06-algorithm-chains-and-pipelines.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false,
    "hide_input": false
   },
   "outputs": [],
   "source": [
    "from preamble import *\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Algorithm Chains and Pipelines"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test score: 0.95\n"
     ]
    }
   ],
   "source": [
    "from sklearn.svm import SVC\n",
    "from sklearn.datasets import load_breast_cancer\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "# load and split the data\n",
    "cancer = load_breast_cancer()\n",
    "X_train, X_test, y_train, y_test = train_test_split(\n",
    "    cancer.data, cancer.target, random_state=0)\n",
    "\n",
    "# compute minimum and maximum on the training data\n",
    "scaler = MinMaxScaler().fit(X_train)\n",
    "# rescale training data\n",
    "X_train_scaled = scaler.transform(X_train)\n",
    "\n",
    "svm = SVC()\n",
    "# learn an SVM on the scaled training data\n",
    "svm.fit(X_train_scaled, y_train)\n",
    "# scale test data and score the scaled data\n",
    "X_test_scaled = scaler.transform(X_test)\n",
    "print(\"Test score: {:.2f}\".format(svm.score(X_test_scaled, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Parameter Selection with Preprocessing "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best cross-validation accuracy: 0.98\n",
      "Best set score: 0.97\n",
      "Best parameters:  {'gamma': 1, 'C': 1}\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import GridSearchCV\n",
    "# illustration purposes only, don't use this code\n",
    "param_grid = {'C': [0.001, 0.01, 0.1, 1, 10, 100],\n",
    "              'gamma': [0.001, 0.01, 0.1, 1, 10, 100]}\n",
    "grid = GridSearchCV(SVC(), param_grid=param_grid, cv=5)\n",
    "grid.fit(X_train_scaled, y_train)\n",
    "print(\"Best cross-validation accuracy: {:.2f}\".format(grid.best_score_))\n",
    "print(\"Best set score: {:.2f}\".format(grid.score(X_test_scaled, y_test)))\n",
    "print(\"Best parameters: \", grid.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false,
    "hide_input": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA1MAAAJWCAYAAAC9At2bAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAiAYgoAAAAAAvg/hCDlsIDfXa8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7ff08724fe80>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mglearn.plots.plot_improper_processing()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Building Pipelines"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.pipeline import Pipeline\n",
    "pipe = Pipeline([(\"scaler\", MinMaxScaler()), (\"svm\", SVC())])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(steps=[('scaler', MinMaxScaler(copy=True, feature_range=(0, 1))), ('svm', SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,\n",
       "  decision_function_shape=None, 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": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pipe.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test score: 0.95\n"
     ]
    }
   ],
   "source": [
    "print(\"Test score: {:.2f}\".format(pipe.score(X_test, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Using Pipelines in Grid-searches"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "param_grid = {'svm__C': [0.001, 0.01, 0.1, 1, 10, 100],\n",
    "              'svm__gamma': [0.001, 0.01, 0.1, 1, 10, 100]}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best cross-validation accuracy: 0.98\n",
      "Test set score: 0.97\n",
      "Best parameters: {'svm__gamma': 1, 'svm__C': 1}\n"
     ]
    }
   ],
   "source": [
    "grid = GridSearchCV(pipe, param_grid=param_grid, cv=5)\n",
    "grid.fit(X_train, y_train)\n",
    "print(\"Best cross-validation accuracy: {:.2f}\".format(grid.best_score_))\n",
    "print(\"Test set score: {:.2f}\".format(grid.score(X_test, y_test)))\n",
    "print(\"Best parameters: {}\".format(grid.best_params_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false,
    "hide_input": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA1MAAAHnCAYAAABdfAfWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzs3Xl8FfW9//H35wRCIotEIoRFFhGLvS0i7grK5hrZBBWr\nlqBiW+0Pr16ol2oL2N5WCtcFtbWgBepSi0WhyiIVBEGFKov0UqhW2ZRFVJBdJPn8/jiTeJKcbMNJ\nchJez8djHuTMvOc735OQmflkvjPH3F0AAAAAgIqJVHcHAAAAAKAmopgCAAAAgBAopgAAAAAgBIop\nAAAAAAiBYgoAAAAAQqCYAgAAAIAQKKYAADWKmeWZ2cnB178zs3vLkw2xne+Z2byw/QQA1H7G50wB\nQO1nZt+TdJekjpL2SFot6Vfu/ma1diwEM8uV1MHdP0pU1szaSNogqY675yWmpwCA2o4rUwBQy5nZ\n3ZIelPRLSU0ltZb0uKQ+JeRTqq53oVglZE2SV7BtAMAxjmIKAGoxM2skaayk2919lrsfdPdcd5/j\n7v8dZEab2Qtm9rSZ7ZY0xMxSzexhM/vEzD42s4fMrG6Qb2JmL5vZLjP73MwWx2zvniC/x8zWmVmP\nOH06x8y2mZnFzBtgZu8FX59tZm8F7X9iZo+aWZ0S3t8UM7s/5vVIM9sa9GGoogVS/rIrzWylmX1p\nZpvMbHRMU/nvYXfQ93PNbIiZLYlZ/wIz+3vQr+Vmdn7MstfN7H4zWxqsP8/MTijfTwkAUFNRTAFA\n7Xa+pHqSZpaR6ytpurs3lvScpPsknSOpk6TTg6/vC7L/JWmLpCaKXun6qSSZ2amS7pB0prs3knSZ\npI1FN+Tuf5e0T1LPmNnXS3om+DpX0n9KOiHof09Jt5f1Rs3sckl3S+olqYOk3kUi+yTd5O7HS8qW\n9EMz6xssuyj4t5G7N3L35fndDdrOkPSKpIeD9/2QpNnB/Nj3METSiYp+z0eU1WcAQM1GMQUAtVsT\nSZ+V4z6gt939ZUly90OSvidprLt/7u6fK3p166Yg+7Wk5pLaBVe58u+7ypWUKuk7ZlbH3Te7+4YS\ntvd8sA2ZWUNJVwbz5O4r3f3vHrVZ0iRJF5fjvV4jaYq7r3P3g5LGxC509zfcfW3w9f8F2yvabknD\n/LIlve/uz7l7nrs/L2m9Cg+VnOLuH7r7V5KmS+pcjj4DAGowiikAqN0+l5RpZmXt77cUed1C0uaY\n15uCeZI0XtKHkuab2b/N7B5JcvcPFb2iNEbSDjN7zsyal7C95yQNCIYOXi1phbtvkSQz6xAMI9wW\nDDv8H0mZ5XivLYq8j02KKY6CoXsLzezToN0flLPd/LY3FZm3SVLLmNfbY74+IKlBOdsGANRQFFMA\nULu9LekrSf3LyBV9tOsnktrEvG4jaaskufs+dx/h7u0VHR54d/69Ue7+vLt3i1n3gbgbc1+naDFy\npaLD456LWfw7SesktQ+GHd6r8j0YYpukk4r0OfZ9PavocMeWQbu/j2m3rEfbbpXUtsi81op+nwAA\nxyiKKQCoxdx9j6TRkh43s35mlm5mdczsCjOLW+gEnpd0n5llmlmmpJ9JelqSzCzbzNoHub2SjkjK\nM7NTzayHmaVKOizpoKTShhc+J+lOSd0kvRAzv6GkPe5+wMw6SvpROd/udEk5ZnaamR0n6edFljeQ\ntMvdvzazcxQMMwzsDPraXvHNkdTBzAabWYqZXSfpNEkvl7NvAIBaiGIKAGo5d39Q0Qcz3CfpU0WH\n792u0h9K8UtJ70paI+m94Ov/CZZ1kPSame2V9Kakx919saIPXXhA0cJkq6IPYhhVyjaeV/TBDwvc\n/YuY+SMk3WBmexS9evR80bdUwvucp+gDIhZKel/SgiKR2yX9wsy+VPR78eeYdQ8G7+9NM/siKLZi\n2/5C0lVB3z4L/s12912l9QkAULvxob0AAAAAEAJXpgAAAAAgBIopAAAAAAiBYgoAAAAAQqCYAgAA\nAIAQKKYAAAAAIASKKQAAAAAIgWIKAAAAAEKgmAIAAACAECimAAAAACAEiikAAAAACKFOdXcAOJak\np6dvP3ToULPq7kdNlZaWtuPgwYNZ1d0PIJmxnwH7SqDqmLtXdx+AY4aZOb9z4ZmZ3N2qux9AMmM/\nA/aVQNVhmB8AAAAAhEAxBQAAAAAhUEwBAAAAQAgUUwAAAAAQAsUUgITYu3evhg8frnbt2qlu3bpK\nSUnRmjVrtHjxYkUiEd1///3V3UUAqFQ5OTmKRCLavHlzwbxNmzYpEono5ptvrsaeAagsFFNAEsnL\ny9PkyZPVvXt3NWnSRKmpqWrWrJlOP/10DRs2TC+//LIkyd3VunVrRSIRrV+/vtQ2Dx48qMaNGyst\nLU2fffZZoWUHDhzQww8/rF69eqlZs2aqV6+eMjIydO655+q+++7Thg0byt33kSNH6rHHHlOnTp30\n05/+VKNHj1ZWVvTJvGYms8IPlurevbsiEXZBQFWryfuZZBdvX1fa/PKKV6QBSA58zhSQJPLy8pSd\nna1XX31VGRkZys7OVqtWrXT48GGtXbtWf/rTn/Svf/1Lffr0kZnplltu0f33368nn3xSEyZMKLHd\n6dOna8+ePbr22muVmZlZMH/ZsmUaNGiQtm3bplatWik7O1stWrTQgQMHtHr1ak2YMEETJkzQsmXL\n1Llz5zL7P3v2bH3rW9/SrFmzCs1v1KiR1q1bV2jb0tGfXACouJq+n6mJWrZsqXXr1un4448P3Qb7\nSyCJuTsTE1MVTdFfufieeeYZNzPv0qWL7927t9jygwcP+qJFiwpeb9myxevUqeNNmzb1r7/+usR2\nu3bt6pFIxBcsWFAwb/369d6oUSOvU6eOjx8/3nNzc4utt2XLFr/++ut98eLFJbYdKxKJeI8ePcqV\ndXfv3r27RyKRcufd3YPvX7X/HJmYknmqzfuZZJeTk+ORSMQ3bdqU0HaHDBlSoXbZVzIxVd1U7R1g\nYjqWptJOcm6//XaPRCL+yCOPlJgp6qqrrvJIJOLTp0+Pu3z9+vVuZt6hQ4dC83v37u2RSMTvu+++\nMrdx+PDhUpd3797dzcwjkYibWcGUX1gtWrTIzczHjh3r7u4bN26Mm49dpyScIDAxlT3Vxv2Mu/vU\nqVPdzHzatGn+yiuv+AUXXOD169f3jIwMHzRokH/wwQfF1hkyZIibmW/YsMEnTpzonTp18vT09GL7\nmnnz5vkVV1zhmZmZXq9ePW/fvr2PHDnSd+/eHbcvf/vb37xr165ev359P+GEE7x///6+fv36uMVU\n/j5v6NChxdo5cOCAP/DAA37WWWd5w4YNvUGDBn7aaaf58OHD/dNPP3V3L7afzJ/atWtX4veKfSUT\nU9VNDPMDkkSTJk3k7nr//ffLvc6wYcM0e/ZsPfnkk7rmmmuKLZ88eXLBUJ18Gzdu1IIFC5Senq6R\nI0eWuY26deuWunzo0KHq0aOHxowZo7Zt2yonJ0eS1LZt27j5xo0ba8yYMZoyZYo2b96sMWPGyN1L\nXQdAYtTU/Uw+M9OMGTM0d+5cXX311erRo4dWr16tGTNmaNGiRXrrrbfUoUOHQnkz0/Dhw7V06VJl\nZ2crOztbKSkpBZmxY8dq7NixatKkia666io1bdpUa9as0YQJEzR37ly9/fbbatCgQUH+L3/5iwYP\nHqx69epp8ODBysrK0tKlS3X++eerU6dO5XofkrR79251795da9asUceOHXXLLbcoNTVVH374oaZO\nnaqBAwfqxBNP1JgxY/TSSy9pzZo1uvPOO9W4cWNJKvgXQDWr7mqOielYmlTKX4xXrVrlqampHolE\n/KabbvIXX3yxzCEdubm53rJlS09JSfGNGzcWWnb48GFv2rSpp6am+o4dOwrmP/30025m3q1bt1Lb\nrqiSriwVvTKVj2F+TEyVM9XW/Uz+lalIJOJz5swptGzixIluZt67d+9C83NyctzMvFWrVnHf58KF\nC93MvGvXrr5nz55Cy6ZNm+Zm5nfffXfBvH379vkJJ5zgqampvnLlykL5u+++u6B/5bkydf3113sk\nEvE77rijWL/2799fqD8VHT7IvpKJqeomHqUFJInOnTvr2WefVVZWlp599lkNHDhQbdu2VWZmpq6+\n+mq98sorxdbJf9xuXl6ennrqqULLZs6cqZ07d6pv375q2rRpwfxt27ZJklq1alW5bwhA0qkN+5le\nvXrpiiuuKDTvjjvuUPv27bVw4UJt2bKl0DIz0z333KPWrVsXa2vixIkyM02aNEkNGzYstOz73/9+\nwfcr36xZs7Rr1y7dcMMNOuOMMwrlR48eXe6HTOzcuVPTp09X8+bNNX78+GLLjzvuuGL9AZCcKKaA\nJDJo0CBt3rxZr776qn7+85+rT58+cnfNmjVLffv21dChQ4utc+uttyoSiWjKlCly94L5+UNvhg0b\nVpVvAUCSq+n7mYsuuqjYvEgkoq5du0qSVq1aVWz52WefHbetZcuWqW7dupo+fXrBcL/Y6fDhw9q5\nc6d27dolSVq5cqXMLG4fGjVqVO4nEr7zzjvKy8vTRRddpPT09HKtAyA5cc8UkGRSUlLUu3dv9e7d\nW1J0KO6MGTM0dOhQ/fGPf9SAAQPUt2/fgnzr1q11ySWXaP78+ZozZ46ys7O1adMmLViwQG3atNGl\nl15aqP3mzZtLkj755JOqe1MAkkpN3s80a9Ys7vz8z7X78ssvS1xW1Oeff67c3NxSP1TczLRv3z5l\nZGQUtF1WH8qye/duSdHHpgOo2bgyBSQ5M9OgQYN01113yd21cOHCYpnbbrtN7q7JkydLkp588klJ\n0b8mF5X/19t3331Xe/furcSeA6gpatJ+ZseOHXHnb9++XZLiDrUr6TOajj/+eGVkZCg3N7fE6ciR\nIzrppJMKtV1WH8qS//AI/qgF1HwUU0ANkT9+PnaITb6+ffsqKytLc+bM0ccff6wpU6YoJSUl7nCd\ntm3bqnfv3jp06FDcsfpFHT58+Og7H0f+07TivR8A1aMm7GcWL15cbF5eXp6WLl0qScXuZSrNeeed\np127dmndunXlynfp0kXuHrcPe/bs0erVq8vVzjnnnKNIJKI33nhDBw8eLDOfv7/Mzc0tV/sAqg7F\nFJAknn/+eb322mtxT2K2b9+uSZMmlThWPyUlRTk5OTpy5IhuuOEGbd26VdnZ2QVDbYqaOHGiGjZs\nqF//+td68MEH4x6gN2/erMGDB2vZsmVH/+biaNKkScF2AFSN2rCfWbhwoWbPnl1o3qOPPqoPP/xQ\nPXv2LLiKVB75V+KGDRtW8NCMWAcOHNDy5csLXvfr108ZGRl67rnntGLFikLZ0aNHxx1iGE9mZqYG\nDx6srVu3asSIEcV+Hvv379eePXsKXrO/BJIX90wBSWL58uV65JFHlJWVpa5du6pdu3aSpA0bNmj2\n7Nk6dOiQ+vfvr4EDB8Zdf9iwYRo3bpyWLFkiM9Ntt91W4rY6duyo+fPna9CgQRo5cqQeeeQR9erV\nSy1atND+/fv13nvv6c0331QkEtGoUaMq5f326tVLL7zwggYMGKArr7xS6enpatOmjW688cZK2R6A\n2rGf6dOnjwYMGKABAwbolFNO0apVqzRv3jxlZmbq8ccfr9D3o2fPnho3bpxGjRqlDh066Morr1S7\ndu20b98+bdq0SYsXL1a3bt00Z84cSVL9+vU1adIkDR48WN26ddN1112n5s2ba+nSpVq7dq0uuugi\nLVmypFzbfuyxx7R27Vo98cQTev3113XZZZcpNTVVH330kebPn6+XX365oKjt1auXxo8fr1tvvVUD\nBw5Uw4YN1bhxY91xxx0Ver8AKkF1P5udielYmlTK5798/PHH/tvf/tavvvpq79ixox9//PFer149\nb9GihWdnZ/tzzz1X4rr5Lr30Uo9EIt6mTRvPy8srM79//35/+OGHvWfPnt6sWTNPTU31xo0b+1ln\nneX33ntvsc+UKU0kEvGePXsWm79o0SKPRCJ+//33F5qfm5vr9957r7dv377gc2/ifU5VLPHZKUxM\nZU61dT8zdepUj0QiPm3aNJ89e7ZfcMEF3qBBA8/IyPBrrrnGP/jgg2Lr5OTkeEpKSpmfz/Tmm2/6\ndddd5y1btvR69ep506ZN/YwzzvARI0b4ihUriuVfe+0179atm9evX99POOEEHzBggP/rX/+Ku72N\nGzd6JBLxm2++uVg7Bw4c8F/96ld++umne/369b1Ro0b+H//xH3733Xf7zp07C2Ufeugh//a3v+1p\naWkeiUS8Xbt2Jb4f9pVMTFU3mTv3KwBVxcyc37nwzEzuHv9OcgCSau9+Ztq0abr55ps1ZcoUff/7\n36/u7iQ19pVA1eGeKQAAAAAIgWIKAADUCLXxihuAmo1iCgAA1AglfV4UAFQX7pkCqlBtvZehqnAf\nAFA29jNgXwlUHa5MAQAAAEAIFFMAAAAAEALFFAAAAACEQDEFAAAAACFQTAEAAABACHWquwPAsSQt\nLW2HmTWr7n7UVGlpaTuquw9AsmM/A/aVQNXh0egAZNEPb5klaae731Ld/QGAMMzsu5JekDTC3V+p\n7v4AqP0Y5gdAko6T1EfS4OruCAAchYskfUvSFdXdEQDHBoopAAAAAAiBYgoAAAAAQqCYAgAAAIAQ\nKKYAAAAAIASKKQAAAAAIgWIKAAAAAEKgmAIAAACAECimAAAAACAEiikAAAAACIFiCgAAAABCoJgC\nAAAAgBAopgAAAAAgBIopAAAAAAiBYgoAAAAAQqCYAgAAAIAQKKYAAAAAIASKKQAAAAAIgWIKAAAA\nAEKgmAIAAACAECimAAAAACAEc/fENWY2VtLFCWsQQFWpL+ms4OvF1dkRAKHNcveHwqxoZi0lPSYp\nI7FdqnKx5yA1fV92UNK97r6yujsCoGQJK6bMLF3SgYQ0BgAAKmqvuzcKs6KZ/VDS7xLcHxy9R919\neHV3AkDJ6iSwrfwhg19JujyB7QKoGs0kHZa0q7o7AqBC0iXNkZRyFG3krztL0sNH3aPqY5JOkfSh\npLxq7svRGCBpuI7uZwqgCiSymMqX6+6LKqFdAABQhJnVT2Bzn9SCY/jr1d2Bo2Vm/1HdfQBQPjyA\nAgAAAABCSOSVqa8k7ZT0eQLbBAAApUvE8XdzkX9Rvfh5ADVEop/m10zSYXfnngsAAKrI0R5/zazg\nXiN3r8n3GtUK/DyAmiOhxRQAAAAAHCu4ZwoAAAAAQqCYAgAAAIAQKKYAAAAAIASKKQAAAAAIgWIK\nAAAAAEKgmAIAAACAECimAAAAACAEiikAAAAACIFiCgAAAABCoJgCAAAAgBAopgAAAAAgBIopAAAA\nAAiBYgoAAAAAQqCYAgAAAIAQKKYAAAAAIASKKQAAAAAIgWIKAAAAAEKgmAIAAACAECimAAAAACAE\niikAAAAACIFiCgAAAABCoJgCAAAAgBDqVHcHgNoiPT19+6FDh5pVdz+OdWlpaTsOHjyYVd39AAAA\ntZ+5e3X3AagVzMz5fap+ZiZ3t+ruBwAAqP0Y5gcAAAAAIVBMAQAAAEAIFFMAAAAAEALFFAAAAACE\nQDEFHEOmTZumSCSiP/7xj9XajxUrVuiSSy7RiSeeqEgkoi5dukiScnJyFIlEtHnz5mrtHwAAQHnw\naHTgGGNWvQ+627t3r6688kodPnxYN910kzIzM9W8efOCvhXt36ZNm9SuXTvl5OToD3/4Q3V0GQAA\nIC6KKQBV6u9//7t27typX//617rnnnsKLXvggQc0atQotWzZspp6BwAAUH4UUwCq1CeffCIzK7ga\nFatZs2Zq1qzw5x7z2V0AACBZcc8UUEX++te/qlevXmrRooXS0tLUsmVLde/eXb/73e+KZXft2qV7\n771X3/3ud1W/fn01btxYnTt31qhRo3Tw4MGC3MqVK3XnnXeqc+fOatKkidLT03XqqadqxIgR2r17\nd4X698knn+jHP/6x2rdvr7S0NGVmZqpfv3569913i2XHjBmjSCSiN954Q88995zOO+88NWzYUCef\nfHKJ7W/atEmRSEQ5OTly94L7o2Lv4Sp6z9TYsWN18skny8w0derUgnwy3PcFAADAlSmgCkyaNEk/\n/OEP1bx5c/Xt21eZmZn69NNPtWbNGk2dOlU/+tGPCrIbN25U9+7dtWXLFp155pm6/fbblZeXp/ff\nf18PP/ywfvSjH6l169aSpMmTJ2vmzJm6+OKLdckllygvL08rVqzQgw8+qHnz5mn58uWqX79+mf1b\nuXKlLr30Uu3evVuXXXaZBg4cqM8++0wzZ85U165dNXPmTF1++eUF+fx7myZMmKDXXntNffr0Uc+e\nPfXll1+WuI3GjRtrzJgxWr16tWbNmqV+/fqpc+fOklTwb9F7pnr06KEvv/xSDz/8sDp37qz+/fsX\nLMtfBwAAoNq4OxMTUwKm6K9TfGeeeaanpaX5Z599VmzZ559/Xuj1+eef75FIxMeNGxc3+9VXXxW8\n3rx5s+fl5RXL/eEPf3Az89/85jeF5k+dOtUjkYhPmzatYN6RI0e8ffv2np6e7kuWLCmU37Ztm7ds\n2dJbtGjhhw8fLpg/ZswYNzNv0KCBv/feeyW+73ji9SFfTk6ORyIR37RpU8G8jRs3upn50KFDy9V+\n8HOo9v8PTExMTExMTLV/YpgfUEXq1KmjlJSUYvNPOOGEgq9XrlypZcuW6YwzztBPfvKTuNnU1NSC\n1yeddFLcp/Pl5OSoUaNGevXVV8vs1+zZs/XRRx/p//2//6euXbsWWpaVlaWf/OQn2r59uxYsWFBs\n3R/84Afq1KlTmdsAAACojRjmB1SBG264QSNGjNC3v/1tDR48WBdffLEuvPBCZWZmFsotW7ZMknTp\npZeWq90jR47oiSee0J///Gf985//1Jdffqm8vLyC5Z988kmZbbz99tuSosMLx44dW2z5Bx98IHfX\nunXrig31O/vss8vVTwAAgNqIYgqoAnfddZdOPPFE/fa3v9Wjjz6qRx55RJJ08cUXa/z48TrzzDMl\nSbt375aZlfvR4Ndee61mzpyp9u3bq3///srKylK9evUkSQ899JC++uqrMtv4/PPPJUl/+ctfSsyY\nmfbt21dsflZWVrn6CQAAUBtRTAFV5MYbb9SNN96oPXv26K233tJLL72kp556SpdffrnWr1+vJk2a\nqHHjxnL3cl1RWrFihWbOnKlLL71Uc+bMUSTyzahdd9e4cePK1a/jjz9eZqa//vWvys7OrtB7qu4P\nAAYAAKhO3DMFVLFGjRrp8ssv1+9//3vl5OToiy++0BtvvCFJOu+88ySpXPc6/fvf/5Yk9enTp1Ah\nJUnLly8v9Aj10px33nly94I+JJv8+8xyc3OruScAAACFUUwBVWDRokVx5+/YsUOSdNxxx0mSunTp\nogsuuECrV6/WAw88UCz/xRdfFAzda9u2bdy2P/30U/34xz8ud9/69eun9u3b6/HHH9fcuXPjZpYt\nW6ZDhw6Vu81EysjIkJkVfPYUAABAsmCYH1AFBgwYoAYNGui8885T27Zt5e5asmSJ3nnnHZ199tnq\n3bt3QfaZZ55Rjx49dO+992rGjBnq3r273F3vv/++XnvtNa1fv16tW7fW2WefrQsvvFAvvviiLrzw\nQnXt2lU7duzQ3Llz1bFjR7Vo0SJuX9y90Os6deroxRdf1OWXX67s7GxdcMEF6ty5s4477jht2bJF\n77zzjjZs2KBt27YpLS0tId+Pon0oTf369XXuuedqyZIluvHGG3XqqacqJSVF/fr103e+852E9AcA\nACAMiimgCowbN06vvvqqVq1apblz5yotLU1t2rTR+PHj9cMf/rDQI9Pbtm2rlStX6je/+Y1mzpyp\nxx9/XGlpaWrbtq1GjBihpk2bSpIikYhefvll3XfffZozZ44effRRtWzZUrfddpvuvfdenXbaaXHv\naYo377vf/a7ee+89Pfjgg3rllVc0depURSIRNW/eXF26dNEvfvGLYk8ePBql3WsVb9kzzzyju+66\nS6+++qqef/55ubtOOukkiikAAFCtrCJ/IQZQMjNzfp+qn5nJ3XkyBgAAqHTcMwUAAAAAIVBMAQAA\nAEAIFFMAAAAAEALFFAAAAACEQDEFAAAAACFQTAEAAABACBRTAAAAABACxRQAAAAAhEAxBQAAAAAh\n1KnuDgC1RVpa2g4za1bd/TjWpaWl7ajuPgAAgGODuXt19wFAEjCzayTtc/e51d0XAACAmoBiCoDM\nrJ6kfZJ2u/uJ1d0fAACAmoB7pgBI0SG/dSQdV90dAQAAqCkopgAAAAAgBIopAAAAAAiBYgoAAAAA\nQqCYAgAAAIAQKKYAAAAAIASKKQAAAAAIoU4iGzMzS3SbAKpE3fx/zaxuqUkAyeqI8+GRAFClEvah\nvUEhtVTSBQlpEAAAVMRMdx9Q3Z0AgGNJIoup+pL2BS+PJKRRAFXFJKUEX/P7C9Qs+b+/B9y9fnV3\nBgCOJZUxJI+dOQAAVaTIHzMBAFWIB1AAAAAAQAgUUwAAAAAQQiKH+R2Q9JqkTxPYJgAAKB3HXwCo\nJgl7AAUAAAAAHEsY5gcAAAAAIVBMAQAAAEAIFFMAAAAAEALFFAAAAACEQDEFAAAAACFQTAEAAABA\nCBRTAAAAABACxRQAAAAAhEAxBQAAAAAhUEwBAAAAQAgUUwAAAAAQAsUUAAAAAIRAMQUAAAAAIVBM\nAQAAAEAIFFMAAAAAEALFFAAAAACEQDEFAAAAACHUqUg4PT19+6FDh5pVVmcAAEBipKWlfXrw4MFC\nx+y6detuP3LkCMfxWqpOnTo7vv7666z81/y8gfIp+rtTEebu5Q+beWZmpl544QV17969zPyiRYt0\nzTXXkCdPnjx58uSrON+jRw+5u8XONzMfM2ZMsfyGDRv0wgsv6JprrlG7du3KbJ98cubHjBlT6Gde\n0s8bQGFFf3cqosLD/JLpQEGePHny5MmTj58vr5pSKJAPlwdQuSpcTCXTgYI8efLkyZMnHz9fHsl2\n4k8+sXkAlS/hD6BI5gMLefLkyZMnTz4q2U78ySc2D6BqJLSYSrYDBXny5MmTJ0++uGQ78Sef2DyA\nqpOwYipdaiHOAAAgAElEQVTZDhTkyZMnT548+eKS7cSffGLzAKpWQoqpZDtQkCdPnjx58uTjS6YT\nf/KJzQOoekddTCXbgYI8efLkyZMnX7JkOfEnn/g8gKp3VMVUsh0oyJMnT548efKlS5YTf/KJzwOo\neqGLqWQ7UJAnT548efLkj16yFgrky84DqHqhiqlkO1CQJ0+ePHny5I9eMhcK5I8+DyDxKlxMJduB\ngjx58uTJkyd/9JLtxJ98YvMAKkeFi6lkOlCQJ0+ePHny5OPnKyLZTvzJJzYPoPJUuJhKpgMFefLk\nyZMnTz5+vryS7cSffGLzACpXhYupZDpQkCdPnjx58uTj58sj2U78ySc2D6DyVbiYKksyH1jIkydP\nnjx58lHJduJPPrF5AFUjocVUsh0oyJMnT548efLFJduJP/nE5gFUnYQVU8l2oCBPnjx58uTJF5ds\nJ/7kE5sHULUSUkwl24GCPHny5MmTJx9fMp34k09sHkDVO+piKtkOFOTJkydPnjz5kiXLiT/5xOcB\nVL2jKqaS7UBBnjx58uTJky9dspz4k098HkDVC11MJduBgjx58uTJkyd/9JK1UCBfdh5A1QtVTCXb\ngYI8efLkyZMnf/SSuVAgf/R5AIlX4WIq2Q4U5MmTJ0+ePPmjl2wn/uQTmwdQOSpcTCXTgYI8efLk\nyZMnHz9fEcl24k8+sXkAlafCxVQyHSjIkydPnjx58vHz5ZVsJ/7kE5sHULkqXEwl04GCPHny5MmT\nJx8/Xx7JduJPPrF5AJWvwsVUWZL5wEKePHny5MmTj0q2E3/yic0DqBoJLaaS7UBBnjx58uTJky8u\n2U78ySc2D6DqJKyYSrYDBXny5MmTJ0++uGQ78Sef2DyAqpWQYirZDhTkyZMnT548+fiS6cSffGLz\nAKreURdTyXagIE+ePHny5MmXLFlO/MknPg+g6h1VMZVsBwry5MmTJ0+efOmS5cSffOLzAKqeuXu5\nw+np6dsPHTrUrBL7AwAAEiAtLW3HwYMHs2Ln1a1bd/uRI0c4jtdSderU2fH1118X/Mz5eQPlU/R3\npyIqVEwBAAAAAKIS8gAKAJXPzF43s4kVXGeDmd1dWX2K2Y6Z2e/N7DMzyzWzi8qxThszyzOzLqVk\nzgwyrRPbYwBVqej+qzz7MzP7h5n9PNHbrixm1szM5pvZPjPLLec6Q8xsbxmZ/zIzbogCklSd6u4A\nUFuZ2euS/uHuwxPU5ABJX1dwnbMk7U/Q9ktzpaQhki6WtEHSF+VcrzyXxrl8DtQ+YfZnpTKzIZIe\nc/eGlb2tEoyQlCWpk6R9FViP/WCSqYTjt8zsYkmvS8p09/IeI5N2O/gGxRRQzcysjrsfKSvn7rsr\n2ra7fx6uVxXWQdI2d19ewfWsMjoDILmF2Z+VgylO0VFJ24rnFEkr3P2jKtoeapb8/5+Vfdyrqu0g\nwDA/oBKY2RRFr9LcEQxTyzWz1mZ2cfD6CjNbbmaHJF1qZieb2Uwz2xYMEVlhZtlF2iw6TGaDmd1r\nZk+Y2ZdmtsXMRhRZp9Awv2Dbw8xserCdD83shiLrnBts/6CZvWNmlwXrxR26F7zXByW1DnIfBfNT\nzexhM9setPW2mV1YxvftcjNbF+QXSzq1yPJGZva0me0IMv82s4T95RBAYcH+YruZWZH5z5nZzODr\nMvdfcdotuj870cxmmdmBYL81NM46d5nZe8E2PjazyWbWKFh2saQ/SKofs8/9eQnbamxm08zsi2B7\nfzOzb8csH2Jme82sZzDUcJ+ZLTSzNqW8nw2S+koaEmz7D8H8k8zsJTPbE0wzzKxlGd+bnwTfyz1m\nNlVSg9LySKySjt/Bsm+b2SvBz2ZH8HvQLGbd75jZa8Exea+ZrQqO+20kLQxiO2P/j8TZfh0zm2hm\nn5jZITPbZGa/ille18zGBcf8/RY9l7g0WFbu7SBxKKaAynGnpLclTZHUTFJzSVtilj8g6V5JHSUt\nV/RgOUdSL0WHiPxF0gwzK1RMxPGfktZIOkPSOEm/MbNzy1jnZ5JeCrbzZ0l/MLNWkmRm9SW9LOmf\nQZv/LWmCSh9iMlzS/ZI+Dt7r2cH88ZKukZQjqbOkf0iaF3vgiWVmJwX9elXS6ZIelfSbIrH/kfQf\nig4rPFXSzZI+KeP9AgjvBUmNJF2SPyPYT/SV9HQwK+z+K9Y0SSdL6impv6TvSypavOQqum/9tqTr\nFd3XPBose0vR/eEBfbPPnVDKts6W1Cf494Ci+6Z6MZl6iu7/ciSdJ6mxpCdK6f9ZkhYouk/NknRn\nUID+VdKJip6cd5fUQtH9XFxmdq2kXyi6n+4i6X1JlX7fKwqJe/w2syxJixU95p6l6P/3+pJmxaz7\nnKStwfLTJY2RdEjSZkkDg8xpQZt3lrL9fpKuVfRq53WS/hWzfKqkbpIGK3o8nCbpr2b23QpuB4ni\n7kxMTJUwKTpmeWKReRdLypPUvxzrvy3ppyW1p+i9Sc8WWef9IutskHR3zOs8Sb+MeZ2i6D1V3wte\n/0DSZ5LqxWSuV/Qk5qJS+vpfkj6KeX2cpK8k3RAzLyLp35LuD163CfrTJXj9K0nri7R7b7Dt1sHr\nWZKerO6fLRPTsTRJmiFpWszrGyXtkpRayjpl7b8KXiv6h5E8SefFLG8t6Yikn5eyjcskHYx5PUTS\nnji52G11CLZ1YczyRpJ2S7o5pp1cSafEZL4Xu60S+vOypD/EvL5E0Xu1ToqZ1y5ou2e8Pkt6U9IT\nRdr9W+z+lalK/s/HO36PlfS3IvMygv9PZwWvv5R0UwltXhz87E8oY9uPFN1OzLKTgzZaFZn/kqL3\nC5Z7O0yJm7gyBVQ9l7QidoaZHWdmvzGztcHQk72SzlT0hKI0a4q83iqpaRnr/KOgI+65knbGrPMt\nSf/n7l/F5Jer4mOv2yt6T+ZbMdvKU/QE69slrNNR0rIi894u8vp3kgab2WozG2/leGoggKP2jKT+\nZpYWvP6epBnuflg6qv1Xvo6Knvy9kz/D3Tcruj8rEAy7mx8Mb9oj6UVJqcEVg/LK31bBvsbd9yi6\nX4zdN33l7v+Oeb012FbjCm5rq7sXjEpw9w1BWyXtB09T2ftBVI8zJV0cDN/bG/w/36zoMb19kHlQ\n0lNmtsDMfmpm3wqxnamSzjCz983sMTO7MrjKKUWvVpqkfxbpx5UxfUAVo5gCqkfRJ+z9r6KX5u+V\ndJGiwwPekZRaRjtFn1DlKvv3Osw6iRT6qVTuPk/RE7TxkppImm1mTyWqYwDimq1oAdLPzE6U1Fvf\nDPGTwu+/iipx3xDcs/KKpLWSBil6UnlzsLii2ynP9os+FCh/WaL2lTydr+aJKPp/sJOi/8fzpw7B\nfLn7WEUL4pckXSBpjZnlVGQj7r5K0ZEb/61o4TRN0vyYPuTpm2GE+dNp+ub3AVWMYgqoPIcVHUZX\nHhdK+qO7z3T3/1P0L5fV8Vem9ZK+U+TegXNV8QP/h4oWbQUPnDCziKTzFb0fK551wbZinV805O5f\nuPuz7n6zpFsUveG7bgX7B6CcgitQLyg6vO86RZ/cuTgmcrT7r/WKno+ckz8jKJ5axGTOklRX0WHL\ny4OrRkUf5FCefe66YFsF+5bgIRbfVbRQS6R1klpYzOfkmdnJir6vkra1TtF7tGIV2w+i0sX7v7RS\n0XuUNrv7R0Wmgj+QuvuH7v6Yu18l6SlJt8a0qTjtFuPu+939RXe/Q1K2pF5mdoqkVYoWWM3j9GFb\nRbeDxKCYAirPRknnWPTDaZvEXKaPN2TufUkDzOyM4CbSpxW9AbqqPafoX72eNLPTzKy3pFHBsnIX\nVO5+QNEheeMs+uTCjorevN1U0m9LWO0JSW0t+gTAU81skKL3cBUws7Fm1s/MTjGz0xT9a/iH7l4V\nnyEDHMueUfQepR9K+lORZUe1/3L39xV98Mzvzew8M+us6M3/B2JiHyh6znKXmbU1s+tV/Mb6jZLS\nzKx3sM9Nj7Otfyv6UIjfm1nXoL/PKHqvS9H3VVSFhju7+2uKDh981qIfQH5WsK133X1RCas9ougf\niG4N9nOjFFNkospsVMzxO5j3uKTjJU03s3PMrF3wf+33ZlbfzNKCYXkXB+udK6mrvimcNyl6HM02\ns8zgQS7FWPSplYPNrGNQQN2g6P/Pj939A0WP01PNbGDQhzMt+sHO/SuyHSQOxRRQeSYo+heif0r6\nVNJJwfx4RcndQeYNRYfUvC1pSZFM0fXitVNWptR13H2fpKsUHc+/UtEnBI5W9CTiUJx1S3OPgqcF\nKvrXtO9Iuszdd5Sw7S2Srlb0hG21oidK9xRp8ytJvwyWL1H0SUp9K9gvABXk7ksUfXJmR0ULgliJ\n2H8NUfSBOQsUfdDMs4qe0OZv/x+K7hPuUvTk9GZFH3wT28e3Ff2jzJ+C/ows4e3kSPp7sJ1lihZ+\nlxe5VzSeMEPz+ip6X+pCRd/bVkU/RDj+BtynK/oEuF/qmysh/xtiuzg6hY7fZtY6uPJzoaJDXudK\n+j9FnyZ5SNFjU66iD6SYoujV1hmKPlDkvyTJ3bcqejz9H0nb9c2TKIvaq+j/3eWS3lV0WOHl7p5/\nDM4JtjFO0SuZLyv6dL9NFdwOEsTcGbYLoGRm1k/Rg0JT59PUAQAACtSp7g4ASC5m9n1JHyn6uVjf\nlfSQpL9SSAEAABRGMQWgqGaKfp5GlqJDBF5R9KlCAAAAiMEwPwAAAAAIgQdQAAAAAEAIFFMAAAAA\nEALFFAAAAACEQDEFAAAAACFQTAEAAABACBRTAAAAABACxRQAAAAAhEAxBQAAAAAhUEwBAAAAQAgU\nUwAAAAAQAsUUAAAAAIRAMQUAAAAAIVBMAQAAAEAIFFMAAAAAEALFFAAAAACEQDEFAAAAACFQTAEA\nAABACBRTAAAAABACxRQAAAAAhEAxBQAAAAAhUEwBAAAAQAgUUwAAAAAQAsUUAAAAAIRAMQUAAAAA\nIVBMAQAAAEAIFFMAAAAAEALFFAAAAACEQDEFAAAAACFQTAEAAABACBRTAAAAABACxRQAAAAAhEAx\nBQA4ppjZr81sUvB1BzP7ImQ7Y8xsYmJ7BwCoSSimAKAGM7O9ZrYnmHLN7EDMvOuPot23zex7iexr\n0O4PzOxviW43LHf/wN1PKCtnZpeZ2QdF1h3j7sMrr3cAgGRXp7o7AAAIz90b5n9tZh9JusXdX6/G\nLpWHJ6ohM0tx99xEtVfappTAfgMAageuTAFA7WHB9M0Ms4iZ/czMPjSzT83saTNrFCw7zsz+ZGaf\nm9mu4GrU8WY2QdLZkp4MrnCNL7ahEtYNlmWY2TQz22Zmm8zs58H8zpIeltQ9uHq2Ne6biLZ1v5m9\nG7T9gpk1DJZ9y8y+NrNbzWyzpNnB/G5mtizIv2tmF8S0197MlprZl2Y2W1JGzLJvmdnXMa+bxPT9\n8+A9niDpRUknx1z1y4gdLhisO9DM1prZF2Y238xOiVm2zcz+08z+EfTxaTPjD5oAUMNRTAFA7TZS\nUm9JF0hqJelrSQ8Fy26VlCKpuaQmkn4s6bC7j5D0jqJXuRq5+8g47cZdN1j2rKRdktpKOkdSPzO7\nyd1XS/pPSYvcvaG7tyil3zdJul5SS0n1JD0YsywlaPfUoO02kl6SNMrdMyTdJ2lmfnEn6QVJi4J+\n/m/QdqzYK07Tg39PldRM0uPu/oWkAZI+CvrdyN13xTZgZt+VNEXSDyU1lfSGpFlmFnucHSiph6RT\nJJ0nKeHDKAEAVYtiCgBqtx9I+m933+HuhyX9QtLgYNnXkk6U1MHd89x9hbsfjFnXVLK465pZa0nd\nJP2Xu3/l7jskPapoYVQRU4L7mQ5IGh3TZyla/PzM3Q+5+1eShkiakT+80d3nSfqnpEvNrIOkjpLu\nd/cj7r5Q0rx4GzSzdpIulPQjd98b5JeWs7/XSXrR3Ze4+xFJv1L0+3NWTOZBd//M3T+XNEdS53K2\nDQBIUgwxAIDa7SRJc8ws/+qLSVIwdO0pSVmS/mJm9SU9Lek+dy/PvUFPKnrlJn/dP0r6maQ2ktIl\n7TSz/O2ZpA9KaKckW2K+3iTpuPyhfpLygiItXxtJ15vZNTHvsY6kFpI+lbQzKCRj22sQZ5utJH0a\nFHAV1SJoV5Lk7nlm9omiV9byxfb5gKJXygAANRjFFADUbh9LutrdV5WwfLSk0WbWVtJ8Sf8n6U8q\n42EL7v61pDGSxsSsu1bS25L2lvKEvPI+xOGkmK/bSDrg7nuDAq1oG1skTXb3O4s2YmanSso0s9SY\ngqq1pHiPQ98iqamZHRenoCqr31uDfuZvN6JoIfVxGesBAGowhvkBQO32e0njzKyVJJlZUzO7Kvi6\nl5mdZtEKZZ+kI5Lyn4y3Q9LJJTVa0rruvlHSMjP7jZk1sKhTzOzCmHZPKsfDF3Is+hlQDRQt+J6P\n3XyR7DRJ15hZz+CBG+nB103d/X1J/5L0MzOra2Y9JF1e9O1IUtD3NyQ9ZmaNgny3mH43Da7CxfNn\nSQPMrGvw3kZJ+kzSijLeJwCgBqOYAoDaI97Vk3GS/iZpoZl9KWmppDOCZS0lzZK0R9IaSa+4e/4D\nGB6SNCR4ot0Dcdotbd3rJTWWtF7S54oWQk2DZfMkbZT0afA0vpI8regVsi2KFmojSnqf7r5B0Yc7\njFW0gNkgabi+OcZdK6ln0JcRQduFmoj5+npJqYoOS9ym6AMl5O7vSfqrpE3B0/oaF+nDPyTdImmS\nokMLe0jq5+558foMAKgdrHxD4wEAqBpm9rakR939ueruCwAApeHKFAAAAACEQDEFAEg2DJkAANQI\nDPMDAAAAgBC4MgUAAAAAIVBMAQAAAEAIFFMAAAAAEALFFAAAAACEQDEFAAAAACFQTAEAAABACHWq\nuwMAar/09PTthw4dalbd/aip0tLSdhw8eDCruvsBAInCcQG15djG50wBqHRm5uxrwjMzubtVdz8A\nIFE4LqC2HNsY5gcAAAAAIVBMAQAAAEAIFFMAAAAAEALFFAAAAACEQDEFAEdh7969Gj58uNq1a6e6\ndesqJSVFa9as0eLFixWJRHT//fdXdxcBACiQk5OjSCSizZs3F8zbtGmTIpGIbr755mrsWc1EMQWg\n2uXl5Wny5Mnq3r27mjRpotTUVDVr1kynn366hg0bppdfflmS5O5q3bq1IpGI1q9fX2qbBw8eVOPG\njZWWlqbPPvus0LIDBw7o4YcfVq9evdSsWTPVq1dPGRkZOvfcc3Xfffdpw4YN5e77yJEj9dhjj6lT\np0766U9/qtGjRysrK/qkVzOTWeEHFXXv3l2RCLteAChNTT4uJLt4x6bS5pdXvCLtWMDnTAGoVnl5\necrOztarr76qjIwMZWdnq1WrVjp8+LDWrl2rP/3pT/rXv/6lPn36yMx0yy236P7779eTTz6pCRMm\nlNju9OnTtWfPHl177bXKzMwsmL9s2TINGjRI27ZtU6tWrZSdna0WLVrowIEDWr16tSZMmKAJEyZo\n2bJl6ty5c5n9nz17tr71rW9p1qxZheY3atRI69atK7Rt6egPVgBQ29X040JN1LJlS61bt07HH398\n6DaO2eObuzMxMTFV6hTd1cT3zDPPuJl5ly5dfO/evcWWHzx40BctWlTwesuWLV6nTh1v2rSpf/31\n1yW227VrV49EIr5gwYKCeevXr/dGjRp5nTp1fPz48Z6bm1tsvS1btvj111/vixcvLrHtWJFIxHv0\n6FGurLt79+7dPRKJlDvv7h58/6r958jExMSUqKk2HxeSXU5OjkciEd+0aVNC2x0yZEiF2q0tx7Zq\n7wATE1Ptn0o7aN5+++0eiUT8kUceKTFT1FVXXeWRSMSnT58ed/n69evdzLxDhw6F5vfu3dsjkYjf\nd999ZW7j8OHDpS7v3r27m5lHIhE3s4Ipv7BatGiRm5mPHTvW3d03btwYNx+7TklqywGHiYmJKX+q\njccFd/epU6e6mfm0adP8lVde8QsuuMDr16/vGRkZPmjQIP/ggw+KrTNkyBA3M9+wYYNPnDjRO3Xq\n5Onp6cWODfPmzfMrrrjCMzMzvV69et6+fXsfOXKk7969O25f/va3v3nXrl29fv36fsIJJ3j//v19\n/fr1cYup/GPU0KFDi7Vz4MABf+CBB/yss87yhg0beoMGDfy0007z4cOH+6effuruXuy4lj+1a9eu\nxO9VbTm2McwPQLVq0qSJ3F3vv/9+udcZNmyYZs+erSeffFLXXHNNseWTJ08uGPqRb+PGjVqwYIHS\n09M1cuTIMrdRt27dUpcPHTpUPXr00JgxY9S2bVvl5ORIktq2bRs337hxY40ZM0ZTpkzR5s2bNWbM\nmPwTihLXAYBjUU09LuQzM82YMUNz587V1VdfrR49emj16tWaMWOGFi1apLfeeksdOnQolDczDR8+\nXEuXLlV2drays7OVkpJSkBk7dqzGjh2rJk2a6KqrrlLTpk21Zs0aTZgwQXPnztXbb7+tBg0aFOT/\n8pe/aPDgwapXr54GDx6srKwsLV26VOeff746depUrvchSbt371b37t21Zs0adezYUbfccotSU1P1\n4YcfaurUqRo4cKBOPPFEjRkzRi+99JLWrFmjO++8U40bN5akgn9rtequ5piYmGr/pFL+Arlq1SpP\nTU31SCTiN910k7/44otlDhHIzc31li1bekpKim/cuLHQssOHD3vTpk09NTXVd+zYUTD/6aefdjPz\nbt26ldp2RZV0Zanolal8DPNjYmJiqr3HhfwrU5FIxOfMmVNo2cSJE93MvHfv3oXm5+TkuJl5q1at\n4r7PhQsXupl5165dfc+ePYWWTZs2zc3M77777oJ5+/bt8xNOOMFTU1N95cqVhfJ33313Qf/Kc2Xq\n+uuv90gk4nfccUexfu3fv79Qfyo6fLC2HNt4pBSAatW5c2c9++yzysrK0rPPPquBAweqbdu2yszM\n1NVXX61XXnml2Dr5j2/Ny8vTU089VWjZzJkztXPnTvXt21dNmzYtmL9t2zZJUqtWrSr3DQEAjkpt\nOC706tVLV1xxRaF5d9xxh9q3b6+FCxdqy5YthZaZme655x61bt26WFsTJ06UmWnSpElq2LBhoWXf\n//73C75f+WbNmqVdu3bphhtu0BlnnFEoP3r06HI/ZGLnzp2aPn26mjdvrvHjxxdbftxxxxXrz7GI\nYgpAtRs0aJA2b96sV199VT//+c/Vp08fubtmzZqlvn37aujQocXWufXWWxWJRDRlypT8v3JK+mYo\nx7Bhw6ryLQAAEqimHxcuuuiiYvMikYi6du0qSVq1alWx5WeffXbctpYtW6a6detq+vTpBcP9YqfD\nhw9r586d2rVrlyRp5cqVMrO4fWjUqFG5n0j4zjvvKC8vTxdddJHS09PLtc6xiHumACSFlJQU9e7d\nW71795YUHYI8Y8YMDR06VH/84x81YMAA9e3btyDfunVrXXLJJZo/f77mzJmj7Oxsbdq0SQsWLFCb\nNm106aWXFmq/efPmkqRPPvmk6t4UACC0mnxcaNasWdz5+Z9D+OWXX5a4rKjPP/9cubm5pX4IvJlp\n3759ysjIKGi7rD6UZffu3ZKij01HybgyBSApmZkGDRqku+66S+6uhQsXFsvcdtttcndNnjxZkvTk\nk09Kiv51sqj8vwa+++672rt3byX2HABQGWrScWHHjh1x52/fvl2S4g61K+kzmo4//nhlZGQoNze3\nxOnIkSM66aSTCrVdVh/Kkv/wCP4IWTqKKQBJLX88duyQjXx9+/ZVVlaW5syZo48//lhTpkxRSkpK\n3OEfbdu2Ve/evXXo0KG4Y7+LOnz48NF3Po78pzPFez8AgLLVhOPC4sWLi83Ly8vT0qVLJanYvUyl\nOe+887Rr1y6tW7euXPkuXbrI3eP2Yc+ePVq9enW52jnnnHMUiUT0xhtv6ODBg2Xm849vubm55Wq/\ntqCYAlCtnn/+eb322mtxD4rbt2/XpEmTShz7nZKSopycHB05ckQ33HCDtm7dquzs7IKhG0VNnDhR\nDRs21K9//Ws9+OCDcXf4mzdv1uDBg7Vs2bKjf3NxNGnSpGA7AIDiasNxYeHChZo9e3aheY8++qg+\n/PBD9ezZs+AqUnnkX4kbNmxYwUMzYh04cEDLly8veN2vXz9lZGToueee04oVKwplR48eHXeIYTyZ\nmZkaPHiwtm7dqhEjRhT7eezfv1979uwpeH2sHt+4ZwpAtVq+fLkeeeQRZWVlqWvXrmrXrp0kacOG\nDZo9e7YOHTqk/v37a+DAgXHXHzZsmMaNG6clS5bIzHTbbbeVuK2OHTtq/vz5GjRokEaOHKlHHnlE\nvXr1UosWLbR//3699957evPNNxWJRDRq1KhKeb+9evXSCy+8oAEDBujKK69Uenq62rRpoxtvvLFS\ntgcANU1tOC706dNHAwYM0IABA3TKKado1apVmjdvnjIzM/X4449X6PvRs2dPjRs3TqNGjVKHDh10\n5ZVXql27dtq3b582bdqkxYsXq1u3bpozZ44kqX79+po0aZIGDx6sbt266brrrlPz5s21dOlSrV27\nVhdddJGWLFlSrm0/9thjWrt2rZ544gm9/vrruuyyy5SamqqPPvpI8+fP18v/v717D5OiPNM/fj/N\nCI2DOApyXM6Ji3hCkIhAAqIiP1kFY1ATiU5YN9Fo1mhYjUs2iyZrRFzFGI0mK4IhbhKjjkckwYTI\nFYUgeMqqwQMHRUCRgCgHhXl+f1TN0NPTc+ia6uli5vu5rrqGrr6r3neYnnrrmX6r+tFHq4vak08+\nWbNmzdJFF12ks88+WwcddJDKysp06aWX5vX97neKfW92FhaWlr+ons8Teeedd/yOO+7wL37xiz5w\n4EA/+OCDvV27dt6jRw+fMGGC33fffXVuW2XcuHGeSqW8T58+XllZ2WD+448/9tmzZ/vYsWO9a9eu\n3rZtWy8rK/Pjjz/ep0+fXuszSuqTSqV87NixtdYvXrzYU6mUX3fddTXW792716dPn+4DBgyo/hyV\nXHAFm6UAABx9SURBVJ9TlUkt5LM4WFhYWKqWljouzJ0711OplM+bN88ff/xxHzFihHfo0MEPOeQQ\nnzx5sr/++uu1tikvL/c2bdo0+PlMf/7zn/3cc8/1nj17ert27bxLly5+3HHH+bRp03zFihW18osW\nLfLPf/7zXlpa6oceeqifddZZ/re//S1ne2vWrPFUKuVTp06ttZ8dO3b49ddf78cee6yXlpZ6x44d\n/cgjj/Qrr7zS33///RrZW265xQcNGuTpdNpTqZT369evzu+npYxtFnwvAFA4ZuYca6IzM7l77iuT\nAWA/1FLHhXnz5mnq1Km65557dMEFFxS7O4nWUsY2rpkCAAAAgAgopgAAAICYtMR33FA3iikAAAAg\nJnV9XhRaJq6ZAlBwLXVufHNpKfPKAaAK4wJaytjGO1MAAAAAEAHFFAAAAABEQDEFAAAAABFQTAEA\nAABABBRTAAAAABBBSbE7AKDlS6fTm8ysa7H7sb9Kp9Obit0HAIgT4wJaytjGrdEBtGpmdoak/u5+\na7H7AgBo+cyst6RLJM129xZRULRmTPMD0NrNljTbzLoUuyMAgFbhnyV9V9JXit0RNB3FFIDWrl34\ntW1RewEAaC3aZX3FfoxiCgAAAAAioJgCAAAAgAgopgAAAAAgAoopAAAAAIiAYgoAAAAAIqCYAgAA\nAIAIKKYAAAAAIAKKKQAAAACIgGIKAAAAACKgmAIAAACACCimAAAAACACiikAAAAAiIBiCgAAAAAi\noJgCAAAAgAgopgAAAAAgAoopAAAAAIiAYgoAAAAAIqCYAgAAAIAIKKYAAAAAIIKSQuzUzI6UdGoh\n9g0AMesZfv22mb1T1J4AQMNedvenmrIDM+so6VxJpfF0CXn6dtVXM9tV1J60Xh9K+pW772jqjszd\nY+hP1k7N/irpyNh3DAAA0LrtldTZ3bdG3YGZfVfSj+LrErBfusTd72zqTgryzpSksvDrHEnbC9QG\nAMRhjKQ+kuYVuR8A0JCLFLyb1EFS5GJK+87TnpG0vKmdQt66KXhn8BeSthS5L63RKElDte/3oEkK\nVUxV+U93Z9oMAABAE5nZlxTv1LxH3f2GGPeHxjuv2B1orczsBgXFVCy4AQUAAAAARFCoYqrqYi4u\nqgMAAIhHXOdXO7K+Aq1JrK//Qk3z+7qkPu6+uUD7BwAAaG3iOr/6mYKCbH7TuwTsd2J9/Rfkbn4A\nAAAA0NJxzRQAAAAAREAxBQAAAAARUEwBAAAAQAQUUwAAAAAQAcUUAAAAAERAMQUAAAAAEVBMAQAA\nAEAEFFMAAAAAEAHFFAAAAABEQDEFAAAAABFQTAEAAABABBRTAAAAABABxRQAAAAAREAxBQAAAAAR\nUEwBAAAAQAQUUwAAAAAQAcUUAAAAAERAMQUAAAAAEVBMAQAAAEAEFFMAAAAAEAHFFAAAAABEQDEF\nAAAAABGUFLsDAPZv7du337hr166uxe4HpHQ6vWnnzp3dit0PAABaC3P3YvcBwH7MzJzjSDKYmdzd\nit0PAABaC6b5AQAAAEAEFFMAAAAAEAHFFAAAAABEQDEFAAAAABFQTAFo8ebNm6dUKqV77723qP1Y\nsWKFTj31VB122GFKpVIaMmSIJKm8vFypVErr1q0rav8AAEB+uDU6gFbBrLg3udu+fbtOP/10ffLJ\nJ/rqV7+qzp07q3v37tV9y+7f2rVr1a9fP5WXl2vOnDnF6DIAAGgAxRQANIO//OUvev/99/WjH/1I\nV199dY3nbrjhBl1zzTXq2bNnkXoHAACioJgCgGawfv16mVn1u1GZunbtqq5da37uMZ/dBQBA8nHN\nFICCeuSRR3TyySerR48eSqfT6tmzp8aMGaOf/vSntbJ///vfNX36dB199NEqLS1VWVmZBg8erGuu\nuUY7d+6szq1cuVKXX365Bg8erE6dOql9+/Y6/PDDNW3aNG3dujWv/q1fv16XXXaZBgwYoHQ6rc6d\nO2vixIl67rnnamVnzJihVCqlp59+Wvfdd5+GDx+ugw46SP37969z/2vXrlUqlVJ5ebncvfr6qMxr\nuLKvmbr22mvVv39/mZnmzp1bnU/CdV8AAGAf3pkCUDA/+9nPdPHFF6t79+4688wz1blzZ7333nt6\n6aWXNHfuXF1yySXV2TVr1mjMmDF6++23NXToUH3zm99UZWWlVq1apdmzZ+uSSy5R7969JUk///nP\nVVFRodGjR+vUU09VZWWlVqxYoZtvvllPPvmkli1bptLS0gb7t3LlSo0bN05bt27VaaedprPPPlub\nN29WRUWFRo0apYqKCo0fP746X3Vt00033aRFixbpjDPO0NixY7Vt27Y62ygrK9OMGTP0wgsv6OGH\nH9bEiRM1ePBgSar+mn3N1EknnaRt27Zp9uzZGjx4sCZNmlT9XNU2AAAgAdydhYWFJfISHEZyGzp0\nqKfTad+8eXOt5z744IMaj0888URPpVI+c+bMnNndu3dXP163bp1XVlbWys2ZM8fNzG+88cYa6+fO\nneupVMrnzZtXvW7Pnj0+YMAAb9++vS9ZsqRGfsOGDd6zZ0/v0aOHf/LJJ9XrZ8yY4WbmHTp08Bdf\nfLHO7zuXXH2oUl5e7qlUyteuXVu9bs2aNW5m/rWvfa3RbYQ/i6K/JlhYWFhYWFrLwjQ/AAVVUlKi\nNm3a1Fp/6KGHVv975cqVWrp0qY477jhdddVVObNt27atftyrV6+cd+crLy9Xx44dtXDhwgb79fjj\nj+utt97St771LY0aNarGc926ddNVV12ljRs36qmnnqq17Te+8Q0dc8wxDbYBAABaNqb5ASiY888/\nX9OmTdOgQYN03nnnafTo0Ro5cqQ6d+5cI7d06VJJ0rhx4xq13z179ujOO+/Ur3/9a73yyivatm2b\nKisrq59fv359g/t49tlnJQXTC6+99tpaz7/++utyd7366qu1pvoNGzasUf0EAAAtG8UUgIK54oor\ndNhhh+mOO+7QbbfdpltvvVWSNHr0aM2aNUtDhw6VJG3dulVm1uhbg59zzjmqqKjQgAEDNGnSJHXr\n1k3t2rWTJN1yyy3avXt3g/v44IMPJEm//e1v68yYmT766KNa67t169aofgIAgJaNYgpAQU2ZMkVT\npkzRhx9+qGeeeUYPPfSQ7r77bo0fP16vvfaaOnXqpLKyMrl7o95RWrFihSoqKjRu3Dg98cQTSqX2\nzVZ2d82cObNR/Tr44INlZnrkkUc0YcKEvL6nYn8AMAAASAaumQLQLDp27Kjx48frrrvuUnl5ubZs\n2aKnn35akjR8+HBJatS1Tm+88YYk6YwzzqhRSEnSsmXLatxCvT7Dhw+Xu1f3IWmqrjPbu3dvkXsC\nAADqQjEFoGAWL16cc/2mTZskSQceeKAkaciQIRoxYoReeOEF3XDDDbXyW7ZsqZ6617dv35z7fu+9\n93TZZZc1um8TJ07UgAEDdPvtt2vBggU5M0uXLtWuXbsavc84HXLIITKz6s+eAgAAycM0PwAFc9ZZ\nZ6lDhw4aPny4+vbtK3fXkiVLtHz5cg0bNkynnHJKdXb+/Pk66aSTNH36dD3wwAMaM2aM3F2rVq3S\nokWL9Nprr6l3794aNmyYRo4cqQcffFAjR47UqFGjtGnTJi1YsEADBw5Ujx49cvbF3Ws8Likp0YMP\nPqjx48drwoQJGjFihAYPHqwDDzxQb7/9tpYvX67Vq1drw4YNSqfTsfx/ZPehPqWlpTrhhBO0ZMkS\nTZkyRYcffrjatGmjiRMn6qijjoqlPwAAoGkopgAUzMyZM7Vw4UI9//zzWrBggdLptPr06aNZs2bp\n4osvrnHL9L59+2rlypW68cYbVVFRodtvv13pdFp9+/bVtGnT1KVLF0lSKpXSo48+qu9973t64okn\ndNttt6lnz576+te/runTp+uII47IeU1TrnVHH320XnzxRd1888167LHHNHfuXKVSKXXv3l1DhgzR\nD37wg1p3HmyK+q61yvXc/PnzdcUVV2jhwoX61a9+JXdXr169KKYAAEgIy+cvpQCQzcyc40gymJnc\nnbtjAADQTLhmCgAAAAAioJgCAAAAgAgopgAAAAAgAoopAAAAAIiAYgoAAAAAIqCYAgAAAIAIKKYA\nAAAAIAKKKQAAAACIgGIKAAAAACIoKXYHAOzf0un0JjPrWux+IPhZFLsPAAC0Jubuxe4DABSNmR0k\nqYO7byh2XwAAwP6FaX4AWruFklaZWWmxOwIAAPYvFFMAWrvekjpIOqTYHQEAAPsXiikAAAAAiIBi\nCgAAAAAioJgCAAAAgAgopgAAAAAgAoopAAAAAIiAYgoAAAAAIqCYAgAAAIAISgqxUzObIunLhdg3\nAMSsZ/j1cTN7p6g9AYCGrXT3/yh2JwAEzN3j36nZGkl9Yt8xAAAAurv7xmJ3AkCB3pnK2G+5pM0F\nagMA4tBJUpmkN4vdEQBowL2SDlXhzt8A5KnQv4xPuTvTZgAAAJrIzHYWuw8AauIGFAAAAAAQQaGK\nqXcl7ZS0tUD7BwAAaG04vwISplA3oOgkqczduQYBAAAgBpxfAclTkGIKAAAAAFo6rpkCAAAAgAgo\npgAAAAAgAoopAAAAAIiAYgoAAAAAIqCYAgAAAIAIKKYAAAAAIAKKKQAAAACIgGIKAAAAACKgmAIA\nAACACCimAAAAACACiikAAAAAiIBiCgAAAAAioJgCAAAAgAgopgAAAAAgAoopAAAAAIiAYgoAAAAA\nIijJJ9y+ffuNu3bt6lqozgAAgHik0+n3du7cWWPMPuCAAzbu2bOHcbyFKikp2fTpp592q3rMzxto\nnOzfnXyYuzc+bOadO3fW/fffrzFjxjSYX7x4sSZPnkyePHny5MmTb+b8SSedJHe3zPVm5jNmzKiV\nX716te6//35NnjxZ/fr1a3D/5JOZnzFjRo2feV0/bwA1Zf/u5CPvaX5JGijIkydPnjx58rnzjbW/\nFArko+UBFFbexVSSBgry5MmTJ0+efO58YyTtxJ98vHkAhRf7DSiSPLCQJ0+ePHny5ANJO/EnH28e\nQPOItZhK2kBBnjx58uTJk68taSf+5OPNA2g+sRVTSRsoyJMnT548efK1Je3En3y8eQDNK5ZiKmkD\nBXny5MmTJ08+tySd+JOPNw+g+TW5mEraQEGePHny5MmTr1tSTvzJx58H0PyaVEwlbaAgT548efLk\nydcvKSf+5OPPA2h+kYuppA0U5MmTJ0+ePPmmS2qhQL7hPIDmF6mYStpAQZ48efLkyZNvuiQXCuSb\nngcQv7yLqaQNFOTJkydPnjz5pkvaiT/5ePMACiPvYipJAwV58uTJkydPPnc+H0k78Scfbx5A4eRd\nTCVpoCBPnjx58uTJ5843VtJO/MnHmwdQWHkXU0kaKMiTJ0+ePHnyufONkbQTf/Lx5gEUXt7FVEOS\nPLCQJ0+ePHny5ANJO/EnH28eQPOItZhK2kBBnjx58uTJk68taSf+5OPNA2g+sRVTSRsoyJMnT548\nefK1Je3En3y8eQDNK5ZiKmkDBXny5MmTJ08+tySd+JOPNw+g+TW5mEraQEGePHny5MmTr1tSTvzJ\nx58H0PyaVEwlbaAgT548efLkydcvKSf+5OPPA2h+kYuppA0U5MmTJ0+ePPmmS2qhQL7hPIDmF6mY\nStpAQZ48efLkyZNvuiQXCuSbngcQv7yLqaQNFOTJkydPnjz5pkvaiT/5ePMACiPvYipJAwV58uTJ\nkydPPnc+H0k78Scfbx5A4eRdTCVpoCBPnjx58uTJ5843VtJO/MnHmwdQWHkXU0kaKMiTJ0+ePHny\nufONkbQTf/Lx5gEUXt7FVEOSPLCQJ0+ePHny5ANJO/EnH28eQPOItZhK2kBBnjx58uTJk68taSf+\n5OPNA2g+sRVTSRsoyJMnT548efK1Je3En3y8eQDNK5ZiKmkDBXny5MmTJ08+tySd+JOPNw+g+TW5\nmEraQEGePHny5MmTr1tSTvzJx58H0PyaVEwlbaAgT548efLkydcvKSf+5OPPA2h+5u6NDrdv337j\nrl27uhawPwAAIAbpdHrTzp07u2WuO+CAAzbu2bOHcbyFKikp2fTpp59W/8z5eQONk/27k4+8iikA\nAAAAQCCWG1AAKDwz+6OZ/TjPbVab2ZWF6lNGO2Zmd5nZZjPba2ZfaMQ2fcys0syG1JMZGmZ6x9tj\nAM0p+/jVmOOZmb1sZt+Pu+1CMbOuZvY7M/vIzPY2cpsLzWx7A5nvmBkXRAEJVVLsDgAtlZn9UdLL\n7v6vMe3yLEmf5rnN8ZI+jqn9+pwu6UJJoyWtlrSlkds15q1x3j4HWp4ox7N6mdmFkn7i7gcVuq06\nTJPUTdIxkj7KYzuOgwlTgPFbZjZa0h8ldXb3xo6RiW0H+1BMAUVmZiXuvqehnLtvzXff7v5BtF7l\n7bOSNrj7sjy3s0J0BkCyRTmeNYIpR9FRoLZy+YykFe7+VjO1h/1L1euz0ONec7WDENP8gAIws3sU\nvEtzaThNba+Z9Taz0eHj/2dmy8xsl6RxZtbfzCrMbEM4RWSFmU3I2mf2NJnVZjbdzO40s21m9raZ\nTcvapsY0v7DtfzGz34TtvGlm52dtc0LY/k4zW25mp4Xb5Zy6F36vN0vqHebeCte3NbPZZrYx3Nez\nZjaygf+38Wb2apj/k6TDs57vaGa/MLNNYeYNM4vtL4cAagqPFxvNzLLW32dmFeG/Gzx+5dhv9vHs\nMDN72Mx2hMetr+XY5gozezFs4x0z+7mZdQyfGy1pjqTSjGPu9+toq8zM5pnZlrC935vZoIznLzSz\n7WY2Npxq+JGZ/cHM+tTz/ayWdKakC8O254Tre5nZQ2b2Ybg8YGY9G/i/uSr8v/zQzOZK6lBfHvGq\na/wOnxtkZo+FP5tN4e9B14xtjzKzReGYvN3Mng/H/T6S/hDG3s98jeRov8TMfmxm681sl5mtNbPr\nM54/wMxmhmP+xxacS4wLn2t0O4gPxRRQGJdLelbSPZK6Suou6e2M52+QNF3SQEnLFAyWT0g6WcEU\nkd9KesDMahQTOXxb0kuSjpM0U9KNZnZCA9v8h6SHwnZ+LWmOmf2DJJlZqaRHJb0S7vO7km5S/VNM\n/lXSdZLeCb/XYeH6WZImSyqXNFjSy5KezBx4MplZr7BfCyUdK+k2STdmxf5L0pEKphUeLmmqpPUN\nfL8AortfUkdJp1atCI8TZ0r6Rbgq6vEr0zxJ/SWNlTRJ0gWSsouXvQqOrYMkfVnBsea28LlnFBwP\nd2jfMfemetoaJumM8OsOBcemdhmZdgqOf+WShksqk3RnPf0/XtJTCo6p3SRdHhagj0g6TMHJ+RhJ\nPRQc53Iys3Mk/UDBcXqIpFWSCn7dK2rIOX6bWTdJf1Iw5h6v4PVeKunhjG3vk/Ru+PyxkmZI2iVp\nnaSzw8wR4T4vr6f9iZLOUfBu57mS/pbx/FxJn5d0noLxcJ6kR8zs6DzbQVzcnYWFpQCLgjnLP85a\nN1pSpaRJjdj+WUn/Xtf+FFyb9MusbVZlbbNa0pUZjysl/TDjcRsF11R9JXz8DUmbJbXLyHxZwUnM\nF+rp63ckvZXx+EBJuyWdn7EuJekNSdeFj/uE/RkSPr5e0mtZ+50ett07fPywpP8p9s+WhaU1LZIe\nkDQv4/EUSX+X1LaebRo6flU/VvCHkUpJwzOe7y1pj6Tv19PGaZJ2Zjy+UNKHOXKZbX02bGtkxvMd\nJW2VNDVjP3slfSYj85XMturoz6OS5mQ8PlXBtVq9Mtb1C/c9NlefJf1Z0p1Z+/195vGVpVle87nG\n72sl/T5r3SHh6+n48PE2SV+tY5+jw5/9oQ20fWt2OxnP9Q/38Q9Z6x9ScL1go9thiW/hnSmg+bmk\nFZkrzOxAM7vRzP4vnHqyXdJQBScU9Xkp6/G7kro0sM3L1R1x3yvp/Yxt/lHSX919d0Z+mfKfez1A\nwTWZz2S0VangBGtQHdsMlLQ0a92zWY9/Kuk8M3vBzGZZI+4aCKDJ5kuaZGbp8PFXJD3g7p9ITTp+\nVRmo4ORvedUKd1+n4HhWLZx297twetOHkh6U1DZ8x6CxqtqqPta4+4cKjouZx6bd7v5GxuN3w7bK\n8mzrXXevnpXg7qvDfdV1HDxCDR8HURxDJY0Op+9tD1/n6xSM6QPCzM2S7jazp8zs383sHyO0M1fS\ncWa2ysx+Ymanh+9ySsG7lSbplax+nJ7RBzQziimgOLLvsPffCt6any7pCwqmByyX1LaB/WTfocrV\n8O91lG3iFPmuVO7+pIITtFmSOkl63MzujqtjAHJ6XEEBMtHMDpN0ivZN8ZOiH7+y1XlsCK9ZeUzS\n/0n6koKTyqnh0/m205j2s28KVPVcXMdK7s63/0kpeA0eo+A1XrV8Nlwvd79WQUH8kKQRkl4ys/J8\nGnH35xXM3PiugsJpnqTfZfShUvumEVYtR2jf7wOaGcUUUDifKJhG1xgjJd3r7hXu/lcFf7ksxl+Z\nXpN0VNa1Ayco/4H/TQVFW/UNJ8wsJelEBddj5fJq2FamE7ND7r7F3X/p7lMl/bOCC74PyLN/ABop\nfAfqfgXT+85VcOfOP2VEmnr8ek3B+cjnqlaExVOPjMzxkg5QMG15WfiuUfaNHBpzzH01bKv62BLe\nxOJoBYVanF6V1MMyPifPzPor+L7qautVBddoZap1HETB5XotrVRwjdI6d38ra6n+A6m7v+nuP3H3\nf5J0t6SLMvapHPutxd0/dvcH3f1SSRMknWxmn5H0vIICq3uOPmzItx3Eg2IKKJw1kj5nwYfTdsp4\nmz7XlLlVks4ys+PCi0h/oeAC6OZ2n4K/ev2PmR1hZqdIuiZ8rtEFlbvvUDAlb6YFdy4cqODi7S6S\n7qhjszsl9bXgDoCHm9mXFFzDVc3MrjWziWb2GTM7QsFfw9909+b4DBmgNZuv4BqliyX9b9ZzTTp+\nufsqBTeeucvMhpvZYAUX/+/IiL2u4JzlCjPra2ZfVu0L69dISpvZKeExt32Ott5QcFOIu8xsVNjf\n+Qqudcn+vrLlNd3Z3RcpmD74Sws+gPz4sK3n3H1xHZvdquAPRBeFx7lrlFFkotmsUcb4Ha67XdLB\nkn5jZp8zs37ha+0uMys1s3Q4LW90uN0JkkZpX+G8VsE4OsHMOoc3cqnFgrtWnmdmA8MC6nwFr893\n3P11BeP0XDM7O+zDUAs+2HlSPu0gPhRTQOHcpOAvRK9Iek9Sr3B9rqLkyjDztIIpNc9KWpKVyd4u\n134aytS7jbt/JOmfFMznX6ngDoH/qeAkYleObetztcK7BSr4a9pRkk5z9011tP22pC8qOGF7QcGJ\n0tVZ+9wt6Yfh80sU3EnpzDz7BSBP7r5EwZ0zByooCDLFcfy6UMENc55ScKOZXyo4oa1q/2UFx4Qr\nFJycTlVw45vMPj6r4I8y/xv259/q+HbKJf0lbGepgsJvfNa1orlEmZp3poLrUv+g4Ht7V8GHCOdu\nwP03Cu4A90PteyfkvyO0i6apMX6bWe/wnZ+RCqa8LpD0VwV3k9ylYGzaq+CGFPcoeLf1AQU3FPmO\nJLn7uwrG0/+StFH77kSZbbuC1+4ySc8pmFY43t2rxuDysI2ZCt7JfFTB3f3W5tkOYmLuTNsFUDcz\nm6hgUOjifJo6AABAtZJidwBAspjZBZLeUvC5WEdLukXSIxRSAAAANVFMAcjWVcHnaXRTMEXgMQV3\nFQIAAEAGpvkBAAAAQATcgAIAAAAAIqCYAgAAAIAIKKYAAAAAIAKKKQAAAACIgGIKAAAAACKgmAIA\nAACACP4/49xAuzumVSwAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7ff084c73c18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mglearn.plots.plot_proper_processing()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "rnd = np.random.RandomState(seed=0)\n",
    "X = rnd.normal(size=(100, 10000))\n",
    "y = rnd.normal(size=(100,))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_selected.shape: (100, 500)\n"
     ]
    }
   ],
   "source": [
    "from sklearn.feature_selection import SelectPercentile, f_regression\n",
    "\n",
    "select = SelectPercentile(score_func=f_regression, percentile=5).fit(X, y)\n",
    "X_selected = select.transform(X)\n",
    "print(\"X_selected.shape: {}\".format(X_selected.shape))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cross-validation accuracy (cv only on ridge): 0.91\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.linear_model import Ridge\n",
    "print(\"Cross-validation accuracy (cv only on ridge): {:.2f}\".format(\n",
    "      np.mean(cross_val_score(Ridge(), X_selected, y, cv=5))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cross-validation accuracy (pipeline): -0.25\n"
     ]
    }
   ],
   "source": [
    "pipe = Pipeline([(\"select\", SelectPercentile(score_func=f_regression,\n",
    "                                             percentile=5)),\n",
    "                 (\"ridge\", Ridge())])\n",
    "print(\"Cross-validation accuracy (pipeline): {:.2f}\".format(\n",
    "      np.mean(cross_val_score(pipe, X, y, cv=5))))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### The General Pipeline Interface"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def fit(self, X, y):\n",
    "    X_transformed = X\n",
    "    for name, estimator in self.steps[:-1]:\n",
    "        # iterate over all but the final step\n",
    "        # fit and transform the data\n",
    "        X_transformed = estimator.fit_transform(X_transformed, y)\n",
    "    # fit the last step\n",
    "    self.steps[-1][1].fit(X_transformed, y)\n",
    "    return self"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def predict(self, X):\n",
    "    X_transformed = X\n",
    "    for step in self.steps[:-1]:\n",
    "        # iterate over all but the final step\n",
    "        # transform the data\n",
    "        X_transformed = step[1].transform(X_transformed)\n",
    "    # fit the last step\n",
    "    return self.steps[-1][1].predict(X_transformed)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![pipeline_illustration](images/pipeline.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Convenient Pipeline creation with ``make_pipeline``"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.pipeline import make_pipeline\n",
    "# standard syntax\n",
    "pipe_long = Pipeline([(\"scaler\", MinMaxScaler()), (\"svm\", SVC(C=100))])\n",
    "# abbreviated syntax\n",
    "pipe_short = make_pipeline(MinMaxScaler(), SVC(C=100))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Pipeline steps:\n",
      "[('minmaxscaler', MinMaxScaler(copy=True, feature_range=(0, 1))), ('svc', SVC(C=100, cache_size=200, class_weight=None, coef0=0.0,\n",
      "  decision_function_shape=None, degree=3, gamma='auto', kernel='rbf',\n",
      "  max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
      "  tol=0.001, verbose=False))]\n"
     ]
    }
   ],
   "source": [
    "print(\"Pipeline steps:\\n{}\".format(pipe_short.steps))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Pipeline steps:\n",
      "[('standardscaler-1', StandardScaler(copy=True, with_mean=True, with_std=True)), ('pca', PCA(copy=True, iterated_power='auto', n_components=2, random_state=None,\n",
      "  svd_solver='auto', tol=0.0, whiten=False)), ('standardscaler-2', StandardScaler(copy=True, with_mean=True, with_std=True))]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.decomposition import PCA\n",
    "\n",
    "pipe = make_pipeline(StandardScaler(), PCA(n_components=2), StandardScaler())\n",
    "print(\"Pipeline steps:\\n{}\".format(pipe.steps))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Accessing step attributes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "components.shape: (2, 30)\n"
     ]
    }
   ],
   "source": [
    "# fit the pipeline defined above to the cancer dataset\n",
    "pipe.fit(cancer.data)\n",
    "# extract the first two principal components from the \"pca\" step\n",
    "components = pipe.named_steps[\"pca\"].components_\n",
    "print(\"components.shape: {}\".format(components.shape))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Accessing attributes in grid-searched pipeline."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "pipe = make_pipeline(StandardScaler(), LogisticRegression())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "param_grid = {'logisticregression__C': [0.01, 0.1, 1, 10, 100]}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=5, error_score='raise',\n",
       "       estimator=Pipeline(steps=[('standardscaler', StandardScaler(copy=True, with_mean=True, with_std=True)), ('logisticregression', LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
       "          intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,\n",
       "          penalty='l2', random_state=None, solver='liblinear', tol=0.0001,\n",
       "          verbose=0, warm_start=False))]),\n",
       "       fit_params={}, iid=True, n_jobs=1,\n",
       "       param_grid={'logisticregression__C': [0.01, 0.1, 1, 10, 100]},\n",
       "       pre_dispatch='2*n_jobs', refit=True, return_train_score=True,\n",
       "       scoring=None, verbose=0)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(\n",
    "    cancer.data, cancer.target, random_state=4)\n",
    "grid = GridSearchCV(pipe, param_grid, cv=5)\n",
    "grid.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best estimator:\n",
      "Pipeline(steps=[('standardscaler', StandardScaler(copy=True, with_mean=True, with_std=True)), ('logisticregression', LogisticRegression(C=0.1, class_weight=None, dual=False, fit_intercept=True,\n",
      "          intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,\n",
      "          penalty='l2', random_state=None, solver='liblinear', tol=0.0001,\n",
      "          verbose=0, warm_start=False))])\n"
     ]
    }
   ],
   "source": [
    "print(\"Best estimator:\\n{}\".format(grid.best_estimator_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Logistic regression step:\n",
      "LogisticRegression(C=0.1, class_weight=None, dual=False, fit_intercept=True,\n",
      "          intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,\n",
      "          penalty='l2', random_state=None, solver='liblinear', tol=0.0001,\n",
      "          verbose=0, warm_start=False)\n"
     ]
    }
   ],
   "source": [
    "print(\"Logistic regression step:\\n{}\".format(\n",
    "      grid.best_estimator_.named_steps[\"logisticregression\"]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Logistic regression coefficients:\n",
      "[[-0.389 -0.375 -0.376 -0.396 -0.115  0.017 -0.355 -0.39  -0.058  0.209\n",
      "  -0.495 -0.004 -0.371 -0.383 -0.045  0.198  0.004 -0.049  0.21   0.224\n",
      "  -0.547 -0.525 -0.499 -0.515 -0.393 -0.123 -0.388 -0.417 -0.325 -0.139]]\n"
     ]
    }
   ],
   "source": [
    "print(\"Logistic regression coefficients:\\n{}\".format(\n",
    "      grid.best_estimator_.named_steps[\"logisticregression\"].coef_))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Grid-searching preprocessing steps and model parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_boston\n",
    "boston = load_boston()\n",
    "X_train, X_test, y_train, y_test = train_test_split(boston.data, boston.target,\n",
    "                                                    random_state=0)\n",
    "\n",
    "from sklearn.preprocessing import PolynomialFeatures\n",
    "pipe = make_pipeline(\n",
    "    StandardScaler(),\n",
    "    PolynomialFeatures(),\n",
    "    Ridge())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "param_grid = {'polynomialfeatures__degree': [1, 2, 3],\n",
    "              'ridge__alpha': [0.001, 0.01, 0.1, 1, 10, 100]}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=5, error_score='raise',\n",
       "       estimator=Pipeline(steps=[('standardscaler', StandardScaler(copy=True, with_mean=True, with_std=True)), ('polynomialfeatures', PolynomialFeatures(degree=2, include_bias=True, interaction_only=False)), ('ridge', Ridge(alpha=1.0, copy_X=True, fit_intercept=True, max_iter=None,\n",
       "   normalize=False, random_state=None, solver='auto', tol=0.001))]),\n",
       "       fit_params={}, iid=True, n_jobs=-1,\n",
       "       param_grid={'polynomialfeatures__degree': [1, 2, 3], 'ridge__alpha': [0.001, 0.01, 0.1, 1, 10, 100]},\n",
       "       pre_dispatch='2*n_jobs', refit=True, return_train_score=True,\n",
       "       scoring=None, verbose=0)"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid = GridSearchCV(pipe, param_grid=param_grid, cv=5, n_jobs=-1)\n",
    "grid.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.colorbar.Colorbar at 0x7ff084c737b8>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAcMAAADyCAYAAAAx165iAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nX9m78CxjwAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7ff084c3e978>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.matshow(grid.cv_results_['mean_test_score'].reshape(3, -1),\n",
    "            vmin=0, cmap=\"viridis\")\n",
    "plt.xlabel(\"ridge__alpha\")\n",
    "plt.ylabel(\"polynomialfeatures__degree\")\n",
    "plt.xticks(range(len(param_grid['ridge__alpha'])), param_grid['ridge__alpha'])\n",
    "plt.yticks(range(len(param_grid['polynomialfeatures__degree'])),\n",
    "           param_grid['polynomialfeatures__degree'])\n",
    "\n",
    "plt.colorbar()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best parameters: {'polynomialfeatures__degree': 2, 'ridge__alpha': 10}\n"
     ]
    }
   ],
   "source": [
    "print(\"Best parameters: {}\".format(grid.best_params_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test-set score: 0.77\n"
     ]
    }
   ],
   "source": [
    "print(\"Test-set score: {:.2f}\".format(grid.score(X_test, y_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Score without poly features: 0.63\n"
     ]
    }
   ],
   "source": [
    "param_grid = {'ridge__alpha': [0.001, 0.01, 0.1, 1, 10, 100]}\n",
    "pipe = make_pipeline(StandardScaler(), Ridge())\n",
    "grid = GridSearchCV(pipe, param_grid, cv=5)\n",
    "grid.fit(X_train, y_train)\n",
    "print(\"Score without poly features: {:.2f}\".format(grid.score(X_test, y_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "pipe = Pipeline([('preprocessing', StandardScaler()), ('classifier', SVC())])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "param_grid = [\n",
    "    {'classifier': [SVC()], 'preprocessing': [StandardScaler(), None],\n",
    "     'classifier__gamma': [0.001, 0.01, 0.1, 1, 10, 100],\n",
    "     'classifier__C': [0.001, 0.01, 0.1, 1, 10, 100]},\n",
    "    {'classifier': [RandomForestClassifier(n_estimators=100)],\n",
    "     'preprocessing': [None], 'classifier__max_features': [1, 2, 3]}]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best params:\n",
      "{'classifier': SVC(C=10, cache_size=200, class_weight=None, coef0=0.0,\n",
      "  decision_function_shape=None, degree=3, gamma=0.01, kernel='rbf',\n",
      "  max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
      "  tol=0.001, verbose=False), 'preprocessing': StandardScaler(copy=True, with_mean=True, with_std=True), 'classifier__C': 10, 'classifier__gamma': 0.01}\n",
      "\n",
      "Best cross-validation score: 0.99\n",
      "Test-set score: 0.98\n"
     ]
    }
   ],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(\n",
    "    cancer.data, cancer.target, random_state=0)\n",
    "\n",
    "grid = GridSearchCV(pipe, param_grid, cv=5)\n",
    "grid.fit(X_train, y_train)\n",
    "\n",
    "print(\"Best params:\\n{}\\n\".format(grid.best_params_))\n",
    "print(\"Best cross-validation score: {:.2f}\".format(grid.best_score_))\n",
    "print(\"Test-set score: {:.2f}\".format(grid.score(X_test, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Summary and Outlook"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [Root]",
   "language": "python",
   "name": "Python [Root]"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}