aFXFHPVVRUsHr1aoqKirjrrrt48cUXwy5JIiZnSwTN7Gag0t2/mXDsN8Dz7n5PwrFdwPfc/Zn47YHAMWCIux9NuubVwNXxmxOAnb37LPLOcOBQ2EVEjMakPY1HKo1JqqiMyRh3H5HNib2RQfHvK4fSi8rPTZRoTFJpTFJpTFJFZUwyyqFcdgQwYuvZE7WQ+i5a8v1a4v9NeZvC3VcBqwDM7JWerM0vRBqTVBqT9jQeqTQmqQpkTALPIFAOdUbjkUpjkkpjkkpjkirfxiSXa6v+CCRvTHF6/Hhn9zsdOA78qfdKExGRAqcMEhGRnMjlBGsdUGVmEwHMbApwDvCsmf23mY2L3+8XwHfMrG2TgeuAx7wQdqkTEZGwKINERCQncrZE0N2PmtkC4H4zc2LLLuYDpwBjgMHxuz4EfAZ42cxagNeAazN4iFXBV533NCapNCbtaTxSaUxS5f2Y5CCDoADGKWAaj1Qak1Qak1Qak1R5NSY53QdLRERERESkkKm/tYiIiIiISEA0wRIREREREQlI3k+wzGymmb1qZtvN7BUzmxZ2TWEzs1Izu97Mms3sG2HXEwVmtsjMfh//GdluZt8Lu6awmdnt8TF5Of47dE3YNUWBmZ1rZkfM7Nawawmbme0ws/8xsy3xr2fDrimKlEPtKYNSKYNSKYPSUw59LF9zKJf7YAXOzIYAjwFfdPfNZjYLWGdmn3b3D8OtLlTfBRzYEnYhUWBmxcA4YLq7Hzez04C3zGydu+8Nubww1QEXuHuzmY0A3jGzje7+TtiFhSX+N+WnwH+EXUtEDASq3b017EKiSjnUIWVQAmVQWsqgDiiHUuRlDuX7O1hfAHa5+2YAd98EvA/MDrOosLn7z9z9TlI31eyT3P2ku1/v7sfjhw4DTUBxiGWFzt3vcvfm+M2xxPb6ORJeReEysyLgQeAW4GDI5UTFMOA5M9tqZr8ys0lhFxRByqEkyqD2lEEdUwalUg51KC9zKN8nWGcAu5OO7Y4fF0nnn4BH3f3/wi4kbGY2zszeBJ4ArnT3o2HXFKIfA0+5+4thFxIhp7r7DOB8YvtIPW1myZv19nXKIekuZVCcMiiFcihVXuZQvk+wjNRXyFrI/+clvcTMfgycRub72hQ0d3/T3ccRe7X9ITM7N+yawmBmXwFOd/d7w64lSty9If7fVnd/GPhfYu/YyMeUQ5IxZVB7yqCPKYc6lq85lNefwQL+CMxJOnY68OsQapGIM7OfAGcCX3H3prDriRJ332pmm4GLgNqw6wnBPGC8mbV9ZmQ0xD5o7O4LwisrcoqBY2EXETHKIcmIMig9ZRCgHMpUXuRQvr/Ctg6oMrOJAGY2BTgHeCrUqiRSzKzIzH4O/AWwQMEGZjbRzL5mZha/fRowFXgl3MrC4e7fcffz3X2au08D7gPu68uhZmYXmtn5CbfnA+OJLeWRjymHpFPKoFTKoFTKoVT5nEN5/Q6Wux81swXA/WbmxJZlzHf3P4VcmkTLfGARsT/c/xX/ew6wxN2fDq2qcL0H/A1wo5k1A2XAUndX1y9pcxy4y8xGAieIffj8Yv19bU85JBlQBqVSBkkm8jaHzN3DrkFERERERKQg5PsSQRERERERkcjQBEtERERERCQgmmCJiIiIiIgERBMsERERERGRgGiCJSIiIiIiEhBNsEQiwswuNrOXzOy4me0ys8VJ399iZtd285pte4xcEG8h3Z1zHzAz7SgvItIHKINEgqMJlkgvMLMlZuYJX9cl3XYzuzLh/pOBtcBPgTHE9ky5ycyuzvDxHjazloSvZjNrBa7p5JwzzOwpM2s0s91m9o2ePWsREYkCZZBIuDTBEukdK4GBwGTAgZ8DXwfGxo+/DhxNuP8C4HF3f8jdD7v7JuAnwBUZPt43gYr4VzlwLrFN+dZ1dGczKwceB/YBE4DbgAfMbEbGz1BERKJKGSQSopKwCxApRO7eDDSbWRlQ5+7NZvZLYI67v2dmw4GDCacUASeTL9ONx2sFWgHiO57/CrjV3femOeUvgU8C33X3JuAtM6sBfgC8kOnjiohI9CiDRMKld7BEetdwYGviATMbADQC7yQc/jXwRTP7qpn1N7NpwN8Dj2b6QGZWZGZXAC8B6919ZSd3nwK8GA+2Ns/Ej4uISGFQBomEQO9giQTMzN4ltoY98VjbK4G/Szj8Qfzzv3/l7mvj689vB9YAHwD3Av+aweP1B5YRW+KxH/h2fHlHZwYBh5OOHSK2dERERPKUMkgkfJpgiQTM3cdmed56YH0W59Wb2ZvApcAuADNL/t1+D7g84fZeUl8pHB4/LiIieUoZJBI+TbBEepGZLQB+kXT4eXe/OM39iwFLc7lLgIaOvuHu98XP/yDNuSXAJ4BH4rd3Aj8wswp3b4wf+zxQm+Z8ERHJM8ogkXBogiXSu/4T+EzSsRNmZu7e0QeITwDFnVzvHuDv0n3T3Ud2dNzMJtF+Hf4GYks5/t3MlgPTgb8G5nTy2CIikl+UQSIh0ARLpJfE18GPTjjkQAtQBhwws08nvHIXu4N7iZn9MzDA3a+KX6c/cByodPc/dPGYGXV9cvcmM5sLrCL2SuJe4Fvuru5NIiIFQBkkEh5NsER6SfI6eDM7F7gRqAYuTw62BE3E9hFp0/Z72pLhQ18A/D6D+t4htiRDREQKjDJIJDyaYIn0EjP7BLHNFi8itnZ9PLFNGO8GRpvZMWBv4jKN+AeDW4CKhA8JtwWdm1mxuyfvVZKsmDS/22bWGt+vRERECpgySCQ82gdLpBeY2RLgD8ByYssxvk+sLe1k4Cjwt8T2Cjk/4ZyxQDOxVxi/HP//ZmLr1AHeBJ7L4OFfIvZB5I6+Ovxgs4iIFA5lkEi4rOPPOIqIiIiIiEh36R0sERERERGRgGiCJSIiIiIiEhBNsERERERERAKiCZaIiIiIiEhANMESEREREREJiCZYIiIiIiIiAdEES0REREREJCCaYImIiIiIiATk/wFYgMkVWsuOtgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed2548b2b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Bad models\n",
    "x0 = np.linspace(0, 5.5, 200)\n",
    "pred_1 = 5*x0 - 20\n",
    "pred_2 = x0 - 1.8\n",
    "pred_3 = 0.1 * x0 + 0.5\n",
    "\n",
    "def plot_svc_decision_boundary(svm_clf, xmin, xmax):\n",
    "    w = svm_clf.coef_[0]\n",
    "    b = svm_clf.intercept_[0]\n",
    "\n",
    "    # At the decision boundary, w0*x0 + w1*x1 + b = 0\n",
    "    # => x1 = -w0/w1 * x0 - b/w1\n",
    "    x0 = np.linspace(xmin, xmax, 200)\n",
    "    decision_boundary = -w[0]/w[1] * x0 - b/w[1]\n",
    "\n",
    "    margin = 1/w[1]\n",
    "    gutter_up = decision_boundary + margin\n",
    "    gutter_down = decision_boundary - margin\n",
    "\n",
    "    svs = svm_clf.support_vectors_\n",
    "    plt.scatter(svs[:, 0], svs[:, 1], s=180, facecolors='#FFAAAA')\n",
    "    plt.plot(x0, decision_boundary, \"k-\", linewidth=2)\n",
    "    plt.plot(x0, gutter_up, \"k--\", linewidth=2)\n",
    "    plt.plot(x0, gutter_down, \"k--\", linewidth=2)\n",
    "\n",
    "plt.figure(figsize=(12,2.7))\n",
    "\n",
    "plt.subplot(121)\n",
    "plt.plot(x0, pred_1, \"g--\", linewidth=2)\n",
    "plt.plot(x0, pred_2, \"m-\", linewidth=2)\n",
    "plt.plot(x0, pred_3, \"r-\", linewidth=2)\n",
    "plt.plot(X[:, 0][y==1], X[:, 1][y==1], \"bs\", label=\"Iris-Versicolor\")\n",
    "plt.plot(X[:, 0][y==0], X[:, 1][y==0], \"yo\", label=\"Iris-Setosa\")\n",
    "plt.xlabel(\"꽃잎 길이\", fontsize=14)\n",
    "plt.ylabel(\"꽃잎 폭\", fontsize=14)\n",
    "plt.legend(loc=\"upper left\", fontsize=14)\n",
    "plt.axis([0, 5.5, 0, 2])\n",
    "\n",
    "plt.subplot(122)\n",
    "plot_svc_decision_boundary(svm_clf, 0, 5.5)\n",
    "plt.plot(X[:, 0][y==1], X[:, 1][y==1], \"bs\")\n",
    "plt.plot(X[:, 0][y==0], X[:, 1][y==0], \"yo\")\n",
    "plt.xlabel(\"꽃잎 길이\", fontsize=14)\n",
    "plt.axis([0, 5.5, 0, 2])\n",
    "\n",
    "save_fig(\"large_margin_classification_plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Sensitivity to feature scales"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saving figure sensitivity_to_feature_scales_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA1gAAADfCAYAAADrwnJVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzs3Xd8VFX+//HXSSOEkAAJnYSEEGoogVRRwYq6CiKKAmv57Sq4umtZK+K6u7DYy9e1o4sVEUWlqIsrKKKYSQgESEAgtEBooaf38/vjZq5zkxlIIMmkfJ4+8iBz58ydcyXk3PecprTWCCGEEEIIIYQ4dx7uroAQQgghhBBCtBQSsIQQQgghhBCinkjAEkIIIYQQQoh6IgFLCCGEEEIIIeqJBCwhhBBCCCGEqCcSsIQQQgghhBCinkjAEkKglFJN6TxCCCFaPmkzREslAUsIB0qpB5RSp90cTinVXinVt9pXf6VUJ6XUGKWUVkr1UkrdppQqd/LaF5RSh5RSxUqpZKXUxdXK3HamOlSV6+OkHo716enkNflKqfJqXxp4xaGM/RrCXLxvhJNz2M/zXF2vQwghWjtpe87c9lQ7X3BV2epffz7Ta4VoDF7uroAQTUxX4OgZykwE3nV4XAF4AncAO1y9SCnlCXxT9R53A/uBScAKpdRYrfV3dazrFqCNk+OVGB+e3IND41Uljt8+WKkEBgKLgB/r8L77gOEOjyuAaGA+sKYO5xFCCGGQtqcWlFLdgXYY132NkyKHlFJ9q77fqbWWD/mEW0jAEsLqQsBTKRWktT7mosz7wEdV3+uqr3Lg+BnOPQE4Hxiktf616phNKdULeAEYWse6BgKOwys0RgO2AugCvF39BVrrLY6PlVKPAHuBZbV9U611KZBR7TwPADnA8tqeRwghhEnantp5G/hdLcu2BYrreH4h6oUMERSiilIqDogHyoCXXZXThvKqrwqgN0Zjs+0Mb9EPyHdo4OxSgP51ra/WukRrXWz/wvhEcTEwCJhQdcwlpdR1wM3Aw1rrYqVURtWQjR/qUg+l1EjgNmCO1rpIKXW06jzvnv6VQgghpO2pfdujtb5aa63sX1Xv3RkIcDxe9SXhSriN9GAJASilOmIEguXAg0CqUmqz1vopJ2Uvx/jEzxPwA2KAAqAUqDH23EEm4K+UGqC13upwPI4zN5Bnqv+FwDwgHBijtc44Q/mRGJ+GKsCn6vA1GI1VXNVztXnfEOALjP8Xe6sOJ2D8bpkAPFmnCxFCiFZE2h7g7Nqe6RhDEQdQ1VmglMoBlgKPaa2P1PlihKhH0oMlWj2l1CCMceCewG1a683ALcAspdSzSinvai8JA0Zg/GL3A0ZjjAnfzm/DN5z5AmOO0hdKqfFKqTil1LPA9cBDZ1n3vkqpRcAqIAlYDXxZNWHaz8VrpgI/YTToU4B3lFIPAnuqGt+9zl7n5DxDqt7vEPB3YL5S6lqt9Y6q8xw8m2sSQojWQNqes257JgBvVH31xghnQcDvgYuAD87mmoSoTxKwRKullBqolJoPbMKYXDxGa30YQGv9GcYk4OnAOqXU5KqJwmit52qtp2it/x/wK8ancFGAN3Cpq/erGtJxFUbjMhejoRkNXKa1/rYO9fZWSk1USn0NbAV6AedprW/WWl8E/BV4AMhWSr2slOpT9brYqtd8BLwDTNZaLwD+HzAb+LmWqze1UUo9DNgwJlZforWeBTyP0YD/WykVUNvrEUKI1kTanrNrexyUYCzwcRw4VjUv+GTV46Kq54VwKwlYojVrCwzGWIHpEq31IccntdZfYqyWtwljjpFlvw6lVAzGSklPaa03a63LMX7pu6S1zgUeA0ZVvfdUIMtxmVvgMPDMaU7jBTwK5AMXaa0TgEyl1PNKqf5a6w8wPul8EGPs/amq1z2CMd7+Gq31PVX1RWv9McZQk52coddJKeUB/AI8gdEwXqW1zq86z9+BycAVQPvTnUcIIVoxaXuoW9tT7Vq+wfj/ci9wXCmVjxGqPgW+A26t7bmEaChKVrAUou6UUmP57Zf5DfalYJVSYzAm6oZgfKL4jtbaq9prz8f4BPG0qibw1rY+fTHG2V+ptXa6kp9SyhcotzduVcc8ge6AL8Yk6ENV9fsKGKq1rjFkQykVDRzRWme7eB9PrXWFUuo24N26XIcQQgjXWmvbU1W+s4tqtMWYh+YsZOZprQtqez1C1BfpwRKiDpRSbZVS7wD/BT4DbqzrPhta65+drHbkuCrSzQ1R96oVn8rBCElKqWUYn0Tuw2ggDyqlTgDTgCHOwlXVedLs4Uopdb5Sqmu150/7SaoQQoi6kbaHEIxeLmdfu4BsF8/9pSGuSYgzkVUERauklOoC1HWeUKXWepdS6ghws9Z6/lm+d60+RXTxWj8g1MlT9mO9lFIDnDy/X2udV3WOvsDPVV9XAulALtAJSAT+BfyklBqotS46Q5V+whjm8o6T5wowNrQUQgiBtD2cfduTheuh59uATzAWW6pO5mMJt5CAJVqrZ6n7OO0SwFdrPaOe6nABxqdudRHH6fcKqbHBY5XJGA0QGJOb/TAa6hyHMoeBxUqpIozJ0EMw9kk5K1WTtT8729cLIUQLJG3PWbQ9Vb11+c7eQBl7aJXZ5wML0RRIwBKt1R+A2+v4mvqesNgGY/y5U0qp9vZP/swKaL2KahOez8JqjJWW5imlnsP4FDEP41PEeOCfwAFgcy3P183FJ5d2+6tfhxBCtFLS9tRf2yNEkyUBS7RKWutKoNLN1VhxhudvABbV95tqrTOrNoecBfyP3zZ7BKOx+xqYWIeJwbOrvlxpkOsQQojmRtqeem17hGiyZBVBIVoxpZQXDis5AYfqOnFaCCGEqAtpe0RLJwFLCCFEq6WUmg7cBZRhfKL+ptb6dffWSgghRHMmQwSFEEK0SlV760QCo7TW+UqpnsAOpdQSrbWsgCmEEOKsSA+WEEIIgbkh6mFOsw+cEEIIcSbNNmAFBwfrsLAwd1dDCCFEA1u3bt1RrXXnhn4fpdSbgIfWelq149MwNkHFx8dnpK+vLwUFBVRUWPfU7tSpE+Hh4QBUVlZSVlZGmzZtGrraQgghGsC5tD3NdohgWFgYqamp7q6GEEKIBqaUymqE9/gX0BOYWP05rfVcYC5ATEyMTk1NpbKyku3bt2Oz2cyvO+64g7vvvhuAr7/+mquvvprg4GASEhLMr9jYWAIC6rrPrBBCiMZ2Lm1Psw1YQgghRH1QSj0PRGAsEV1am9d4eHgwYMAABgwYwG233Vbj+RMnTtClSxdycnL46quv+Oqrr+zvRVRUFGvXrjV7t7TWKHWuWwwJIYRoKiRgCSGEaJWUUh7A60BH4AatdXl9nfv3v/89U6dOZc+ePZZerrS0NEpLSy1DB4cNG0a3bt3MXq74+HiCgoLqqypCCCEamQQsIYQQrdVVwHQgFfjZoRfpca31mTZjPSOlFOHh4YSHhzN58mQAioqKOHDggFnm0KFDpKenk56eznfffWcej4yMJDExkXvuuYeRI0eea1WEEEI0IglYQgghWiWt9VdAo47Na9u2LREREebjbt26sW/fPpKTk81ertTUVDIzM8nMzOTWW281yy5YsICNGzeaPV3dunVrzKoLIYSoJQlYQgghhBv16tWLXr16MXGisb5GWVkZGzduxGazERsba5ZbuHAhS5YsMR/37t2bxMREEhISuPDCC4mOjm70ugshhKhJApYQQgjRhHh7exMTE0NMTIzl+F/+8heioqKw2WwkJyeTlZVFVlYWn3zyCRMnTmTRokUA5Obm8t///peEhARCQ0NlAQ0hhGhkErCEEEKIZuCSSy7hkksuAaCiooItW7aYwwpHjx5tlrPZbNx0000AdO/e3bJM/MiRI2nXrp1b6i+EEK2FBCwhhBCimfH09GTIkCEMGTKEO+64w/Kcr68vV111FTabjYMHD/Lll1/y5Zdfmq87cuQIHTt2BCAnJ4fOnTtLL5cQQtQjCVhCCCFEC3LhhRdy4YUXorUmMzPTskx8QUGBGa4AzjvvPE6ePFljM+TAwEA3XoEQQjRvSmvt7jqclZiYGJ2amuruagghhGhgSql1WuuYM5dseM297SkvL8fLy/hsNT8/n8jISA4dOmQpo5Ri0KBBPPnkk4wbN84d1RRCCLc7l7ZHerCEEEKIVsIergD8/f05cOAAWVlZll6u9evXs3nzZstmyG+++SaLFi0iISGBxMRE4uPjCQ4OdsclCCFEk9eoAUsp9RBwM1AEtAU+0Vo/qZSKAt4C2mPsSTKjan8SIQQwfz7MnAl790JoKMyZA1OnurtWQojmTilFWFgYYWFh5sIYxcXFbNiwgaioKLPcypUrzS+7vn37kpCQwGWXXcYtt9zS6HUXQoimqtECllLqAuA+YKjW+phSqhOwXim1DngNmKm1XqiUGgisUUoN1VpnN1b9hGiq5s+HadOgsNB4nJVlPAYJWUKI+ufr60tCQoLl2CuvvMKUKVPMXq61a9eyY8cOduzYQW5urhmw8vPzmT17tjmfq3v37u64BCGEcKvG7ME6UvVnR+AYEAB4AMUYvVmfAmitf1VKrQYmAK80Yv2EaJJmzvwtXNkVFhrHJWAJIRpDt27dmDBhAhMmTACMzZAzMjKw2WyEhISY5dauXcuzzz5rPg4NDbUsoBETE4O3t3ej118IIRpTowUsrfVWpdQ0YK1S6gjQBfgD0APYpa2rbewE+lQ/R9Xrp4HxS1uI1mDv3rodF0KIhubt7U10dDTR0dGW47169eLxxx83N0Peu3cve/fu5dNPPwUgOzubnj17ApCcnEzXrl3p3bu3LBMvhGhRGnOIYD/gTeAyrXWqUqov8BXwAVBRrXg54Fv9HFrrucBcMFZyatgaC9E0hIYawwKdHRdCiKYkMjKS2bNnA8ZmyFu3bsVms5GUlMSePXvMcAVw2223sXXrVrp27Wr2cCUmJhITEyObIQshmrXGHCI4DvhZa50KoLXeoZT6ErgUqH6rGAqsb8S6CdFkzZljnYMF4OdnHBdCiKbK09OTwYMHM3jwYP74xz9anisvL6dv374cOXKEw4cPs2TJEpYsWWK+7sUXX+See+4BoLS0FG9vb+nlEkI0Gx6N+F7bgfOVUt0BlFIBwFjgc0Arpa6sOt6n6vjiRqybEE3W1Kkwdy707g1KGX/OnSvzr4QQzZeXlxfLli3jyJEjZGZm8uGHH3LXXXcxYsQIAPr0+W2WwFtvvUVwcDBXXXUVs2bN4rvvvuPUqVPuqroQQpxRo240rJT6K3ArxjLtfsDXwOPAYIyVBO3DAmdprZed7lzNfbNHIYQQtSMbDbcuhYWFeHp6mvtw3Xnnnbz11luWMkopBg4cyBVXXMELL7zgjmoKIVq4ZrPRsNb6ReBFJ09tAi5ozLoIIYQQounx8/OzPH7jjTd47LHHzLlc9s2Qt2zZQu/evc1yRUVFjBs3jri4OHNOV+fOnRu7+kII0bgBSwghhBCiLpRShIaGEhoayqRJkwAoKSlhw4YNlnLr1q1jxYoVrFixwjwWERFhhq3JkycTFBTUqHUXQrROjTkHSwghhBDinLVp04b4+Hji4+PNY0OGDGHx4sU88sgjjB49Gj8/P3bu3Mn8+fP5y1/+Ql5enll20aJFfP755xw4cMAd1T8r8+dDWBh4eBh/zp/v7hoJIVyRHiwhhBBCNHuBgYGMHz+e8ePHA8ZKhfbNkNPT0y3DCWfNmkV6ejoAISEhls2QR4wYga9vjZ1i3Gr+fOtqsllZxmOQBY+EaIoadZGL+iQTjYUQonWQRS5EfdJaM3v2bH7++WeSk5PJzc21PP/QQw/x7LPPAnDs2DFOnTpFeHi4W5eJDwtzvh9i796wZ09j10aI1qHZLHIhhBBCCOFOSimeeOIJACorK83NkO1f5513nln2s88+409/+hNdunSx9HLFxsbi7+/faHXeu7dux4UQ7iUBSwghhBCtkoeHB4MGDWLQoEH84Q9/qPF8UVERwcHB5OTksHTpUpYuXWq+btSoUaxevdosq7VusF6u0FDnPVihoQ3ydkKIcySLXAghhBBCOHH//feTk5PDjh07+Oijj/jzn/9MTEwMHh4e5j5dAKWlpfTs2ZMrr7ySf/7zn3z77becOHGi3uoxZw5UW70ePz/juBCi6ZEeLCGEEEIIF5RSREREEBERwdSqFSUKCws5duyYWWbTpk0cPHiQgwcPsnz5cvP4gAEDSEhIYMaMGfTr1++s62BfyGLmTGNYYGioEa5kgQshmiZZ5EIIIUSTJotciKZOa012drZlLte6desoKSkBYPv27URGRgLwyiuvcPDgQXM+V5cuXdxZdSGEC7LIhRBCCNEISktL3V0F0QQppQgJCSEkJIQbbrgBMDZD3rhxI2vXrqVv375m2ffee4/169ebj/v06WOGrYsvvpjBgwc3ev2FEPVLerCEEEI0aU2pB0sppXv27GlZUW7kyJG0bdvW3VUTzcTXX3/NmjVrsNlspKSkUFBQYD5333338dJLLwGwf/9+fvnlFxISEggJCXFXdYVotc6l7ZGAJYQQoklryICllPIG7gWeAm7WWn9yuvJeXl66oqKi+jGGDx9uCV19+vRx675JonkoLy9n8+bN5rDCG2+8kSuuuAKAefPm8cc//hEACfVCND4JWEIIIVqsBg5YdwFtgWuB184UsGJiYvRHH31kmWuTnp5OZWWlpVxwcLB5M5yYmEhsbCzt27dviEsQLdSyZct47bXXSE5O5uTJk5bnAgMDOXbsGJ6engAcOHCA7t27S6gXoh5JwBJCCNFiNcYQQaXUKuDN2gSs6m1PXl4eqampZuBKSkriyJEj1c9PVFSUpRdiwIABeHjIbini9CorK9m+fbsl1AcFBbFy5UrA6AULCAjA39+/xmbIEuqFOHsSsIQQQrRY7g5YSqlpwDSA0NDQkVnOdnx1oLVmz549JCUlmTfEaWlplJeXW8oFBgYSHx9v3hDHx8fTqVOnersm0XJVVFSYvVe7d+8mPj7eZah//fXXOf/8891RTSGaNQlYQgghWix3ByxHZ9v2FBUVkZaWZunlys7OrlGuX79+JCYmmqErKioKLy9Z8Fecntaa3bt3W3q57KE+IyPDXJnwiSeewGazmT9jEuqFcE0ClhBCiBarJQQsZ7Kzs0lOTjZviFNTUykuLraU8fPzIzY21jL0q1u3bvXy/qJlKyoqYv369SQmJppDURMTE7HZbJZy/fr1IyEhgfHjx3Pddde5o6pCNEmyD5YQQgjRzPTq1YtevXoxceJEAMrKyti4caOlF2Lnzp38+OOP/Pjjj+brwsLCLIFr+PDhtGnTxl2XIZqotm3bMmrUKMuxRYsWWX6+UlNT2b59O9u3b6d9+/ZmwMrKyuLNN980f8a6du3qjksQotmSHiwhhBBNWkvtwaqNI0eOkJycbM7nSklJIT8/31KmTZs2jBgxwhK6QkJCZEU5cUZlZWVs2rQJm83GsGHDzLlaH374IbfccotZzjHUJyYmMnLkSHMOmBAtlQwRFEKIlqi8HPbtgwMHoLQUfHygRw8ICYFWNC+nKW007O62p6Kigi1btlh6IbZs2VKjXPfu3WvcEPv5+bmhxqI52rRpEwsXLnQa6tu0aUNubi4+Pj4ArFq1ij59+kioFy2OBCwhhGhJtIaMDMjMNB47bmxr/9Q4MhKioqAV3NBIwDq9kydPkpKSYgldJ06csJTx9PRk2LBhll6uvn37yg2xOCN7qLf3opaWlvLRRx+Zz3Xs2JG8vDwJ9aLFkYAlhBAthdawZg3k5FiDVXWentClC4wa1eJDlgSsutFak5mZaVkmftOmTTU2Qw4KCrIErri4OAICAtxUa9EcHTt2jJtvvtllqP/444+ZNGkSAMXFxbRp00ZCvWg2WmXA6tGjhy4oKMDDwwNPT088PT3N73v37s2aNWvMsgkJCeTl5Tkte9ddd5njjFevXs3s2bMtzzt+P2/ePHPTvueee44tW7aYZRzLRUdHc9tttwFw4sQJXnjhBafn8/Dw4IYbbiAsLAwAm83Gxo0bnZZt374911xzjXlNy5cvp6KiwmnZ8PBwQkJCzPffs2ePy2sKDQ01x1GfOHGC8vJyp9fk5eUl462FaAzp6UbP1enClZ2np9GTNWRIw9fLjSRgnbv8/HzWrVtnLhFvs9k4fPiwpYxSikGDBlmWiR84cKBshizOyFWoX7duHcOHDwfgkUce4T//+Y+llys2NlZCvWiymlXAUko9APw/oAhQwIVALPASxqqGpcCftdY2lycBunXrpqs3DnZ9+vRh586d5uMOHTpw6tQpp2WffvppHnnkEQAWLlzITTfd5PI9jx8/TseOHQG45JJL+P77752Wu+GGG/j0008B2LlzJ3379nV5zu+++45LL70UMH75PPvss07LhYeHs2vXLvNxx44dOXnypNOyTz31FI8++igAn332mfnp0Zmu6dJLLzV3hq/u+uuv57PPPgOMTQ379+/vMjQuWrSI0aNHA/DMM88wd+5cp2VDQkJYtmyZ+R5XXXUVBQUFNcKgp6cnt912m7nSls1m4+WXX3YZGp9//nnatWsHwNy5c9mxY4fTeg4ePNhcMenkyZO89957Lq9p7Nix9OjRA4D09HS2b9/utGy7du0sqzalpKSgtXZ6TV26dCEoKAiAwsJCjh075vKa/P39zU/9KisrUUrJp4AtUXk5LF1qCVfzfwph5oIh7D3mR2hQIXMmpzP1gn2/vcbTE8aNa9FzsiRg1T+tNVlZWZZhhevXr6esrMxSLiAggLi4OEtPl/33lhCnk5+fT9u2bc0PZ8ePH8/SpUstZZRSDB48mEmTJvG3v/3NHdUUwqVms0y7UuqfQBtguNa6XCnVAfAFvgCu1lonKaXGAEuUUuFa60JX5+rZsyfbtm2joqKCiooKKisrzT+r33gmJydTXl7utKy9pwdg9OjRfPvttzXK2P+037QDPPTQQ0ydOrXGOSsqKiyBqmPHjsyePdvpe1dUVNC7d2+zbHx8PNOmTXNatnPnzpZruvzyy8nPz3daNjQ01CwXEBDAsGHDXF6TY69UQEAAQUFBTq/JcaPLioqKGo2wowqHm8OjR49agqGj6ithrVmzhtzcXKdlL7jgAvP7PXv28Mknrhf6euqpp8zvFy5c6DIIT5o0yQxYR48e5f7773d5zhUrVpgB68MPP+S5555zWq56uL/ssstcXtMzzzzDww8/DMCyZctOG+5PnDhBhw4dzHN+//33lgBo//P666/n3XffBYwgPGrUKJehcd68eSQmJgLwyiuv8NFHHzkt27NnT95//32zLlOnTqW4uNhpaJw0aRJXXnklAOvXr+fdd991GRofe+wx2rZtC8CCBQvYt2+f03pGRkaaH0Lk5eWxZMkSl+dMSEggODgYMD7cyM7OdlrO19eXQYMGmddk/xl1FW7t9SwvL6esrMxSpl6D7r59lofzfwph2lsxFJYa//6yjrZj2lvG73pLyNq3D8LD668eosVTShEWFkZYWJj5u6e4uNiyGbLNZmPv3r2sWLGCFStWmK+NjIy0BK4hQ4bg7e3trksRTZS/v7/l8eLFi8nKyrL0cqWlpZGRkWG2RWAsEX/77bdLqBfNWqMFrKowNRZYBqxRShUDTwMBwDatdRKA1nqVUuogcElVWVfnIzAwsFbv3b9//1qV69atW603cLziiitqVa5Tp048/vjjtSp73XXX1XqTv4ULF9aq3NixYxk7dmytyn7xxRe1KhcREUFJSYnL0GgPAgCPPvood955p9OyXtU+cf/vf/9LaWmpJdjZyw4cONAsl5iYyPz5812GRsdJtdOmTWPs2LFOQ2NUVJRZLjAwkHvvvdflNdnDFUBUVBTXXXed07Ldu3e3XFNMTAx5eXlOr8mxwfD19aVnz54ur8lxiI6917mysrLGnArHTUrLyso4ePCgy79Hx7JZWVmkpKQ4LVe9B3bp0qU1wrHdoEGDzIC1fft2Xn31VZfv/9BDD5nBZe7cuaxatcppuRtvvNEMWIcOHeLmm292ec6VK1dy8cUXA/DGG2/wwgsvOC0XERHBjh07zMfR0dEug/Czzz7LQw89BBh7yEyePNnyvFLKDGM5OTnm76Vrr72Wn3/+2WloHDduHC+//DJg/L+/5pprjOcKC/GsqMDTwwMPDw/W7QyipPwVYETVu/2HwtIvuONNzWdJR8xynu3a0X3IEF566SWzXvfeey8lJSVOQ+PVV19t9jJnZGTw+eefO62np6cnd955p7nH0jfffMOhQ4eclu3duzdxcXEAFBQU8NNPP7kMwoMGDTKHBR08eJDjx487Lefj4yOb6jYiX19fEhMTLTe7Bw4csCwTn5qaSmZmJpmZmXz44YeAsd9S9c2Qq/8uFMIx1Nt/j9pDvX36BcAvv/xy2lB/yy231AhvQjQ1jTZEUCl1CbAYY/jf+0qpaOB74HlggNb6ZoeynwOrtdYvVzvHNGAaQGho6MisrKxGqbsQTY3W2mkYsw9TBKO3JScn57S9p/aye/fu5eDBg07L+vr6mnujgBGwXAXhmJgYhlTNB9q+fbs5V9BZaHz44YfNG/e33nqLHTt2OC0bExPD7bffDhg34w8++KDlvR2/f+qppxg6dCgAr7/+OgsXLnTZc71kyRLzmoYPH05ubq7Ta/rHP/7B3XffDRgfbNx6663mc9UDbl5entnwjx49mtWrVzv9+5s8eTIff/wxANu2bWPAgAGn+dv+ARhT9f1fMUZT1xQZGcn27dvNx/7+/hQUFDgt+/zzz/PAAw8A8PHHHzN16lSX756bm2ve/IwZM8ay4a2jm266iQULFgBnvqbvv/+eiy66CIAHH3zQZRC2X5MMEWw6ysrKSE9PN3sgkpKSLB9W2IWGhprzbBISEoiOjpbNkEWtHDt2jNWrV1tCfVFREQBeXl7k5uaaH87NmzePjh07kpiYKB/GiHrXLOZgKaUmA49prYc4HFsEpAGRWuvbHI4vBGxaa+d3EkgjJ4RwP621Jez6+PiYQwZPnTpFaWmp09Do5+dn3gwUFxezbds247n166k4fJhKramorOT6F2I5fCoBsPcKbwF20bl9IW+Xvjj+AAAgAElEQVRNX2u8prKSyg4daDdihGUhnHnz5lFSUuI03I4ZM4bY2FjA2O9m0aJFLoPws88+a94YP/PMM2zdutVp2fPPP5/77rsPgH379nH77be7DMJvvPEGI0YYvXLPPvss7733ntP3Dg8P58cff5SA1cQdPXqU5ORkM3QlJyeTl5dnKePj40N0dLSll6t3794yl1SckWOoP3jwILNnzwaM379BQUHm6oW9e/e2/HxJqBfnqrkErEuAWVrrUQ7HPgUygDFa64sdjicBL2qtP3N1PmnkhBAtzu7dkJZmLnJRfQ4WgJ9POXOnp/42B8vTE6KjW/QcLAlYzUtFRQVbt261zLXZsmUL1e83unXrZrkhjomJscx1FuJ0ioqKmDNnjrkZcvVQ//LLL3PPPfcAsH//fsrLywkNDZVQL2qtuQQsHyALuEFr/bNSqj+wBmOu1UrgIq11ulIqDvgWCNdaO18mD2nkhBAtkKwi6JQErObv1KlTrF271rJM/PHjxy1lPD09GTJkiGWZ+MjISLkhFmdUUVHBr7/+ahm6+v777xMTY/zamDlzJk8++aSEelEnzSJgAVTNu3oH8AAqgSe01l8rpS4CngU0UA48YF/0whVp5IQQLZLsg1WDBKyWR2vNjh07LCsWbty40bIKLRgLRcXHx5vzueLi4mq9wJUQdo8//jhvvPGG01B/7bXXsmjRIvOY1lpCvQCaUcCqT8HBwXr8+PEuV6ly17GGfj/5Ry9EC6c1rFkDOTmnD1mentClC4waBS3894IErNahoKDA3AzZ3gtx6NAhSxmlFAMHDrT0QgwaNMiy5YgQzlQP9UlJSWzatImbb77Z3N5k//79DB061PLzVe+hvrzc2FrjwAEoLQUfH+jRA0JCWvRIhOaoVQYspVTzrPg5Uko1yeDXVANpQ9VLgq5oUFpDRobRkwXWoOXlZTwfGQlRUS0+XIEErNZKa82+ffsswwrXr19PaWmppVz79u0tmyHHx8fX2DtSCGcKCgrIy8szFx1asmQJ1157raWMUopBgwaRkJDAv/71r7NfrfB0v9ftHxC0ot/rzUGjBCyl1P+Ay4CJWusvHI4r4F3gVuAZrfWjZ1ORugoLC9OPP/6401Wq3HWsod+v+pLQwr2aYvBrqoG0oY457g/WIjl+0llWBt7erfKTTglYwq6kpIQNGzZYeiGcbdkSERFhWSZ+6NChshmyOCOtNXv37rUMXbWHeg8PD06ePGluW/H3v/+dioqK2oV6GZnQLDVWwBoGrAe2AUO01hVVx1/A2Jzlba31tLOpxNlojY2c1toMWk0p+DXVQNqQ9RJNR1MMfk01kDZUvRp6+LIELHE6Bw8etCwTv3btWgoLCy1lfH19iYmJMXu5EhMTLRvIC+FKSUkJaWlpbNu2jVtvvRUw7se6detGTk6OWS4iIsIM9JdffjmRkZG/nUTm1jZLjTZEUCn1HkZP1f/TWr+nlHoMmAN8CkzWWjfanWerbORk3G6T4WzPnqYQ/JpqIG2oY6LpsAethgh+a9askYAlaq28vJyMjAzLMvGOm3DbhYSEWObajBgxAl9fXzfUWDQ3FRUVLFu2zGWod9zMfWtGBls++ICEiAh6dOoEyOqwzUVjBqxeQCZwGHgeeAVjSfVxWuvS0722vrWqRk7G7QrhlOMmv00p+DXVQNpQ9WqEubwSsMQ5OXbsGCkpKeawwuTkZHJzcy1lvL29GT58uGWZ+LCwMJlzK86ovLzc3AzZZrNx7733mpup//O++/jHyy8DEBIURPeOw0jbfQVlFaOAEYBvq9zfsDlo1EUulFJPAfZ5Vr8Al2mtC6uVuRB4EBgJ9KCqx+tsKuhKq2nkZNyuEKKJsw9fbqjgd+GFF0rAEvWqsrKSrVu3WubaZGRk1PiwoEuXLpZhhTExMfj7+7up1qI5+uiJJ/hgyRKSMzPJLSqq9uwA4FcAegcXsHrWB4QEBRmhvkcP455OuE1jB6y/Ai9UPRyotd7qpMxVwPkYc7Y+AO6SgHWWZNyuEKKVkzlYojHk5uaamyHbv44ePWop4+HhwZAhQyxDC/v169fyF9wRZ++HH+DoUSoqK9m6fz9RDwQAyYANiMa4TQZjcFg3ugYGkhAZSUJ0NAlTpkiod6PGHCI4GZiP/acA3tRa/+kMr8kH/iwB6yyUl8PSpWa4WnNHEGWnav4S9w6sZNTbx4wHMm5XCNHCSMAS7qC1ZteuXZZl4jdu3Eh5ebmlXMeOHc3NkO37JnXs2NFNtRZNzpo1xtz5KmF3XUXW0XZVjyoB476uW+AqyirHcSwvz/Jye6hfsGABAwcObKRKCzi3tqfWd+FVvVLvA5uBi4HVwO1KqZed9WI1tOLiYpYvX+50UrSfnx/R0dFm2S1btqC1djp5ukOHDgQEBABQWlpKQUGBy4nWjT4Oe98+y0Nn4crp8X37ZNyuEEIIcQ6UUkRERBAREcHUqVMBKCwsZP369ZZl4g8cOMDy5ctZvny5+doBAwZYlokfPHgwnp6yGXKr1KMHHD5sflg+Z3I6096KobDUC3u48vMp5/lbfJly/jvsPHwY244d2E6cwLZlCxs2bGDjxo10797dPOXtt9/O/v37JdQ3YbXqwVJKnQ/8DzgEjNJaH1RKXQ98BizRWl97mtc2SA9W9+7ddfUd3u3CwsLYvXu3+bhTp06cOHHCadknn3ySGTNmAPDZZ58xadIkl+959OhRgoKCALj66qv54YcfnIaxcePGMXfuXACys7O5+OKLXa6W9eqrrxIfHw/A22+/zYIFC34rk5uLZ9XeC907dOCmFTPNurzESxRTjEfVf70uLcVDKTw9PBg3diyX338/AOnp6Xz88ccuQ+M999xD27ZtAVi8eDH79+93WjYsLIwLL7wQMDbm++6771xe0/Dhw81/6Pv27ePw4cNOz+nr60vv3r3Nazp06BBKKafn9fHxwUt65YRolaQHSzRVWmuys7MtwwrXrVtHSUmJpZy/vz+xsbGWoYVdunRxU61Fo6o2GgnqtopgYWEhmzdvJjY21ny6V69e7N+/3/I29lB/0003MXbs2Aa/rNagQXuwqva/+go4hbGgxUEArfUipVQqMF4pdYHW+qezqcDZatOmDZdffrnTSdGOKR+MH7pTp045nVAdGBholvPy8qJDhw4uV9dy/PSpqKioxj4bdqdOnTK/Ly4uJtO++p8TeQ5dwZmZmfzwww9Oy/Xp2pWbHB6vZCUFFPx2YMVv34b07s3lVd//+uuvPP300y7ff/r06WbAeuWVV/j++++dlrv++uvNgHXo0CEmTJjg8pzfffcdl156qXnO5557zmm58PBwdu3aZT4eOHAgJ0+edFr26aef5pFHHgGMIHzTTTe5DI07duwwA96UKVNYs2aNWcax3BVXXGHWbf/+/dxwww0uQ+OTTz5p9orOnz+fpUuXOi3bpUsXnnzySbPejz32GCUlJU7LXnbZZYyqmsC6bds2vvrqK5fX9Pvf/542bdoA8MMPP5CTk+P0mnr06MGwYcMA42d03bp1Lq8pPDzcHNd9/Phx8vLynJb18vIye3kBysrKGmXvIyGEaOqUUoSEhBASEsINN9wAGKNhHDdDttls7N69mx9++MHSxvfp08cSuIYNG4aPj4+7LkU0FC8vY368w3z6qRfsswYqR/b59FUfKvv5+VnCFcAvv/xSI9Rv3bqVrVu30r9/fzNgrV+/ns8//9zcDFlCfeM5bcBSSvXFWIZdA2O11jurFZkBfAc8ByQ0SA1dCA4O5ttvv61V2V9++aVW5SZMmHDa4ODY2/f1119TXl7udAUs+40wQGhoKNu2bXO5Wlb//v3NstOmTWPs2LG/Bbv0dCqrJka29fGB3+7b+St/pZRSKqv+i7wjl4rKSiq1JjEuziwXFRXFnDlznL53ZWWlZc+P8ePH079/f6dl4xzO6efnx/jx411eU6eqfR4AevbsyYgRI5wG4ZCQEMv/386dO+Pj4+P0vN7e3mY5+zFXG/46TjY+dOgQe/fudVpu8ODB5veFhYUkJSU5LQfw0EMPmd+npaXx6aefOi0XERFhCVivvfZajaWA7dq1a2cGrLS0NB588EGX7z9x4kTz52r27Nkug/ikSZNYuHAhYPSeXnDBBS7PuWLFCi655BLACLCugnBERAQ7duwwHwcHB5vXpJSyhLE5c+Zwf1Xv6eLFi5k+fbrL0Lh27VozuE2fPp3U1FSnofHiiy/mb3/7G2BsKDp9+nSXofHRRx81/16/+OILVq5c6bRscHCw5f/3888/T1lZWY26enp6ct5555nhevfu3axevdrlNV111VXmz+r69es5efKk03oGBQXRp08fwLgZ27Vrl8trCgoKMm+6SkpKKC8vr1FWgm7jOX78OM8//zwJCQmMHDnS/IBKCDsfHx/i4uKIi4vjnnvuAeDw4cPmZshJSUmsXbuWXbt2sWvXLj7++GPA2Ax55MiRltDVq1cvd16KqC9RUXDqVO1XhI6KOu3pQkNDCQ0NNUddlZSUsHHjRmw2GxdffLFZbvny5ZZ7Enuotw9dHTlypLQfDaTOqwjW+Q1kkYuzt3s3pKWZ/xhXTerssuiYT48Y37SCvRPsex+5Co0dO3Y0Q9bhw4cpKipy2isZEBBAWFgYYPT2pKWluTxvfHy8OTx048aNbN261Wlo9Pf356abfutrfPXVVykuLnZaT8cerLS0ND788EOX1/Tqq6+aN3KzZs0iIyPD6TWNHj3aHPK6d+9epkyZ4vKa/vOf/5jDU+fMmcPcuXOdlg0PD2fdunXmNXXs2JFTp0453fvoueeeM4PLggULmDJlisu/x9zcXNq3bw/AmDFj+PHHH52Wu+mmm1iwYAEA27dvt3woUd3333/PRRddBMCDDz7ICy+84LRc3759LT3LAQEBlt7ks72mU6dOmaHxoosuYtWqVU7L3XjjjXzyySeA0XPdr1+/c7ompRT9+vVj69bfpsP26tXL6ZxST09PZs6cybRp0wD45ptveOihh1xu+Pvtt9+aPZ0PP/wwmzZtclr2/PPP59577wWMf3czZsxwGRrvuusuIiMjAfj222/55ZdfnJbt1KkTf/jDH5rUEMGAgABt/1nx8vJi+PDh5s3wqFGjzN8pQpxOeXk5mzdvtvRCOP77tevZs6dlmfgRI0ZIqG+uTrenqZeX8Xw972lqs9lYsmQJNpuNlJQUy8ir0NBQsrKyzMfffPMNQ4cOlVDvoFEWuagLpZQ/0LfqoQcQqpQaDhzXWjvvThA1hYQYAauKd2Cly1UEa7yuBVNK4eXlVas5WV27dq3VOdu2bct5551Xq7LDhg0zh+GdyZ///OdalYuOjrYszHI6TzzxRK3KhYaG8vPPP9eq7MyZM5k5c+aZC4I5n9G+95FjGHP8O5kwYQIHDx50GRrbtWtnlp07dy65ublOQ6PjkIYePXqwZMkSl6HRcYWlCRMmEB4e7rRshw4dLNf0wAMPUFRUZHlfe9nhw4eb5cLCwrj55ptdXpNjT2t0dDRKKafXZO+9AuMm3d5zfKYecS8vL9q2bWspU1lZida6RuA9efIkBQUFOJOfn2/5+9yyZYuLv21rz31ycjKrV692Ws7x2k+cOMG7777r8pzjx483A9b//vc/XnzxRafl+vXrxx/+8AeX53GH4OBgJk+ebO6blJqaSmpqKq+++iq33nor7733HmD8P1i/fj1xcXHmBwlC2Hl5eZltyfTp0wGjd9S+GbLNZiM5OZn9+/fz+eef8/nnn5uvcwz1CQkJ9OnTR3ohmgOljG10Bg40FiM7cADKysDb21gIIySk3leAtv+MgBHqMzIyzJ+v4OBgs9zx48f53e9+B0iory8N0oOllBoDOBvD9L7W+rb6eI9W0YMFsg+WEOK07OGqesC1zzt1Fho7dOhgzj89efIk2dnZLkNjQkKCOf80JSWFY8eOOS0bEhJCYmKiec4vvvjCZWicNGmS+Snp//73P3755Ren7x0cHMyjjz7apHqwHNuevLw8UlNTzRuWSZMmmavNffHFF0ycOBGlFFFRUZYb4gEDBsi+SeKMKisr2b59u7lEvD3UVx8e37lzZ8vPV2xsrIR6USc7d+7k7rvvJjk5ucZceC8vL1auXGlZ6MzPz69VhPpG3Wi4qWg1AUtrYw+F2o7bHTWq3rqWhRCiKWjIgKWUGg28hDGioxRjSLvNVfnatj1Llixhzpw5pKWl1dg3qVOnThw4cMDsnSwqKpJPiEWtVA/1SUlJHDlyxFLGw8ODwYMHm/NsEhIS6N+/v4R6cUauQv3hw4fNHq8pU6awYsUKy1yumJiYFhnqJWC1dG4YtyuEEE1FQwUspVQHYCdwtdY6qWr0xUIgXGvtdJnYurY99vmdjjfEnTp1YuPGjYDRA9mjRw8CAwMtvRBRUVH1szVFeflvw5FKS8HHp8GGI4nGp7Vm9+7dlrlczkJ9YGCgZTPk+Ph4y4JUQrhSUFBgGdIfGxtL9d+BHh4eREVFMW3aNO6+++7GrmKDkYDVWjg2lA08blcIIZqKBgxYNwL3aq3Pczi2Afib1nqZs9fUR9uTl5dnftp76NAhwsPDKS4utpRp164dsbGxzJo167Srgbp0ug/m7FuOyAdzLVJRUZFlM2SbzUZ2dnaNcv3792+YUC9aNHuod+zl2rBhA+Xl5cyaNctc9ddms/H3v//d7OmKi4trdqFeApYQQogWqwED1gxgkNb6ZodjnwOrtdYvOxybBkwDCA0NHem48lZ9KC0tZdOmTZYb4p07jV1Rfv75Z3O10ddff52ffvrJvCEePny4ZREUkwwtF9VkZ2eTnJxs3hSvW7fOZah3DF21XShKtG72UN+jRw/Cq1axfu6553j44Yct5RxD/e23397kA70ELCGEEC1WAwasx4B+josvKaUWAjat9UvOXtNYbU9OTg7Jyclceuml5vysK6+8kuXLl5tl2rRpw4gRI0hISODyyy/niiuuMJ6QxZHEGZwu1DsKCwuzzLUZPny4bIYsauXw4cP8/PPPTkN9t27dOHDggLlQxosvvkhERESTC/USsIQQQrRYDRiwbgFu01pf7HAsCXhRa/2Zs9e4s+3ZsmULa9asMW+IHZfXnzJlCvPnz4fyco7Nn8+8FStIiIyk9IUYPHP9apzLO7CSUW8fMx54esK4cTLUvJWzh3r7z1dKSoplSwewhnr7V0hISKtYUU6cG8dQX1ZWxv333w8Ye2J26NDB3BIkPDzc8vPlzlAvAUsIIUSL1YABKxBjkYuLtNbpSqk44FuMRS5OOntNU2p7Tp48ae6bFB0dzTXXXAO7d7Ns7lzGPf00AJ54EkEEgxz+60EPFKpVbVAv6q6iooItW7ZY5tr8+uuvNcr16NHDckM8cuRI/PxqhnohnDl69CgvvviiGeqr79/45Zdfcu211wKwa9cuvL296dWrV6OEeglYQgghWqwGXqb9IuBZQAPlwANa6yRX5Zt827NmDet//pm3V67ElpnJpj17qcS6b9JSltKe9oz59Ajpe/fSOziYgL59jblYQpyGY6hPSkoiOTnZ3HzeztPTk2HDhlmWiY+IiJBeLnFGFRUVbN682bLq6vfff0/37t0BuPXWW/nggw8aLdQ3u4CllBoM/AT8W2v9j7ruQwLNoJETQghRL5rqRsNN0g8/wNGj5sP/TvJnG9vYUvVfHnm8jLF+x5hPjxDypz+x//hxBoeFkXDJJeYNy8CBA2XfJHFGlZWVZGZmWuZybdq0qcZmyEFBQebPVmJiIrGxsQQEBLip1qK5mj59Op9++qnTzZCnTZvGa6+9BmAONzzXUN+sAlbVviOLgc3AEeD/qOM+JNAMGjkhhBD1QgJWHaxZY2zlUWXVpM4ui8Z8sI9LZ89m/a5dlFVbECMgIIA33niDKVOmAMYNi/RAiNrIz8+vsRlyTk6OpYxSisGDB1t6ISTUi9pwDPX24avp6enMnDmTWbNmAZCcnMzVV19t+fmKi4ur82bIzSZgKaU8gC+BZ4DLqw7/Sh33IYFm0MgJIYSoFxKw6mD3bkhLM1cQPF3Ass/BKq6oIM3bG9veveZN8d69e/nhhx8YM2YMAE8//TTz5s2zrCg3ZMiQJr/MsnA/rTVZWVmWG+K0tDTKysos5QICAmpshhwUFOSmWovmJD8/n9LSUnOfrddff73GhsdKKaKiokhISOCll16ybJ7sSnMKWE8CB7TWryql/lF1uIRa7ENSdbxB9yIRQgjR9EjAqoPycli61AxYa+4IouxUzV6BM60ieODAAYKCgsx9tiZOnMgXX3xhOYefnx8xMTFcffXVPPTQQw10QaIlKi4uJi0tzdLLtW/fvhrlIiMjJdSLOtNas2fPHsvPV1paGuXl5QQFBXHkyBGzR/7uu+82h7BWD/XNImAppSYCE7TWv696/I+qp0qp4z4k0AwaOSGEEPVCAlYdNcA+WGVlZTX2TdqxYwcAN9xwA59++ikAJ06c4E9/+pPZCxEdHe18M2Qhqtm/f79lmfjU1FSKioosZeyh3nHol30BBCFOp6ioiLS0NA4dOsR1110HQEFBAYGBgVQ4/K7s168fCQkJjB49mj/+8Y/NImC9A0QD9j7hXlV/9gR+qMs+JNBMGjkhhBDnTAJWHWltzMXKyTl9yPL0hC5djNUDz2J+1ZEjR0hJSaFDhw6MqlqBcPny5Vx55ZVmGR8fH6Kjo82b4auvvhp/f/86v5dofcrKykhPT7csE28P9Y569+5tCVwS6kVtFRcX88033zgN9RdccAE//fRT0w9YNd74tx6sl6jjPiTQTBo5IYQQ50wC1lnQGjIyjJ4ssAYtLy/j+chIiIo6q3DlysGDB/n6668tmyE73mfk5OTQubMxL2zZsmUEBgYycuTIWs2HEOLo0aNmL1dSUhIpKSnk5eVZythDveMy8aGhobJIizgjx576oKAgJk+e3HwDVtUy7XXahwSaUSMnhBDinEjAOgfl5bBvn7GyYFkZeHtDjx4QEmKZc9VQTp06xdq1a0lKSmL37t3MmzfPfK5Pnz7s3r0bT09Phg4daplr07dvX7khFmdUUVHBr7/+ahm6Wj3UA3Tr1s2yTLyEelEbzWIOVn1rdo2cEEKIsyIBq+UpLy/nnnvuMfdNqqg2lPHJJ59kxowZgLG5rVKKwMBAd1RVNDOnTp0yN0O2fx0/ftxSxjHU278iIyMl1AsLCVhCCCFaLAlYLVtBQUGNfZM+/PBDLrvsMgCef/55Hn74YQYOHGjp5Ro4cCCenp5urr1o6rTW7Nixw7JMvLNQ36lTpxr7Jkmob90kYAkhhGixJGC1LlprtNbmprMzZszgxRdfpLS01FKuffv2/O53v2PBggXuqKZoxgoKCli3bp0l1B86dMhSRiklob6Vk4AlhBCixZKAJUpKStiwYYNlRbmsrCzGjx/P4sWLAcjLyyMmJoa4uDizF2Lo0KF4e3u7ufaiqdNas9dho22bzcb69eudhnrHn6+EhASCg4PdVGvR0CRgCSGEaLEkYAlnDh48SH5+PpGRkQB8//33XHLJJZYyvr6+5r5J999/Pz169HBHVUUzVFJSYtkM2R7qq+vbt68lcEmobzkkYAkhhGixJGCJ2igrKyMjI8My1ybTvkw9xka29oD1+uuvU1xcTGJiItHR0fj6+rqr2qIZOXjwoGWZ+LVr19bYDLlt27Y1NkOWYN88ScASQgjRYknAEmfr6NGjpKSkkJ6eziOPPGIeHzBgANu2bQPA29vbshnyBRdcQK9evdxVZdGMlJeXk56ebunl2r59e41yISEhlrlcEuqbBwlYQgghWiwJWKI+aa157733+OWXX7DZbGzevNmyb9Ljjz/O7NmzAcjOziYzM5PY2Fj8/f3dVWXRjBw7dszs5bLZbCQnJ5Obm2spUz3UJyYm0rt3b1kmvomRgCWEEKLFkoAlGlJubi5r1641b4jvu+8+cy7Xyy+/zH333YeHhwdDhgyx3BBHRkaaKx0K4UplZSVbt261LNBSPdQDdO3a1TKsMCYmRkK9m0nAEkII0WJJwBLu8u677/Lqq6+ycePGGvsmRUZGWoaD5efnyw2xqJXqoT4pKYljx45Zynh4eDjdDFlCfeORgCWEEKLFkoAl3K2wsLDGvknR0dF8/fXX5vMdOnSosaLc4MGDZd8kcUZaa3bu3GmZy7Vhw4Yaob5jx47Ex8ebc7ni4uLo0KGDm2rd8knAEkII0WJJwBJNjdaawsJC2rVrB8CGDRuIj4+vsW+Sv78/cXFx/N///R9DhgxxR1VFM+Us1B88eLBGOftmyBLq658ELCGEEC2WBCzRHJSUlLBx40bLDfGePXsAyMrKIjQ0FIAnnniCnTt3mjfEw4YNw8fHx401F82B1prs7GzLNgTr1q1zGertP1/x8fF06dLFTbVu3iRgCSGEaLEkYInm6tChQ6SmpvK73/3OXCEuKiqKzZs3m2V8fX0ZOXIkCQkJXHPNNYwePdpd1RXNTPVQb7PZ2L17d41yffr0MYcV2jdDllB/ZhKwhBBCtFgSsERLUr2Xy74fF8DDDz/MM888A8DOnTtZvHgxCQkJjBgxgrZt27qryqIZOXTokGWZ+JSUFAoLCy1lHEO9PXj17NnTTTVuuiRgCSGEaLEkYImW7Pjx46SkpGCz2bj88ss577zzAHjjjTe46667AGPfpOHDh1vm2oSHh8u+SeKMysvL2bx5s2WZeMdQb9erVy/Lz5eEeglYQgghWjAJWKI1Wr16NfPnz8dms5GRkUFlZaX5XEBAACdOnDCX7F67di0DBgygffv27qquaEYcQ31SUhLJycmcOnXKUkZCvQQsIYQQLZgELNHa5eXlWfZN8vf35+OPPwaMeTgBAQGUl5cTFRVluSHu37+/7JskzqiyspJt27ZZ5nJVD/UAnTt3tgwrjImJadGhXgKWEEKIFksClhCu7d27l+uvv560tKtJ5K4AABi7SURBVDTKy8stzwUGBrJo0SIuvfRSwLiRlsAlaqN6qLfZbBw5csRSxsPDo0WHeglYQgghWiwJWEKcWVFREevXr7fcEGdnZ5OZmUnfvn0BuPvuu1m5cqVlRbnBgwfj5eXl5tqLpk5rze7duy3LxG/YsKFGqO/QoQPx8fGWZeI7duzoplqfGwlYQgghWqyGClhKKW/gXuAp4Gat9Sdneo20PaI5yc7OpmfPnua8mYSEBJKTky1l2rVrR2xsLJMnT2batGnuqKZopqqH+qSkJPbv31+jXP/+/ZtlqD+XtqfpX50QQgjRMO4ANGBzd0WEaAi9evWyPF69enWNfZN27drFqlWriI6ONstt27aNf/7zn+YN8fDhw2XfJFFD27ZtGTVqFKNGjTKP2TdDtn+lpqaybds2tm3bxnvvvQf8Furt87ni4+Pp2rWrm66iYUgPlhBCiCatoYcIKqVWAW9KD5ZojXJyckhOTqZ3794MHToUgHfeeYc77rjDLNOmTRtGjBhh9kKMHz9eApeoldLSUjZt2mRZJn7Xrl01yoWHh1vmcjWFUC9DBIUQQrRY59TIKeUDrHby1IVa69KqMquQgCWEKSsri2+//da8If7111/N59q1a8fJkyfNIV4ff/wxoaGhjBw5stXvmyRqxx7q7cMKU1JSKCgosJRp06aNuRmy/atXr16Nukx8swlYSqnpwF1AGeCD0aC9rpSKAt4C2gMKmKG1/up055JGTgghWgd392AppaYB0wBCQ0NHZmVlNVRVhGiSTp48SUpKCklJSRQWFvLMM88AUFZWRmBgIEVFRXh5eTFs2DDLXJs+ffq0qn2TxNmpqKhg8+bNlqGFjqHerkePHpZl4hs61DeLgKWU8gSeAf6htc5XSvUEdgB9gR+BmVrrhUqpgcAaYKjWOtvV+SRgCSFaujXd1lB2uKzGce+u3ow6NMrJK1omdwcsR9L2CPGbEydOMGPGDGw2G+np6TX2TXr//fe55ZZbAKPXwtfXl4CAAHdUVTQzJ06cMDdDtn+dPHnSUsYe6u2BPjExsV5DfbMIWDXeWClf4DAwDvgY6KWrKqOUWgys1Fq/4ur10sgJIVq6VWqVy+fG6DGNVg93k4AlRNOXn59fY9+kVatWMXDgQADuu+8+/v3vfzN48GBLL8SAAQNazL5JouFUVlaSmZlpWSbeWagPDg62DCuMjY0961DfXFcR/D9gIdAD2KWtSW8n0McttRJCCCGEEHXi7+/PRRddxEUXXQQY+yY5ysvLw8vLi4yMDDIyMnjnnXcACAgI4NZbb+Xf//53o9dZNB8eHh7079+f/v37c+uttwJGqE9NTbUsE5+Tk8NXX33FV18ZM42UUgwePNgydLUxQr1bApZS6l9AT2AicD1QUa1IOeDr5HWO4+AbuJZCCCFaA61bUXegEI2k+jCt//znP7z22mukpaVZVpTbt2+fJYxt27aNcePGWYZ9RUVFNYt9k0Tj8vf3Z8yYMYwZMwYwQv2ePXssvahpaWlmqH/77bcBI9TbN0O2LxPfqVOneq1bow8RVEo9D0QAN2qtS5VSFwLvaa37OJRZAKzXWj/n6jwyTEMI0dLJEEFDQw8RrAtpe4SoX/v376eyspKQkBDAmLd12223Wcr4+fmZ+ybNmDGDwMBAN9RUNEfFxcVOQ311/fr1swwtHDJkCN7e3k1/DpZSygN4HegITNVal1cd9wS2A3/WWv9XKdUHSAXitdaZrs4njZwQoqWTgGWQgCVE61FWVlZjM+SdO3cCxtLdubm55v5I//rXv/D39ycxMZHhw4fTpk0bd1ZdNBP79++3LBOfmppKcXGxpYyfnx+FhYXNImBdDSzDCE+OQwIfB3KA1/htWOAsrfWy051PGjkhREsnqwgaJGAJ0bodOXKE5ORksrOzufPOOwFjae8OHTqQn58PgI+PDyNGjDB7IMaMGUPXrl3dWW3RTJSVlbFp0yZLqN+xYwdA0w9Y9U0aOSGEaB0kYAkhqispKeGDDz4wb4i3bNlief6tt95i2rRpgDGv6/Dhw4wcOZJ27dq5o7qimTly5P+3d//BUdd3Hsef7yQkRFGwIHJRQH6sLQgBCYb1UqUnUjTjeDNYq527Xm/mjFjs1cN2plNHr3ced/VG6+mcbcGiUpU721On1jp3tchJUbJgBMWAYPBMMaVSFAERMRDe98d3WTcbAhvY3e93N69HJgP73S/Z1/eT5fvZ936++/nsZPjw4UU5i6CIiIiISJ9VVVXR1NREU1MTECyGfGSa+ObmZhoaPh3lX7JkCXfffTfl5eXU1tZ2+6xNLBbTYsjSw5lnnnlS/14FloiIiIgUtSFDhjB79mxmz57d476amhqmTp3Khg0bWL9+PevXr+fHP/4xAJdccgkrV64EglnoPvzwQy2GLCdNBZaIiIiIlKwFCxawYMEC9u3bxyuvvNJtsdojCyEDtLe3M27cOCZOnNhtlGvixIm5Wzfp0CF45x3Yvh06O6GyEmpqYORI0FT0JUO/SREREREpeYMGDWLmzJnMnDkTCEas0meP27JlCxUVFWzcuJGNGzfy4IMPAsG6SfX19Tz88MOcc845J/bg7tDaCm3JCbK70uZ727ED1q+HWAwmTQJdslj0VGCJiIiISL9jZlRXV6duX3755ezdu5dXX301NXlGc3Mz27ZtY+XKlQwdOjS1b1NTEwcOHEiNctXW1jJgwICjP5A7vPQS/PGP3QurI45sa2uDPXugoUFFVpFTgSUiIiIiAgwcODBVNB2xfft23njjjVQxdvjwYZ544gl2797NY489BkB1dTXTp08nHo/zpS99ifr6+k9/aGtr78VVuq6uYL/WVpg8OefHJoWjAktEREREpBc1NTXU1NR02/b88893Wzepra2NVatWsWrVKkaNGpUqsFoSCVYtXkx8/HimjRlDy/wRHNzT8/NcAwYfpuEn7wdFVlsbTJigz2QVMf3mRERERESyVFZWxrRp05g2bRrz588H4L333mPt2rUkEgkuu+yy1L5PPfII31+6FIDKigrGHYoxMe1rBCMAehZd77wDY8YU5Hgk91RgiYiIiIichGHDhtHY2EhjY2O37RePHcvOSy+lua2NTR0dvJH8epInGc1olrI0te+qN97ggjFjGDRwYDDLoAqsoqUCS0REREQkD66oq+OK0aMB2LN/P0v++j02Jb9GMjK130528uXvfY8yM2pHjyZeW0t87lzi8TjnnXeeFkMuMiqwRERERETyobIy9dfBp5xCXfIr0252Uzd2LK+2t6e+F/3ylwCcccYZvPDCC9TW1gLQ1dVFeXl5YfLLCVGBJSIiIiKSDzU1wTpXx5lBMEaMljvvZP8nn/BKezuJjz4i8eabNDc3s2PHDsaNG5fad+7cuWzdurXHYsgquqJDBZaIiIiISD6MHBksIpw0YPDhXmcRBDilqoqLJ07k4quugooK3J0dO3Zw6qmnpvZdt24dHR0dbNq0iYceeggIFlGur6+nqamJ6667Ls8HJcejAktEREREJB8qKiAWC6Ze7+oKpmI/lvLyYP/kFO1mxogRI7rtsnXr1m6LIScSCdrb21mxYgVXXnllar9Vq1axePHi1CjXlClTel8MWXJKBZaIiIiISL5MmgR79hx/seHychg+PNj/GKqqqpgxYwYzZszg5ptvBuDdd98lkUgwZcqU1H7Lly9n2bJlLFu2DAgWUa6rq+Oiiy4iHo8zd+5cTZ6RJ+buYWc4IdOnT/eWlpawY4iISJ6Z2SvuPj3sHKC+R0ROkDu0tgYjWdC90KqoCO6PxYLiKkdFT1tbW7cFkbds2ZK6b/z48bQdyQIsXryYyZMnM23aNAYOHJiTxy92J9P3aARLRERERCSfzGDyZJgwIVhEePt2OHgQBgwIJsIYOTJ1WWCuxGIxYrEYN954IwC7du1izZo1JBIJBg0alNrv3XffTe0zYMAApk6dmhrlisfjnHvuuRrp6iONYImISKRpBEtEJH+2bdvGwoULSSQStLa2klkbLF++nFmzZgHQ0dHBkCFDuhVopUojWCIiIiIi0mejRo3igQceAGDv3r28/PLLqcsK165dy/Tpn9YYN998M7/4xS+YNGkS8Xg8NdJ13nnnUVbWc3bE/koFloiIiIiIcPrppzNr1qzUiJW7d7s8sLOzk7KyMjZs2MCGDRtShdmQIUO45ZZbuP3220PJHTUqsEREREREpIfMz14988wzfPzxx6xbt47m5mYSiQTNzc1s376dqqqq1H4rVqzgpptu6jbKdf755/ebxZBVYImIiIiISFaqq6tpaGigoaEhta2jo4PKysrU7ebmZjZv3szmzZtZunQpECyGfOGFFxKPx7njjjuoyPGkHlGiSS5ERCTSNMmFiEhx6ezs7LEY8ttvvw3A6NGjaW9vT+27YMECxo4dm1oMOb1QC5MmuRARERERkUiorKykvr6e+vp6vvnNbwKwY8cO1qxZw0cffZTa7/333+fee+9N3a6qqqKuri41Rfyll17K0KFDC57/ZEWmwDKzmcC/EWTqBL7h7olwU4mISKkys3nAfOAgUAkscvcfhZtKRKQ0nXXWWVx11VXdtlVUVLBo0aLUKNfmzZtZvXo1q1evBuDZZ5+lsbERgHXr1rF//37q6uqorq4ueP6+iESBZWZDgKeAK9292cy+ADxtZmPcfX+46UREpNSYWTkQAxrcfZ+ZnQ1sNbOn3f33IccTEekXBg8ezLx585g3bx4AH3zwAWvXrk1NnlFfX5/a96677uLxxx+noqKCqVOnpka54vE4Y8eOjdRiyJEosIA5wBZ3bwZw9xfM7A/ALOCZUJOJiEjJcfcu4Ntpm94nuHqif0xxJSISQWeccQZz5sxhzpw5Pe6LxWJMmTKF119/nZaWFlpaWrj//vsBuPbaa3n88ccBOHjwIAcOHOC0004raPZ0kZjkwsy+C0x096+mbXsS+K2735e27QbghuTNSUBrQYNGyzDgvbBDhETH3j/152OH/n38n3X3E+opzawS+O1R7rrE3TvT9lsElLn7DZk7lljfU8zPo2LODsWdX9nDU8z5izk7nETfE5URLAO6MrYdArotCe3uDwAPAJhZS1RmlQpDfz5+HbuOvT/qz8dvZic8bV+yiIof5+cvBM4Gru7lZ5RM31PM+Ys5OxR3fmUPTzHnL+bscHJ9T9nxdymIDmBUxrZRye0iIiJ5YWZ3A+cDV6ePaImIiJyoqBRYTwO1ZjYZwMzqgc8Bvwk1lYiIlCQzK0teFjgSuEbFlYiI5EokLhF09z1mdg3wkJk5weWBje6++xj/7IHCpIus/nz8Ovb+qT8fO/Tv48/HsTcC84AW4MW02aduc/flBc5SSMWcv5izQ3HnV/bwFHP+Ys4OJ5E/EpNciIiIiIiIlIKoXCIoIiIiIiJS9FRgiYiIiIiI5EjRFVhmNtPM1pnZBjNrMbNjTr9basxsgJl928wOmtl1YecpJDObZ2avJX/vG8xsftiZCsXM/il57GuTz/+vh52p0MzsfDPbZWb/EHaWQjKz183sZTNLJL9XhJ2pkMzsW2bWmmyDFjM7pcCPn/V5x8xuMrPNyby/MrOzCpn1KHmy6i/MrC75fyuR9v2tQmbtJVe2+S15jtxiZpvM7DEzO7WQWY+SKavXKlFq+2wyR7Gtk7myyR6Zts6UzXM9wm2fTfYot31W5/i+nt8jMclFtsxsCPAUcKW7N5vZF4CnzWyMu+8PN13BNAEOJMIOUkhmVg7EgAZ332dmZwNbzexpd/99yPEK4QNgursfNLMzgbfN7H/c/e2wgxVC8v/+D4H/DDtLCE4Dprj74bCDFJqZ/SNQBUx190PJ58GBAj5+1uedZH/0HaDO3Xcm3wh4ELiyUHmPItv+YhjwpLs35T9Sn2Sb/2sEk5ZMdfePzexh4F+Bb+Q531H18bVKJNq+D5kj1dbQp+yRaOteZPNcj1zbJ2WTPZJtn+05/kTO78U2gjUH2OLuzQDu/gLwB2BWmKEKyd1/5O4/oOfCzCXN3bvc/dvuvi+56X2gEygPMVbBuPs97n4wefNcYB+wK7xEhWNmZcBPgVuBnSHHCcNngJVmtt7Mfm5mU8MOVAjJF01zgA+Bl8xsJXBRIQvNPp53rgUec/cjz9H7gMvNbHABoh5VH/qLYcAXzWxN8vufzey0AkQ8pj7kvxZY7O4fJ2/fB3wlr+GOrS+vVaLS9tlmjlpbQ/bZo9LWPWT5XI9i22ebPZJt34dzfJ/P78VWYI0F3srY9lZyu/Qv9wI/c/dtYQcpFDOLmVkb8GvgL919T9iZCmQh8Bt3Xx12kJCc5e4XA3UEawYuN7PMhdlLUR3BAsAd7j4D+DvgP8xsXK4fyMwqMy5dOfJdmbHrsc473fond/8A2EPwhkje9CH7sTwFnJts5yuAMcAjeQmcIUf5M18bvAV8Jt/FbW/ZCdbxzPa1SmhtnyHb11ehtPVxZJs9Km19oqLY9tkqlrbv7Rzf5/N7UV0iCBg9K+RDFF+hKCfBzBYCZwNXh52lkNy9DYiZ2QXAs2Y22903hp0rn8zsamCUu98adpawHHm3Mjlys8zM/orgHdufhBos/4YD7e7+UwB3X29mzwOXE1wumjPJRYaP+XneLM47ofRP2WTP4md8nPb3XcnPRrxjZtXp9+VDLvLTs+0PJf8Mpe3N7FayfC6E2fYZsn3+htLWx5FV9gi19YmKYttnpRja/jjn+D6f3yP/S8nQAWS+czsquV36ATO7m+Bd7auTnVu/4+7rgWbgz8LOUgBXABPS3hm+HrjezP4r5FxhKgf2hh2iAP5Iz+M8TAiXR2d53unWP1kwGcdQirN/Kif4rNsnYQfJUuZrg1EEl1HvDifOSb1WCavts80ctbaGE29vPc/DE6m2z+Ic3+fze7EVWE8DtWY2GcDM6gmG4n8TairJOzMrM7NFwEjgmv5UXJnZZDP7splZ8vbZwAygJdxk+efu17t7nbvH3T0OLAGWuPs1YWcrBDO70Mzq0m43AhMILhMtdauAsWb2eQAz+yxwKfBcoQIc67xjZuVm9ryZXZzc9CjwF2mX69wEvJR2zX5kmNlQM3vJzGLJ29clP/OGmVUA3wcejerEKpn5Cdr++rTLCv8WeMrdPZyEvb9WiXDb95Z5RcTbGrLMHqG2zkoRPM97FeHneWbOo57jc3F+L6pLBN19j5ldAzxkZk4wPNfo7sVYvUvfNALzCIqKF5O1BsBt7r48tFSF8TvgRuA7ZnYQqARud/d+NZNkP7UPuMfMRhC807cL+GJ/OOe5e2eyoFxiwUQnh4Gvufv/FTBGr+cd4EWCzxF8Jpn3f83shwQTkhwEtgNRXUrjFGA0cOTFQjXwvJkdJpgNbCXw9yFly0Zm/keA8cBaMzsEbCLEmdWO9VrFzEYSwbbvLTMRb2voU/ZItHUfRL7tj6FY2r63c/xCTvL8bhEsfEVERERERIpSsV0iKCIiIiIiElkqsERERERERHJEBZaIiIiIiEiOqMASERERERHJERVYIiIiIiIiOaICS0REREREJEdUYImIiIiIiOSICiwREREREZEcUYElIiIiIiKSIyqwREREREREckQFlkiEmNlzZuZmNjdju5nZ0uR9d4aVT0RESo/6HpHcMncPO4OIJJnZFGAdsAWY7O5dye0/AG4BfuLuN4QYUURESoz6HpHc0giWSIS4+2vAo8AE4KsAZnYrQQf3c+DG8NKJiEgpUt8jklsawRKJGDM7B2gDdgB3A/8O/Bq4yt07w8wmIiKlSX2PSO5oBEskYty9A7gXGE3Qwa0G5h6tgzOz+Wb2tpkdMLNXzOziAscVEZESoL5HJHdUYIlE0860v/+Nu+/P3MHMrgXuA/4FuICgM/xvMxtVmIgiIlJi1PeI5IAuERSJGDP7CrCM4DKNEcAid//6UfZbA2xw96a0bW3AE+7+3ULlFRGR4qe+RyR3NIIlEiFm1gj8FNgI1AKbgevN7HMZ+1UCdcBzGT/iOeBPCxBVRERKhPoekdxSgSUSEWb2eeAJoAP4orvvBG4HKoDM9UeGAeUE7zSmO/LOo4iIyHGp7xHJPRVYIhGQXIPkV8AeYLa7/wHA3Z8AWoA/7+VDxJnX+NpRtomIiPSgvkckP1RgiYTMzMYTTIXrwBx3fytjlyPXtN+Vtu09oIue7xgOp+c7iyIiIt2o7xHJH01yIVKkkh80fs3db0jb9ibwpD5oLCIi+aC+R+T4KsIOICIn7B7gUTNbC7wE3AjUAItCTSUiIqVMfY/IcajAEilS7v4zMxsK3Ab8CdAKNLr778JNJiIipUp9j8jx6RJBERERERGRHNEkFyIiIiIiIjmiAktERERERCRHVGCJiIiIiIjkiAosERERERGRHFGBJSIiIiIikiMqsERERERERHJEBZaIiIiIiEiOqMASERERERHJERVYIiIiIiIiOaICS0REREREJEf+H5RrOOVsnvAxAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed1c34a0b8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Xs = np.array([[1, 50], [5, 20], [3, 80], [5, 60]]).astype(np.float64)\n",
    "ys = np.array([0, 0, 1, 1])\n",
    "svm_clf = SVC(kernel=\"linear\", C=100)\n",
    "svm_clf.fit(Xs, ys)\n",
    "\n",
    "plt.figure(figsize=(12,3.2))\n",
    "plt.subplot(121)\n",
    "plt.plot(Xs[:, 0][ys==1], Xs[:, 1][ys==1], \"bo\")\n",
    "plt.plot(Xs[:, 0][ys==0], Xs[:, 1][ys==0], \"ms\")\n",
    "plot_svc_decision_boundary(svm_clf, 0, 6)\n",
    "plt.xlabel(\"$x_0$\", fontsize=20)\n",
    "plt.ylabel(\"$x_1$  \", fontsize=20, rotation=0)\n",
    "plt.title(\"스케일 조정 전\", fontsize=16)\n",
    "plt.axis([0, 6, 0, 90])\n",
    "\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "scaler = StandardScaler()\n",
    "X_scaled = scaler.fit_transform(Xs)\n",
    "svm_clf.fit(X_scaled, ys)\n",
    "\n",
    "plt.subplot(122)\n",
    "plt.plot(X_scaled[:, 0][ys==1], X_scaled[:, 1][ys==1], \"bo\")\n",
    "plt.plot(X_scaled[:, 0][ys==0], X_scaled[:, 1][ys==0], \"ms\")\n",
    "plot_svc_decision_boundary(svm_clf, -2, 2)\n",
    "plt.xlabel(\"$x_0$\", fontsize=20)\n",
    "plt.title(\"스케일 조정 후\", fontsize=16)\n",
    "plt.axis([-2, 2, -2, 2])\n",
    "\n",
    "save_fig(\"sensitivity_to_feature_scales_plot\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Sensitivity to outliers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saving figure sensitivity_to_outliers_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed25fe9fd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_outliers = np.array([[3.4, 1.3], [3.2, 0.8]])\n",
    "y_outliers = np.array([0, 0])\n",
    "Xo1 = np.concatenate([X, X_outliers[:1]], axis=0)\n",
    "yo1 = np.concatenate([y, y_outliers[:1]], axis=0)\n",
    "Xo2 = np.concatenate([X, X_outliers[1:]], axis=0)\n",
    "yo2 = np.concatenate([y, y_outliers[1:]], axis=0)\n",
    "\n",
    "svm_clf2 = SVC(kernel=\"linear\", C=10**9)\n",
    "svm_clf2.fit(Xo2, yo2)\n",
    "\n",
    "plt.figure(figsize=(12,2.7))\n",
    "\n",
    "plt.subplot(121)\n",
    "plt.plot(Xo1[:, 0][yo1==1], Xo1[:, 1][yo1==1], \"bs\")\n",
    "plt.plot(Xo1[:, 0][yo1==0], Xo1[:, 1][yo1==0], \"yo\")\n",
    "plt.text(0.3, 1.0, \"불가능!\", fontsize=24, color=\"red\")\n",
    "plt.xlabel(\"꽃잎 길이\", fontsize=14)\n",
    "plt.ylabel(\"꽃잎 폭\", fontsize=14)\n",
    "plt.annotate(\"이상치\",\n",
    "             xy=(X_outliers[0][0], X_outliers[0][1]),\n",
    "             xytext=(2.5, 1.7),\n",
    "             ha=\"center\",\n",
    "             arrowprops=dict(facecolor='black', shrink=0.1),\n",
    "             fontsize=16,\n",
    "            )\n",
    "plt.axis([0, 5.5, 0, 2])\n",
    "\n",
    "plt.subplot(122)\n",
    "plt.plot(Xo2[:, 0][yo2==1], Xo2[:, 1][yo2==1], \"bs\")\n",
    "plt.plot(Xo2[:, 0][yo2==0], Xo2[:, 1][yo2==0], \"yo\")\n",
    "plot_svc_decision_boundary(svm_clf2, 0, 5.5)\n",
    "plt.xlabel(\"꽃잎 길이\", fontsize=14)\n",
    "plt.annotate(\"이상치\",\n",
    "             xy=(X_outliers[1][0], X_outliers[1][1]),\n",
    "             xytext=(3.2, 0.08),\n",
    "             ha=\"center\",\n",
    "             arrowprops=dict(facecolor='black', shrink=0.1),\n",
    "             fontsize=16,\n",
    "            )\n",
    "plt.axis([0, 5.5, 0, 2])\n",
    "\n",
    "save_fig(\"sensitivity_to_outliers_plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Large margin *vs* margin violations"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is the first code example in chapter 5:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "     steps=[('scaler', StandardScaler(copy=True, with_mean=True, with_std=True)), ('linear_svc', LinearSVC(C=1, class_weight=None, dual=True, fit_intercept=True,\n",
       "     intercept_scaling=1, loss='hinge', max_iter=1000, multi_class='ovr',\n",
       "     penalty='l2', random_state=42, tol=0.0001, verbose=0))])"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "from sklearn import datasets\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.svm import LinearSVC\n",
    "\n",
    "iris = datasets.load_iris()\n",
    "X = iris[\"data\"][:, (2, 3)]  # petal length, petal width\n",
    "y = (iris[\"target\"] == 2).astype(np.float64)  # Iris-Virginica\n",
    "\n",
    "svm_clf = Pipeline([\n",
    "        (\"scaler\", StandardScaler()),\n",
    "        (\"linear_svc\", LinearSVC(C=1, loss=\"hinge\", random_state=42)),\n",
    "    ])\n",
    "\n",
    "svm_clf.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1.])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "svm_clf.predict([[5.5, 1.7]])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now let's generate the graph comparing different regularization settings:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "     steps=[('scaler', StandardScaler(copy=True, with_mean=True, with_std=True)), ('linear_svc', LinearSVC(C=100, class_weight=None, dual=True, fit_intercept=True,\n",
       "     intercept_scaling=1, loss='hinge', max_iter=1000, multi_class='ovr',\n",
       "     penalty='l2', random_state=42, tol=0.0001, verbose=0))])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scaler = StandardScaler()\n",
    "svm_clf1 = LinearSVC(C=1, loss=\"hinge\", random_state=42)\n",
    "svm_clf2 = LinearSVC(C=100, loss=\"hinge\", random_state=42)\n",
    "\n",
    "scaled_svm_clf1 = Pipeline([\n",
    "        (\"scaler\", scaler),\n",
    "        (\"linear_svc\", svm_clf1),\n",
    "    ])\n",
    "scaled_svm_clf2 = Pipeline([\n",
    "        (\"scaler\", scaler),\n",
    "        (\"linear_svc\", svm_clf2),\n",
    "    ])\n",
    "\n",
    "scaled_svm_clf1.fit(X, y)\n",
    "scaled_svm_clf2.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Convert to unscaled parameters\n",
    "b1 = svm_clf1.decision_function([-scaler.mean_ / scaler.scale_])\n",
    "b2 = svm_clf2.decision_function([-scaler.mean_ / scaler.scale_])\n",
    "w1 = svm_clf1.coef_[0] / scaler.scale_\n",
    "w2 = svm_clf2.coef_[0] / scaler.scale_\n",
    "svm_clf1.intercept_ = np.array([b1])\n",
    "svm_clf2.intercept_ = np.array([b2])\n",
    "svm_clf1.coef_ = np.array([w1])\n",
    "svm_clf2.coef_ = np.array([w2])\n",
    "\n",
    "# Find support vectors (LinearSVC does not do this automatically)\n",
    "t = y * 2 - 1\n",
    "support_vectors_idx1 = (t * (X.dot(w1) + b1) < 1).ravel()\n",
    "support_vectors_idx2 = (t * (X.dot(w2) + b2) < 1).ravel()\n",
    "svm_clf1.support_vectors_ = X[support_vectors_idx1]\n",
    "svm_clf2.support_vectors_ = X[support_vectors_idx2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saving figure regularization_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed1cee7828>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,3.2))\n",
    "plt.subplot(121)\n",
    "plt.plot(X[:, 0][y==1], X[:, 1][y==1], \"g^\", label=\"Iris-Virginica\")\n",
    "plt.plot(X[:, 0][y==0], X[:, 1][y==0], \"bs\", label=\"Iris-Versicolor\")\n",
    "plot_svc_decision_boundary(svm_clf1, 4, 6)\n",
    "plt.xlabel(\"꽃잎 길이\", fontsize=14)\n",
    "plt.ylabel(\"꽃잎 폭\", fontsize=14)\n",
    "plt.legend(loc=\"upper left\", fontsize=14)\n",
    "plt.title(\"$C = {}$\".format(svm_clf1.C), fontsize=16)\n",
    "plt.axis([4, 6, 0.8, 2.8])\n",
    "\n",
    "plt.subplot(122)\n",
    "plt.plot(X[:, 0][y==1], X[:, 1][y==1], \"g^\")\n",
    "plt.plot(X[:, 0][y==0], X[:, 1][y==0], \"bs\")\n",
    "plot_svc_decision_boundary(svm_clf2, 4, 6)\n",
    "plt.xlabel(\"꽃잎 길이\", fontsize=14)\n",
    "plt.title(\"$C = {}$\".format(svm_clf2.C), fontsize=16)\n",
    "plt.axis([4, 6, 0.8, 2.8])\n",
    "\n",
    "save_fig(\"regularization_plot\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "# Non-linear classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saving figure higher_dimensions_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsoAAAEYCAYAAABFiND5AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAFgBJREFUeJzt3X+MZWd5H/Dvg42rgJsuLWQ3gIzjQMqPBlJ5Gmgx8TUJuLFC0mKhQisU1JA1on8YpY5UrERytVSKFMKPRFBrI8VYkYKSolTEpAJDyuA4uyaM2QUTiSU2qtHCLHEg43gheNi7b/+YWXt25931rGdnzr13Ph9ptDPvPeeeR8+899V3z5x7brXWAgAAnO4pQxcAAACTSFAGAIAOQRkAADoEZQAA6BCUAQCgQ1AGAIAOQRkAADoEZQAA6BCUAQCg4+Lz2fiZz3xmu/zyyzd1wO985zt5+tOfvqnnmEX60qcvffrSd6H6cu+99/5ta+1ZF6CkqWOd3zr6st6RI0cyHo/z4he/eOhSJo750rfd6/x5BeXLL788CwsLT76qJPPz8xmNRpt6jlmkL3360qcvfReqL1X14OarmU7W+a2jL+uNRqMsLS1tes7NIvOlb7vXeZdeAABAh6AMAAAdgjIAAHQIygAA0CEoAwBAh6AMAAAdgjIAAHQIygBTrKqeWlU3VdX3q+qNZzz2X6vqS1X1uapaqKqnbVUde/YkVStf11wzeuz7PXu26ojATjHk+nJeHzgCwMT55SQtyT1rB6vqvyf5R0l+orV2oqp2JfneVhXxzW+e3zjARg25vjijDDDFWmsfbK39VpLxqbHVUHxtkkeS/EVVfSbJv26tnRyoTICp5IwywOy5MslLkvzP1trLq+pfJvm/VTXXWnvgzI2ram+SvUmye/fuzM/PP4lDjs76yJN7vtlz/PhxvTjD0tJSxuOxvnSYL2uNzvrIVvdIUAaYPT+U5P+11m5Pktbaoar6syT/NskHzty4tbY/yf4kmZuba6PR6IIWc6Gfb1rNz8/rxRl27dqVpaUlfekwXzZmq3vk0guA2fM3Sf7+jLGTWXN5BgBPTFAGmD1/nuSKqroqSarqnyd5dZI7t+qAu3ef3zjARg25vgjKADOmtbac5Lok76+qQ0n+IMkvtta+ulXHPHYsaW3l69Ofnn/s+2PHtuqIwE4x5PriGmWAGdBaG53x86GsvKkPgCfJGWUAAOgQlAEAoENQBgCADkEZAAA6BGUAAOgQlAEAoENQBgCADkEZAAA6BGUAAOgQlAEAoENQBgCADkEZAAA6BGUAAOgQlAEAoENQBgCADkEZAAA6BGUAAOgQlAEAoENQBgCADkEZAAA6BGUAAOgQlAEAoENQBgCADkEZAAA6BGUAAOgQlAEAoENQBgCADkEZAAA6BGUAAOgQlAEAoENQBgCADkEZAAA6BGUAAOgQlAEAoENQBgCADkEZAAA6BGUAAOgQlAEAoENQBgCADkEZAAA6BGWAKVZVT62qm6rq+1X1xjXjN1TVF6pqoaq+WFVvH7JOgGl08dAFALApv5ykJbnn1EBVXZTkBUle2Vo7XlXPSXJ/VX20tfb1geoEmDqCMsAUa619MEmq6nVrxsZJblqz2beSLCe5aHurA5hugjLA7Htfkj9srX2t92BV7U2yN0l2796d+fn5TR3s+PHjm36OWaQv6y0tLWU8HutLh/nSt919EZQBZlhVvSvJc5Jcf7ZtWmv7k+xPkrm5uTYajTZ1zPn5+Wz2OWaRvqy3a9euLC0t6UuH+dK33X0RlAFmVFW9O8mPJrm+tbY8dD0A00ZQBpgxVfWUJB9M8owkb2itnRi4JICp5PZwALPnuiQ3JLkiyd1Vdc/q188MXBfAVHFGGWAGtNZGa77/WJIarhqA2eCMMgAAdAjKAADQISgDAECHoAwAAB2CMgAAdAjKAADQISgDAECHoAwAAB2CMgAAdAjKAADQISgDAECHoAwAAB2CMgAAdAjKAADQISgDAECHoAwAAB2CMgAAdAjKAADQISgDAECHoAwAAB2CMgAAdAjKAADQISgDAECHoAwAAB2CMgAAdAjKABOiqu6sqlZVrz9jvKrqQ6uP/cZQ9QHsNIIywOT41SQnk7yrqi5aM/7uJL+Y5Hdba/9tkMoAdiBBGWBCtNa+kOT3k7woyZuTpKpuTvIrSf4oyduGq2567NmTVK3/2rNn6Mpgengdrbh46AIAOM2vJfkPSW6pqkuT/I8kn0jy5tbayUErmxLf/Ob5jQPreR2tcEYZYIK01o4meV+S5yX5nSQHkry+tba8druqemdVfa6q/r6qHqqqO6rqXwxQMsDMEpQBJs9Da77/pdbadzvbjJJ8MMm/SfLqJCeSfKqq/unWlwewMwjKABOkqt6UlTfvHVsdurG3XWvt2tbaba21L7XW7svKNc3PSvLK7akUYPYJygAToqquS3J7kr9K8tIkX07y1qp64QZ2/8dZWdP/busqBNhZBGWACVBVVyX5SJKjSV7bWnsoya9n5U3XG7l38vuTHE5ycMuKnBK7d5/fOLCe19EKd70AGFhVvSzJx5I8nOQ1rbXFJGmtfaSqFpL8QlW9qrX252fZ/z1JrkpyVWttvF11T6pjx554G+DcvI5WOKMMMKCqen5Wbv/WklzbWnvgjE3eufrvb55l//cmeVOSV7fWvrplhQLsQM4oAwyotXZ/krPewr+19qkk1Xusqt6f5I1JRq21L29NhQA7l6AMMIWq6gNZudPFv0vyd1V1Kmwfb60dH64ygNnh0guA6fT2rNzp4s+SLK75umnIogBmybYE5bWfF37NNaNBPy98Ej+7fPGRxdx4+MYcOz78lfOLjyzm6g9dPTG16Eu/lqH7Mkmvo0laX7ZTa63O8nXLqW2q6ler6otV9dnVf28esGSAqbMtQXmSPi98kmo5Zd9d+3Lfw/dl32f2DVfEmlru/trdE1OLvvRrGbovk/Q6mqRaJklVvSrJO5Jc01p7eVY+yW9vVb1m0MIApsh5XaN85MiRjEajJ3GY+bM+8uSebzPmz/rI9teSPHrJo/nLV/xl2kUtt3721hx6/6FcsnzJttextpaTF52cmFr0pV/L8H2ZP+sjO/01PUFOfQz2M5J8K8kPZuXkyA7/LwTAxnkz38AevPzBtLQkSUvLg897MC/46xeoRS0TXwuTrbX25aram+RzVfVQkh9K8p9ba188c9vV7fYmye7duzM/P7+pYx8/fnzTzzGL9GW9paWljMdjfekwX/q2uy/VWtvwxnNzc21hYeH8D9K9sdGK8zj8BTFJtSw+spgrfvuKfO/E9x4b+4GLfyBfvfGr2XPp9l5gqRa1nI9Jeh1tRS1VdW9rbe7J7T0ZqurHsvJGv3/fWltYvV/zx5K8pbV2z9n2e7Lr/Frz8/M7/Wx+l76sNxqNsrS0lMOHDw9dysQxX/ouVF82us6768WA9t21LyfbydPGxm08yLWnalELM+fnk9zdWltIHrtf8/9O8h8HrQpgimxLUJ6kzwufpFoOHj2Y5fHyaWPL4+UcOHpALWqZ6Fom6XU0SbVMmK8kuaqqfjhJquoHk1yb5MigVQFMkW25Rnnt54UP/aeESfrs8kM3HHrs+6H7sraWoelL3yT1ZZJeR5O0vkyS1tqfrF5u8fGq+ockT0vyp0luHbYygOnhzXwAM6q19p4k7xm6DoBp5RplAADoEJQBAKBDUAYAgA5BGQAAOgRlAADoEJQBAKBDUAYAgA5BGQAAOgRlAADoEJQBAKBDUAYAgA5BGYDHfeMbSdXGvvbuXb//3r0ZXXPNxva/5Zb1+7/udRs//v796/e/8sqN73/HHev3f/azN77/vfeu3/8c26/ryze+8eR7X7X+2Pfeu/F9n/3s9fvfccfG97/yyvX779+/8f1f97r1+99yy6bn3ob3n4K5d87X0XnOvXVfUzz3ruz97jc7985BUAYAgA5BGQAAOqq1tuGN5+bm2sLCwqYOOD8/n9FotKnnmEX60qcvffrSd6H6UlX3ttbmNl/R9LHObx19WW80GmVpaSmHDx8eupSJY770bfc674wyAAB0CMoAANAhKAMAQIegDAAAHYIyAAB0CMoAANAhKAMAQIegDAAAHYIyAAB0CMoAANAhKAMAQIegDAAAHYIyAAB0CMoAANAhKAMAQIegDAAAHYIyAAB0CMoAANAhKAMAQIegDAAAHYIyAAB0CMoAANAhKAMAQIegDAAAHYIywIyrqpdU1ber6paha9mpFh9ZzI2Hb8yx48eGLoUpYL5MDkEZYIZV1a4kH0jy4aFr2cn23bUv9z18X/Z9Zt/QpTAFzJfJISgDzKiqekqS25PcnOShgcvZsRYfWcxth29LS8tth29zlpBzMl8my8VDFwDAlnlXkk+21g5U1WvPtlFV7U2yN0l2796d+fn5TR30+PHjm36OWfLer7w3J8YnkiTfH38/b/vw2/KOF7xj4Komw9LSUsbjsfmyhvlybtu9vgjKADOoqq5Pcllr7eYn2ra1tj/J/iSZm5tro9FoU8een5/PZp9jViw+spg7/+LOnGgrwedEO5E7/+bO3PqmW7Pn0j0DVze8Xbt2ZWlpyXxZZb48se1eX1x6ATCbfjbJi6rqnqq6J8lbk7y1qv7XwHXtKPvu2peT7eRpY+M2du0pXebL5HFGGWAGtdbeuvbnU3e8aK3dMkQ9O9XBowezPF4+bWx5vJwDRw8MVBGTzHyZPIIyAGyRQzcceux7l6TwRMyXySMoA+wAziQDnD/XKAMAQIegDAAAHYIyAAB0CMoAANAhKAMAQIegDAAAHYIyAAB0CMoAANAhKAMAQIegDAAAHYIyAAB0CMoAANAhKAMAQIegDAAAHYIyAAB0CMoAANAhKAMAQIegDAAAHYIyAAB0CMoAANAhKAMAQIegDAAAHYIyAAB0CMoAANAhKAMAQIegDAAAHYIyAAB0CMoAANAhKAPADrL4yGKu/tDVOXb82NClTBR9oUdQBoAdZN9d+3L31+7Ovs/sG7qUiaIv9AjKALBDLD6ymNsO35aT7WRuO3ybs6er9IWzEZQBYIfYd9e+nGwnkyTjNnb2dJW+cDaCMgDsAKfOmi6Pl5Mky+NlZ0+jL5yboAwAO8Das6anOHuqL5yboAwAO8DBowcfO2t6yvJ4OQeOHhioosmgL5zLxUMXAABsvUM3HBq6hImkL5yLM8oAANAhKAMAQIegDAAAHYIyAAB0CMoAM6qqrq6qz1fVF6tqoapeMXRNsNajlzya+191v3sWM7EEZYAZVFW7kvxxkv/SWntpkpuSfLSqnjZsZfC4By9/MN/5Z99xz2ImltvDAcyma5Mcaa0dTJLW2nxVLSb56SR3nG2nI0eOZDQaberAS0tL2bVr16aeYxbpy+keveTRLL58Mank1s/emkPvP5RLli8ZuqyJYb70bXdfBGWA2XRFkgfOGHtgdfw0VbU3yd4keepTn5qlpaVNHXg8Hm/6OWaRvpzu6MuOPvb9yZzMV374K3nuF547YEWTxXzp2+6+CMoAs6mSjM8YO5HOJXettf1J9ifJ3NxcW1hY2NSB5+fnN31Wehbpy+MWH1nMFb99xcqMTJKLku/+2Hfz8Q98PHsu3TNobZPCfOm7UH2pqg1t5xplgNl0NMllZ4xdtjoOg9p3176cbCdPGxu3sWuVmTiCMsBs+miSl1bVjydJVf1kkhcm+eSgVUGSg0cPZnm8fNrY8ng5B44eGKgi6HPpBcAMaq09XFVvSPJ7VdWy8kfu61prLnpkcIduOJQkGY1GWVpayuHDhweuCPoEZYAZ1Vr7dJJ/NXQdANPKpRcAANAhKAMAQIegDAAAHYIyAAB0CMoAANAhKAMAQEe11ja+cdVDSR7c5DGfmeRvN/kcs0hf+vSlT1/6LlRfntdae9YFeJ6pY53fUvrSpy99+tK3rev8eQXlC6GqFlprc9t60CmgL3360qcvffoyGfwe+vSlT1/69KVvu/vi0gsAAOgQlAEAoGOIoLx/gGNOA33p05c+fenTl8ng99CnL3360qcvfdval22/RhkAAKaBSy8AAKBDUAYAgI5Bg3JVvaSqvl1VtwxZx6Soqhuq6gtVtVBVX6yqtw9d01Cq6uqq+vxqHxaq6hVD1zQJzJFzs6ZMHr+T03kNP84632eOnNt2rykXb8dBeqpqV5IPJPnwUDVMkqq6KMkLkryytXa8qp6T5P6q+mhr7esDl7etVufGHyf5udbawaoaJfloVf1Ia+27w1Y3HHPk3Kwpk8fv5HRew4+zzveZI+c2xJoyyBnlqnpKktuT3JzkoSFqmDSttXFr7abW2vHVoW8lWU5y0YBlDeXaJEdaaweTpLU2n2QxyU8PWdTQzJGzs6ZMHr+T9byGT2Od7zBHzm6oNWXLzihX1SVJ7uo89FNJbknyydbagap67VbVMInO1ZfW2vKan9+X5A9ba1/bnsomyhVJHjhj7IHVcR63k+fImd6VHbqmDMk632ed3xDr/Mbs5DlypkHW+S0LyquLwbrrjarq+iSXtdZu3qpjT7Kz9WWtqnpXkuckuX5bipo8lWR8xtiJePPpY8yRx+30NWVI1vk+6/yGWOefgDnyuCHXlCGuUf7ZJC+qqntWf35usnJxdmvtDQPUM1Gq6t1JfjTJ9WecedhJjib5mTPGLkvykQFqmTjmyDrWlMnjd3IOXsNJrPPnZI6sM9iaMvgHjpx612Jr7ZZBCxnY6rU3H0zyjCT/qbV2YuCSBlNV/yQrf4K7prV2X1X9ZJJPJPmR1trSsNUNxxzZGGvK5PE7WeE1/DjrfJ85sjHbuaYMdtcL1rkuyQ1JFpLcXVWnxn+ttfapwaoaQGvt4ap6Q5Lfq6qWlT/HXbeTF89V5ghMN6/hVdb5szJHJszgZ5QBAGASuWgeAAA6BGUAAOgQlAEAoENQBgCADkEZAAA6BGUAAOgQlAEAoENQBgCADkEZAAA6BGUAAOgQlBlUVd1ZVa2qXn/GeFXVh1Yf+42h6gNgc6zzTLNqrQ1dAztYVb0syeeTHEny46218er4byX5lSS/21rbO2CJAGyCdZ5p5owyg2qtfSHJ7yd5UZI3J0lV3ZyVxfOPkrxtuOoA2CzrPNPMGWUGV1XPTfLXSb6Z5N1JfifJJ5L8fGttecjaANg86zzTyhllBtdaO5rkfUmel5XF80CS15+5eFbVT1XVn1TV11evaXvL9lcLwPmyzjOtBGUmxUNrvv+l1tp3O9tcmuRLSW5M8g/bUhUAF4p1nqkjKDO4qnpTVv4Ud2x16Mbedq21/9Nau7m19pEkJ7erPgA2xzrPtBKUGVRVXZfk9iR/leSlSb6c5K1V9cJBCwPggrDOM80EZQZTVVcl+UiSo0le21p7KMmvJ7k4iXtqAkw56zzTTlBmEKv31fxYkoeTvKa1tpgkq39uW0jyC1X1qgFLBGATrPPMAkGZbVdVz8/KbYFakmtbaw+csck7V//9zW0tDIALwjrPrLh46ALYeVpr9yfZc47HP5Wktq8iAC4k6zyzQlBmalTVpUmev/rjU5JcVlU/keTbrbWvDVcZABeCdZ5J45P5mBpVNUry6c5Dt7fW3rK91QBwoVnnmTSCMgAAdHgzHwAAdAjKAADQISgDAECHoAwAAB2CMgAAdAjKAADQISgDAECHoAwAAB2CMgAAdAjKAADQ8f8BC/5SyG3bOI4AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed1cd5cb38>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X1D = np.linspace(-4, 4, 9).reshape(-1, 1)\n",
    "X2D = np.c_[X1D, X1D**2]\n",
    "y = np.array([0, 0, 1, 1, 1, 1, 1, 0, 0])\n",
    "\n",
    "plt.figure(figsize=(11, 4))\n",
    "\n",
    "plt.subplot(121)\n",
    "plt.grid(True, which='both')\n",
    "plt.axhline(y=0, color='k')\n",
    "plt.plot(X1D[:, 0][y==0], np.zeros(4), \"bs\")\n",
    "plt.plot(X1D[:, 0][y==1], np.zeros(5), \"g^\")\n",
    "plt.gca().get_yaxis().set_ticks([])\n",
    "plt.xlabel(r\"$x_1$\", fontsize=20)\n",
    "plt.axis([-4.5, 4.5, -0.2, 0.2])\n",
    "\n",
    "plt.subplot(122)\n",
    "plt.grid(True, which='both')\n",
    "plt.axhline(y=0, color='k')\n",
    "plt.axvline(x=0, color='k')\n",
    "plt.plot(X2D[:, 0][y==0], X2D[:, 1][y==0], \"bs\")\n",
    "plt.plot(X2D[:, 0][y==1], X2D[:, 1][y==1], \"g^\")\n",
    "plt.xlabel(r\"$x_1$\", fontsize=20)\n",
    "plt.ylabel(r\"$x_2$\", fontsize=20, rotation=0)\n",
    "plt.gca().get_yaxis().set_ticks([0, 4, 8, 12, 16])\n",
    "plt.plot([-4.5, 4.5], [6.5, 6.5], \"r--\", linewidth=3)\n",
    "plt.axis([-4.5, 4.5, -1, 17])\n",
    "\n",
    "plt.subplots_adjust(right=1)\n",
    "\n",
    "save_fig(\"higher_dimensions_plot\", tight_layout=False)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed1cd32400>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.datasets import make_moons\n",
    "X, y = make_moons(n_samples=100, noise=0.15, random_state=42)\n",
    "\n",
    "def plot_dataset(X, y, axes):\n",
    "    plt.plot(X[:, 0][y==0], X[:, 1][y==0], \"bs\")\n",
    "    plt.plot(X[:, 0][y==1], X[:, 1][y==1], \"g^\")\n",
    "    plt.axis(axes)\n",
    "    plt.grid(True, which='both')\n",
    "    plt.xlabel(r\"$x_1$\", fontsize=20)\n",
    "    plt.ylabel(r\"$x_2$\", fontsize=20, rotation=0)\n",
    "\n",
    "plot_dataset(X, y, [-1.5, 2.5, -1, 1.5])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "     steps=[('poly_features', PolynomialFeatures(degree=3, include_bias=True, interaction_only=False)), ('scaler', StandardScaler(copy=True, with_mean=True, with_std=True)), ('svm_clf', LinearSVC(C=10, class_weight=None, dual=True, fit_intercept=True,\n",
       "     intercept_scaling=1, loss='hinge', max_iter=1000, multi_class='ovr',\n",
       "     penalty='l2', random_state=42, tol=0.0001, verbose=0))])"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.datasets import make_moons\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.preprocessing import PolynomialFeatures\n",
    "\n",
    "polynomial_svm_clf = Pipeline([\n",
    "        (\"poly_features\", PolynomialFeatures(degree=3)),\n",
    "        (\"scaler\", StandardScaler()),\n",
    "        (\"svm_clf\", LinearSVC(C=10, loss=\"hinge\", random_state=42))\n",
    "    ])\n",
    "\n",
    "polynomial_svm_clf.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saving figure moons_polynomial_svc_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed1ccf3ac8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_predictions(clf, axes):\n",
    "    x0s = np.linspace(axes[0], axes[1], 100)\n",
    "    x1s = np.linspace(axes[2], axes[3], 100)\n",
    "    x0, x1 = np.meshgrid(x0s, x1s)\n",
    "    X = np.c_[x0.ravel(), x1.ravel()]\n",
    "    y_pred = clf.predict(X).reshape(x0.shape)\n",
    "    y_decision = clf.decision_function(X).reshape(x0.shape)\n",
    "    plt.contourf(x0, x1, y_pred, cmap=plt.cm.brg, alpha=0.2)\n",
    "    plt.contourf(x0, x1, y_decision, cmap=plt.cm.brg, alpha=0.1)\n",
    "\n",
    "plot_predictions(polynomial_svm_clf, [-1.5, 2.5, -1, 1.5])\n",
    "plot_dataset(X, y, [-1.5, 2.5, -1, 1.5])\n",
    "\n",
    "save_fig(\"moons_polynomial_svc_plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "     steps=[('scaler', StandardScaler(copy=True, with_mean=True, with_std=True)), ('svm_clf', SVC(C=5, cache_size=200, class_weight=None, coef0=1,\n",
       "  decision_function_shape='ovr', degree=3, gamma='auto', kernel='poly',\n",
       "  max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
       "  tol=0.001, verbose=False))])"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.svm import SVC\n",
    "\n",
    "poly_kernel_svm_clf = Pipeline([\n",
    "        (\"scaler\", StandardScaler()),\n",
    "        (\"svm_clf\", SVC(kernel=\"poly\", degree=3, coef0=1, C=5))\n",
    "    ])\n",
    "poly_kernel_svm_clf.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "     steps=[('scaler', StandardScaler(copy=True, with_mean=True, with_std=True)), ('svm_clf', SVC(C=5, cache_size=200, class_weight=None, coef0=100,\n",
       "  decision_function_shape='ovr', degree=10, gamma='auto', kernel='poly',\n",
       "  max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
       "  tol=0.001, verbose=False))])"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "poly100_kernel_svm_clf = Pipeline([\n",
    "        (\"scaler\", StandardScaler()),\n",
    "        (\"svm_clf\", SVC(kernel=\"poly\", degree=10, coef0=100, C=5))\n",
    "    ])\n",
    "poly100_kernel_svm_clf.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saving figure moons_kernelized_polynomial_svc_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed1c475908>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(11, 4))\n",
    "\n",
    "plt.subplot(121)\n",
    "plot_predictions(poly_kernel_svm_clf, [-1.5, 2.5, -1, 1.5])\n",
    "plot_dataset(X, y, [-1.5, 2.5, -1, 1.5])\n",
    "plt.title(r\"$d=3, r=1, C=5$\", fontsize=18)\n",
    "\n",
    "plt.subplot(122)\n",
    "plot_predictions(poly100_kernel_svm_clf, [-1.5, 2.5, -1, 1.5])\n",
    "plot_dataset(X, y, [-1.5, 2.5, -1, 1.5])\n",
    "plt.title(r\"$d=10, r=100, C=5$\", fontsize=18)\n",
    "\n",
    "save_fig(\"moons_kernelized_polynomial_svc_plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saving figure kernel_method_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed1ccaa470>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def gaussian_rbf(x, landmark, gamma):\n",
    "    return np.exp(-gamma * np.linalg.norm(x - landmark, axis=1)**2)\n",
    "\n",
    "gamma = 0.3\n",
    "\n",
    "x1s = np.linspace(-4.5, 4.5, 200).reshape(-1, 1)\n",
    "x2s = gaussian_rbf(x1s, -2, gamma)\n",
    "x3s = gaussian_rbf(x1s, 1, gamma)\n",
    "\n",
    "XK = np.c_[gaussian_rbf(X1D, -2, gamma), gaussian_rbf(X1D, 1, gamma)]\n",
    "yk = np.array([0, 0, 1, 1, 1, 1, 1, 0, 0])\n",
    "\n",
    "plt.figure(figsize=(11, 4))\n",
    "\n",
    "plt.subplot(121)\n",
    "plt.grid(True, which='both')\n",
    "plt.axhline(y=0, color='k')\n",
    "plt.scatter(x=[-2, 1], y=[0, 0], s=150, alpha=0.5, c=\"red\")\n",
    "plt.plot(X1D[:, 0][yk==0], np.zeros(4), \"bs\")\n",
    "plt.plot(X1D[:, 0][yk==1], np.zeros(5), \"g^\")\n",
    "plt.plot(x1s, x2s, \"g--\")\n",
    "plt.plot(x1s, x3s, \"b:\")\n",
    "plt.gca().get_yaxis().set_ticks([0, 0.25, 0.5, 0.75, 1])\n",
    "plt.xlabel(r\"$x_1$\", fontsize=20)\n",
    "plt.ylabel(r\"유사도\", fontsize=14)\n",
    "plt.annotate(r'$\\mathbf{x}$',\n",
    "             xy=(X1D[3, 0], 0),\n",
    "             xytext=(-0.5, 0.20),\n",
    "             ha=\"center\",\n",
    "             arrowprops=dict(facecolor='black', shrink=0.1),\n",
    "             fontsize=18,\n",
    "            )\n",
    "plt.text(-2, 0.9, \"$x_2$\", ha=\"center\", fontsize=20)\n",
    "plt.text(1, 0.9, \"$x_3$\", ha=\"center\", fontsize=20)\n",
    "plt.axis([-4.5, 4.5, -0.1, 1.1])\n",
    "\n",
    "plt.subplot(122)\n",
    "plt.grid(True, which='both')\n",
    "plt.axhline(y=0, color='k')\n",
    "plt.axvline(x=0, color='k')\n",
    "plt.plot(XK[:, 0][yk==0], XK[:, 1][yk==0], \"bs\")\n",
    "plt.plot(XK[:, 0][yk==1], XK[:, 1][yk==1], \"g^\")\n",
    "plt.xlabel(r\"$x_2$\", fontsize=20)\n",
    "plt.ylabel(r\"$x_3$  \", fontsize=20, rotation=0)\n",
    "plt.annotate(r'$\\phi\\left(\\mathbf{x}\\right)$',\n",
    "             xy=(XK[3, 0], XK[3, 1]),\n",
    "             xytext=(0.65, 0.50),\n",
    "             ha=\"center\",\n",
    "             arrowprops=dict(facecolor='black', shrink=0.1),\n",
    "             fontsize=18,\n",
    "            )\n",
    "plt.plot([-0.1, 1.1], [0.57, -0.1], \"r--\", linewidth=3)\n",
    "plt.axis([-0.1, 1.1, -0.1, 1.1])\n",
    "    \n",
    "plt.subplots_adjust(right=1)\n",
    "\n",
    "save_fig(\"kernel_method_plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Phi(-1.0, -2) = [0.74081822]\n",
      "Phi(-1.0, 1) = [0.30119421]\n"
     ]
    }
   ],
   "source": [
    "x1_example = X1D[3, 0]\n",
    "for landmark in (-2, 1):\n",
    "    k = gaussian_rbf(np.array([[x1_example]]), np.array([[landmark]]), gamma)\n",
    "    print(\"Phi({}, {}) = {}\".format(x1_example, landmark, k))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "     steps=[('scaler', StandardScaler(copy=True, with_mean=True, with_std=True)), ('svm_clf', SVC(C=0.001, cache_size=200, class_weight=None, coef0=0.0,\n",
       "  decision_function_shape='ovr', degree=3, gamma=5, kernel='rbf',\n",
       "  max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
       "  tol=0.001, verbose=False))])"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rbf_kernel_svm_clf = Pipeline([\n",
    "        (\"scaler\", StandardScaler()),\n",
    "        (\"svm_clf\", SVC(kernel=\"rbf\", gamma=5, C=0.001))\n",
    "    ])\n",
    "rbf_kernel_svm_clf.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saving figure moons_rbf_svc_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed1cf2cb70>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.svm import SVC\n",
    "\n",
    "gamma1, gamma2 = 0.1, 5\n",
    "C1, C2 = 0.001, 1000\n",
    "hyperparams = (gamma1, C1), (gamma1, C2), (gamma2, C1), (gamma2, C2)\n",
    "\n",
    "svm_clfs = []\n",
    "for gamma, C in hyperparams:\n",
    "    rbf_kernel_svm_clf = Pipeline([\n",
    "            (\"scaler\", StandardScaler()),\n",
    "            (\"svm_clf\", SVC(kernel=\"rbf\", gamma=gamma, C=C))\n",
    "        ])\n",
    "    rbf_kernel_svm_clf.fit(X, y)\n",
    "    svm_clfs.append(rbf_kernel_svm_clf)\n",
    "\n",
    "plt.figure(figsize=(11, 7))\n",
    "\n",
    "for i, svm_clf in enumerate(svm_clfs):\n",
    "    plt.subplot(221 + i)\n",
    "    plot_predictions(svm_clf, [-1.5, 2.5, -1, 1.5])\n",
    "    plot_dataset(X, y, [-1.5, 2.5, -1, 1.5])\n",
    "    gamma, C = hyperparams[i]\n",
    "    plt.title(r\"$\\gamma = {}, C = {}$\".format(gamma, C), fontsize=16)\n",
    "\n",
    "save_fig(\"moons_rbf_svc_plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Regression\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(42)\n",
    "m = 50\n",
    "X = 2 * np.random.rand(m, 1)\n",
    "y = (4 + 3 * X + np.random.randn(m, 1)).ravel()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LinearSVR(C=1.0, dual=True, epsilon=1.5, fit_intercept=True,\n",
       "     intercept_scaling=1.0, loss='epsilon_insensitive', max_iter=1000,\n",
       "     random_state=42, tol=0.0001, verbose=0)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.svm import LinearSVR\n",
    "\n",
    "svm_reg = LinearSVR(epsilon=1.5, random_state=42)\n",
    "svm_reg.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "svm_reg1 = LinearSVR(epsilon=1.5, random_state=42)\n",
    "svm_reg2 = LinearSVR(epsilon=0.5, random_state=42)\n",
    "svm_reg1.fit(X, y)\n",
    "svm_reg2.fit(X, y)\n",
    "\n",
    "def find_support_vectors(svm_reg, X, y):\n",
    "    y_pred = svm_reg.predict(X)\n",
    "    off_margin = (np.abs(y - y_pred) >= svm_reg.epsilon)\n",
    "    return np.argwhere(off_margin)\n",
    "\n",
    "svm_reg1.support_ = find_support_vectors(svm_reg1, X, y)\n",
    "svm_reg2.support_ = find_support_vectors(svm_reg2, X, y)\n",
    "\n",
    "eps_x1 = 1\n",
    "eps_y_pred = svm_reg1.predict([[eps_x1]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saving figure svm_regression_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAoAAAAEYCAYAAADMEEeQAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzsnXlYVdX6xz8bEBCnVNSyBLuJU2LOmUVlWd7K1KxbIXm9t5LM1JzCgSz7qYljDmlpdc0SzUolp+qWebXMFJyHFEwmxQkHUEE4h/P+/tjM45ngHGB9nmc/p7PO3mu/EH7Pu9d6B01EUCgUCoVCoVBUH1wcbYBCoVAoFAqFomJRDqBCoVAoFApFNUM5gAqFQqFQKBTVDOUAKhQKhUKhUFQzlAOoUCgUCoVCUc1QDqBCoVAoFApFNUM5gAqFQqFQKBTVDOUAKhQKhUKhUFQzlAOoqDJomjZR07RvNE07pWmaaJoWZ8UcUsJxvRxMVigUCrPRNM1F07TRmqYd1zTtpqZpiZqmzdU0rZYFcyiNUwDg5mgDFAo78j5wGdgH3GLDPL8CywqNGWyYT6FQKOzBB8BIYD0wF2iT/b6jpmm9RMRk5jxK4xTKAVRUKe4SkVMAmqYdAWpbOc8pEVlpP7MUCoXCNjRNuxsYAawTkWfzjccCC4EXgVVmTqc0TqG2gBX2RdN5WdO0nZqmXcrepojXNG2Tpmk1yvPeOc6fPdA0zV3TNGsdSIVCUUVxoMYFAhowv9D4J0Aa8JIlkymNU2gi4mgbFFUITdOWAsHAWmArkAX4AH8TkYHFnO8CNLDgFpfN2ebIWQEUkeYWzI2maQLcADwBV+AisAZ4W0RSLJlLoVBUPRylcZqm/Qj0ArxEJKPQZzuBliLSyAz7lcYpAOUAKuyIpmn10GPwPhWR18y8pjkQa8Ft7hSRODPmtdYB3A18A5wE6gJPAi8Ah4EeIqICpRWKaoojNU7TtMNAYxFpUsxnXwP/ADxEJLMMe5TGKQAVA6iwLwYgFeisaVo3IA7IKOOp8hzwmAX3OGe9eWUjIvcWGvpC07RDwHTgzexXhUJRPXGkxnkBGSV8djPfOaU6gErjFDmoFUCFXdE0LQB9O+G27KFvROR5B9hh1QpgCXPVAK4De0Wkh63zKRSKyoujNM5eK4AlzK00rhqiVgAVdkPTtGeB5UAEsAl9qySujGtcgTLjVvJxUUSyrLXRGkTEoGlaEuBdkfdVKBTOhYM1Lgloq2maR+EYQOB2INka5w+UxlVXlAOosAuaptUHVgCfi8hwCy5thoXxMZQhuPZG0zRP4A7gj4q8r0KhcB6cQOMigceBbuh1/HLs8gQ6ADssuEcBlMZVT5QDqLAX/kAtINrC6xwSA5i95XEXkCYiCdljDUXkUjGnT0X/t7LRHvdWKBSVEkdr3BpgEjCKfA4gMAQ99i88/8lK4xRloWIAFXZB07Rb0bPKAD4FjgIe6AJ0q4gEVoANgwDf7LcjAHf0avkA8SLyZb5zm6M/lW8XkYezxz4AugPbgAT0QtJPAj2B3UBPEUkv759DoVA4H06icYuA4eidQLaQ1wlkJ/BI/vIxSuMUZaFWABV2QUTOaZrWG3gH+Cd6eYErwHF0sawIXgEeKjQ2Nft1O/AlpfM/oC0wGGiIXt8rBggF5onIzZIvVSgUVRkn0bhR6NvDwcBTQDKwCHjHzDZw/0NpnCIbtQKoUCgUCoVCUc1QreAUCoVCoVAoqhnKAVQoFAqFQqGoZtjVAdQ0rYamaeM0TTNomvZiWeMKhUJR2VA6p1AoqgL2XgEcAghFawmVNK5QKBSVDaVzCoWi0mPXLGARWQKgadrT5owXh6ZpwegZTtSqVatz69at7WmiQqGoDKSnQ0Z2s4P8iWqapr96eEDNmsVeeu3aNaKjo0HvjGBJBwazsFXnlMYpFApbNC4/e/futVrnnK4MjIgsA5YBdOnSRaKiohxskUKhcAhGIyQmQlISGAxQowY0bQrNmoFbQenau3cv+/btY8iQIQD8/vvv3H///fGOMLsslMYpFArAIo0rCU3TrNY5p3MAFQqFAtAF8M479aMEEhISCA0NZeXKldxxxx0MGjQIT09PevRQ/ewVCoWTY4bGlScqC1ihUFQ6UlJSmDhxIi1btuTbb79l4sSJHD16FE9PT0ebplAoFJUCtQKoUCgqHUlJScydO5cXX3yRadOm4ePj42iTFAqFolKhHECFQuH0iAgbN27k119/Zfbs2bRp04bY2Fhuv/12R5umUCgUlZJy2QIWkYdF5CtzxxUKhaIkoqKi6NmzJ/369WPTpk1cu3YNwOHOn9I5hUJRmVExgAqFwik5f/48L730El27duXYsWMsWbKEQ4cOUadOHUebplAoFJUetQWsUCicEldXV7Zu3crEiROZMGECdevWdbRJCoVCUWWoEg5gSkoKycnJZGZmOtoUp8bV1ZU6derQoEEDPDw8HG2OQlEAg8HAsmXL2LRpE5s3b8bb25vY2FiV2QtkZGRw+fJlrl27RlZWlqPNcWrc3d3x9vamXr16jjZFoXBqKr0DePPmTc6fP88dd9xBzZo10XKqaCsKICIYDAZSU1NJSEjAx8dHOYEKp0BE2LBhAyEhIURHR/Pwww9z5coVGjZsqJw/dOcvISGB+vXr07x5c2rUqKF0rgREhPT0dE6fPo2Hh4f6+1EoSqHSxwBevHiRRo0a4eXlpUSxFDRNy30yrl+/PpcvX3a0SQoFSUlJ9OzZk/79+6NpGhs2bOCXX36hYcOGjjbNabh8+TL169fH29sbd3d3pXOloGkaXl5eeHt7c/HiRUebo1A4NVViBfDWW291tBmVirp16xIXF8dtt93maFMUlZ38rYwyM8Hd3axWRgaDgRo1auDt7U1mZiZLlizh1VdfpUaNGhVofOXg2rVrNG/e3NFmVCrq1KnDpUuXHG2GoipgpcaVN5cvXyY2NtamOSq9A2g0GnFz4P+EykiNGjVUHJHCNkTgyBGIidHf5/97On8e9u8HPz9o1y6vuTl6vO6MGTNYu3YtBw8exMvLi507d6pVrVLIyspSjrGFuLm5YTQaHW2GojJjpcaVNwaDgaVLl/Luu+/aHOda6beAAfXlYSHq96WwCRHYuVMXxqysgsIIeWMxMfp52fGnixcvpkWLFsycOZP77ruP9PR0QP09moP6HVmG+n0pbMIKjasonn32WUaMGEGHDh2IiIiwaa4q4QAqFIoK5MgRuHChqCgWJisLLlzgwvbt+Pv7M3z4cNq1a8fevXv54osvVJyfQqFwTizUOI4cKVdzjh8/zo0bNwAYOXIkERER/Pzzz7Rv396meZUDqFAozMdozHsqzib812Y0H/YkLi88R/NhTxL+azMALqSkQFYWjS5d4oEePXITPDp16uQo6xUKhaJ0LNA4IG8lsBxCDi5fvsybb75Ju3btmDdvHgC9evWiX79+dlnlVsFzCkV1wF6BzImJBd6G/9qM4KVdSMvU54hPrsWQjzuz5McfORg/g+gFC2jaqBGfTp4Md95pz59IoVAo8qhAjQte2gWAoIDEgtfZSeMMBgMff/wx7777LikpKQQHBzN06FC7zJ0f5QBWYk6cOMHq1asJDAykVatWjjZH4YzYO5A5KanAHKGr/XOFMYd0Qw1+j/4Xk545SG1PT/38pCTlACqsQumcolQcoHFpmW6ErvbPcwDtrHGvvfYay5cvp1evXsybNw9/f3+7zFsY5QBWUkSEV155hUOHDrF161Z27NihAp8VBckJZC4pliVnLCYGUlLg/vvLFshC3XYSLnkVe5qGD9MDA/MGDAZLLFcoAKVzijJwoMYVGbdR444dO0b9+vW57bbbGDVqFAMGDOCpp54q1793FQNYSfnss8+Ii4vjwIEDxMbGsmLFCkebpHA2yiOQ2d099z9FhCb1Uos9zcc7reCAKmOisAKlc4pSKWeNA/BpmFbsaUXGrdS4S5cuMWLECNq3b897770HQPv27enTp0+5P+woB7ASkpyczMSJE1m+fDl/+9vf+M9//kNISIjq7qHIo7wCmZs2BVdXov76i4enTOHc1WF41ij45OvlbmR64OG8AVdX/TqFwgKUzilKpZw1LofpgYfxci94jT00zmAwsGDBAlq0aMGSJUsIDg5m2rRpFs1hK2oLuBJSuM3R448/zoULFxxokcLpKKdA5niTidAFCwjfsYNGdeuy5NX7qe0RyeQ195BwyQufhmlMDzxccE7QA7EVCgtQOqcolfJK1mjWTI8bzCbn2tDV/nbVuMmTJzNz5kwee+wx5s2bR7t27Sy63h7Y1QHUNK0G8CYwAxgkIl9ljz8EfJB9v0xguIj8Yc97KxSKfJRDIHN6ejqdunUj7cYNJg4YwIS+fanr5QWcYdBDZ4q3w9VVD8CuQt16lM4pFE5AeSVruLnpmpVvdTEoILGow5eDBRp39OhRNE2jbdu2jBw5kgceeKDc4/xKw95bwEMAAXJFT9O0W4B1wBsi0h4YB3ynaVrxkZUKhcJ27BTIbDAY+OabbxARatasyWeffcaJEyd4f/Ro6tapU7oNrq7QuLGefVe1UDqnUDia8kzWaNdO1658W8HFYqbGJScn88Ybb9C+fXvGjx8PQNOmTW2O84uMjLT6WrCzAygiS0RkLpA/IrM3cEJEdmWf8z/gLPCoPe9dHRg6dCiappGUlFTksxMnTuDu7s6bb77pAMsUToeNgcwiwoYNG/D39+f5559n+/btAPTv3x8fX189m87PTxfAwiLp5pb3VGxO1l0lQ+lc+aJ0TmEW5ZmsoWl20bjMzEw++OADWrRowdKlSxk2bBiff/552fcvgzNnzvDPf/6Tbt262TRPRSSB/A34q9DYX9njRdA0LVjTtChN06Lyx38o4L777gNgz549RT4bPXo0devWZcqUKRVslcIpsSGQOSoqip49e+ZWm9+wYQMPPfRQwfk1Dfz9oW9f6NhRv1+jRvprhw76uL+/zc5feDg0bw4uLvpreLhN05UnZuuc0rjSUTqnMIvyTtawg8YtWbKEMWPG0L17dw4ePMiiRYuKtMC0ROPS0tKYOnUqLVu25Msvv8S9kBNsKRURmKNR8EkZwEgJzqeILAOWAXTp0qXiOixXArp37w7owti/f//c8c2bN/P999+zePFi6tev7yjzFM6ElYHMhltvZcBDD3Hz5k2WLFnCkCFDcCsttsXNTY+nKYciz+HhEBwMadkP8PHx+nuAoCC7385WzNY5pXGlo3ROYRbWaFxWFpw+nXe9ObHJFmrc4cOHuXr1KgEBAQQHB9OmTRt69+5d7LnmapyI8NVXXzF+/HgSs5NfBgwYwKxZs2jRooVZdhX7o1l9pfmcBnoVGvMBvi3PmzprsVAR6/W+VatWNGjQoMCTscFgYMyYMbRr147XXnvNHiYqqgIWBDKnpKWx5KefGDtmDO41a7J+/Xr8/PyoW7duRVtdgNDQPGHMIS1NH3dCB7DCda4qahwonVOYiaXJGjmcOwcXL1reIaQMLl68yLvvvsvSpUvp1KkTe/bswcvLq0TnD8zTuN27dzN69Gh27doFQIcOHfjggw94+OGHbba5IraAvwPaa5rmD6BpWjegNfBTBdy7ytG9e3eioqJyRXbBggVER0czf/58XMsKWFVUL8oIZDYYjSz+4QdajBzJpPBwtp47B0Dnzp0d7vwBJCRYNu5glM7ZEaVzCrMwN1mjMFlZeXUBd+7UO4pYSWZmJnPnzsXPz49ly5bxxhtv8MMPP5j1gFaaxp0+fZpBgwbRvXt3du3aRePGjfn000+Jioqyi/MHFbACKCIpmqb9A/iPpmmCvi3ypIhcLef7luf0DqN79+5s2bKFEydO0KBBA6ZOnUr//v159FEVa64oRE4gc6E+mSLChqgoQsLDiU5K4uGuXZn70Ud06tzZsfYWwsdH3xIpbtzZcITOVVWNA6VzCjMpQePMJn+HECv77UZERDBu3DiefPJJ5syZQ5s2bcy+tiSNq1v3Ki1btiQ9PR13d3fGjBnDxIkT7f5gXi4OoIg8XOj9NqBredyrupE/QHrHjh1kZGQwd+5cB1ulcFpyApnbtNELoCYlIRkZTN2wAc3Tkw3r15N6ox8DntVISNAFafp059hinT69YHwMgJeXPu4MKJ0rP5TOKcymsMadPq1v8+Yj/NdmJccG5qwEtmljdr3SQ4cOcfLkSQYMGMBzzz3H9u3befDBBy02vTiN07Q0UlKGAek8++yzzJ49mzvLIcYaVCu4Sse9996Li4sLn332GcuXL2fUqFH87W/FJlQrFLnEnznDq9Onk9yqFS6PPELEzz9z+PhxUm/0JzhYIz5e3wXJCUJ2hmzboCBYtgx8fXWN9/XV3zuDc6ooX5TOKSwmJ1njjjsKbAnndAiJT66FiJbbIaRAmzgo0lmkOC5cuMBrr71Gx44deeuttzAajbi4uFjl/EGext16awZgAuIQeZWOHY+zfft2vv3223Jz/kA5gJWOOnXq0LZtW3bs2EHjxo0JDQ11tEkKJyYlJYUJEybQqlUrwsPD+eMPvXbxHXfcQY0aNUoNQnYGgoIgLg5MJv1VOX/VA6VzCquxoENILjkdQkogIyODOXPm4Ofnx3/+8x9GjBhBZGRk6RUSzCAxMZEtW4I4d84TcOXWW+/js896ERkZabVTaQnKAayE5BR/nDFjBnXK6sagqJaICIsXL6ZFixbMnDmT559/nhMnTtCnT58C51WyRAtFNULpnMIqyqFDyN69e3nrrbd48MEHOXz4MPPnz6dBgwZWm3jjxg2mTJlCq1atWLVqFR4eHkyaNIno6GhefvnlCkt0qjoNOqsJBoOB//3vf3Tp0oXBgwc72hyFk6JpGtu2bcPf3585c+bQqVOnYs+rTIkWiuqD0jmF1RTTISQ+uVaR08rqEHLw4EF27tzJsGHD6NGjB/v376dDhw42mWYymVi1ahUTJkzgzBm9f/rzzz/PzJkzad68uU1zW4NaAaxkzJkzh9jYWBYtWuS0dcAUjiEqKopevXrx559/AvDFF1+wdevWEp0/0IOQvQo9CDtTooW1VOUM2eqA0jmF1djYIeT8+fMEBwfTsWNH/u///o/r168D2Oz87dq1i/vuu49BgwZx5swZOnfuzI4dO1izZo1DnD9QDmCl4PLly6xevZqJEycyefLk3NYyCgVAfHw8L730El27duXQoUPExcUB4OXlVeaXZ2VMtCirdZI962QpKg6lcwq70KxgckdQQCLLXovC1/sGmib4et9g2WtRRQpGZzRuzKxZs/Dz82P58uW8+eab/Pnnn9SuXdsmcxISEhg4cCA9evRgz5493HrrrSxfvpw9e/YQEBBQ4nXmtIhLKxzAbSFqC7gS8OOPPzJw4EAaN27M6NGjCQsLc7RJCidhypQphIWFoWkaEydOZMKECRbXigoKcm6HLz+ltU564IF4QkNDCXeGFGaFxSidU9gFSzuEuLqCnx9nzp9n8uTJ9O7dmzlz5tCyZUubzLh+/TqzZs1i9uzZ3Lx5E09PT8aNG8f48ePLdCrLahEXExPDuHHjclcnrUU5gJWAwMBAAgMDHW2GwkkwGo252WdpaWm88MILTJs2jWaFnnyrIiVlLQ8bdoWMjFZkZGTg7u7Om2++yezZsx1jpMIqlM4p7Ea7dpCSohd5LqUw9P74eNYdOcLU5cv5m6Zx7Ngx7rrrLptubTKZWLlyJRMnTiQpO7P4xRdfJCwsDF9fX7PmKK06w4kT7xAWFoaHhwehoaH88ssvVtuqtoAVikqCiPDdd99x991389NPeoexmTNnsmLFimrh/EHJ2cmpqfXIyMggMDCQEydOMGvWrIo1TKFQOA85HUL8/PQVvkJZteeuXWPIsmV0Dgnh482bSTp7FsBm52/nzp3ce++9DB48mKSkJLp06cJvv/3G6tWrzXb+oPTqDDVr1uSf//wnMTExTJgwwSZ71QqgQlEJiIqKYty4cWzfvp3WrVtTIztjrboFyJeUtezhcYEdO3bnlg5RKBTVnGK6IN28cYP569YxfeVKMjIzGTNmDG+//Ta33HKLTbeKj49n/PjxrFmzBoCmTZsSFhZGUFAQLi6Wr7OVVp1hwoQJdtN9tQKoUDg5o0aNomvXrhw7dowlS5Zw+PDhapvk8Morp3BxuVlgzMPDyGefNVHOn0KhKEpOh5D77+dmt27MW7eOR3v14ujRo8yZM8cm5+/69eu8/fbbtGrVijVr1uDp6cnkyZOJjo5m0KBBVjl/oFdh8PQ0FRjz8DAyfbp9H/qVA6hQOCEpKSkYjXrpgrvvvpuJEydy8uRJXn/9dZurz1dG4uPjCQoK4p137sJkehkXl0RA8PERPvvMjaCg6rUSqlAozGPfvn28+uqrGI1GbrnlFg4fPkxERAR+fn5Wz2kymVixYgUtW7Zk+vTpBcJP/u///o9atYrWHbSExMQwMjIGo2kJFNQ5m6YtgnIAFQonwmAw5HbwWL58OQBDhgzh/ffftzi7tyqQkpLC+PHjC1TMDwlpxqVLdfS+nvFapclgVigUFcfZs2d5+eWX6dKlC9999x3R0dEANGnSxKZ5f/vtN7p168a//vUvzp49S7du3fj9999ZtWoVPjZU0DcajWRmdzFp3rw5Q4Z4cfasR7nqXJVwAFXRV8tQvy/nIyfBo127dgwfPhx/f3+6du3qaLMchsFg4MMPP6RFixbMmjUr9wn7+PHjzJw50+aYncqG+jdrGer3VX3JzMzk/fffx8/Pj5UrVzJ27FhOnjxJ27ZtbZo3Li6OF154gYCAAPbu3cvtt9/Ol19+mVvg2Rb++9//cs8997BgwQJAzxpeunSpzc5qWVR6B9DNzS13q0xhHgaDocJ6DSrM47XXXqN///64uLiwceNGtm7danPl+cpIfkd4xIgRJCcnExAQwO7du1m1apXDKuY7EldXVwyl9ClVFCV/qSRF9cLFxYVVq1bx2GOPcezYMWbPnk29evWsnu/atWuEhobSunVrvv76a2rWrMmUKVM4ceIEL730ktVxfgDHjx/nqaeeonfv3mRkZNjspFpKpf8X4unpyfXr16lfv76jTak0pKamqubqTkB8fDz169enbt26PP/883Tq1IlXX3212n5xRUZGMm7cOHbs2AGAn58fs2bNol+/ftUu2zk/derUITU1FW9vb0ebUmm4du0anp6ejjZDUUHs3buXadOmsWLFCurWrcvOnTttcvogL85v0qRJnDt3DoCgoCBmzJhhl7JbixYtYsyYMXh5eTF79mxGjBiBh4eHzfNaQoWtAGqa1kfTtN2apkVqmrZX07Q+9pi3UaNGXLx4kbS0NLXsXwoiQmZmJsnJyVy5coUGDRo42qRqS0pKChMmTKBVq1bMnDkTgF69ejF06NBq6fzlJHh069aNHTt20LBhQxYuXMjRo0fp379/pXL+ykPnGjRowJUrV0hOTiYzM1PpXCmICGlpaSQnJ9OoUSNHm6MoZ5KSkvj3v/9N165d+f3333P7oNvq/O3YsYMuXbrw8ssvc+7cOe6991527drFypUrbXL+DAZDbveOzp078+qrr+Z29aho5w8qaAVQ07Q7gc+B7iJyUtO0W4FfNU3bLSIXbZnb09OTJk2acO7cOTIyMuxhbpXF1dWVOnXq4OPj45A/tuqOwWBg6dKlvPfeeyQnJzNo0CCGDh3qaLMcRkpKCjNmzGD+/PlkZGTg4eHBm2++ycSJEytljF956ZyHhwc+Pj5cvnyZuLg4skrpbKDQf19NmjRRK4BVmKysLMLCwpgxYwYGg4G33nqL0NBQmxPlYmNjCQkJ4dtvvwXgjjvuYObMmbz44os2bfUCfP/994wZM4aePXuyZMkSevToQY8ePWya02ZEpNwP4DlgR6GxlcC/S7uuc+fOolBUFYKDgwWQnj17yt69ex1tjsNYscIg9eunCGQJxAoESmBgoMTGxtr1PkCUVIC+5RzW6JzSOIXCOv7+97/LgAED5OTJkzbPlZKSIuPHjxd3d3cBxMvLS6ZMmSI3btywar6VK0V8fUU0TeS22zKkffuZAkiLFi1kw4YNNtubH1t0rqKE0Q+4ArTOft8WSAImFHNuMBAFRPn4+Nj1F6VQVDSRkZESHx8vIiJHjx6VDRs2iMlkcrBVjsFkMsno0ZGiaWkCknt4ehpl5Ur7388BDqBZOqc0TqGwnD179sgjjzyS+6CYnp5u85xGo1E+/fRTadKkiQACyEsvvSSJiYlWz7lypYiXlxTQOLghgYGbJCMjw2abC+P0DqBuI88AvwF/AIuB1WoFUFFViYuLk6CgIAFkyJAhjjbH4ezZs0cefPDB7BW/wuKoPy3bm4p2AMUKnVMap1CUzunTp+Wf//ynANK4cWP5+eef7TLvtm3bpEOHDrmOX/fu3eWPP/6weV4fH1OFaZxIJXEAC9wUXIGYnCflkg4ljorKxtWrV2X8+PHi4eEhnp6eMmnSJElJSXG0WQ4jLi5OBg4cmCuy+rZvUXHUNPvf2xEOYP7DHJ1TGqdQlMz7778vXl5e4u7uLuPHj7eLlp48eVIGDBiQq0nNmjWTVatW2bwzYzKZZPPmzRWqcSK26VxFZgHXy351AWYBB0TkeEXdX6GoCN577z1mzpzJ888/z4kTJ5g+fXq17eCRk+mc18EjhDvuKD6D1YYC+k6F0jmFwjZ0n0YnISGBJ554guPHjxMWFmaTlqampjJ+/Hjatm3LunXr8PLyYurUqRw/fpzAwECbqg0cO3aMJ554gqeeego3t7PFnuOUGmet52jpAWwFItGfiBcDtcu6Rj0dK5wdk8kkERERsmfPHhEROX/+fLVO8MjMzJRFixaJt7d37hP2wIEDc+N2iouP8fKSKhEDKFbonNI4hSKP3bt3S48ePWTnzp0iosfo2YrRaJRly5ZJ48aNczVp8ODBcvr0aZvnFhGJiIgQV1dXueWWW+SDDz6Qzz83VJjGidimcxUqjpYeShwVzkxkZKQ89NBDuYHD1ZkcR7hly5a5IhsQEJDrGOcnf4acr69zCmNFHUrjFAqRxMREGTRokADSpEkT2bhxo13m/eWXX+See+7J1aT7779fIiMjbZ43IyNDTp06JSJ6BvG7EjrxAAAgAElEQVS4cePk4sWLuZ9XlMaJKAdQoahQ8id4NGrUSJYsWSKZmZmONsthREZGZid46CLr5+cn69evd3i2s3IAFQrnZ86cOeLl5SUeHh4yceJESU1NtXnOkydPSv/+/XM1ydfXV9asWWOXOL+NGzdKy5YtpW3btmIwGGy21VZs0blK3wtYoaho1qxZw9q1a5k0aRInT57k9ddfp0aNGo42q8KJj4/npZdeomvXrpW+g4dCoag4chwQ0Is69+nTh+PHj/P+++/b1KY0JSWFkJAQ2rRpQ0REBLVq1WLatGn8+eefPP/88zZp0tGjR/n73//O008/jaZpzJ49G1dXV6vncwqs9Rwr4lBPxwpnIDMzUz788ENZt26diOj1pxISEhxslePIn+kMiIeHh4SEhMiVK1ccbVoBUCuACoXTsWvXLrn33nslPDxcRMQuOwVGo1GWLl0qjRo1yl31+9e//iVnzpyxeW4RkR07doiLi4vccsstsmDBAqfa8bFF59QKoEJRAiLCd999R7t27Rg+fDgRERGA3n7QHs3AKxsGg4EPP/yQFi1aMHPmTDIyMggMDOT48ePMnDmzUrZvUygUFUNiYiJBQUHcd999xMfH4+7uDmDzTsHWrVvp2LEjr732GhcvXuSBBx4gMjKS5cuX07RpU6vnzczM5NChQwDcd999vPfee5w8eZKRI0dWnR0faz3HijjU07HCUezbty83waN169ayceNGh8e0OQpLEjycCdQKoELhFHz44YdSs2ZN8fDwkNDQULvE+UVHR0vfvn1ti/MzGEROnRL57TeRX34R+e03Mf31l2xYv178/PzE29tbrl27ZrOt5YktOufmQN9ToXBaYmJiOHbsGB999BGvvvoqbm7V859KVFQUY8eOZceOHQD4+fkxa9Ys+vXrp2L8FApFiZhMJrKysqhRowZNmjTh6aefZtasWfj6+to079WrV5k6dSqLFi3CYDBQu3ZtJk2axOjRo/H09DRvEhE4cgRiYvT3WVkAHE5IYMwXX/DzoUO0vvNOvlixgtq1a9tkrzNTPb/VFIpCpKSkMGPGDBo1asTYsWP5xz/+wRNPPGFTQHJlJj4+nkmTJrFq1SoAGjZsyDvvvMPQoUNzt24UCoWiOHbt2sWoUaPo27cvoaGhPPfcczz33HM2zWk0Gvnkk0945513SE5ORtM0Xn75ZaZNm8Ztt91m/kQisHMnXLiQ6/gB/Hn6NB3eeot6Xl4s/Pe/Gfr3v1Ojbl39/Cr6sKtiABXVmsJxbadOnQL0uJTq6PylpKQwfvz4Ih08cmJflPOnUChKIjExkYEDB9KjRw8SExO566677DLvTz/9RMeOHRk2bBjJyckEBAQQFRXFZ599ZpnzB/rKX7bzl2k08uuffwLQ5o47WPzKK8QsXMiIJ56ghqbp5x05YpefwRlRDqCi2rJjxw7atWvHiBEj8Pf3Z+/evSxevNjRZjmE/I7wrFmzyMjIYODAgYSFJbJmzUwaNLiF5s0hPNzRlioUCmdkxYoVtGrVivXr1zN58mSio6N58cUXbZozOjqavn378vjjj3PkyBGaN2/Ot99+y/bt2+nUqZPlExqNEBODGI18FxnJ3WPG0HNKDHcM/TsuLzxHWMR8fjjQNu/8rCx9m9hotOnncFbUFrCi2mEymXBxccHDwwNXV1c2bNhAnz59KkVMW3g4hIZCQoLeW3L6dAgKsn4+EWHDhg2EhIQQHR0NQEBAAHPnziU6uivBwZCWpp8bHw/Bwfp/23JPhUJRNTCZTKSnp1OrVi3atm1Lv379CAsLsznO75NPbjB2bAbXrrUAFuLh4c2UKS0ZNWqU+XF+xZGYyKG4OEYvX84vR47Q9JY3cHOdx5nL+s5GfHItgpd2ASAoILHAddx5pw0/kZNibfZIRRwqQ05hT+Li4mTgwIEyZMiQ3LGsrCwHWmQZ9u6jm7+VHcV08PD1LXivnMPX124/UrmCygJWVAWKyVSVU6f0cQfy22+/SdeuXeWVV16x25wGg0H+9a//CtwooDk1a2bZpZ3auQ0bxN3NTRrUri2LXn5ZfLyvF69x3tdFvv467/jtN9tvXk7YonNqC1hR5UlJSWHChAm0atWKdevWceutt6L/uwEXF8f9EwgPh+bNwcUFs7ZXQ0PzVuNySEvTxy0hPj6eoKAgunbtyvbt22nYsCGLFi0q0sEjIaH460saVygUdkQEDh+GDRtg/35ISoLkZP11/359/PBh/bwKJD4+nhdffJEHHniApKQkevbsWer55urcTz/9RIcOHfj8cz/Aq8Bn6ekuFutcDhkZGXz33XcANKldm1UjR3Jy4UKG//3vJF7yKvaahMLjBoN1N3dy1Bawokrz008/ERgYyOXLl3nppZeYPn26UxRxDg/H4u1VWx2ynEzn+fPnk5GRgbu7O6NGjWLixInFFnH28dHtKm5coVCUIyVkquaSMxYTAykpcP/9FZKpunbtWl566SU0TeOdd94hJCSEWrVqlXi+OTp3/Phxxo0bx+bNm7OvKl5gLH3wFBEiIiIYN24cp06d4tChQ/i7u/Ns9+655/g0TCM+uaj9Pg0LPWlXlcLPhVArgIoqh4iQmpoKQOvWrenWrRtRUVF88cUXTuH8gXWreSU5XmU5ZDkJHnfddVeBDh4nTpwotYPH9OngVehB2MtLH1coFOVIvkzVUsnKKvdMVZPJxKVLlwDo1q0bL7zwAidOnOC9994r1fmD0nXu8uXLjBo1Cn9/fzZv3kydOnWYOXMmPj7FO7KWPHgeOHCARx55hAEDBlCzZk1+/PFH/P39oWlTyNe/d3rgYbzcCyZ4eLkbmR54OG/A1VW/rgqiHEBFpSf/FsNtt2XQtu00BgwYgIjQrFkztmzZYl3GWDlizWqepQ5ZzhPw3XffzYgRI7h06RIBAQHs2bOHVatW0bx581JtDAqCZcvA11dfXPD11d+rBBCFohzJzlTN7/yF/9qM5sOexOWF52g+7EnCf833IFuOmaq//fYb3bp14/nnn8/V088//9zsB+mS9Cw+XvDz82PBggVkZWURHBxMTEwMISEhvP++ZtODZ1paGo8++iiHDx9myZIlHDhwgMcff1z/sJDdQQGJLHstCl/vG2ia4Ot9g2WvRRVMACnmuqqC2gJWVGoKbzGcO+fBuXNj6d69OyLitJm91myv5jhe5mQBR0ZGMm7cOJs7eAQFKYdPoahQEgs6H+G/NiN4aRfSMvWv64rIVI2LiyMkJIRvvvmG22+/ndGjR1s1T0k6B/FcvnyZnj178sEHH3DPPffkfmKJzuWQkZFBeHg4//rXv/Dy8mLt2rXcc8891K9fv+CJbm7g51fAwQ4KSCzq8OXg6qqfX0U7QakVQEWlprgtBvBi27bH7JrgYWnCRllYu70aFARxcWAy6a+FRTEnwaNbt27s2LGjxAQPhULhpCQlFVj9C13tn+v85ZCW6Uboav+8gaws/TobCQ+HJk3SufNOH779dg4DBnzLiRMnCAoKsko7itM5uEHjxgtYv349W7duLeD85VCWzuUgIqxbt462bdvyyiuvsG3bNgAefvjhos5fDu3aQePGBbaCi8XVVT+vXbvSz6vEVJgDqGnaW5qmHdI0bXf266SKurei6mEwGDh16lSFZKrmrDLGx+ux2TmBzLY4gfbeXi2pg8dff/3F8OHDqVFFg5idDaVzCpvJzCzwtkhGaknjNmSqmkwmFi68RHAwXLhQE3BBxIcffniWiIjS4/xK44knLhMQ8AUQD5jQtAQCA7eRkBBm8wPp/v376dmzJ88++yxeXl7897//5dFHHy37Qk3Tk2b8/HQnr7Aj6OaWt/JXQck1DsPa+jGWHEAAcAZomP2+ARAHPFbadapGlqIwJpNJIiIipGXLltKqVSvx8TGVS626lSv1OTRNxMXFeevhZWZmyqJFi6Rhw4a59fwCAwMlNjbW0aY5jNTUVAkNDa3wOoDW6JzSOEURfvutQA0633KsVbdypUiTJukCpuzDPjqXmZkpCxculPr16wsgLi4uEhwcLOfPn7d8smLIysoSPz8/8fb2liVLlojB2pqI+WssbtvmNDUWLcEWnasoYWydLYwtst83BxKA9sWcGwxEAVE+Pj7l8xtTVEoiIyPlwQcfFEBat24tGzZskJUrTXYtjixSfMHl4g5Ns9/PZikmk0nWr18vfn5+uY7fAw88ILt373acUU7CH3/8IZqmOcIBNEvnlMYpSuXUKZG1a3Mdu5UjdomXu6GgxrkbZOWIXXnO39q1+nUW8MEH58XVNd3uOrdlyxZp3bp1ri49+uijcvDgQcsmKYb09HSZP3++3LhxQ0REDhw4IFeuXLF53sqO0zuAuo08BVwBooGrwICyrlFPx4octm3bJoA0atRIPvroowJPfPlX63x9bXP+RErugOEsK4B79uyRgICAEjt4VDdMJpNs2rRJpk2bljt28uRJh3QCsVTnlMYpimAwFHAAc5xAX+/romkm8fW+XtD5y3EALVi1ioyMFIizq84dPXpUevfunatLLVq0kIiICJt1yWQyyTfffCPNmzcXQFatWmXTfFUNp3cAgZZAItAl+30L4DjQvbTrlDhWb65evSq//vqriIgYjUaZN2+epKSklPt9Nc08B9AerYksIaeVXY7ANmzYUBYuXCiZmZkVa4gTsW/fPnnkkUdyV4XT09NzP3PACqDFOqc0TlEshw4VcQJLPNau1c8vA6PRKMeOHcv975K2fC3dTbl48aK88cYb4urqKoDUq1dP5syZIxkZGTb/Gvbu3Zv7sOvv7y8///yzzXNWNWzRuYpKAukL/CYiUQAichJYDwysoPsrKhH5Cxc/88wzpKen4+rqyujRo6lbt26539+cgqMNG1ZceZT8rexyEjzGjx/PX3/9xYgRI6plgsfZs2cZPHgwnTt35uDBgyxatIhDhw7Z1ijedpTOKeyDnTNVt2/fTpcuXbj//vu5cuUKrq6u+PqWntzg6lp6YprBYGDBggX4+fmxePFiRITXX3+dmJgYxo4di7u7e+m2m8GYMWM4fvw4S5cuZf/+/eYleSjMpqIcwGjgAU3TbgPQNK0u0Bs4UUH3V1QCRAoWLm7fvj0//PADNWvWrFA7ii9dkIeXFyxYUP525DjCLVq0yO3gMXDgQI4fP05YWBj16tUrfyOclMzMTL777jveeustZ8p0VjqnsA92ylQ9deoUzz33HA8//DCXL1/mo48+yu38U5rOeXnBihXFO38iwubNm/H392fUqFFcvXqVXr16ceDAAZYsWUKjRo2s/rHT09OZMWMGSdklbZYvX05MTAzBwcG4luUMKyzH2qVDSw9gDHAQ+AM4BMwAXEu7Rm2PVC/0uBR9K2/jxo0OjWnLH1fYsKF+2CvGsCzyZzqTvd0bEBAge/bsKd8bOzEGg0E++ugjCQoKyv27SE1NLfUaHBMDaJHOKY2rBuTPNP3lF8szTa3MVI2NjRV3d3epVauWTJ06VdLS0oqck6NzIOLqKrkxfyVp3JEjR+Txxx8vEH9sD602mUyyZs0a8fX1FUAWLlxo03zVCVt0rkLF0dJDiWPVJy4uTlbmU5uNGzdan9JfBdizZ09upnOOwKoEj03Spk2bXEe4LMcvB0c4gJYeSuOqMCZTXixf4Xi+nLFDh/Tz7ITRaJRdu3blvl+4cKGcOXPG5nkvXrwow4YNy43zu+WWW+SDDz6wS5xfZGSk3H///QLIPffcI7/88ovNc1YnbNE51QlE4RDyx7UNGzaM1NRUAPr06YNbFW27Uxqqg0dREhIS6NWrF3369CErK4uIiAi2b99OnTp1HG2aQlE6IrBzZ17LsXydPYC8sZgY/Tx99dgmtm3bRufOnQkICCA2NhaAESNG0LRpU6vnzMzMZP78+fj5+bFkyRIA3njjDWJiYhg1apRd4vwWLlxITEwMy5YtY+/evfTs2dPmORXmoRxARYWSP8Fj1qxZvPDCCxw9erRCkjuckeISPKp7B4+s7C/L+vXrc/78eRYtWsSRI0cs7mOsUDiMI0fgwoWijl9hsrL0844csfpWJ0+e5JlnnuGRRx7h6tWrrFy5kubNm1s9H+g7g5s2bcLf35/Ro0dz9epVHnvsMQ4ePMiHH36It7e31XOnp6czbdo09u/fD8C8efOIiYlhyJAhKs6vglEOoKJCiY+PZ/To0bRv356oqChWrFjB9u132LXPbmUgvyNcOMFj5syZ1TLB49q1a7z99tt07doVg8FAnTp1OHToULV1hBWVFKMxb+Uvm/Bfm9F82JO4vPAczYc9SfivzfLOz1kJNBotvtXVq1fp0KEDP/30E9OnT+fPP//khRdesOlB6ciRI/Tu3Zunn36a6OhoWrZsyaZNm/jxxx+5++67rZ5XRPjqq69o3bo1kydPZuPGjQB4e3tX2wUAh2POPjHwMXpMUtNiPmsFZAILrN2HLulQ8TFVgz179sg777yT+/7o0aO5MW3Fdd2wtZOHM1NcBw+V4KEneDRu3FgACQoKskuFf1QMoMIRlHMnD6PRKN9//33u+6+++kqSkpJsNvvChQvy+uuvi4uLS4E4P3vUGd2zZ4/06NFDAOnQoYNs27bN5jkVOrbonLkO4ODsL6v+xXy2BUgG6ltrREmHEsfKTf7CxY0aNSq2D2RJXTes6bJh744g9iLPLpN4eJwVCFQJHtkkJCTkJng8+OCDdnWElQOocAjl2Mt34sTDUqPGGYEsufXWm3bRuIyMDJk7d67Uq1dPAHF1dZXhw4dLcnKyRfOUpr/vvvuuNGnSRD799FMxGo22G63IpSIcwFbZDuD7hcafyh4fZq0BpR1KHCsnqampEhISIh4eHuLp6SmTJk0qsYNHSV03LO0/6awriStXitSsmVXo57sh//znD9W6g8fVq1dFRF/NePbZZ+3SMqowygFUOIRffing2Gla8R03NM1U0AEsZVUsOjpaOnWaI3DdbhqXU26qRYsWubsRvXv3lqNHj1o8V3H6W6NGhowdu1dERG7cuFEhXZyqI7bonKZfXzaapl0C9otIr+z3NYAj2du/HUSkjGhXy+nSpYtERUXZe1pFOXPlyhX8/Px46qmnmDZtGs2aNSvx3ObNIT6+6LivL8TFmX9Pe81jT1JSUvDxMZGaWr/IZ460y5EkJiby9ttvs2XLFmJiYnKL0pYHmqbtFZEu5XYDO6A0rgqycydkFzIGaD7sSeKTaxU5zdf7BnFLtuQNNG2qF3UuRGZmJr6+vpw/vxuRom2KrNGSQ4cOMWbMGLZu3QpA69atmTt3Lk8++aRlE2VTkv7WqXOZ1NQGVs2pMA9bdM6SJJA/gC5aXnTpm+i9L0eVh/OnqDyICN999x3/+Mc/yMrKon79+sTExLBixYpSnT8ovhq9l5c+Xhzh4RSbMJKQUPz5JY2XJ/kTPFJTi0/mcIRdjuTatWuEhobSsmVL1qxZwyuvvIKLi4U5aEYjxMbqX7DbtumvsbFWBc8rFOVG06YFOndMDzyMl3vBv1EvdyPTAw/nDbi66teha5qvr6Bpgq+v8M037qxcuRIoXkst0ZILFy4wdOhQOnbsyNatW6lfvz4LFy7k0KFDVjt/pdlw/bpy/iyigjXOUgewHtBK07TGwGQgQkS2lotlikpBZGQkDz/8MP379+fo0aOcPXsW0Et4mENQkN5v0tdX72bk61ty/8nwcAgO1p80RfTX4GB9vKT+veb09S2NkhzO4hAp2Mru0qVLeHhcKBe7KhPnz5+nRYsWvP/++wwYMIATJ04QFhZmfuafCBw+DBs2wP79+upKcrL+un+/Pn74sF1qqSkUNlPooTcoIJFlr0Xh631Dd+q8b7DstSiCAhKLXBceDq+8kkVCggZoJCRoBAfDuXOP4uNTfGavOVqSkZHBnDlz8PPzY+nSpWiaxsiRI4mJiaFBgxH4+dWwqQpDeelvtcFBGmfJFnAv4Cf0hJAHgZeAtiJyyq4W5UNtjzgvV65cYfjw4axatYrGjRvz3nvv8eqrr5ZrEefStnmnT9edwbS0vHEvr9KbmZdFjsNpzpyRkZGMGzeOHTt2AODn58esWbO4caMfwcGaXe2qDIgIx48fp02bNgBMnjyZvn370rVrV0sn0p+Cy6qp5uoKjRsX6IuqtoAVDuPw4SKlYEoku6dvtIcHHTrcQnp64yKnWKtxObsz48aN46+//gLgiSeeYO7cubRp08YijStMWloas2fPpnHjxtSt+zrBwUJaWp6TWh10zi7YoHFQcVvAuwET8Arwb2B+eTp/Cuck54Ghdu3aHD16lNDQUGJiYhg6dGi5d/AobZvXkpVEcwkNLSiMoL8PDc17HxcXV2oHj6Agze52OTv79++nV69e3HPPPZw6pUvE1KlTLXf+oEIL6ioUdqNdO/3LuqzCxtlf6nL33Tz77LOkpxdfYNkajTt48CCPPvoozzzzDH/99Rdt2rTh+++/Z8uWLbkPZuZoXGFMJhPh4eG0atWKKVOmcODAgWzbtGqlc3bDgRpn9goggKZph4F2wDmgpYhcs5slxaCejp0Hg8HA0qVL+eSTT9i5cye1a9cmKyurQiu3V3Sih4tLySvumibUqXOVGzdGkZX1BR4eHowaNYqJEydWyyLOkJfg8eWXX9KwYUPee+89hgwZYn0RZ6NR3/ooVFA3dLU/CZe88GmYxvTAwwW30lxdoW9fcHNTK4AKxyKif1nHxOjv83/Bu7lhNBpZfuQIgePGUbtOHfbv30/fvu05fbqoplqicefPn2fy5Ml8+umniAgNGjTgvffeK/YhvXSN07dwp0/Pc+T279/P66+/zu7du+nUqRPz588nICDAPMMURbFR48C2FUBLl2z2oDuAE8vb+VM4ByLChg0bCAkJITo6mp49e3L58mVq165d4W17StoCKSlhxFZ8fIp3OAFEtOzs3iXcd183Vq16yub2S5WZlJQU2rVrR0ZGBuPHj2fChAm2O8KJBWOkwn9tRvDSLqRl6rIVn1yL4KW67hUQyMREuPNO2+6tqH4YjfrfTlISZGaCu7uemNGsWe6XrUVoGvj7Q5s2efMaDFCjBv/96y9Gz5zJsWPHkNtvJzg4mI4dOxIWZr3GZWRksGDBAqZNm8a1a9dwc3PjjTfe4N133y0xJrt0jcuLswbdCUxNTSUhIYHPP/+cQYMGWZ7IpSiIgzXO7P972WVfHgaigBU231nh9KSmpuYmeLi6urJp0ya2bt2Kj4Mie8tjm7c0istQLkotkpLeqJbOn9FoZMsWvYxFvXr1+PDDDzlx4gQzZsywzypoUlKBJ+PQ1f65wphDWqYboav98waysgqU4FAoyqS8A/Dd3PQv6/vv58Rtt/F0WBi9Bw8mIyOD9evXM2TIkNxTrdE4EWHdunW0bduW8ePHc+3aNZ588kkOHz7M/PnzS03IM0fj0tJg2LArADz00EPExsYyePBg5fzZAwdrnCWPNeOAO4EgsWTfWFHpuH79OrVr16ZOnTo0bdqUjz76qNwTPMwlKKji4kpy7jNuXAbnztUAtOyjINWtpIuIsHnzZt566y2OHz9OVFQUnTt3ZtCgQfa9UWZmgbcJl4r/pioybjDY1w5F1aWsAPycsZgYSEkpEoBvKW+++Sa7du1i1qxZjBw5Eg8PjyLnWKJx+/fvZ/To0Wzfvh2Atm3bMm/ePHr37m3W9Tn3CQ3Vdaykb/bU1Hp64WBNK9ZmhZU4WONKdeE1TWugaVqgpmkzgKnAPBH5wy53VjgdKSkpTJgwAR8fH06fPo2maaxevbpCEjyckfj4eLZsCeLcOU/AFReX08WeV51KHezbt49HH32Up59+mqysLNavX0+nTp3K52bu7gXe+jRMK/a0IuPWxhwqqh/lHIBvMBhYvHgxidlbfR999BExMTG89dZbNjlS58+f59VXX6Vz585s376dhg0bsnjxYg4ePGi285dDUJAeX2gy6SuOxeHr64Jmg+OrKAEHa1xZa7i9gVXAy8AHwHhrbqJp2ghN0/4odKRqmvaQNfMp7EtO4eIWLVowa9Ysnn766WLj+yypiVeZSUlJYfz48bRq1YpVq1bh4eFBSEgIS5c2tKhodVUjPT2dxx57jEOHDrFw4cLcTOdy+2KwsaBuRaN0rpJhNBYp1RL+azOaD3sSlxeeo/mwJwn/NV9Nv6ws/Xwzi/L++OOP3HPPPQwfPpwvv/wSgDvvvJPGjYuWeTGXmzdvEhYWRosWLfjss89wdXVl9OjRxMTEMGzYMJse1EWE6dPB09NUYLw6aVyF42CNK/WvRURWA6ttvYmILAIW5bzXNO029MLSKv3NwaSnp9O5c2f+/PNPevbsyZw5c4pd0SlcL6pwcHBVICfTecqUKVy6dAmAwMBA3n///dwYv5o187ZLCmfIVUWuXbvGf/7zH0aMGEHNmjWJiIigffv2FZPp3KyZHoOVTU4QdKkZcjnXOQClcw7EmgSOcgrAP378OGPHjmXLli3cddddrF+/nn79+tn044kIa9euJSQkhNjYWAD69OnD3LlzadmypU1zX79+nVmzZhEdHc1XX30FuDBpkpCYqFULjXMoDtY4i8rA2AtN0xYAZ0RkVmnnqRIJ5UdsbCx3ZovY1KlT6dixI0899VSJqzn2KMESHu6czlNOsdSQkBBisks2BAQEMHfuXOtq11UBjEYjn376Ke+++y4XLlxg+/btPPjggxVviBUFdfHXA6YdXQbGHJ1TGmcjpZVayVlZ8fPT6/IV1jY79+wFXeOGDk3m+vUG3HLLNebP92LwYNu26/bt28eoUaP49ddfAWjXrh3z5s3jscces2lek8nEl19+ycSJEzl79iyBgYEsX75cxfhVNDZoHFRsGRibyX4qHgC0KeHzYCAYcFi2aVUmPj6eSZMmsXr1anbv3k3Xrl2ZPHlymdfZ2mvXWVcQC3fwaNmyJTNnzqRfv37VMuYlJ8EjJCSEP//8k5kn6G8AACAASURBVAcffJBNmzY5zhFu104Pvje3Sn67dhVnWymUpnNK42wkZ7XvzBk9W9doLD57oawEDjsF4BsMBj7++GOuXn2KsLC/kZamF3O+erUew4bpC5DWaNzZs2cJDQ3l888/R0Tw9vZm6tSpdknIi4mJYeDAgURFRdGtWze+/fZbevToYdOcCitxoMY5Io97IrBYRK4X96GILBORLiLSpVGjRhVsWtXl6tWruXFt69atY9KkSbRq1crs623t9WhNxXlrMDdOMT4+vkAHD29vbxYtWsSRI0fKN67NyTEajYwZM4asrCwiIiL43//+59hVUE3Tv7j9/HQBLByb6uaW91RsY4amnSlR55TGWUnhci1nz+rOWFm7WCUlcNghAP/777+nffv2jBw5ktmzb7GLxt28eZMZM2bQsmVLli9fjpubG2PGjMntuLRmjZvVsdgmkx7f5+3tjcFg4IsvvmDXrl3K+XMkDtS4Cl0B1DStKfpTceuKvG91x2g04uf3LsnJo4Ewbr89izZt3Khb1/w5bC3CbOsKojmYs8qYkpLCjBkzmD9/PhkZGaqDB3oHj1mzZhEWFkatWrX4/vvv8fHxsb6Dh70ppaCuTYV6ywmlc+WAGf1SS+2gkJPA0aZN3t9K06Zw/nzufNMDDxeIAYSSA/CPHTvG2LFj+eGHH/Dz82PDhg3061d8vT1zNU5E+PbbbwkJCSEuO66mX79+zJ49Gz8/P/1ntHIn5fr164SFhfHzzz+zc+dO6tevz/79+6vtg67T4SiNE5EKO9ADpCeZe37nzp1FYR0mk0m2bt0qJpNJVq4UcXc3iK6i+uHlJbJypWVzrlwp4usromn6qyXX+/pKgfvnHL6+ltlg7T1WrDBI/fopAlkCsQKBEhgYKLGxsfYzoJKRkpIikyZNEk9PT/Hw8JCffvrJ0SbZFSBKKlDfcg5LdE5pnJkcOiSydq3I118Xe6wcsUu8Cmucu0FWjtiVd97atSKnTuXNaTAUmXPliF3i631dNM0kvt7XC16fM4fBINOnT5d69erJvHnzJCMjQ0Rs07ioqCh54IEHBBBA/P395eeffy5yXmn3KE6fs7KyZPny5XLbbbcJIEFBQZKSkmL7/w+F02CLzlVYEoimabcDkUArMbONXKUKkLZ3GyErCQ/PX7g4gTFjLrF2becK7aFbkl3FrSDas5NHyX0tBU1LRyQvlsfTM4tPP3V1iiSUisZkMrF06VLeffddLl68SFBQENOnT8e3pCJglRRHJIFYqnOVSuPAMTpnRr/U6xluXLpWNHmhzAQOMwPwDUYjS376iTs7dKDvyJGkp6dz48YNvL2982yyQuOSkpIIDQ1lxYoViAiNGjVi2rRpvPLKK8WW4iqtd6+XV8F716xpokmTycTFvc+9997L/Pnz6d69e6k/p6LyUSmSQETkDOCYAl3lSWlZaOfP67EqJWWh2ZkFCy4ydmxdsrJyhLA5H3/sWyQuJYeK7GBRuOJ8eWQBl9zXMquA8wdw86YroaHOkYVc0Wiaxtdff02bNm3YvHlztc10Lg+UzpWDzplRrkVfOCtKmR0UygjAFxG27N/P2C++4ERSEq+8/DJ9gZo1a1KzZs0C51qicenp6cybN48ZM2Zw48YNatSowZtvvsnbb79daihKSRrn6lo0xjo93YVz50YSHn43gYGBartXUQSHlIExF6d/OjYjLgXIy94pxyB1k8mEh8dZjMbbi719ceZV5ApgRVDcEzjcALworoWbpunV76sD+/btY/LkySxdupQ77riDlJQU6tatW6W/FBxdBsYcnF7jwPE6Z2a5luIwq4RLCc7tsdOnGf3FF/z3wAFa+voyZ+FC+jz9tE3/ZkSEr7/+mpCQEBKyn8D79+/P7NmzadGiRZnXl7TKWNJDvtNqnJPsmFUFbNE51c3ZFsq5jVBZGAwGPvnkE27evImLiwtZWcUvPGRlFW34XRWruz/11FUeeWQ1EA+YgHj69NlAs2bFP+RUhwociYmJDB48mC5durB7926OHz8OQL169aq086ewIw7WOXPLtRReBTS7g0JOAH7fvtCxo35Oo0YcuHqVPadOMW/2bA5HR/N03742/ZuJjIzkgQce4MUXXyQhIYH27duzdetW1q9fb5bzB/pq4rJl+sO7pumvH39somHDYotqOJ/GSaFM7qQkvZRPUpL+fsMG/XMnXpiqSigH0FrKuY1QaYgIERER3H333QQHB7Nu3ToAfHyKFydf36KiYc/YO0eTv5Xdpk0DgeYMHDiI2Fhh48ZAZsxwqRYOcH5EhHfeeYeWLVuyZs0aQkJC+Ouvv+jVq5ejTVNUJhyoc7mYWa6lYe0MfL1voGmCr/cNlr0WZVEHhUyTiQ8iIlh84AA8/DCBU6bwV2wso8eNw72QDZZw5swZBg8eTLdu3fj9999p1KgRy5YtY9++fTzyyCMWz5e/d++BA1f54IMuXLoUjItLeoHznE7jclaSc/6eCj9Q5IzFxOjnKSew3FEOoLWUEJcSn1wLES23jVABcSzmOkuJjIzkoYce4plnnsHN7f/bu/O4KKv9geOfAwiKqaVo5QJ6FdywW+q1+pm4plczbXEJKStT3NK0VFTKpcQ99zRtVUG73rxqanVvda2rXQ1Qb4IraiBmKmqI5sIy5/fHADKsw8wwMzDf9+s1L5uHZ545PM18+T7nOed7PNixYwfBwcGA8cteVKKTN2gkJlaM5C9vIjxmzBguX75MUFAQMTExREVF5S7fVthVc0VKgPPKqfOllOLcuXM888wzHD9+nLlz57psmRthBQfFORNmrpe69OX/kbjySwx/+5zElV+aJn85ddQKub2otWb79u0EBgby+uuv88MPPwDG71DNmjUtbvaNGzd45513CAgIYN26dXh6ejJx4kQSEhIYNmxYoZM8zHXtmnF+UY0aNXjooYeIiurNunWVnTvGObonWRQgCaClzp0z+SCHb2xlUj8K4Ea6B+Eb7yzZQlaWyViW0tJa89prr3Hs2DFWrVrFoUOHTJZvc6VEJ28inJCQQEBAAFu2bOH777+nbduCwyHKWwJsbkHrHFprduzYQatWrcgZU7Z69WqioqIq3OxeYUcOiHMF5Ou1C+mQzJrhsSX39uUoZgWFo0eP0r17d/r06YObmxs7d+7kb3/7m1XN1VqzceNGmjVrxrRp07hx4wbPPPMMR44cYf78+VZdiKWlpTF58mQaNGhAUlISSik++ugjBg0aREiIct4Y5ww9yaIAGW1pKRstI1SSq1evMn/+fF577TXq1KnD+vXrqV27NtWLqOIcEuJkX3wby1nKbsOGDQDUqlWLGTNmMHz4cOcpXGyl0hZ7PXDgABMmTGDXrl0EBARwI/uF1vQwCAHYLc4Vy8PD2HuXJ4EI6ZBcdMKX93VaFzs7OTU1lQMHDrB06VJGjhxpdQz56aefGD9+PHv37gXgwQcfZPHixXTq1Mmq42ZlZfHJJ58QHh7OxYsXGTx4cPlas9eMmdyhq40X7ib/X5OTIXvNemF70gNoKRssI1ScnHFtjRs3Zs6cOXz99dcANG7cuMjkryK7evVq7lJ2GzZswMvLi7CwME6dOsWrr75aYZI/KN2yeaNHj6ZNmzYcOnQodym7oKAg+zRUVHxlHOfMFhho7MUz56LG0xPuvx8efNA4saNVq9zkLz09nUWLFhEWFgbAo48+ypkzZxg7dqxVMeTs2bO88MILPPLII+zdu5d7772XDz74gNjYWKuTv/T0dB5++GGGDRtGkyZNiI6OZu3atdx3331WHdeunKEnWRQgCaClzByXYtYstHy2bduWO67tz3/+M/v372fw4ME2a3p5kneCx/z587l9+zaDBg3i2LFjFXZcW0nL5l2/fj1nxQn8/PwqbCIsnEAZxrlSMXe91GbNjEnfY48Ze46yx/xprfniiy9o2bIlb7zxBkeOHCErOyGpWtW8kjKFuXHjBm+//TZNmzYlMjIST09PJk+ezIkTJxg6dKhVvfApKSkAeHp60rt3bz777DP27NlTPut2OkNPsihAEkBLWToupZhZaDk2btyYO8Hj22+/5aGHHrJly8uF/BM8Ll26RIcOHYiOjiYqKooTJ04Q4VRT3GynqNINDRpo3n//fRo3bsyWLVsAmDRpUoVNhIUTKMM4V2pFlGuhbt1Ce/tynDx5kscff5y+fftSqVIlvvrqK7Zv325Vcqa1ZsOGDTRt2pTp06dz48YN+vXrx9GjR5kzZ45Vd2nS0tIICwujQYMG7Nu3D4AZM2YwcODA8lu6yVl6koUJGQNoqdKOS1EKGjcudBZaUlIS4eHhTJkyhZYtW7Jq1SqqVauGh4sWxIyJiWHChAn85z//ASAgIIB58+bRt29flFJorQkJCeHSpUsEBwfzpz/9ycEttq2IiILFXr28MsnMnMLIkQsJCgqikYyLEfZgwzhn0zY1amT22DAPDw+OHj3KsmXLGDFihE3G+Y0bNy43OXvooYdYsmSJ1UMvsrKy+Pjjj3nzzTe5ePEiL774Ir5OV8jPQnXrGleMyf4MRQTHmYwBBDv1JAsT0gNojdKMS9EaTp40KXKZd1zb5s2bOXjwIAD33HOPSyZ/SUlJhISE0K5dO/7zn//g4+PDihUriI+P56mnnsq9+t2+fTuXLl0CqJC9gPlnc1eteonbtwdz111fsHXrVr7//nuX7BUWDmJlnLO327dvs3DhQgYOHAhAw4YNSUxMZMyYMVYlf8nJyTz//PM88sgj7Nu3j/vuu4+PP/6Y2NhYq5M/rTWdOnUiNDSUgIAAYmNj+fTTT6lbURIgZ+pJFrlkKThr5V1GyGAoOehllyRYc/gwU8PDuXLlCi+88AIRERHUr1/fPm12MqmpqcyZM4elS5dy+/ZtvLy8GDduHFOmTClwa1NrTZs2bTh16hQZGRmkp6dz4sSJCtcLePbsWWrXro2Xlxeff/4558+fr1Azne1BloKzIQvjXFkuf1mwiZpt27YxYcIETp06xRNPPMGmTZvwzl8ctZT++OMPFixYwPz587l58yZeXl68/vrrTJkyhWrVqll17KSkJHx9fVFK8eGHH1K9enX69+9ffm/1FicurkApmCLl1G1s1arkfV2cLAXnSDnjUho3LnFXrTU6MxMuXuR0dHTuBI+1a9e6ZPKXkZHB8uXLTSZ4hISEFFu4ePv27Rw8eJDWrVtTtWpVKlWqVKF6AdPS0ggPD8ff358VK1YA0K9fP5ngIRyrFHEOMP6Rv3DBbsV8z5w5Q9euXXn66afx9PTk66+/ZseOHVYlfwaDgcjISJo2bcrMmTO5efNm7ji/2bNnW5X8Xb16lYkTJ+Lv759bd3Do0KEMGDCgYiZ/YH5PcjF1G4Vtud59xrKQmQmnTplcFUftbkD4xlacueyNb60bvNL5W76Nn8KUp5/mrw8+yDs9euDx9NMoF/yjnnOlPmnSJBKyF2APCgpi4cKFJc5wS0lJ4bHHHsPf35+jR4/y5ptvsn//fns0u0xlZmbywQcfMH36dFJSUggJCaFfv36ObpYQd2RmGm/vFhPnIoLj7tzGMxjgxAlo3rzMxgRqrVFKcffdd3PhwgVWrFjB8OHDrR5Cs3fvXsaNG0d0dDQArVu3ZsmSJXTo0MGq42ZlZfHhhx/y1ltvcenSJV5++WU6duxo1THLjZyZ3Dk9yWDaG2hG3UZhW5IA2oIZRS6n/b0b1aps5/qtWwBU8vCAs2ddrshlYRM85s+fTx8zF1p/5ZVXGDJkCCNHjgQgPDwcZx7GYK6XXnqJqKgogoKC2LlzZ/ks9SAqtuRkY1KXzaxivgZDmRTzvX37NkuXLmXLli3s3r2b6tWrExcXh5ubdTe1zpw5w+TJk9m4cSMA9913H7Nnz+bFF1+0+tgATz/9NNu3bycoKIjFixfTunVrq49ZruT0JDdvbvxcnDtnLPVSqZJxwkeDBmU7gUiYkDNtC2YUuYSqVHJbQb9Hthuf5hS5dJEEMDExkalTp+YGVh8fH2bMmEFoaGipb23mTxTL6y2TAwcO0KBBA2rXrs2YMWPo37+/2YmwEHZ39qzJ0+KK+ZoM5j90yLg2maen1X/ktdZs2bKFiRMncvr0aZ544glSU1Px8fGxKkH7448/mDdvHgsWLODWrVt4eXkxYcIEJk+ezF133WXxccFYhqZ+/fpUrlyZ4cOHM3jwYJ599lnX/p6Xcia3KBt2HQOolHpDKRWvlIpRSsUqpawbness8hS5zMjM5Mylwn+tK394ma536AJFLlNTUwkLC6NZs2Zs3LgRLy8vJk2axMmTJxk9erRLjmtLTk5m8ODBtGnThnnz5gHw8MMP55a5EeVbhY1z16+bPE0qIs4V2J6eDpcuGS94Dx6EL76waJbwxYsX6dKlC88++yxVqlThX//6Fzt27MDHx6dUx8nLYDCwbt06AgICeOedd7h16xYDBw7k2LFjzJo1y6rkLzU1lTfeeIMWLVqwdOlSAJ544gn69esn33PhFOzWA6iUmgl4AQ9qrTOVUncDt+z1/mXK09M4ri0mhrCoKO6563GuXC8sKCnTq+MKnPxkZGSwevVqZsyYweXLlwEIDg5m9uzZNGzY0LGNc5C0tDTmzZvHokWL0FozadIkpk6d6uhmCRuq0HEu3wWru5smy1AwkXF3Kyaxy7lTkpAAV6+aNUs4MzMTDw8PatasiZubGytXrmTYsGFWj/P78ccfGTduHDmzsNu2bcuSJUto3769VcfNGc87bdo0Ll++zMsvv8yLL75o1TGFKAt26QHMDoI9gGvAj0qpH4BHtdaG4l9ZPsSkpNBp5kyeXrgQdzc3hnXbBRQeBHOXurFTkcuoKGjYENzcjP9GRZXt++VfwePy5csEBQURExPDhg0bXDb5A3jjjTeYPXs2zz77LMePH2fevHmygkcFUtHjXP4L1sKSv+K2m+6UBRcvFjtL+NatW8ybN4+mTZuSmpqKh4cH3377LSNHjjRJ/kob45KSknjuued47LHHiI2NpW7duqxdu5affvrJ6uQPYNiwYYwaNYqWLVuyf/9+Pvroo/K1bq9wHVrrMn8AXTEGxReznz8E/A40LmTfUCAWiPX19dXObtq0aRrQdWrU0O8PG6YzNm7UetMmXavaLW28x2H68PO5rvWmTVpv3qwj12VqPz+tldLaz0/ryEjbti0yUmtvb9P39/a2/fvk+Omnn3SHDh00xuxXBwQE6K1bt2qDwWDR8Y4ePapHjhypmzRpoqtUqaKrVaummzZtqgcMGKCHDh2q7733Xhv/BrZlMBj09u3b9ZEjR7TWWv/yyy86Ojrawa1yHUCstkN8y3mYG+fKW4zL9Z//GGNX9sPP53rxMW7TJh05Zq/287mulTJoP5/rOnLMXpNj6M2btc7IMHkbg8GgP//8c92oUSMN6D59+uizZ88W2qTSxLhr167pN998U1euXFkDunLlyvrNN9/U165ds/rUHD9+XF+4cEFrrXVsbKzevHmzxXFPiNKwJs7ZKzAGA3H5tn0OjC7udW3atCn8N87I0Pr0aa337NH63/82/nv6dIFAUlZSU1P11atXtdZaf/PNNzo8PFxf/e9/jcEsT+Dz9swwDUyeGcYAuHmzjpyTVObJmZ9fweAMxu22lJiYqIODg3MTv1q1aunly5fr9PR0i4+5a9cuXblyZe3l5aWfffZZPXnyZD1mzBjds2dP3bRpUz18+HCnTgD379+vO3furAEdGhrq6Oa4JAckgKWOc0XGOK0dHucKOH1a67//3bwYZ8bPcxPA06dz3+L69es6KChIAzowMFB/8803xTbJnBiXlZWlP/30U33//ffnxqjnnntOJyUlWX1Krly5osePH689PDz06NGjrT6eEKVlTZyz1xjAi0Bavm0GwIyS4HloXXQNoQsXjAOMy7CGUEZGBu+//z4zZ85kyJAhzJ8/n27dutGtWzdj23780XhbIysrd5xfgRpZnc5BnTqEj29gstYrGNd+DQ83LgVmC2fOlG57aZVmBY/SCg8PJyMjg+jo6EJLJYwYMcKq45eV5ORkwsPDWb9+fe5SdqGhoY5ulrCPChHnitSggfG986wJDIXEuDzbS5wlnF0N4eZ991GlShWqVq3Kn/70JwYNGsQrr7xS4ji/kmLcnj17GD9+fO44v3bt2rFkyRIeffRRS85ArszMTNasWcO0adO4cuUKQ4cO5a233rLqmELYm70SwN3An5RSj2mt9yilmgJdgMlmHyFfglWABYOLzX9r08LFnTt35rnnnjPdqZAilyaLpucrcnkmufC22So5A/D1haSkwrdbo7AJHoMGDWL27Nn4+flZd/Bsly5dokaNGrRo0cImx7OXlStXsmnTJsLCwmySCItypVzHuRJ5eBjjV57lvExiXD65452L2X4rPZ3Fn37Ku3378tNPP9G4cWM++eQTs5tUVIyrWzeTgQND2LRpEwD16tVj7ty5DBo0yCb1/KZMmcLChQvp1KkTixcv5sEHH7T6mELYm10SQK11ulKqF/ChUsoN41Xxi1rr02YfJD6+6KCYV97BxTZaRzA8PJw5c+bQvHlzduzYQa9evQqfxl+KIpdllZzlFREBoaGY9DR6exu3WyJ/IgzQoUMH3n33XZsXLl60aBFDhgyhdevW9OzZk2rVqtG1a1erK/HbWs6Mv6ZNm9KlSxcmT57MiBEjbJYIi/KjvMc5swQGQmoqnD9f4q6+tW6QdKlqodu11ny+bx8TIyNJSkmhb9++Fs3qLSzGeXikc/78MDZt2kSVKlWYNGkSEydOpGrVgm0pjWPHjuHu7o6/vz9jxoyhffv2UrpJlGt2KwOjtT4ItLHoxZmZBRaRLnYJoqws4/5WLEGUmJiIu7s7DRo04Pnnn8fX15ehQ4eaF6TMKHJp6+SsMDm3ksPDjT2Lvr7G41tyizkmJoY33niD3bt3A6VfwaM0tNZcuHABPz8/YmJiOHr0KADNmze36ftYQ2vNzp07mThxIseOHWPYsGF06dKFGjVqSK+fCytvca7UlIIaNYy3onXxdfwiguNMVgoB8PbM5J3nDtH17bfZdfgwD/j58V1kJF0sHPeS87KpUzVnzoCb21kyM8OAjYSEhDBnzhwaNGhQ7DFKcuXKFd5++23ee+89evfuzZYtW/D19cXXllfrQjiAXQtBW6yIpdaSLlVFa5W7BJFJkeVCXmeOvIWLw8LCAGjRogUjRowo9RVqceUJQkJgzRrw8zPGVD8/43Nbjf/L+z6JicYVmRITS3/8xMREQkJCaNeuHbt376ZWrVosX76c+Pj4Mrv6HTt2LK+88gpt27YlPj6eW7duobVm4MCBNn8vS/zvf/+ja9euPPnkkxgMBrZu3crq1asd3SxR3tkxzlmsiPWAG47qhdvAfjQc1Su3fSEdklkzPBY/nz9QSlO/1nXWDI/lhaCzBDVvzprQUA4sWEAXK7/XDRr8Bx+ftoAbBoMv7dqdYu/evURGRlqV/GVkZLBixQr8/f1Zvnw5Q4YMke+5qFDKRwJoxlJrOYOLc+UstWamjIwMli9fjr+/PwsWLGDgwIHMnz/f4iZHRRl7+JKSjLEyKcn4PH8SaE1yVpbyJsIbNmzAy8uLsLAwTp06xauvvlpmK3hcvHiRlStX0qNHD1auXEnLli3x8vIqk/ey1L59+4iLiyvzRFi4GDvEOauVMkkN6ZDM0SVbeWfgIH6/7oOvz78AmDFgAMN69MC9WTOLey9/+eUX+vfvT8eOHTlw4AD16tUjMjKSvXv38sgjj1j3ewLLli1jzJgx/PnPf+bgwYOsXr2aOnXqWH1cIZxF+VgLOM9Sa2De4GKgVEutRUREMHPmTDp37szChQutXqQ7PJwyn+VbFvLOdC6rCR7FuXjxIgaDgbS0NLKysnB3dzf5+c2bN6lSpUqZtyOvnBU8GjduzJAhQxg6dCjBwcFyq1fYlh3inNVKkaQOeuwMm/buZVJkJGcuXeLpdu2oV7OmcSd3d6hTxzimsJTS0tKYM2cOixYtIj09nSpVqhAWFsaECROsHud39OhRrl69yiOPPEJoaCgBAQH07t1bLvBEhVQ+EkBPT5OnxQ0uNlFCL1VMTAxubm60adOG0aNH07ZtW5544gmbfNnLugSLrRU2wSMoKIh3332Xtm3b2q0dTZs2JSAggL1799KiRQsef/xxatSowaVLlzh8+DABAQF8/PHHdmlLzgSP6dOnk5KSwmuvvQaAh4eHJH/C9sooztlUKZLUJ+fNY+eBA/zZz49PR42ic2BggWoIpZnBnJWVxaeffkp4eDgXLlwAICQkhLlz51K/fn3LfyeM4/xmzJjBypUradu2Lfv27aNatWo8+eSTVh1XCGdWPm4B161rvGLMFhEch7dnpsku3p6ZRATH3dlQzFJriYmJDBo0iHbt2jFt2jQAateubdMrvaLGBzvjuOHo6Gg6duzI008/TUJCAgEBAWzdupXvv//erskfQKVKlfjuu+8YNmwY6enprFmzhiVLlvDtt99y//338/LLL9ulHd9//z2tWrVi1KhRNG/enJiYGJYsWWKX9xYuysZxrkwUkqQWxrfWDZ76y1/4YPhw9s+bR+fWrY3tfPBB6NPHOHO5FLE2JxYNHTqUCxcu8Mgjj7Bv3z4iIyOtSv4yMjJYtmwZTZo04b333mPYsGFs377d4uMJUZ6Ujx7AnAKk2UoqQGryujzyFi5WShEeHs6kSZPKpMn2mOVrrcTERKZMmcJnn30GgI+PDzNmzCA0NLTMxviZo379+qxZs8Yh7621RinFrVu3cid4lMVMZyEKsFGcK1N16xpnAGffBi5spq+nx+3sdnY1bnB3hwceKLYqQlFOnz7NxIkT+cc//gFAgwYNmDt3LsHBwTb5Tv7973/ntddeo1u3bixatIhW9iypI4SDlY8EsJQFSHF3N+6fb3Dx2rVrWbBgAc8//zwRERFWlwcoji1LsNhaZNQyjQAAIABJREFUYSt4vPbaa0ydOtVlb23mrOBRr1495syZw1//+lcOHz5sUW0yISxiozhXpvIlqYMeO8PeEyd4/5vOZBnqUcXzAhHPHSKkQ1rB15VCWloaERERLFmyhPT0dLy9vZk8eTJvvPEG3t6F33Y215EjRzh9+jS9e/dm4MCB1K5dm27duslFnnA55eevW2CgsfJ9SUVS8wwuzhnX5ubmRp8+fRgxYgQdO3a0W9X2kBDnSPhyFLaCR0hICBERES5buDgtLY25c+eyePFitNa5pX8ASf6E/VkQ5+wqX5L6wvLlRO3Zw0ONGrH4xRfpmH/lnlImqVlZWXzyySeEh4dz8eJFAAYPHszs2bOpV6+eVU2/fPkyM2bMYNWqVTRs2JCePXvi7u7O448/btVxhSivys9fuEKWWjMJkPkGF8fExuYWLu7evTt9+vTBy8vLJZfsyUmEw8LCOHHiBGCc4LFw4UKbr+BRnnz55Ze89NJLpKSkuHwiLJxEKeOcXdcCznb27rupWb063mlphHToQJfAQF7s1An3/EuslTJJ3bVrF+PHj+fnn38G4P/+7/9YsmSJ1TEqIyODlStXMmPGDNLS0hgxYgQzZ84sUGFACFdTfhJAMGuptaRff2VKSAgbN26kTp06rFq1iqFDhzq65Q5jzxU8ygOtNTdv3sTb2xs/Pz8CAwOZN2+eSyfCwsmUYklJe7px4wYLFy5k3rx5hE2axLRnnqFnzg+tSFJPnTrFxIkT2bJlCwC+vr7Mnz+fAQMG2CRG7d27l3HjxvH444+zaNEiAu3dayqEs9JaO+2jTZs2urQ2b96sK1eurMPDw/XVq1dL/fqK4pdfftHBwcEa0ID28fHRK1as0Onp6Y5umlU2bdqkg4ODdWhoqL733nv1mjVr9IgRI8x67f79+3Xnzp31gAEDyriVwhkAsdoJ4lhxD0tinL1lZWXpyMhIXb9+fQ3o/v3769OnTxt/mJGh9enTWu/Zo/WuXcZ/T582bi9Bamqqnjhxoq5UqZIGdNWqVfWsWbP0jRs3rG5zfHy8/uijj3Kf//TTT9pgMFh9XCGcjTVxzuEBsLiHOcExPT1dL1u2TC9evFhrrbXBYNDnzp0rxemrWFJTU/WkSZO0l5eXBrSXl5eeNGmSTk1NdXTTbCIyMlIDunv37rp27dq6evXqun///sW+5syZM/qFF17QgK5Vq5Zevny5/DFwAZIA2sarr76qAd26dWv9ww8/WH28zMxMvXr1al27du3cC9TBgwfrX3/91epjp6Sk6FGjRmk3Nzd977336uvXr1t9TCGcmUsmgAaDQW/ZskX7+/trQD/55JMu/Uc9PT1dL1++XNeqVSs3qA4aNEgnJiY6umk2lZGRof39/XWtWrV01apVNaB//vnnIvf/4osvdOXKlStcIixKJgmg5ZKTk/X58+e11lofPHhQf/zxxzorK8vq43733Xf6gQceyI1R7du31zExMVYf9/bt23rRokW6Ro0a2t3dXb/66qv60qVL5r04by/mv/9dql5MIRzN5RLAQ4cO6Q4dOmhAN2/eXO/YsaNcJH+RkVr7+WmtlPHfyEjrj5mTCAcEBOQG1aCgIJsEVWe1bt263N/12WefLfDzjIwMffbsWa211hcvXtRDhgypcImwKJkkgKV3/fp1PX36dF2lShX9yiuvWHSMwuLciRMndN++fXO/t35+fvqzzz6zPm5nJ2/HN2zQHu7uuke7dvrwP/9pXvJmMGh96JDWmzcbH5s23XnkbDt0yLifEE7KmjinjK93Tm3bttWxsbEFtu/fv5/evXszffp0hg4dWi7KdURFFV4Yes0ay0vFFDbBY8GCBTz55JMVeoJHZmYmPj4+XL16lZ9//pkHHngAMF7M7Nixg0mTJlGtWjX27duHW/6ZicJlKKX2a63tu5RNKRUV4+zNYDCwYcMGJk+ezK+//srAgQOZO3cuDRs2LNVxCotzHh7pGAyvYDBEUrVqVaZMmcLrr79u3ZreWhO/ZQvbPv+c8H79ICuLY7/+SrN69e6splLcJBSt4ccfzS+30769Q2ZcC1ESa+JcufjrmJqaSlhYGCNHjgSgTZs2JCYmMmLEiHKR/IGxIPSNfKsm3bhh3F5aSUlJhISE0K5dO3bv3k2tWrVYvnw58fHxLjG718PDgyVLlvDKK6/kJn8HDhyga9eu9OnTB4PBwNSpUyv8eRDCVmbNmsULL7zAfffdx+7du/nss89KnfxB4XEuM9MTg+EdhgwZQkJCAuHh4VYlfykXLzLy6af5c//+vPvFF5zPrmnaLKdOYFaW8ZGQYEzyCuvkiI8vOfnLOdbFi8b9hahgnDoB1FqzfPlymjRpwoIFC7h9+zYGgwEALy8vB7eudM6cKd32wuQkwk2bNmXDhg14eXkRFhbGqVOnePXVV8ts+baoKGjYENzcjP9GRZXJ25TKSy+9xIcffggY6/m1adOGuLg4VqxYQXx8PE899ZQkgEIUIzk5Obcu6NChQ/n000+Jjo7mscces/iYRcUzpfz46KOPuP/++y0+dnp6Ou+++y5NGjfmg+3bGd29OwnLlnHf3XcX/oKikrfMTJPVVgCidjeg4aheuA3sR8NRvYja3cD0OAkJxtcJUZFYeu+4tA8gDogB9mU//l3Sa3Jmsnbp0kUfOHDApvfN7c3PT2vjpajpw8+v5Nfevn1bL1u2zKIJHtaOO4yM1Nrb27TN3t62Gb9ojatXr+Z+Jm7duqVnzZolEzyECRwwBrC0cc4RYwCvX7+up02bpqtUqaJ79Ohhs+MeP35cV6lyweI4V5JLly7pu+++W/d86CF9ZNGi3PF6kWP2aj+f61opg/bzua4jx+wtOJ4v75jA06dNxvxFjtmrvT0zTGOcZ4bpcTZvNr5OCCdjTZyzZ2BMBNxK8xpvb2+9c+fOcjHBoySWJFIGg0H/4x//yJ3pDOgOHTro6OjoMnvP/KxJXMtCRkaGXrVqla5Tp4729fXVGTJTTxTBQQlgqeKcPRPArKwsvW7dOl23bl0N6Oeee84mk6N+//13PX78eO3h4aEhWMMfNrtg/Pnnn/Xo0aNzZyAn//ij9cnbnj0mCaKfz/XCY5zPddNEcs8eq8+VELZmTZyz5y3gmsAPSqmDSqlNSqlC12RTSoUqpWKVUrE+Pj706tWrQtzKCwkxTvjw8zOOJfbzK34CSHR0NB07duSZZ54hISGBgIAAtm7dyg8//GD2qhW2GHdoi1vXtqC1Zvv27bRq1YqRI0fSrFkzPv/883IzBlS4jBLjXN4Yl5KSYreGrVmzhsGDB1OvXj1+/PFHNm7caNXSh5mZmaxatYomTZqwePFisrKyGDKkCu+9l252nCvkoPDLL1zcsYMRffrw0EMPsTEqihNHjgBQX2uTW7fhG1txI900BtxI9yB8Y6s7G7KyjKup5EhPN9n/zGXvQptSYHtGhpm/hBDlgz3/et6rtb6plHIDgoFvlVKttdYmqYTWeg2wBowz5OzYvjIXElJyIExKSmLKlCls3LgRAB8fH2bMmEFoaGipx/jZInnz9YWkpMK329OuXbvo06dPbiLsCpNdRLlUYpyzZ4w7c+YMv/32Gw8//DCDBw+mRo0aDBw40OrZ8d988w3jx4/n8OHDAHTs2JHFixfz0EMPATBqVCkPqDXEx5N+5AjLvvySd/7+d26kp/Nqjx5MHziQmidOGPe5fdv097MkefP0NPmRb60bJF2qWuAYvrXyXT2X0RhrIRzFbj2AWuub2f8atNZRwH6gh73e39nlneCxcePG3AkeJ0+eZPTo0RZN8CgqSStN8hYRYSxXk5e3t3G7RbKv8PnxR9i1y/jvL78UOsA6OTmZbdu2AdC5c2c+++wz4uPj6du3ryR/wik5S5y7fv06b731Fk2bNmXYsGE5Q2oIDg62Kvk7fvw4vXv3pnv37hw+fJhGjRqxefNmdu3alZv8lZrOLsmSkAAGA6v/9S86NG9O3MKFLH35ZWp6e9+ZiHH9uslLCyRpRW3PGz/r1r1TKgaICI7D29M0/nh7ZhIRHHdng7u78XXmKEWME8KRHHn/zB1Ic+D7O4WMjAzef/99Zs6cyeXscgbPP/88s2bNsur2DBiTtMJqD5YmecvpsQwPN/Yc+voaX1/q2oXZV/gkJBif5y2/cOECHDyYW7cr7do15s6dy+LFi7nrrrvo3r07lStXZuDAgaV8UyEczq5xzmAwsH79eqZMmcJvv/1GcHAwc+fOtfqC6ffff+ftt99mxYoVZGZmUq1aNcLDwxk3bpzVFRl+3ryZOUuW8NHw4VStXJl9ERHUqlat4I5ZWZBdBSJHRHAcoavbmtwGLjF5a9DAGG+yhXRIBoy3k89c9sa31g0iguNyt5u8rjiliHFSU1A4A7skgEqpvwAGrfX+7Oe9gObAP+3x/s5Ia83WrVsJCwsjITtgBAUF8e6779K2beE1HaOiSpeI2Sp5M+fWdbFyrvCLqruVvS3z2DE++PRTpq9fT0pKCiEhIURERFhXMFYIO3GGOPfFF1/w0ksv0a5dOzZv3syjjz5q1fEyMzNZvXo106ZN48qVKyilGDp0KLNmzeLee++16tgXLlzgrfBwPvz4Y2redRfxyck87O/P1/9rUXQypk3vmFuUvHl4GBOxPKVgQjokF3xNDnd34/7FjTc2M8aRkABXr0phaeEU7NUDeB1YpJS6D7gNXAG6a61T7fT+TsWSFTzyV9hPSjI+h5KTQKuSN1sws+hqfGIioxYtokPr1uzcudPsyS5COAmHxLnExESOHDlCr1696NOnD9u2baN3795Wj/P75z//yeuvv86R7AkYnTp1YvHixTz4YKHz98yWlZXFu+++y6xZs7h58ybje/fmrWee4e6qVYna3cCkRy/pUlVCVxsviItK0CxK3gIDjYmYuSuBBAYW/0tZUli6Vavi9xWijNllDKDW+qjWuqfW+iGt9SNa615a68P2eG9bsrYgcmJiIoMGDcpdwcPHxye3cHFJkxrMndHrdEWbSyi6en9od0KWXQLgwYYNiZkzhx+mTuUvlo4nEsJB7B3nrl27Rnh4OM2aNWPYsGGkp6fj5uZGnz59LE7+oqKgXr0MlDLw17825ciRP9O4cWO2bNnCv//9b6uTPwA3Nze+/vprOnbsyOF163j3hRe4u6pxEoZZs3opoXDznTcqOnlTytgL5+9vTPLyjAkEjAljTvJYUm+dFJYW5ZTU0DCTpT1wYJzgMWfOHJYuXcrt27fx8vJi3LhxTJkyhRo1apj1/ubM6LWmjWUm2fTKPP8V/vnUGmzY8yKdW8QytNsF2jZubAy2ycnQqJEjWiyEUzMYDKxdu5apU6dy/vx5QkJCmDNnDp75ZreW1po11xk92pPMzJzjNKRSpbW89RY89ZR1M2APHjxIeHg4a9asoX79+mzfvp2qVasaJ0lcupS7nzmzes3uJSxpDV+ljL1wzZsb4825c8bZwpUqGccMNmhQ/G3fHCXEuCLbJzFOOJhTLwXnTCypqZeRkZG7lN38+fO5ffs2ISEhHD9+nLlz55qd/IF5M3ptud6wzZw7Z3JlPGVDYIErfPBm1j/a3Hmav26XECJXTEwMQ4YMoWHDhuzbt4/IyEgalDRBoRg5cWrkyCt5kr+cn1Vi+nTLk7/z588zdOhQ2rRpQ0xMDMePHwcwJn9QaEmWwuTdbm4vIW5u5o2z8/AwJmLt20OnTsZ/GzUyL/mDAjHOotqEQjiAJIBmKk1NPa01W7ZsoWXLlowdO5bLly8TFBRETEwMkZGRFs3uNacci7MUbTaRr+jq2csF622BFF0Voji//PILa9euBeDhhx/mhx9+4L///S8PP/ywVcf96quveOCBBxg7diwGQ/1C97E0fsyfPx9/f3/WrVvHuHHjSEhIoGvXrqY7WVCSxekKN0thaVFOSQJYhPxj6WrWLHy//D1zMTExha7g8f333xc5u9cc5qwkYou6f7amK1Vie2wsoatXo7XG18eCul1CuKhr164xZcoUmjdvztixY0lNNc4nCQoKsqq0y5EjR/jzn+fRq1dzjh07jIfHWapVKzwhsTR+HDt2jC5dunD48GEWLVrE3XffXXCnfD2XIR2SWTM8Fj+fP1BK4+fzB2uGx5rcOrWo9l9ZsqAXE5AYJxyu3CSA9pzckDOWLinJOLs/KQmuXSv4fc3bA5d/gketWrVYvny5TQsXh4RAYqKxFFZiYsFxfTYv2mylAwcO0HXSJPrMn88PR46QkpZm+6KrQlQgd+Kcplata9SvP4m5c+cyYMAADh8+XHgSVQqXL19m7NixBAbO5tChV4GGgBuZmfW4fdur2BhXkgMHDtC5c2diY2MBWL16Ndu2bcPf37/oF+WUZMnTCxjSIZnElV9i+NvnJK78ssAMX6eLIWVdWFqIMlIuEsDCErLQ0LJLAgsbS5eeDtWrF+yBe+KJVCZNmlRgBY9Tp07x6quvWrSCh6VKu95wWbl8+TKDBw+mTZs2xCUksGLoUOLffZc6NWqYdYUPlFx0VYgKxjTOKa5cqcb164uYOTOBdevWUb9+4bdozZGRkcGyZcvw9/dn+fLlaB0BmA7HKCrGlRQ/fvvtN4YMGULbtm2Jj4/nt99+AzA/9gUGGids5J+JWwSniyEW9GLatX1CFEFp7bzL7bZt21bHxsbSsGHh69H6+Rl7wmzNza1AvVHAGBRzCtGnp6ezevVqkxU8Bg0axOzZs61ewaO80lqjlOKPP/7ggQceoH///saZzmfOFCiTUKSc0gtSI0tYSSm1X2tt+bgLO8iJcWAsv3LuXMGkyZo4p7Xmq6++4vXXX8+dgNGtWze+++5faF3wrkTeGAcYS5XkzJBNTzfe7swzQ3bJkiW89dZb3L59m9dee40333yzVJPb8jS06FU0PDzuPDfn75UjYkhcnMQ44RDWxLly0QNo6eQGS28bFzeWLmeCR2BgYIEJHlFRUS6Z/GVmZrJq1So6dOhAeno6VatW5ejRo3dmOpt7hW9u0VUhKpC0tDQmT57MuXOFfz+Ki3PFxbjDhw/Ts2dPnnjiCY4fP46/vz/btm3jX//6F76+hQ9JyY19WhuTmi++MC5hdu6csVzLuXPoAwfQ27ZBXBzX0tLo1q0bR44cYcGCBZYlf3CnJEufPvDQQ8Yks3Zt478PPgh9+8J99zlvDJEYJ8qhCtsDmL8mHhjHs+S9pVHU0mpFvTYs7CTffPMye/bsAYwreMybN89mY/zKG601O3bsYNKkSRw7doygoCA+++wz7r///sJ2Lv4KX2tZJ1PYVHnoAWzYsKG+efMmFy9epGrVFP74w6fAPkXFuaLi1LvvphEfP5X333+frKwsqlR5hcqVF5GaWg1fX5U7pq/I+Dio6GXN9p8+zbhPP2X8E0/wzP/9HwYfH9w6dLDPd9bZY4izt09USFbFOa210z7atGmjtdY6MlJrb2+tjd8g48Pb27i9KH5+pvvnPPz8tFnHjIw07quU1vXqZehHHlmmAQ1oHx8fvXz5cp2enl50Ayq4lJQU3blzZw3ogIAAvXXrVm0wGEp+YUaG1qdPa71nj9a7dhn/PX3auF0IGwJitRPEseIeHh4eun379jo6OrrUca6oGKdUkga0m5ub7tr1I12liqHQY+aNcX5+ed7n0CGtN2/WetOm3Mev77+vX+rUSSuldJ0aNfTfxo0z/mzzZuP+9uTsMcTZ2ycqFGvinMMDYHGPnARQ62KCVRGUKio4Gn9eUoKotda///67njhxovb09NSA9vLy0mFhYTo1NdWiNlUEt27d0lprnZWVpbt3765XrFjh0omwcF7lIQEMDAw0uXAqTUwpKsZBlu7WrZuOi4szK86ZyMgokPy91GmNViRqyNLVq1zQH4T+YPJzvXmzJDdCOIg1ca5c3AK2REm3jYub6HH7dgbvv/++yQSP559/nlmzZuWO8TPnFnNFkpaWxty5c1m7di1xcXHULKowohBOojzcAi6LGFe79g0uXKiCUsqsCW0mfvkFDh5EZ2Zi0JrPfvRjyKrWpOdZIcTbM9N0Vqu7u3HcnixrJoTdVfhJIJYoqSZeURM9fHxumKzg0bFjR2JiYli/fr3JBA+nXHatDORM8GjSpAlz5syhc+fOZEgFeyEcKiUlhSZNPgb+MNnu7a1ZvNg7d0xyqYvDnztHzPHjdJg2jcU7dhC+sZVJ8geyrJkQFUWFTQBLqolXWILo5naLlJShuSt4bNu2jV27dhW6godTLrtmY2lpabRq1YpRo0bRvHlzoqOjiYyM5N5773V004RwSenp6SxatAh/f3++++4VlBrOXXddNtab84M1a5TJHYjSFIf/9ddfefHtt2k3dSoJ589z3913y7JmQlRgFTYBhOJXzshJEOvVywQMQCIGwxB8fL7JXcGjT58+Rc7udcZl12zlXPbVfPXq1enZs2fuUnZ/+ctfHNwyIVyT1povvviCli1b8sYbb3D16lV69OhBfPxUrl2rhcGgCl0dyNzi8GvXriUgIIDPdu0irG9fEpYu5fmgIFnWTIgKrEIngMVJTU3l558nkZJSFXDHy6sZYWG+nDx50qwVPJxt2TVbSE5OZvDgwTRq1IgTJ04AsGjRIpctcyOEM4iLi+Pxxx+nb9++nDx5kmbNmrFz506+/vprWrRoUeLri7oQ1lpz8+ZNAJo0aULPnj05+s03zB08mOrZwU2WNROi4rJ7AqiUaqmUuqKUmmHv9wbjLZTly5fTpEkTFixYQHp6OiEhIRw/fvxO4WIzOMuya7aQlpbG1KlTCQgIYNOmTYwfP15u8wphIVvFuJSUFEaOHMmDDz7Id999xz333MOyZcs4dOgQvXr1sqqN0dHRtG/fngkTJgDQvn17Pv/8c/7Uvr3JfrKsmRAVl4c930wpdTfwHrDRnu8LxqvdrVu3EhYWRkJ2oc6goCAWLlxo8a3NkJDymfDldfPmTVq0aMGvv/5KSEgIERERLrmaiRC2YIsYl3OR+vbbb5OWloa7uztjx45l+vTpVs++P3v2LFOnTmX9+vXce++9hIaGmu7g4WEsVpxnWbOQDskFE74cOcuaedj1T4kQwgbs1gOolHID1gJTgZRi9gtVSsUqpWJTUorcrVj5l0d6++2TBAUF8cwzz+RO8HDlcW1aa/bt2wdAlSpVeOutt4iJiSEyMlKSPyEsZG2M01qzbds2WrZsyYQJE0hLS6Nnz57ExcWxdOlSk+TPkmUuN2/eTNOmTdm0aROTJ08mISGBl156qeCOsqyZEC7BnreAZwHfaK3/W9xOWus1Wuu2Wuu2tWvXLvWb5NTnS0oy1r9KSoLp0+9nz54G+Pj4sGLFCuLj4112XNuBAwfo2rUrjz76KP/5z38AGD58eKEznYUQpWJxjDt06BDdunXjqaee4uTJkzRv3pyvvvqKL7/8kubNm5u8vrAYFxpaeBJoMBhITU0FoHXr1vTt25ejR48yZ84cqlWrVngDlYL27Y09e+7uBRNBD487PX/t28uyZkKUU3bpt1dKPQv4aq2nlvV7FVafD6pSvfp7nDzpZvli5eXcmTNnePPNN1m/fn1uIvzoo486ullCVAiWxriMjAxCQ0P56KOPMBgM1KxZk5kzZzJ8+PAiJ6IVV4M075CUffv2MX78eO655x6+/PJLGjVqxIYNG8z9haBVK2jeHJKTjXX+MjKMs33r1jWO+ZPbvkKUa/b6BvcEmiul9mU/rw/GwdJa6/62epP09HSSkioBBa9Ir127BxfN/cjMzKR9+/akpKQwefJkJk+e7LKJsBBlxKIYFx8fz6FDh0oe55eZmZuInTnzfxQW43JqkCYnJzNlyhSioqK4//77GT58uHHZJ0t66jw8jCt8yCofQlQ4dkkAtdZD8z7PmR2ntZ5ho+OzZcsWwsLCgG+AhgX2qQj1+UojIyODv/3tbwQHB+Ph4cGHH35Is2bNZIyfEGXA0hhnMBh44oknWLhwIc2aNSvswBAfb5yUAZCVhW+tGyRdqlpgV19f+Oabb+jbty9aa958803CwsK46667LPythBAVWbmvAxgdHU1QUBDPPvssJ0+e5L77luPlla9uVTmvz1caWmu2b9/OAw88wAsvvMCOHTsA6NGjhyR/QjgZf39/duzYUXTy9+OPd2bkZs/KLaw2XxWvTCIiNO3atSMkJIRjx47xzjvvSPInhCiSQxJArfUMa3v/EhMTCQ4O5uGHH2bPnj2549rOnJnLRx95VIj6fKW1f/9+unTpQp8+fXLL3vTp08fRzRLC5Zgb46pXr170D+Pj4eLF3MQvh0ltPjSeHr9Ss+p4Brb4HzVq1OCDDz6Qiz0hRInK3Sje1NRUZs+ezdKlS0lPT8fLy4tx48bRpMk0Zs3yZswY462QiAjXSPpyGAwGBg8ezMWLF1mxYgWhoaElrmYihHBSmZkmtfgAonY3IHxjK85c9qbuPdfxrbWApEszqVXtHmYHD8Lt1CnjxA1bTs7IM/aQ9HTw9JRJIEJUEOXmG5yens7q1auZOXMmly9fBmDQoEHMnj2bPXv8CA29MzMupywCVOwkMC0tjSVLljB+/HiqVavGpk2bqF+/vkzwEKK8SzYtvBy1uwGhq9tyI90Ysn+9Uo1fr0zkqb/czfoxtbircmVjUcDkZNtM2Chk7GGuCxfg4EFjGZjAQCkDI0Q55fRjAHMmeAQGBjJ27FguX75MUFAQMTExREVF4efnV2xZhIooMzOTVatW0aRJE6ZPn87XX38NQMuWLSX5E6IiOHfOJOmaurFVbvJ3R1UO/jLMmPyBcf9z56x/7yLGHubK2ZaQYNxPa+vfUwhhd06dAP7xxx8mK3g0bdqUbdu28f3335sULs4pf5BfUdvLq5wJHq1atWLUqFG0aNGCmJgY+ve3WSUdIYQzSE/P/c8fjx3jzKUqhe525rK36YaMDOvfu4ixhwVkZRn3i4+3/j2FEHbn1AngsWPnKDlkAAALPElEQVTHTCZ4xMXF0adPnwL1rIoq8VLWpV8sWY7JWsuWLcud4LFr1y5ZwUOIisjTk6SUFJ5bsoTHpk3D3e1sobv51sp368Pacb9FjD1sOKoXbgP70XBUL6J2N7izf05PYGZmIQcTQjgzp04AlVJMnjyZkydPMnr06CInNUREGEu95FXWpV9KsxyTNZKTk3n55ZdJSkpCKUVUVBRxcXEuu5SdEK7gVEYGTceNY1tMDNP69WNNaFKB0i/enplEBMfd2eDubpygYY0ixh4mXaqK1oqkS1UJXd3WNAks5HVCCOfn1AlgYGAgc+bMKXFcW0iIsdSLPUu/lPW4w7S0NKZOnUpAQAAbN24kOjoagDp16sjsXiEqIIPBwIEDBwBo/NhjvPPcc5xYupSZAwYwpMv5O6VflMbP5w/WDI8lpEO+xKtBg0KOXAr5xh6GFzL28Ea6B+EbW93ZYKuxh0IIu3LqWcCenp5m7xsSYt8Zv2U57nD16tW89dZbpKSk8PzzzxMREYGvqy1lIoQL2bNnD+PGjSMuLo6EhAR8fX2ZOHGiye3YkA7JBRO+HO7uxlm51pZmyTP2EAoZY1jUdluMPRRC2JVT9wA6s7Icd7h//36aN29OTEwM69evl+RPiArq9u3bDBgwgA4dOnDhwgU++eQT6tevb/xhYCDUqWNM7orj7m7cLzDQ+gblu+guMMawqO1yV0KIckcSQAvZctzhgQMH6NKlC/v2GdeRX7ZsWYGZzkKIiufIkSPs3LmTmTNncvz4cQYNGoSbW3ZYVgratzf27Lm7F0wEPTzu9Py1b2+benx165q8T2HLzpXJ2EMhhN059S1gZ5Zzuzk83Hjb15LVR5KTkwkPD2f9+vX4+Phw/vx5ACrn1PUSQlRovr6+7N69+06vX35KGVf3aN78zoocGRnGHreyWJGjQQNjkedsObecc1Yg8a11g4jgONuPPRRC2J3STlzEs23btjo2NtbRzSgTs2fP5p133kFrzfjx45k8ebIUcRbChpRS+7XWTt2N7pQxLi6uQCmYIuX0QLZqVfK+QgibsybOSQ+gHWVmZuLu7o5SCjc3N/r16ycTPIQQziUwEK5eLbkYtC3HHgoh7E7GANpB3hU8/v73vwMQFhYmEzyEEM7HEWMPhRB2Jz2AZWz//v1MmDCB77//nqZNm1KzZk0AKeIshHBe9h57KISwO/kGl6EpU6Ywd+5cfHx8eO+99xg2bJgUcRZClB8eHtCokfEhhKhQ7HYLWCn1jlLqZ6VUtFLqgFJqpL3e257S0tK4desWAG3atCEsLIyTJ08yatQoSf6EqOBcJc4JIco/e44B/B1oq7VuB/QAFiilKsxlZUZGBitXrqRJkyYsXboUgH79+jF37lyZ3SuE66jQcU4IUXHYLQHUWi/SWuesF9QQuA5csdf7l5WcCR4PPPAAo0ePpkWLFnTt2tXRzRJCOEBFjXNCiIrHrnUAlVL+wJdALWCA1vrbQvYJBUKznwYC8XZroPPzAS45uhFORM6HKTkfpppqravZ+01LinMS44oln2FTcj4KknNiyuI455BC0Eqph4CdwONa68PF7Bfr7IVc7UnOhyk5H6bkfJhy9PkwJ845uo3ORs6HKTkfBck5MWXN+XBIHUCt9UFgL9DZEe8vhBBlTeKcEMKZ2SUBVEq1UkoNUNnF75RS9YCHASdbA0kIISwjcU4IUZ7Yqw5gEjACCFNKZQCewFta630lvG5NmbesfJHzYUrOhyk5H6bsfT4siXPy/8yUnA9Tcj4KknNiyuLz4ZAxgEIIIYQQwnFkLWAhhBBCCBcjCaAQQgghhIuRBFAIIYQQwsU4PAFUSnXMXjPzkFIqVin1SCH7qOw1No8rpY4opSKVUlUd0d6yZub5aKOUuqKU2pfn8YYj2lvWlFKVlFITlFIZSqnnitjHlT4f5pwPV/p8DM9eezc2+zszqoj9Riuljiml4pVSO5RS99qxjRLj8pAYZ0pinCmJcQWVWZzTWjvsAdwNXAYezX7eCbgAeOfb7yVgP1Al+/knwApHtt3B56MH8IGj22unczIKeAPYDTxXxD4u8fkoxflwic8H4A4sBO7Kfl4PuAnUy7dfJ+AMUDv7+Qxgh53aKDHOsvPhEp/h7N9VYlzpz4crfT7KLM45ugewB3Bca70XQGv9PfAbkH8x3YHAaq31zeznS4FgezXSjsw9Hz5Ad6XUT9mPCKWU3Ze8sget9Uqt9btAVjG7ucrnw9zz4RKfD611ltZ6gtb6evamy0A6xoCZ10AgUmudkv18KfBXpVQNOzRTYpwpiXH5SIwzJTHOVFnGOUcngH8CTuXbdip7e3H7nQJq2imA25O55+MfQEOt9cNAT6ARsK7sm+e0XOXzYS5X/XwsAf6mtT6Tb7vJ50Nr/TtwFWhohzZJjDMlMc4yrvL5MJcrfz5sFufsVQi6KIqCWX4mBRPT/PtlZv/r6ATW1sw6H3muAtFaX8ke+5CslKqS92cuxFU+H2Zxxc+HUmoWxlsjzxb2Y8yLM2VBYpwpiXGWcZXPh1lc9fNh6zjn6A/PWcA33zbf7O3F7ecLXAdSy65pDmHu+cjPHbgF3C6LRpUDrvL5sFSF/nwopRYCLYFntdbphexi8vlQSnkDtSj5e2ULEuNMSYyzjKt8PixV4T8fZRHnHJ0AbgMeUEq1AlBKtQOaAf9WSv2olPLP3m89MFQp5Zn9fAzwD5090rECMet8KKWeU0rdnf3fHsAcYL3W2uCgdtuVUqqWi34+CpX/fLjK50Mp5aaUeh9oAPTPCYpKKXel1HdKqQ7Zu64HQvLcLhsN/JhnrExZkhhnSmKcGSTGmXLVGAdlG+ccegtYa31VKdUf+FgppTF2V/YCvAE/IOcXWQc0AaKVUpnAEeBVBzS5TJXifFQBvlNKGQAN/ABMc0CTHcUlPx/FcNXPRy9gOBAL7FFK5WyfhXFMUE0ArfUupdR7wA/KuEbvOaDQ8hK2JjHOlMQ4s7nk56MYrvz5KLM4J2sBCyGEEEK4GEffAhZCCCGEEHYmCaAQQgghhIuRBFAIIYQQwsVIAiiEEEII4WIkARRCCCGEcDGSAAohhBBCuBhJAIUQQgghXIwkgEIIIYQQLkYSQCGEEEIIFyMJoBBCCCGEi5EEUAghhBDCxUgCKJyWUqqKUuqsUuqMUsor388+VEplKaWKXexaCCGclcQ44UiSAAqnpbW+CUwHGgCjcrYrpeYArwBjtNafOah5QghhFYlxwpGU1trRbRCiSEopd+BnoA7wJ2AosBiYrrV+25FtE0IIa0mME44iCaBwekqp3sB24DugC7BCaz3Wsa0SQgjbkBgnHEFuAQunp7XeARwAugJ/A17Lv49SarRSKlopdUsp9b2dmyiEEBaTGCccwcPRDRCiJEqpAcCD2U+v6cK7rX8D5gJ/AR61V9uEEMJaEuOEI0gCKJyaUqo7sB7YAmQAQ5RSi7XWR/Pup7X+R/b+vvZvpRBCWEZinHAUuQUsnJZS6mHgH8CPQAjwJmAA5jiyXUIIYQsS44QjSQIonJJSqjmwEzgBPKW1vq21PgV8BPRVSrV3aAOFEMIKEuOEo0kCKJxO9i2OfwFXgZ5a67Q8P34buAnMd0TbhBDCWhLjhDOQMYDC6Witz2AsjFrYz34DvO3bIiGEsB2JccIZSAIoKgSllAfGz7MH4KaUqgwYtNbpjm2ZEEJYT2KcsDVJAEVF8SbGJZVy3AR+ADo5pDVCCGFbEuOETclKIEIIIYQQLkYmgQghhBBCuBhJAIUQQgghXIwkgEIIIYQQLkYSQCGEEEIIFyMJoBBCCCGEi5EEUAghhBDCxUgCKIQQQgjhYv4fZRC8rvuW0KEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed1cd73da0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_svm_regression(svm_reg, X, y, axes):\n",
    "    x1s = np.linspace(axes[0], axes[1], 100).reshape(100, 1)\n",
    "    y_pred = svm_reg.predict(x1s)\n",
    "    plt.plot(x1s, y_pred, \"k-\", linewidth=2, label=r\"$\\hat{y}$\")\n",
    "    plt.plot(x1s, y_pred + svm_reg.epsilon, \"k--\")\n",
    "    plt.plot(x1s, y_pred - svm_reg.epsilon, \"k--\")\n",
    "    plt.scatter(X[svm_reg.support_], y[svm_reg.support_], s=180, facecolors='#FFAAAA')\n",
    "    plt.plot(X, y, \"bo\")\n",
    "    plt.xlabel(r\"$x_1$\", fontsize=18)\n",
    "    plt.legend(loc=\"upper left\", fontsize=18)\n",
    "    plt.axis(axes)\n",
    "\n",
    "plt.figure(figsize=(9, 4))\n",
    "plt.subplot(121)\n",
    "plot_svm_regression(svm_reg1, X, y, [0, 2, 3, 11])\n",
    "plt.title(r\"$\\epsilon = {}$\".format(svm_reg1.epsilon), fontsize=18)\n",
    "plt.ylabel(r\"$y$\", fontsize=18, rotation=0)\n",
    "#plt.plot([eps_x1, eps_x1], [eps_y_pred, eps_y_pred - svm_reg1.epsilon], \"k-\", linewidth=2)\n",
    "plt.annotate(\n",
    "        '', xy=(eps_x1, eps_y_pred), xycoords='data',\n",
    "        xytext=(eps_x1, eps_y_pred - svm_reg1.epsilon),\n",
    "        textcoords='data', arrowprops={'arrowstyle': '<->', 'linewidth': 1.5}\n",
    "    )\n",
    "plt.text(0.91, 5.6, r\"$\\epsilon$\", fontsize=20)\n",
    "plt.subplot(122)\n",
    "plot_svm_regression(svm_reg2, X, y, [0, 2, 3, 11])\n",
    "plt.title(r\"$\\epsilon = {}$\".format(svm_reg2.epsilon), fontsize=18)\n",
    "save_fig(\"svm_regression_plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(42)\n",
    "m = 100\n",
    "X = 2 * np.random.rand(m, 1) - 1\n",
    "y = (0.2 + 0.1 * X + 0.5 * X**2 + np.random.randn(m, 1)/10).ravel()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVR(C=100, cache_size=200, coef0=0.0, degree=2, epsilon=0.1, gamma='auto',\n",
       "  kernel='poly', max_iter=-1, shrinking=True, tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.svm import SVR\n",
    "\n",
    "svm_poly_reg = SVR(kernel=\"poly\", degree=2, C=100, epsilon=0.1)\n",
    "svm_poly_reg.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVR(C=0.01, cache_size=200, coef0=0.0, degree=2, epsilon=0.1, gamma='auto',\n",
       "  kernel='poly', max_iter=-1, shrinking=True, tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.svm import SVR\n",
    "\n",
    "svm_poly_reg1 = SVR(kernel=\"poly\", degree=2, C=100, epsilon=0.1)\n",
    "svm_poly_reg2 = SVR(kernel=\"poly\", degree=2, C=0.01, epsilon=0.1)\n",
    "svm_poly_reg1.fit(X, y)\n",
    "svm_poly_reg2.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saving figure svm_with_polynomial_kernel_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed1cd5cb38>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(9, 4))\n",
    "plt.subplot(121)\n",
    "plot_svm_regression(svm_poly_reg1, X, y, [-1, 1, 0, 1])\n",
    "plt.title(r\"$degree={}, C={}, \\epsilon = {}$\".format(svm_poly_reg1.degree, svm_poly_reg1.C, svm_poly_reg1.epsilon), fontsize=18)\n",
    "plt.ylabel(r\"$y$\", fontsize=18, rotation=0)\n",
    "plt.subplot(122)\n",
    "plot_svm_regression(svm_poly_reg2, X, y, [-1, 1, 0, 1])\n",
    "plt.title(r\"$degree={}, C={}, \\epsilon = {}$\".format(svm_poly_reg2.degree, svm_poly_reg2.C, svm_poly_reg2.epsilon), fontsize=18)\n",
    "save_fig(\"svm_with_polynomial_kernel_plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Under the hood"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "iris = datasets.load_iris()\n",
    "X = iris[\"data\"][:, (2, 3)]  # petal length, petal width\n",
    "y = (iris[\"target\"] == 2).astype(np.float64)  # Iris-Virginica"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saving figure iris_3D_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed1a4d4630>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "\n",
    "def plot_3D_decision_function(ax, w, b, x1_lim=[4, 6], x2_lim=[0.8, 2.8]):\n",
    "    x1_in_bounds = (X[:, 0] > x1_lim[0]) & (X[:, 0] < x1_lim[1])\n",
    "    X_crop = X[x1_in_bounds]\n",
    "    y_crop = y[x1_in_bounds]\n",
    "    x1s = np.linspace(x1_lim[0], x1_lim[1], 20)\n",
    "    x2s = np.linspace(x2_lim[0], x2_lim[1], 20)\n",
    "    x1, x2 = np.meshgrid(x1s, x2s)\n",
    "    xs = np.c_[x1.ravel(), x2.ravel()]\n",
    "    df = (xs.dot(w) + b).reshape(x1.shape)\n",
    "    m = 1 / np.linalg.norm(w)\n",
    "    boundary_x2s = -x1s*(w[0]/w[1])-b/w[1]\n",
    "    margin_x2s_1 = -x1s*(w[0]/w[1])-(b-1)/w[1]\n",
    "    margin_x2s_2 = -x1s*(w[0]/w[1])-(b+1)/w[1]\n",
    "    ax.plot_surface(x1s, x2, np.zeros_like(x1),\n",
    "                    color=\"b\", alpha=0.2, cstride=100, rstride=100)\n",
    "    ax.plot(x1s, boundary_x2s, 0, \"k-\", linewidth=2, label=r\"$h=0$\")\n",
    "    ax.plot(x1s, margin_x2s_1, 0, \"k--\", linewidth=2, label=r\"$h=\\pm 1$\")\n",
    "    ax.plot(x1s, margin_x2s_2, 0, \"k--\", linewidth=2)\n",
    "    ax.plot(X_crop[:, 0][y_crop==1], X_crop[:, 1][y_crop==1], 0, \"g^\")\n",
    "    ax.plot_wireframe(x1, x2, df, alpha=0.3, color=\"k\")\n",
    "    ax.plot(X_crop[:, 0][y_crop==0], X_crop[:, 1][y_crop==0], 0, \"bs\")\n",
    "    ax.axis(x1_lim + x2_lim)\n",
    "    ax.text(4.5, 2.5, 3.8, \"결정 함수 $h$\", fontsize=15)\n",
    "    ax.set_xlabel(r\"꽃잎 길이\", fontsize=15, labelpad=15)\n",
    "    ax.set_ylabel(r\"꽃잎 폭\", fontsize=15, rotation=25, labelpad=15)\n",
    "    ax.set_zlabel(r\"$h = \\mathbf{w}^t \\cdot \\mathbf{x} + b$\", fontsize=18, labelpad=10)\n",
    "    ax.legend(loc=\"upper left\", fontsize=16)\n",
    "\n",
    "fig = plt.figure(figsize=(11, 6))\n",
    "ax1 = fig.add_subplot(111, projection='3d')\n",
    "plot_3D_decision_function(ax1, w=svm_clf2.coef_[0], b=svm_clf2.intercept_[0])\n",
    "\n",
    "save_fig(\"iris_3D_plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Small weight vector results in a large margin"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saving figure small_w_large_margin_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed1c3de048>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_2D_decision_function(w, b, ylabel=True, x1_lim=[-3, 3]):\n",
    "    x1 = np.linspace(x1_lim[0], x1_lim[1], 200)\n",
    "    y = w * x1 + b\n",
    "    m = 1 / w\n",
    "\n",
    "    plt.plot(x1, y)\n",
    "    plt.plot(x1_lim, [1, 1], \"k:\")\n",
    "    plt.plot(x1_lim, [-1, -1], \"k:\")\n",
    "    plt.axhline(y=0, color='k')\n",
    "    plt.axvline(x=0, color='k')\n",
    "    plt.plot([m, m], [0, 1], \"k--\")\n",
    "    plt.plot([-m, -m], [0, -1], \"k--\")\n",
    "    plt.plot([-m, m], [0, 0], \"k-o\", linewidth=3)\n",
    "    plt.axis(x1_lim + [-2, 2])\n",
    "    plt.xlabel(r\"$x_1$\", fontsize=16)\n",
    "    if ylabel:\n",
    "        plt.ylabel(r\"$w_1 x_1$  \", rotation=0, fontsize=16)\n",
    "    plt.title(r\"$w_1 = {}$\".format(w), fontsize=16)\n",
    "\n",
    "plt.figure(figsize=(12, 3.2))\n",
    "plt.subplot(121)\n",
    "plot_2D_decision_function(1, 0)\n",
    "plt.subplot(122)\n",
    "plot_2D_decision_function(0.5, 0, ylabel=False)\n",
    "save_fig(\"small_w_large_margin_plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1.])"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.svm import SVC\n",
    "from sklearn import datasets\n",
    "\n",
    "iris = datasets.load_iris()\n",
    "X = iris[\"data\"][:, (2, 3)] # petal length, petal width\n",
    "y = (iris[\"target\"] == 2).astype(np.float64) # Iris-Virginica\n",
    "\n",
    "svm_clf = SVC(kernel=\"linear\", C=1)\n",
    "svm_clf.fit(X, y)\n",
    "svm_clf.predict([[5.3, 1.3]])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Hinge loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saving figure hinge_plot\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed1cf260b8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "t = np.linspace(-2, 4, 200)\n",
    "h = np.where(1 - t < 0, 0, 1 - t)  # max(0, 1-t)\n",
    "\n",
    "plt.figure(figsize=(5,2.8))\n",
    "plt.plot(t, h, \"b-\", linewidth=2, label=\"$max(0, 1 - t)$\")\n",
    "plt.grid(True, which='both')\n",
    "plt.axhline(y=0, color='k')\n",
    "plt.axvline(x=0, color='k')\n",
    "plt.yticks(np.arange(-1, 2.5, 1))\n",
    "plt.xlabel(\"$t$\", fontsize=16)\n",
    "plt.axis([-2, 4, -1, 2.5])\n",
    "plt.legend(loc=\"upper right\", fontsize=16)\n",
    "save_fig(\"hinge_plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Extra material"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fed1ce63940>]"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed1ce63eb8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X, y = make_moons(n_samples=1000, noise=0.4, random_state=42)\n",
    "plt.plot(X[:, 0][y==0], X[:, 1][y==0], \"bs\")\n",
    "plt.plot(X[:, 0][y==1], X[:, 1][y==1], \"g^\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LibSVM]0 0.1 0.1996922492980957\n",
      "[LibSVM]1 0.01 0.20404458045959473\n",
      "[LibSVM]2 0.001 0.23720145225524902\n",
      "[LibSVM]3 0.0001 0.43198180198669434\n",
      "[LibSVM]4 1e-05 0.6850612163543701\n",
      "[LibSVM]5 1.0000000000000002e-06 0.6412525177001953\n",
      "[LibSVM]6 1.0000000000000002e-07 4.8058247566223145\n",
      "[LibSVM]7 1.0000000000000002e-08 0.6779589653015137\n",
      "[LibSVM]8 1.0000000000000003e-09 0.6802883148193359\n",
      "[LibSVM]9 1.0000000000000003e-10 0.6725904941558838\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed1c462438>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import time\n",
    "\n",
    "tol = 0.1\n",
    "tols = []\n",
    "times = []\n",
    "for i in range(10):\n",
    "    svm_clf = SVC(kernel=\"poly\", gamma=3, C=10, tol=tol, verbose=1)\n",
    "    t1 = time.time()\n",
    "    svm_clf.fit(X, y)\n",
    "    t2 = time.time()\n",
    "    times.append(t2-t1)\n",
    "    tols.append(tol)\n",
    "    print(i, tol, t2-t1)\n",
    "    tol /= 10\n",
    "plt.rcParams['font.family'] = 'stixgeneral'\n",
    "plt.semilogx(tols, times)\n",
    "plt.rcParams['font.family'] = 'NanumBarunGothic'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Linear SVM classifier implementation using Batch Gradient Descent"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Training set\n",
    "X = iris[\"data\"][:, (2, 3)] # petal length, petal width\n",
    "y = (iris[\"target\"] == 2).astype(np.float64).reshape(-1, 1) # Iris-Virginica"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.],\n",
       "       [0.]])"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.base import BaseEstimator\n",
    "\n",
    "class MyLinearSVC(BaseEstimator):\n",
    "    def __init__(self, C=1, eta0=1, eta_d=10000, n_epochs=1000, random_state=None):\n",
    "        self.C = C\n",
    "        self.eta0 = eta0\n",
    "        self.n_epochs = n_epochs\n",
    "        self.random_state = random_state\n",
    "        self.eta_d = eta_d\n",
    "\n",
    "    def eta(self, epoch):\n",
    "        return self.eta0 / (epoch + self.eta_d)\n",
    "        \n",
    "    def fit(self, X, y):\n",
    "        # Random initialization\n",
    "        if self.random_state:\n",
    "            np.random.seed(self.random_state)\n",
    "        w = np.random.randn(X.shape[1], 1) # n feature weights\n",
    "        b = 0\n",
    "\n",
    "        m = len(X)\n",
    "        t = y * 2 - 1  # -1 if t==0, +1 if t==1\n",
    "        X_t = X * t\n",
    "        self.Js=[]\n",
    "\n",
    "        # Training\n",
    "        for epoch in range(self.n_epochs):\n",
    "            support_vectors_idx = (X_t.dot(w) + t * b < 1).ravel()\n",
    "            X_t_sv = X_t[support_vectors_idx]\n",
    "            t_sv = t[support_vectors_idx]\n",
    "\n",
    "            J = 1/2 * np.sum(w * w) + self.C * (np.sum(1 - X_t_sv.dot(w)) - b * np.sum(t_sv))\n",
    "            self.Js.append(J)\n",
    "\n",
    "            w_gradient_vector = w - self.C * np.sum(X_t_sv, axis=0).reshape(-1, 1)\n",
    "            b_derivative = -C * np.sum(t_sv)\n",
    "                \n",
    "            w = w - self.eta(epoch) * w_gradient_vector\n",
    "            b = b - self.eta(epoch) * b_derivative\n",
    "            \n",
    "\n",
    "        self.intercept_ = np.array([b])\n",
    "        self.coef_ = np.array([w])\n",
    "        support_vectors_idx = (X_t.dot(w) + t * b < 1).ravel()\n",
    "        self.support_vectors_ = X[support_vectors_idx]\n",
    "        return self\n",
    "\n",
    "    def decision_function(self, X):\n",
    "        return X.dot(self.coef_[0]) + self.intercept_[0]\n",
    "\n",
    "    def predict(self, X):\n",
    "        return (self.decision_function(X) >= 0).astype(np.float64)\n",
    "\n",
    "C=2\n",
    "svm_clf = MyLinearSVC(C=C, eta0 = 10, eta_d = 1000, n_epochs=60000, random_state=2)\n",
    "svm_clf.fit(X, y)\n",
    "svm_clf.predict(np.array([[5, 2], [4, 1]]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0, 60000, 0, 100]"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed1cd87198>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(range(svm_clf.n_epochs), svm_clf.Js)\n",
    "plt.axis([0, svm_clf.n_epochs, 0, 100])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[-15.56780998] [[[2.28129013]\n",
      "  [2.71597487]]]\n"
     ]
    }
   ],
   "source": [
    "print(svm_clf.intercept_, svm_clf.coef_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[-15.51721253] [[2.27128546 2.71287145]]\n"
     ]
    }
   ],
   "source": [
    "svm_clf2 = SVC(kernel=\"linear\", C=C)\n",
    "svm_clf2.fit(X, y.ravel())\n",
    "print(svm_clf2.intercept_, svm_clf2.coef_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[4, 6, 0.8, 2.8]"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed1c406ef0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "yr = y.ravel()\n",
    "plt.figure(figsize=(12,3.2))\n",
    "plt.subplot(121)\n",
    "plt.plot(X[:, 0][yr==1], X[:, 1][yr==1], \"g^\", label=\"Iris-Virginica\")\n",
    "plt.plot(X[:, 0][yr==0], X[:, 1][yr==0], \"bs\", label=\"Not Iris-Virginica\")\n",
    "plot_svc_decision_boundary(svm_clf, 4, 6)\n",
    "plt.xlabel(\"꽃잎 길이\", fontsize=14)\n",
    "plt.ylabel(\"꽃잎 폭\", fontsize=14)\n",
    "plt.title(\"MyLinearSVC\", fontsize=14)\n",
    "plt.axis([4, 6, 0.8, 2.8])\n",
    "\n",
    "plt.subplot(122)\n",
    "plt.plot(X[:, 0][yr==1], X[:, 1][yr==1], \"g^\")\n",
    "plt.plot(X[:, 0][yr==0], X[:, 1][yr==0], \"bs\")\n",
    "plot_svc_decision_boundary(svm_clf2, 4, 6)\n",
    "plt.xlabel(\"Petal length\", fontsize=14)\n",
    "plt.title(\"SVC\", fontsize=14)\n",
    "plt.axis([4, 6, 0.8, 2.8])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[-14.062485     2.24179316   1.79750198]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[4, 6, 0.8, 2.8]"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed1cd13f98>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.linear_model import SGDClassifier\n",
    "\n",
    "sgd_clf = SGDClassifier(loss=\"hinge\", alpha = 0.017, max_iter = 50, random_state=42)\n",
    "sgd_clf.fit(X, y.ravel())\n",
    "\n",
    "m = len(X)\n",
    "t = y * 2 - 1  # -1 if t==0, +1 if t==1\n",
    "X_b = np.c_[np.ones((m, 1)), X]  # Add bias input x0=1\n",
    "X_b_t = X_b * t\n",
    "sgd_theta = np.r_[sgd_clf.intercept_[0], sgd_clf.coef_[0]]\n",
    "print(sgd_theta)\n",
    "support_vectors_idx = (X_b_t.dot(sgd_theta) < 1).ravel()\n",
    "sgd_clf.support_vectors_ = X[support_vectors_idx]\n",
    "sgd_clf.C = C\n",
    "\n",
    "plt.figure(figsize=(5.5,3.2))\n",
    "plt.plot(X[:, 0][yr==1], X[:, 1][yr==1], \"g^\")\n",
    "plt.plot(X[:, 0][yr==0], X[:, 1][yr==0], \"bs\")\n",
    "plot_svc_decision_boundary(sgd_clf, 4, 6)\n",
    "plt.xlabel(\"꽃잎 길이\", fontsize=14)\n",
    "plt.ylabel(\"꽃잎 폭\", fontsize=14)\n",
    "plt.title(\"SGDClassifier\", fontsize=14)\n",
    "plt.axis([4, 6, 0.8, 2.8])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Exercise solutions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. to 7."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "See appendix A."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 8."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "_Exercise: train a `LinearSVC` on a linearly separable dataset. Then train an `SVC` and a `SGDClassifier` on the same dataset. See if you can get them to produce roughly the same model._"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's use the Iris dataset: the Iris Setosa and Iris Versicolor classes are linearly separable."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import datasets\n",
    "\n",
    "iris = datasets.load_iris()\n",
    "X = iris[\"data\"][:, (2, 3)]  # petal length, petal width\n",
    "y = iris[\"target\"]\n",
    "\n",
    "setosa_or_versicolor = (y == 0) | (y == 1)\n",
    "X = X[setosa_or_versicolor]\n",
    "y = y[setosa_or_versicolor]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LinearSVC:                    [0.28480668] [[1.05542422 1.09851637]]\n",
      "SVC:                          [0.31933577] [[1.1223101  1.02531081]]\n",
      "SGDClassifier(alpha=0.00200): [0.32] [[1.12293103 1.02620763]]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.svm import SVC, LinearSVC\n",
    "from sklearn.linear_model import SGDClassifier\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "C = 5\n",
    "alpha = 1 / (C * len(X))\n",
    "\n",
    "lin_clf = LinearSVC(loss=\"hinge\", C=C, random_state=42)\n",
    "svm_clf = SVC(kernel=\"linear\", C=C)\n",
    "sgd_clf = SGDClassifier(loss=\"hinge\", learning_rate=\"constant\", eta0=0.001, alpha=alpha,\n",
    "                        max_iter=100000, random_state=42)\n",
    "\n",
    "scaler = StandardScaler()\n",
    "X_scaled = scaler.fit_transform(X)\n",
    "\n",
    "lin_clf.fit(X_scaled, y)\n",
    "svm_clf.fit(X_scaled, y)\n",
    "sgd_clf.fit(X_scaled, y)\n",
    "\n",
    "print(\"LinearSVC:                   \", lin_clf.intercept_, lin_clf.coef_)\n",
    "print(\"SVC:                         \", svm_clf.intercept_, svm_clf.coef_)\n",
    "print(\"SGDClassifier(alpha={:.5f}):\".format(sgd_clf.alpha), sgd_clf.intercept_, sgd_clf.coef_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's plot the decision boundaries of these three models:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fed1ce5af60>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Compute the slope and bias of each decision boundary\n",
    "w1 = -lin_clf.coef_[0, 0]/lin_clf.coef_[0, 1]\n",
    "b1 = -lin_clf.intercept_[0]/lin_clf.coef_[0, 1]\n",
    "w2 = -svm_clf.coef_[0, 0]/svm_clf.coef_[0, 1]\n",
    "b2 = -svm_clf.intercept_[0]/svm_clf.coef_[0, 1]\n",
    "w3 = -sgd_clf.coef_[0, 0]/sgd_clf.coef_[0, 1]\n",
    "b3 = -sgd_clf.intercept_[0]/sgd_clf.coef_[0, 1]\n",
    "\n",
    "# Transform the decision boundary lines back to the original scale\n",
    "line1 = scaler.inverse_transform([[-10, -10 * w1 + b1], [10, 10 * w1 + b1]])\n",
    "line2 = scaler.inverse_transform([[-10, -10 * w2 + b2], [10, 10 * w2 + b2]])\n",
    "line3 = scaler.inverse_transform([[-10, -10 * w3 + b3], [10, 10 * w3 + b3]])\n",
    "\n",
    "# Plot all three decision boundaries\n",
    "plt.figure(figsize=(11, 4))\n",
    "plt.plot(line1[:, 0], line1[:, 1], \"k:\", label=\"LinearSVC\")\n",
    "plt.plot(line2[:, 0], line2[:, 1], \"b--\", linewidth=2, label=\"SVC\")\n",
    "plt.plot(line3[:, 0], line3[:, 1], \"r-\", label=\"SGDClassifier\")\n",
    "plt.plot(X[:, 0][y==1], X[:, 1][y==1], \"bs\") # label=\"Iris-Versicolor\"\n",
    "plt.plot(X[:, 0][y==0], X[:, 1][y==0], \"yo\") # label=\"Iris-Setosa\"\n",
    "plt.xlabel(\"꽃잎 길이\", fontsize=14)\n",
    "plt.ylabel(\"꽃잎 폭\", fontsize=14)\n",
    "plt.legend(loc=\"upper center\", fontsize=14)\n",
    "plt.axis([0, 5.5, 0, 2])\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Close enough!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 9."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "_Exercise: train an SVM classifier on the MNIST dataset. Since SVM classifiers are binary classifiers, you will need to use one-versus-all to classify all 10 digits. You may want to tune the hyperparameters using small validation sets to speed up the process. What accuracy can you reach?_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First, let's load the dataset and split it into a training set and a test set. We could use `train_test_split()` but people usually just take the first 60,000 instances for the training set, and the last 10,000 instances for the test set (this makes it possible to compare your model's performance with others): "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import fetch_mldata\n",
    "\n",
    "mnist = fetch_mldata(\"MNIST original\")\n",
    "X = mnist[\"data\"]\n",
    "y = mnist[\"target\"]\n",
    "\n",
    "X_train = X[:60000]\n",
    "y_train = y[:60000]\n",
    "X_test = X[60000:]\n",
    "y_test = y[60000:]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Many training algorithms are sensitive to the order of the training instances, so it's generally good practice to shuffle them first:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(42)\n",
    "rnd_idx = np.random.permutation(60000)\n",
    "X_train = X_train[rnd_idx]\n",
    "y_train = y_train[rnd_idx]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's start simple, with a linear SVM classifier. It will automatically use the One-vs-All (also called One-vs-the-Rest, OvR) strategy, so there's nothing special we need to do. Easy!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LinearSVC(C=1.0, class_weight=None, dual=True, fit_intercept=True,\n",
       "     intercept_scaling=1, loss='squared_hinge', max_iter=1000,\n",
       "     multi_class='ovr', penalty='l2', random_state=42, tol=0.0001,\n",
       "     verbose=0)"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lin_clf = LinearSVC(random_state=42)\n",
    "lin_clf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's make predictions on the training set and measure the accuracy (we don't want to measure it on the test set yet, since we have not selected and trained the final model yet):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8263333333333334"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score\n",
    "\n",
    "y_pred = lin_clf.predict(X_train)\n",
    "accuracy_score(y_train, y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Wow, 82% accuracy on MNIST is a really bad performance. This linear model is certainly too simple for MNIST, but perhaps we just needed to scale the data first:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = StandardScaler()\n",
    "X_train_scaled = scaler.fit_transform(X_train.astype(np.float32))\n",
    "X_test_scaled = scaler.transform(X_test.astype(np.float32))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LinearSVC(C=1.0, class_weight=None, dual=True, fit_intercept=True,\n",
       "     intercept_scaling=1, loss='squared_hinge', max_iter=1000,\n",
       "     multi_class='ovr', penalty='l2', random_state=42, tol=0.0001,\n",
       "     verbose=0)"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lin_clf = LinearSVC(random_state=42)\n",
    "lin_clf.fit(X_train_scaled, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9224"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred = lin_clf.predict(X_train_scaled)\n",
    "accuracy_score(y_train, y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "That's much better (we cut the error rate in two), but still not great at all for MNIST. If we want to use an SVM, we will have to use a kernel. Let's try an `SVC` with an RBF kernel (the default).\n",
    "\n",
    "**Warning**: if you are using Scikit-Learn ≤ 0.19, the `SVC` class will use the One-vs-One (OvO) strategy by default, so you must explicitly set `decision_function_shape=\"ovr\"` if you want to use the OvR strategy instead (OvR is the default since 0.19)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,\n",
       "  decision_function_shape='ovr', degree=3, gamma='auto', kernel='rbf',\n",
       "  max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
       "  tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "svm_clf = SVC(decision_function_shape=\"ovr\")\n",
    "svm_clf.fit(X_train_scaled[:10000], y_train[:10000])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.94615"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred = svm_clf.predict(X_train_scaled)\n",
    "accuracy_score(y_train, y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "That's promising, we get better performance even though we trained the model on 6 times less data. Let's tune the hyperparameters by doing a randomized search with cross validation. We will do this on a small dataset just to speed up the process:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 10 candidates, totalling 30 fits\n",
      "[CV] gamma=0.001766074650481071, C=8.852316058423087 .................\n",
      "[CV] gamma=0.001766074650481071, C=8.852316058423087 .................\n",
      "[CV] gamma=0.006364737055453384, C=1.8271960104746645 ................\n",
      "[CV] gamma=0.001766074650481071, C=8.852316058423087 .................\n",
      "[CV] .. gamma=0.001766074650481071, C=8.852316058423087, total=   0.8s\n",
      "[CV] .. gamma=0.001766074650481071, C=8.852316058423087, total=   0.8s\n",
      "[CV] gamma=0.006364737055453384, C=1.8271960104746645 ................\n",
      "[CV] gamma=0.006364737055453384, C=1.8271960104746645 ................\n",
      "[CV] .. gamma=0.001766074650481071, C=8.852316058423087, total=   0.8s\n",
      "[CV] gamma=0.051349833451870636, C=9.875199193765326 .................\n",
      "[CV] . gamma=0.006364737055453384, C=1.8271960104746645, total=   1.0s\n",
      "[CV] gamma=0.051349833451870636, C=9.875199193765326 .................\n",
      "[CV] . gamma=0.006364737055453384, C=1.8271960104746645, total=   1.0s\n",
      "[CV] . gamma=0.006364737055453384, C=1.8271960104746645, total=   1.0s\n",
      "[CV] gamma=0.051349833451870636, C=9.875199193765326 .................\n",
      "[CV] gamma=0.05991666578466177, C=6.59992909281409 ...................\n",
      "[CV] .. gamma=0.051349833451870636, C=9.875199193765326, total=   1.0s\n",
      "[CV] gamma=0.05991666578466177, C=6.59992909281409 ...................\n",
      "[CV] .. gamma=0.051349833451870636, C=9.875199193765326, total=   1.0s\n",
      "[CV] gamma=0.05991666578466177, C=6.59992909281409 ...................\n",
      "[CV] .. gamma=0.051349833451870636, C=9.875199193765326, total=   1.0s\n",
      "[CV] gamma=0.003596490522533181, C=9.053435975487119 .................\n",
      "[CV] .... gamma=0.05991666578466177, C=6.59992909281409, total=   1.0s\n",
      "[CV] gamma=0.003596490522533181, C=9.053435975487119 .................\n",
      "[CV] .... gamma=0.05991666578466177, C=6.59992909281409, total=   1.0s\n",
      "[CV] gamma=0.003596490522533181, C=9.053435975487119 .................\n",
      "[CV] .... gamma=0.05991666578466177, C=6.59992909281409, total=   1.0s\n",
      "[CV] gamma=0.004002330992905356, C=2.701062804458301 .................\n",
      "[CV] .. gamma=0.003596490522533181, C=9.053435975487119, total=   0.9s\n",
      "[CV] gamma=0.004002330992905356, C=2.701062804458301 .................\n",
      "[CV] .. gamma=0.003596490522533181, C=9.053435975487119, total=   0.9s\n",
      "[CV] gamma=0.004002330992905356, C=2.701062804458301 .................\n",
      "[CV] .. gamma=0.003596490522533181, C=9.053435975487119, total=   0.9s\n",
      "[CV] gamma=0.017596957507461645, C=3.2711787843881437 ................\n",
      "[CV] .. gamma=0.004002330992905356, C=2.701062804458301, total=   0.9s\n",
      "[CV] gamma=0.017596957507461645, C=3.2711787843881437 ................\n",
      "[CV] .. gamma=0.004002330992905356, C=2.701062804458301, total=   0.9s\n",
      "[CV] gamma=0.017596957507461645, C=3.2711787843881437 ................\n",
      "[CV] .. gamma=0.004002330992905356, C=2.701062804458301, total=   0.9s\n",
      "[CV] gamma=0.01573529056426603, C=6.848991127746501 ..................\n",
      "[CV] . gamma=0.017596957507461645, C=3.2711787843881437, total=   1.0s\n",
      "[CV] gamma=0.01573529056426603, C=6.848991127746501 ..................\n",
      "[CV] . gamma=0.017596957507461645, C=3.2711787843881437, total=   1.0s\n",
      "[CV] gamma=0.01573529056426603, C=6.848991127746501 ..................\n",
      "[CV] . gamma=0.017596957507461645, C=3.2711787843881437, total=   1.0s\n",
      "[CV] gamma=0.03834647526105027, C=2.893035364914488 ..................\n",
      "[CV] ... gamma=0.01573529056426603, C=6.848991127746501, total=   1.0s\n",
      "[CV] gamma=0.03834647526105027, C=2.893035364914488 ..................\n",
      "[CV] ... gamma=0.01573529056426603, C=6.848991127746501, total=   1.0s\n",
      "[CV] gamma=0.03834647526105027, C=2.893035364914488 ..................\n",
      "[CV] ... gamma=0.01573529056426603, C=6.848991127746501, total=   1.0s\n",
      "[CV] gamma=0.008808538172595842, C=5.336260835426313 .................\n",
      "[CV] ... gamma=0.03834647526105027, C=2.893035364914488, total=   1.0s\n",
      "[CV] gamma=0.008808538172595842, C=5.336260835426313 .................\n",
      "[CV] ... gamma=0.03834647526105027, C=2.893035364914488, total=   1.0s\n",
      "[CV] gamma=0.008808538172595842, C=5.336260835426313 .................\n",
      "[CV] ... gamma=0.03834647526105027, C=2.893035364914488, total=   1.0s\n",
      "[CV] .. gamma=0.008808538172595842, C=5.336260835426313, total=   1.0s\n",
      "[CV] .. gamma=0.008808538172595842, C=5.336260835426313, total=   1.0s\n",
      "[CV] .. gamma=0.008808538172595842, C=5.336260835426313, total=   0.9s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Done  30 out of  30 | elapsed:   10.9s finished\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "RandomizedSearchCV(cv=None, error_score='raise',\n",
       "          estimator=SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,\n",
       "  decision_function_shape='ovr', degree=3, gamma='auto', kernel='rbf',\n",
       "  max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
       "  tol=0.001, verbose=False),\n",
       "          fit_params=None, iid=True, n_iter=10, n_jobs=-1,\n",
       "          param_distributions={'gamma': <scipy.stats._distn_infrastructure.rv_frozen object at 0x7fed1ce5c198>, 'C': <scipy.stats._distn_infrastructure.rv_frozen object at 0x7fed1c334d68>},\n",
       "          pre_dispatch='2*n_jobs', random_state=None, refit=True,\n",
       "          return_train_score='warn', scoring=None, verbose=2)"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "from scipy.stats import reciprocal, uniform\n",
    "\n",
    "param_distributions = {\"gamma\": reciprocal(0.001, 0.1), \"C\": uniform(1, 10)}\n",
    "rnd_search_cv = RandomizedSearchCV(svm_clf, param_distributions, n_iter=10, verbose=2, n_jobs=-1)\n",
    "rnd_search_cv.fit(X_train_scaled[:1000], y_train[:1000])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=8.852316058423087, cache_size=200, class_weight=None, coef0=0.0,\n",
       "  decision_function_shape='ovr', degree=3, gamma=0.001766074650481071,\n",
       "  kernel='rbf', max_iter=-1, probability=False, random_state=None,\n",
       "  shrinking=True, tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rnd_search_cv.best_estimator_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.856"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rnd_search_cv.best_score_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This looks pretty low but remember we only trained the model on 1,000 instances. Let's retrain the best estimator on the whole training set (run this at night, it will take hours):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=8.852316058423087, cache_size=200, class_weight=None, coef0=0.0,\n",
       "  decision_function_shape='ovr', degree=3, gamma=0.001766074650481071,\n",
       "  kernel='rbf', max_iter=-1, probability=False, random_state=None,\n",
       "  shrinking=True, tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rnd_search_cv.best_estimator_.fit(X_train_scaled, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.99965"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred = rnd_search_cv.best_estimator_.predict(X_train_scaled)\n",
    "accuracy_score(y_train, y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ah, this looks good! Let's select this model. Now we can test it on the test set:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9709"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred = rnd_search_cv.best_estimator_.predict(X_test_scaled)\n",
    "accuracy_score(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Not too bad, but apparently the model is overfitting slightly. It's tempting to tweak the hyperparameters a bit more (e.g. decreasing `C` and/or `gamma`), but we would run the risk of overfitting the test set. Other people have found that the hyperparameters `C=5` and `gamma=0.005` yield even better performance (over 98% accuracy). By running the randomized search for longer and on a larger part of the training set, you may be able to find this as well."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 10."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "_Exercise: train an SVM regressor on the California housing dataset._"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's load the dataset using Scikit-Learn's `fetch_california_housing()` function:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Downloading Cal. housing from https://ndownloader.figshare.com/files/5976036 to /home/haesun/scikit_learn_data\n"
     ]
    }
   ],
   "source": [
    "from sklearn.datasets import fetch_california_housing\n",
    "\n",
    "housing = fetch_california_housing()\n",
    "X = housing[\"data\"]\n",
    "y = housing[\"target\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Split it into a training set and a test set:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Don't forget to scale the data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "scaler = StandardScaler()\n",
    "X_train_scaled = scaler.fit_transform(X_train)\n",
    "X_test_scaled = scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's train a simple `LinearSVR` first:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LinearSVR(C=1.0, dual=True, epsilon=0.0, fit_intercept=True,\n",
       "     intercept_scaling=1.0, loss='epsilon_insensitive', max_iter=1000,\n",
       "     random_state=42, tol=0.0001, verbose=0)"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.svm import LinearSVR\n",
    "\n",
    "lin_svr = LinearSVR(random_state=42)\n",
    "lin_svr.fit(X_train_scaled, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's see how it performs on the training set:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9499688222172291"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import mean_squared_error\n",
    "\n",
    "y_pred = lin_svr.predict(X_train_scaled)\n",
    "mse = mean_squared_error(y_train, y_pred)\n",
    "mse"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's look at the RMSE:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9746634404845752"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.sqrt(mse)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this training set, the targets are tens of thousands of dollars. The RMSE gives a rough idea of the kind of error you should expect (with a higher weight for large errors): so with this model we can expect errors somewhere around $10,000. Not great. Let's see if we can do better with an RBF Kernel. We will use randomized search with cross validation to find the appropriate hyperparameter values for `C` and `gamma`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 10 candidates, totalling 30 fits\n",
      "[CV] gamma=0.07969454818643928, C=4.745401188473625 ..................\n",
      "[CV] gamma=0.07969454818643928, C=4.745401188473625 ..................\n",
      "[CV] gamma=0.07969454818643928, C=4.745401188473625 ..................\n",
      "[CV] gamma=0.015751320499779724, C=8.31993941811405 ..................\n",
      "[CV] ... gamma=0.015751320499779724, C=8.31993941811405, total=   5.9s\n",
      "[CV] gamma=0.015751320499779724, C=8.31993941811405 ..................\n",
      "[CV] ... gamma=0.07969454818643928, C=4.745401188473625, total=   6.1s\n",
      "[CV] gamma=0.015751320499779724, C=8.31993941811405 ..................\n",
      "[CV] ... gamma=0.07969454818643928, C=4.745401188473625, total=   6.2s\n",
      "[CV] gamma=0.002051110418843397, C=2.560186404424365 .................\n",
      "[CV] ... gamma=0.07969454818643928, C=4.745401188473625, total=   6.4s\n",
      "[CV] gamma=0.002051110418843397, C=2.560186404424365 .................\n",
      "[CV] ... gamma=0.015751320499779724, C=8.31993941811405, total=   5.7s\n",
      "[CV] gamma=0.002051110418843397, C=2.560186404424365 .................\n",
      "[CV] ... gamma=0.015751320499779724, C=8.31993941811405, total=   5.8s\n",
      "[CV] gamma=0.05399484409787431, C=1.5808361216819946 .................\n",
      "[CV] .. gamma=0.002051110418843397, C=2.560186404424365, total=   5.4s\n",
      "[CV] gamma=0.05399484409787431, C=1.5808361216819946 .................\n",
      "[CV] .. gamma=0.002051110418843397, C=2.560186404424365, total=   5.6s\n",
      "[CV] gamma=0.05399484409787431, C=1.5808361216819946 .................\n",
      "[CV] .. gamma=0.002051110418843397, C=2.560186404424365, total=   5.5s\n",
      "[CV] .. gamma=0.05399484409787431, C=1.5808361216819946, total=   5.4s\n",
      "[CV] gamma=0.026070247583707663, C=7.011150117432088 .................\n",
      "[CV] gamma=0.026070247583707663, C=7.011150117432088 .................\n",
      "[CV] .. gamma=0.05399484409787431, C=1.5808361216819946, total=   5.4s\n",
      "[CV] gamma=0.026070247583707663, C=7.011150117432088 .................\n",
      "[CV] .. gamma=0.05399484409787431, C=1.5808361216819946, total=   5.5s\n",
      "[CV] gamma=0.0870602087830485, C=1.2058449429580245 ..................\n",
      "[CV] ... gamma=0.0870602087830485, C=1.2058449429580245, total=   5.3s\n",
      "[CV] gamma=0.0870602087830485, C=1.2058449429580245 ..................\n",
      "[CV] .. gamma=0.026070247583707663, C=7.011150117432088, total=   5.8s\n",
      "[CV] gamma=0.0870602087830485, C=1.2058449429580245 ..................\n",
      "[CV] .. gamma=0.026070247583707663, C=7.011150117432088, total=   5.8s\n",
      "[CV] gamma=0.0026587543983272693, C=9.324426408004218 ................\n",
      "[CV] .. gamma=0.026070247583707663, C=7.011150117432088, total=   6.1s\n",
      "[CV] gamma=0.0026587543983272693, C=9.324426408004218 ................\n",
      "[CV] ... gamma=0.0870602087830485, C=1.2058449429580245, total=   5.3s\n",
      "[CV] gamma=0.0026587543983272693, C=9.324426408004218 ................\n",
      "[CV] ... gamma=0.0870602087830485, C=1.2058449429580245, total=   5.6s\n",
      "[CV] gamma=0.0023270677083837795, C=2.818249672071006 ................\n",
      "[CV] . gamma=0.0026587543983272693, C=9.324426408004218, total=   5.6s\n",
      "[CV] gamma=0.0023270677083837795, C=2.818249672071006 ................\n",
      "[CV] . gamma=0.0026587543983272693, C=9.324426408004218, total=   5.5s\n",
      "[CV] gamma=0.0023270677083837795, C=2.818249672071006 ................\n",
      "[CV] . gamma=0.0026587543983272693, C=9.324426408004218, total=   5.5s\n",
      "[CV] gamma=0.011207606211860567, C=4.042422429595377 .................\n",
      "[CV] . gamma=0.0023270677083837795, C=2.818249672071006, total=   5.6s\n",
      "[CV] . gamma=0.0023270677083837795, C=2.818249672071006, total=   5.4s\n",
      "[CV] gamma=0.011207606211860567, C=4.042422429595377 .................\n",
      "[CV] gamma=0.011207606211860567, C=4.042422429595377 .................\n",
      "[CV] . gamma=0.0023270677083837795, C=2.818249672071006, total=   5.4s\n",
      "[CV] gamma=0.003823475224675185, C=5.319450186421157 .................\n",
      "[CV] .. gamma=0.011207606211860567, C=4.042422429595377, total=   5.6s\n",
      "[CV] gamma=0.003823475224675185, C=5.319450186421157 .................\n",
      "[CV] .. gamma=0.011207606211860567, C=4.042422429595377, total=   5.5s\n",
      "[CV] gamma=0.003823475224675185, C=5.319450186421157 .................\n",
      "[CV] .. gamma=0.011207606211860567, C=4.042422429595377, total=   5.5s\n",
      "[CV] .. gamma=0.003823475224675185, C=5.319450186421157, total=   5.4s\n",
      "[CV] .. gamma=0.003823475224675185, C=5.319450186421157, total=   5.3s\n",
      "[CV] .. gamma=0.003823475224675185, C=5.319450186421157, total=   5.3s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Done  30 out of  30 | elapsed:  1.1min finished\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "RandomizedSearchCV(cv=None, error_score='raise',\n",
       "          estimator=SVR(C=1.0, cache_size=200, coef0=0.0, degree=3, epsilon=0.1, gamma='auto',\n",
       "  kernel='rbf', max_iter=-1, shrinking=True, tol=0.001, verbose=False),\n",
       "          fit_params=None, iid=True, n_iter=10, n_jobs=-1,\n",
       "          param_distributions={'gamma': <scipy.stats._distn_infrastructure.rv_frozen object at 0x7fed1cd7f400>, 'C': <scipy.stats._distn_infrastructure.rv_frozen object at 0x7fed1cd7f470>},\n",
       "          pre_dispatch='2*n_jobs', random_state=42, refit=True,\n",
       "          return_train_score='warn', scoring=None, verbose=2)"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.svm import SVR\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "from scipy.stats import reciprocal, uniform\n",
    "\n",
    "param_distributions = {\"gamma\": reciprocal(0.001, 0.1), \"C\": uniform(1, 10)}\n",
    "rnd_search_cv = RandomizedSearchCV(SVR(), param_distributions, n_iter=10, verbose=2, random_state=42, n_jobs=-1)\n",
    "rnd_search_cv.fit(X_train_scaled, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVR(C=4.745401188473625, cache_size=200, coef0=0.0, degree=3, epsilon=0.1,\n",
       "  gamma=0.07969454818643928, kernel='rbf', max_iter=-1, shrinking=True,\n",
       "  tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rnd_search_cv.best_estimator_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now let's measure the RMSE on the training set:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5727524770785356"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred = rnd_search_cv.best_estimator_.predict(X_train_scaled)\n",
    "mse = mean_squared_error(y_train, y_pred)\n",
    "np.sqrt(mse)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Looks much better than the linear model. Let's select this model and evaluate it on the test set:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.592916838552874"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred = rnd_search_cv.best_estimator_.predict(X_test_scaled)\n",
    "mse = mean_squared_error(y_test, y_pred)\n",
    "np.sqrt(mse)"
   ]
  },
  {
   "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": "3.5.4"
  },
  "nav_menu": {},
  "toc": {
   "navigate_menu": true,
   "number_sections": true,
   "sideBar": true,
   "threshold": 6,
   "toc_cell": false,
   "toc_section_display": "block",
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}