二_Random_Forest.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#  概述\n",
    "## 集成算法\n",
    ">通过在数据上构建多个模型，集成所有模型的结果，而不是单个机器学习算法。\n",
    "\n",
    "|集成算法的目标|\n",
    "| :-------|\n",
    "| 集成算法会考虑多个评估器的建模结果，汇总得到一个结果上。以此老获取比单个模型更好的回归或者分类表现。|\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "多个模型集成成为的模型叫做模型评估器(ensemble estimator), 组成模型评估器的单个模型叫做基评估器\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "常见的集成算法有：  \n",
    "+ 装袋法(bagging)\n",
    "+ 提升法(Boosting)\n",
    "+ stacking\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![](images/2_1.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 随机森林与决策树的对比"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-06T11:35:53.063613Z",
     "start_time": "2019-12-06T11:35:53.048765Z"
    }
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.datasets import load_wine"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-06T11:36:06.161727Z",
     "start_time": "2019-12-06T11:36:06.152351Z"
    }
   },
   "outputs": [],
   "source": [
    "wine = load_wine()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-06T11:36:09.546519Z",
     "start_time": "2019-12-06T11:36:09.532042Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.423e+01, 1.710e+00, 2.430e+00, ..., 1.040e+00, 3.920e+00,\n",
       "        1.065e+03],\n",
       "       [1.320e+01, 1.780e+00, 2.140e+00, ..., 1.050e+00, 3.400e+00,\n",
       "        1.050e+03],\n",
       "       [1.316e+01, 2.360e+00, 2.670e+00, ..., 1.030e+00, 3.170e+00,\n",
       "        1.185e+03],\n",
       "       ...,\n",
       "       [1.327e+01, 4.280e+00, 2.260e+00, ..., 5.900e-01, 1.560e+00,\n",
       "        8.350e+02],\n",
       "       [1.317e+01, 2.590e+00, 2.370e+00, ..., 6.000e-01, 1.620e+00,\n",
       "        8.400e+02],\n",
       "       [1.413e+01, 4.100e+00, 2.740e+00, ..., 6.100e-01, 1.600e+00,\n",
       "        5.600e+02]])"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "wine.data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-06T11:36:38.091952Z",
     "start_time": "2019-12-06T11:36:38.083111Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
       "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
       "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1,\n",
       "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
       "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
       "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2,\n",
       "       2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,\n",
       "       2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,\n",
       "       2, 2])"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "wine.target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-06T11:37:50.893510Z",
     "start_time": "2019-12-06T11:37:50.882879Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.model_selection import cross_val_score\n",
    "import matplotlib.pyplot as plt\n",
    "plt.rcParams['figure.dpi'] = 150"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-06T11:44:18.679057Z",
     "start_time": "2019-12-06T11:44:17.972694Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 900x600 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 创建决策树\n",
    "clf = DecisionTreeClassifier(random_state=0)\n",
    "clf_s = cross_val_score(clf, wine.data, wine.target, cv=10)\n",
    "\n",
    "# 创建随机森林\n",
    "rfc = RandomForestClassifier(n_estimators=100, random_state=0)\n",
    "rfc_s = cross_val_score(rfc, wine.data, wine.target, cv=10)\n",
    "\n",
    "plt.plot(range(1, 11), clf_s, '-r',label='Decision tree')\n",
    "plt.plot(range(1, 11), rfc_s, '-b', label='Random forest')\n",
    "\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-06T11:45:29.866043Z",
     "start_time": "2019-12-06T11:45:29.857599Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                       max_features=None, max_leaf_nodes=None,\n",
       "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                       min_samples_leaf=1, min_samples_split=2,\n",
       "                       min_weight_fraction_leaf=0.0, presort=False,\n",
       "                       random_state=None, splitter='best')"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfc.base_estimator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-06T11:50:33.294589Z",
     "start_time": "2019-12-06T11:50:32.837928Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random Forest:0.9671171171171171\n",
      "Decision Tree:0.860051984757867\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 900x600 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 下面代码跟上面的结果一样，只是比较好玩\n",
    "label = 'Random Forest'\n",
    "\n",
    "for model in [RandomForestClassifier(n_estimators=100), DecisionTreeClassifier()]:\n",
    "    s = cross_val_score(model, wine.data, wine.target, cv=5)\n",
    "    print(\"{}:{}\".format(label, s.mean()))\n",
    "    plt.plot(range(1, 6), s, label=label)\n",
    "    plt.legend()\n",
    "    label = 'Decision Tree'\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 随机森林与决策书在十组交叉验证下的效果对比"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-06T11:55:50.904765Z",
     "start_time": "2019-12-06T11:55:49.938760Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 900x600 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "rfc_list = []\n",
    "clf_list = []\n",
    "\n",
    "for i in range(10):\n",
    "    rfc = RandomForestClassifier(n_estimators=25, random_state=i)\n",
    "    rfc_s = cross_val_score(rfc, wine.data, wine.target, cv=5)\n",
    "    rfc_list.append(rfc_s.mean())\n",
    "    \n",
    "    clf = DecisionTreeClassifier(random_state=i)\n",
    "    clf_s = cross_val_score(clf, wine.data, wine.target, cv=5)\n",
    "    clf_list.append(clf_s.mean())\n",
    "    \n",
    "plt.plot(range(1, 11), clf_list, '-r',label='Decision tree')\n",
    "plt.plot(range(1, 11), rfc_list, '-b', label='Random forest')\n",
    "\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> 可以看出，随机森林一般都要比决策树好。单个决策树的波动与随机森林大多一致，因为随机森林是在决策树的基础上构建的。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 重要参数"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## n_estimators 的学习曲线"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-06T12:38:37.865657Z",
     "start_time": "2019-12-06T12:09:23.681857Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "max score: 0.9888888888888889, index：98\n"
     ]
    },
    {
     "ename": "TypeError",
     "evalue": "range() takes no keyword arguments",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-27-e026befe1310>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m     12\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfigure\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfigsize\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m12\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m5\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     13\u001b[0m plt.plot(range(1, times+1, score_list, '-b',\n\u001b[0;32m---> 14\u001b[0;31m               label='difference of n_estimator\\'s performance'))\n\u001b[0m\u001b[1;32m     15\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlegend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     16\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshow\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mTypeError\u001b[0m: range() takes no keyword arguments"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 1800x750 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# receiving the return of cross_val_score mean\n",
    "score_list = list()  \n",
    "\n",
    "times = 200\n",
    "for t in range(1, times+1):\n",
    "    rfc = RandomForestClassifier(n_estimators=t, n_jobs=-1)\n",
    "    rfc_s = cross_val_score(rfc, wine.data, wine.target, cv=6)\n",
    "    score_list.append(rfc_s.mean())\n",
    "    \n",
    "print('max score: {}, index：{}'.format(max(score_list),\n",
    "                                       score_list.index(max(score_list))))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-06T12:40:25.067272Z",
     "start_time": "2019-12-06T12:40:24.846489Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1800x750 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12, 5))\n",
    "plt.plot(range(1, times+1), score_list, '-b',\n",
    "              label='difference of n_estimator\\'s performance')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "由上图我们可以知道，n_estimators达到一定的个数后，随机森林的效果会围绕一个值波动，所以，并非n_estimators 越大越好。越大也意味着要浪费更多的时间运行"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## random_state\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    ">先思考一个问题，随机森林用了什么方法保证了集成的效果要比单个分类器来的好呢？\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "随机森林的本质是装袋法，装袋集成算法是对集评估器的预测结果进行平均或者少数服从多数的原则进行集成评估结果的。  \n",
    "例如，对于构建了25个集成评估器来说，只有在有13棵树或以上判断错误，随机森林才会判断错误。对于单棵树而言，其准确率在85%浮动(这里假设为80%)，则每棵树判断错误的概率为20%(设为$\\epsilon$), 那么判断错误的概率为：\n",
    "$$e_{random_forest} = \\displaystyle \\sum_{i=13}^{25}C_{25}^{i}\\epsilon^i(1-\\epsilon)^{25-i}$$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "其中，i是判断错误的次数，也是判错的树的数量，ε是一棵树判断错误的概率，（1-ε）是判断正确的概率，共判对\n",
    "25-i次。采用组合，是因为25棵树中，有任意i棵都判断错误."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T01:21:25.103958Z",
     "start_time": "2019-12-07T01:21:25.083957Z"
    }
   },
   "outputs": [],
   "source": [
    "import numpy as np \n",
    "from scipy.special import comb\n",
    "import math\n",
    "\n",
    "def proba_of_error(n_estimators, epsilon):\n",
    "    \"\"\"\n",
    "    计算随机森林算法错误的概率\n",
    "    \n",
    "    Parameters\n",
    "    ----------\n",
    "    n_estimators : int\n",
    "        基评估器的个数\n",
    "    \n",
    "    epsilon : float\n",
    "        单个基评估器发送错误的概率\n",
    "    \n",
    "    Returns\n",
    "    -------\n",
    "    error : 随机森林算法发生错误的概率\n",
    "    \"\"\"\n",
    "    \n",
    "    middle = math.ceil(n_estimators / 2)\n",
    "    errors = [comb(n_estimators, i) * \n",
    "              epsilon**i * \n",
    "              ((1 - epsilon)**(n_estimators - i)) \n",
    "              for i in range(middle, n_estimators+1)]\n",
    "    return sum(errors)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-06T12:55:17.611083Z",
     "start_time": "2019-12-06T12:55:17.600026Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0003690480345558283"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "proba_of_error(25, 0.2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "可见，判断错误的几率非常小，这让随机森林在红酒数据集上的表现远远好于单棵决策树。  \n",
    "\n",
    "那现在就有一个问题了：我们说袋装法服从多数表决原则或对基分类器结果求平均，这即是说，我们默认森林中的\n",
    "每棵树应该是不同的，并且会返回不同的结果。设想一下，如果随机森林里所有的树的判断结果都一致（全判断对\n",
    "或全判断错），那随机森林无论应用何种集成原则来求结果，都应该无法比单棵决策树取得更好的效果才对。但我\n",
    "们使用了一样的类DecisionTreeClassifier，一样的参数，一样的训练集和测试集，为什么随机森林里的众多树会有\n",
    "不同的判断结果？"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "这是因为sklearn中的分类树DecisionTreeClassifier本身带有随机性，这样使得每棵决策树都不一样。   \n",
    "我们知道随机森林的集评估器就算决策树，当随机森林(RandomForestClassifier)中的参数设置了random_state，这不是意味每一个基评估器的random_state都设置一样。  \n",
    "\n",
    "随机森林的random_state的设置了生成的森林随机模式，而不是森林中的一颗树。\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## rfc.estimators_ 查看森林中树的情况"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-06T13:04:04.292186Z",
     "start_time": "2019-12-06T13:04:04.265016Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X_train, X_test , y_train, y_test = train_test_split(wine.data,\n",
    "                                                     wine.target, \n",
    "                                                     test_size=0.3)\n",
    "\n",
    "rfc = RandomForestClassifier(n_estimators=25, random_state=1)\n",
    "rfc = rfc.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-06T13:05:13.489433Z",
     "start_time": "2019-12-06T13:05:13.471887Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=1791095845, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=2135392491, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=946286476, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=1857819720, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=491263, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=550290313, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=1298508491, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=2143362693, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=630311759, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=1013994432, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=396591248, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=1703301249, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=799981516, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=1666063943, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=1484172013, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=729053692, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=1704103302, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=1870626073, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=166716594, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=1487394068, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=1800426750, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=1345499493, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=795511698, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=105433556, splitter='best'),\n",
       " DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                        max_features='auto', max_leaf_nodes=None,\n",
       "                        min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                        min_samples_leaf=1, min_samples_split=2,\n",
       "                        min_weight_fraction_leaf=0.0, presort=False,\n",
       "                        random_state=878115723, splitter='best')]"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfc.estimators_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-06T13:05:41.614545Z",
     "start_time": "2019-12-06T13:05:41.607707Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<bound method BaseEstimator.get_params of DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,\n",
       "                       max_features='auto', max_leaf_nodes=None,\n",
       "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                       min_samples_leaf=1, min_samples_split=2,\n",
       "                       min_weight_fraction_leaf=0.0, presort=False,\n",
       "                       random_state=1791095845, splitter='best')>"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfc.estimators_[0].get_params"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-06T13:06:20.882982Z",
     "start_time": "2019-12-06T13:06:20.875230Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1791095845\n",
      "2135392491\n",
      "946286476\n",
      "1857819720\n",
      "491263\n",
      "550290313\n",
      "1298508491\n",
      "2143362693\n",
      "630311759\n",
      "1013994432\n",
      "396591248\n",
      "1703301249\n",
      "799981516\n",
      "1666063943\n",
      "1484172013\n",
      "729053692\n",
      "1704103302\n",
      "1870626073\n",
      "166716594\n",
      "1487394068\n",
      "1800426750\n",
      "1345499493\n",
      "795511698\n",
      "105433556\n",
      "878115723\n"
     ]
    }
   ],
   "source": [
    "for i in range(len(rfc.estimators_)):\n",
    "    print(rfc.estimators_[i].random_state)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "我们可以观察到，当random_state固定时，随机森林中生成是一组固定的树，但每棵树依然是不一致的，这是\n",
    "用”随机挑选特征进行分枝“的方法得到的随机性。并且我们可以证明，当这种随机性越大的时候，袋装法的效果一\n",
    "般会越来越好。用袋装法集成时，**基分类器应当是相互独立的，是不相同的。**  \n",
    "\n",
    "但这种做法的局限性是很强的，当我们需要成千上万棵树的时候，数据不一定能够提供成千上万的特征来让我们构\n",
    "筑尽量多尽量不同的树。因此，除了random_state。我们还需要其他的随机性"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## bootstrap & oob_score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "随机森林的随机性体现在数据的随机选择，要让每一个基分类器不一样，一种直观的理解就是用不同的数据集去训练。装袋法的本质就是通过有**放回的抽样**来形成不同的数据集，而bootstrap就是用来控制抽样计数的参数。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**RandomForest默认的boostrap为True**，者意味者我们做有放回的抽样，一般不会改为False"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![](images/2_2.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "在一个含有n个样本的原始训练集中，我们进行随机采样，每次采样一个样本，并在抽取下一个样本之前将该样本\n",
    "放回原始训练集，也就是说下次采样时这个样本依然可能被采集到，这样采集n次，最终得到一个和原始训练集一\n",
    "样大的，n个样本组成的自助集。由于是随机采样，这样每次的自助集和原始数据集不同，和其他的采样集也是不\n",
    "同的。这样我们就可以自由创造取之不尽用之不竭，并且互不相同的自助集，用这些自助集来训练我们的基分类\n",
    "器，我们的基分类器自然也就各不相同了"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "虽然数据采集是有放回地抽样，但是可以证明我们只能抽取到其中的63%左右的数据，所以还存在一些没有被抽取到的数据，我们称这些数据为**out-of-bag data(简称，oob)**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "假设有x个样本，n个自助集，则每个样本被抽到的概率为\n",
    "$$1 - (1 - \\dfrac{1}{x})^n = 1 - \\dfrac{1}{e} ≈ 63\\%$$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-06T16:51:55.947336Z",
     "start_time": "2019-12-06T16:51:55.911124Z"
    }
   },
   "source": [
    "所以说，当我们使用**随机森林训练模型的时候我们可以不用划分训练集和测试集，可以用oob数据进行测试**。但是，但样本的数量和n_estimators不是很大，就可能存在没有数据在oob。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "如果我们想利用oob数据进行测试的话，就需要在Randomforest类传入oob_score，训练完毕后可以用rfc.oob_score_查看训练的分数。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-06T16:57:46.324561Z",
     "start_time": "2019-12-06T16:57:46.310972Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.datasets import load_wine\n",
    "\n",
    "\n",
    "wine = load_wine()\n",
    "X = wine.data\n",
    "y = wine.target\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-06T16:59:24.160891Z",
     "start_time": "2019-12-06T16:59:23.973766Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9831460674157303"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfc = RandomForestClassifier(n_estimators=100, random_state=1, oob_score=True)\n",
    "\n",
    "rfc = rfc.fit(X, y)\n",
    "rfc.oob_score_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 重要属性和接口"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T01:06:37.716107Z",
     "start_time": "2019-12-07T01:06:37.686715Z"
    }
   },
   "source": [
    "重要属性：  \n",
    "+ rfc.feature_importances_\n",
    "+ rfc.estimators_\n",
    "+ rfc.oob_score_  \n",
    "\n",
    "重要接口：  \n",
    "+ rfc.fit()\n",
    "+ rfc.apply()\n",
    "+ rfc.predict()\n",
    "+ rfc.predict_proba\n",
    "\n",
    "随机森林的predict_proba接口，这个接口返回每个测试样本对应的被分到每一类标签的概率，标签有几个分类\n",
    "就返回几个概率。如果是二分类问题，则predict_proba返回的数值大于0.5的，被分为1，小于0.5的，被分为0。\n",
    "传统的随机森林是利用袋装法中的规则，平均或少数服从多数来决定集成的结果，而sklearn中的随机森林是平均\n",
    "每个样本对应的predict_proba返回的概率，得到一个平均概率，从而决定测试样本的分类"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T01:13:56.623908Z",
     "start_time": "2019-12-07T01:13:56.434797Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import cross_val_score\n",
    "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=30)\n",
    "rfc = RandomForestClassifier(n_estimators=100, random_state=1)\n",
    "rfc = rfc.fit(X_train, y_train)\n",
    "\n",
    "rfc.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T01:14:00.485121Z",
     "start_time": "2019-12-07T01:14:00.448943Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('alcohol', 0.12988572648792832),\n",
       " ('malic_acid', 0.025614603032081032),\n",
       " ('ash', 0.009374887213216),\n",
       " ('alcalinity_of_ash', 0.029612359086332075),\n",
       " ('magnesium', 0.03505767398116663),\n",
       " ('total_phenols', 0.07041673485845597),\n",
       " ('flavanoids', 0.1450328353202198),\n",
       " ('nonflavanoid_phenols', 0.010118014544751239),\n",
       " ('proanthocyanins', 0.02551487911220493),\n",
       " ('color_intensity', 0.13849106193422286),\n",
       " ('hue', 0.0797571330183984),\n",
       " ('od280/od315_of_diluted_wines', 0.0943386658839796),\n",
       " ('proline', 0.20678542552704318)]"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[*zip(wine.feature_names, rfc.feature_importances_)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T01:15:04.549595Z",
     "start_time": "2019-12-07T01:15:04.520238Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[11, 13,  5, ...,  4, 12,  4],\n",
       "       [ 2,  6,  1, ..., 10,  7,  3],\n",
       "       [ 2,  8,  1, ..., 13,  7,  3],\n",
       "       ...,\n",
       "       [13, 13,  5, ...,  4, 12, 11],\n",
       "       [ 2,  8,  1, ..., 13,  7,  6],\n",
       "       [ 2, 15,  1, ..., 11, 12,  3]], dtype=int64)"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfc.apply(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T01:16:21.816909Z",
     "start_time": "2019-12-07T01:16:21.781006Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test result:  [2 1 1 0 1 1 2 2 2 1 1 0 2 1 1 2 0 0 2 0 2 1 2 0 1 0 2 2 1 1]\n",
      "true result:  [2 1 1 0 1 1 2 2 2 1 1 0 2 1 1 2 0 0 2 0 2 1 2 0 1 0 2 2 1 1]\n"
     ]
    }
   ],
   "source": [
    "print(\"test result: \", rfc.predict(X_test))\n",
    "print('true result: ', y_test)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T01:16:52.541741Z",
     "start_time": "2019-12-07T01:16:52.500473Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.  , 0.07, 0.93],\n",
       "       [0.  , 1.  , 0.  ],\n",
       "       [0.  , 1.  , 0.  ],\n",
       "       [0.98, 0.02, 0.  ],\n",
       "       [0.01, 0.99, 0.  ],\n",
       "       [0.  , 0.99, 0.01],\n",
       "       [0.  , 0.01, 0.99],\n",
       "       [0.18, 0.19, 0.63],\n",
       "       [0.03, 0.22, 0.75],\n",
       "       [0.07, 0.92, 0.01],\n",
       "       [0.  , 1.  , 0.  ],\n",
       "       [0.99, 0.01, 0.  ],\n",
       "       [0.  , 0.02, 0.98],\n",
       "       [0.02, 0.95, 0.03],\n",
       "       [0.01, 0.96, 0.03],\n",
       "       [0.  , 0.36, 0.64],\n",
       "       [0.67, 0.29, 0.04],\n",
       "       [0.99, 0.01, 0.  ],\n",
       "       [0.  , 0.06, 0.94],\n",
       "       [0.93, 0.06, 0.01],\n",
       "       [0.13, 0.19, 0.68],\n",
       "       [0.06, 0.94, 0.  ],\n",
       "       [0.01, 0.14, 0.85],\n",
       "       [0.97, 0.02, 0.01],\n",
       "       [0.18, 0.71, 0.11],\n",
       "       [0.95, 0.04, 0.01],\n",
       "       [0.  , 0.01, 0.99],\n",
       "       [0.  , 0.11, 0.89],\n",
       "       [0.  , 1.  , 0.  ],\n",
       "       [0.17, 0.6 , 0.23]])"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfc.predict_proba(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Bagging的另外一个重要条件"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. 基评估器要尽可能地相互独立\n",
    "2. 数据判断的准确率要超过50%，如果准确率低于50%的话，那么随机森林的效果比单个决策树来的还要差。(如果低于50%也说明当前的数据根本不时候决策树和随机森林)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T01:26:21.178072Z",
     "start_time": "2019-12-07T01:26:20.763370Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 900x600 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "plt.rcParams['figure.dpi']=150\n",
    "# 设置x轴\n",
    "x = np.linspace(0, 1, 20) \n",
    "# 用于接收误差\n",
    "y = []\n",
    "\n",
    "for epsilon in np.linspace(0, 1, 20):\n",
    "    e = proba_of_error(25, epsilon)\n",
    "    y.append(e)\n",
    "    \n",
    "plt.plot(x,y,\"o-\",label=\"when estimators are different\")\n",
    "plt.plot(x,x,\"--\",color=\"red\",label=\"if all estimators are same\")\n",
    "plt.xlabel(\"individual estimator's error\")\n",
    "plt.ylabel(\"RandomForest's error\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "由上图可以看出，当单个基评估器的误差很小的时候，随机森林的误差要比单个基评估器的误差来的小。但是，但基评估器的误差超过50%的时候，随机森林的误差会更大。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# RandomForestRegressor"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "随机森林回归的参数基本与决策树回归一样，只是多了一个n_estimators"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T01:34:21.832049Z",
     "start_time": "2019-12-07T01:34:21.816710Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.datasets import load_boston"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T01:37:21.763963Z",
     "start_time": "2019-12-07T01:37:17.567518Z"
    }
   },
   "outputs": [],
   "source": [
    "# 加载数据\n",
    "boston = load_boston()\n",
    "\n",
    "# 实例化模型\n",
    "regressor = RandomForestRegressor(n_estimators=100, random_state=1)\n",
    "# 交叉验证\n",
    "score = cross_val_score(regressor, \n",
    "                        boston.data,\n",
    "                        boston.target,\n",
    "                        cv=8, \n",
    "                        scoring='neg_mean_squared_error')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T01:37:36.163084Z",
     "start_time": "2019-12-07T01:37:36.146701Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "21.997858562468984"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(score * -1).mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "注意在这里，如果不填写scoring = \"neg_mean_squared_error\"，交叉验证默认的模型\n",
    "衡量指标是R平方，因此交叉验证的结果可能有正也可能有负。而如果写上scoring，则衡量标准是负MSE，交叉验\n",
    "证的结果只可能为负."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T01:40:21.814854Z",
     "start_time": "2019-12-07T01:40:21.801471Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['explained_variance', 'r2', 'max_error', 'neg_median_absolute_error', 'neg_mean_absolute_error', 'neg_mean_squared_error', 'neg_mean_squared_log_error', 'accuracy', 'roc_auc', 'balanced_accuracy', 'average_precision', 'neg_log_loss', 'brier_score_loss', 'adjusted_rand_score', 'homogeneity_score', 'completeness_score', 'v_measure_score', 'mutual_info_score', 'adjusted_mutual_info_score', 'normalized_mutual_info_score', 'fowlkes_mallows_score', 'precision', 'precision_macro', 'precision_micro', 'precision_samples', 'precision_weighted', 'recall', 'recall_macro', 'recall_micro', 'recall_samples', 'recall_weighted', 'f1', 'f1_macro', 'f1_micro', 'f1_samples', 'f1_weighted', 'jaccard', 'jaccard_macro', 'jaccard_micro', 'jaccard_samples', 'jaccard_weighted'])"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 查看评分标准：  \n",
    "import sklearn\n",
    "sklearn.metrics.SCORERS.keys()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 用随机森林填补缺失值，并且与填补0均值作比较。\n",
    "在sklearn中，我们可以使用sklearn.impute.SimpleImputer来轻松地将均\n",
    "值，中值，或者其他最常用的数值填补到数据中，在这个案例中，我们将使用均值，0，和随机森林回归来填补缺\n",
    "失值，并验证四种状况下的拟合状况，找出对使用的数据集来说最佳的缺失值填补方法。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 256,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T06:19:31.400752Z",
     "start_time": "2019-12-07T06:19:31.394174Z"
    }
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np \n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.datasets import load_boston\n",
    "from sklearn.impute import SimpleImputer\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "\n",
    "from sklearn.model_selection import cross_val_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 257,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T06:19:32.260406Z",
     "start_time": "2019-12-07T06:19:32.241758Z"
    }
   },
   "outputs": [],
   "source": [
    "# 加载数据\n",
    "boston = load_boston()\n",
    "\n",
    "X_full = boston.data\n",
    "y_full = boston.target\n",
    "\n",
    "feature_names = boston.feature_names\n",
    "\n",
    "# 获取样本的信息\n",
    "n_samples = X_full.shape[0]\n",
    "n_features = X_full.shape[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 258,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T06:19:33.022938Z",
     "start_time": "2019-12-07T06:19:33.011531Z"
    }
   },
   "outputs": [],
   "source": [
    "\"\"\"设置缺失值\"\"\"\n",
    "\n",
    "# 设置缺失值的比例\n",
    "missing_rate = 0.5\n",
    "# 获取缺失值的数量\n",
    "missing_sample_counts = int(np.floor(n_samples * n_features * missing_rate))\n",
    "\n",
    "# 生成缺失值的索引\n",
    "rng = np.random.RandomState(1)\n",
    "\n",
    "missing_sample_index = rng.randint(0, n_samples, missing_sample_counts)\n",
    "missing_features_index = rng.randint(0, n_features, missing_sample_counts)\n",
    "\n",
    "X_missing = X_full.copy()\n",
    "y_missing = y_full.copy()\n",
    "\n",
    "X_missing[missing_sample_index, missing_features_index] = np.nan\n",
    "# 将其转换成DataFrame方便索引\n",
    "X_missing = pd.DataFrame(X_missing)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 259,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T06:19:33.783496Z",
     "start_time": "2019-12-07T06:19:33.768047Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 506 entries, 0 to 505\n",
      "Data columns (total 13 columns):\n",
      "0     318 non-null float64\n",
      "1     321 non-null float64\n",
      "2     297 non-null float64\n",
      "3     290 non-null float64\n",
      "4     328 non-null float64\n",
      "5     279 non-null float64\n",
      "6     300 non-null float64\n",
      "7     304 non-null float64\n",
      "8     297 non-null float64\n",
      "9     302 non-null float64\n",
      "10    305 non-null float64\n",
      "11    321 non-null float64\n",
      "12    308 non-null float64\n",
      "dtypes: float64(13)\n",
      "memory usage: 51.5 KB\n"
     ]
    }
   ],
   "source": [
    "X_missing.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 260,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T06:19:34.552256Z",
     "start_time": "2019-12-07T06:19:34.541843Z"
    }
   },
   "outputs": [],
   "source": [
    "\"\"\"用0填补缺失值\"\"\"\n",
    "# 实例化模型\n",
    "inputer_constant = SimpleImputer(missing_values=np.nan, strategy='constant', fill_value=0)\n",
    "# 训练\n",
    "X_missing_fill_with_zero = inputer_constant.fit_transform(X_missing)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 261,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T06:19:35.282332Z",
     "start_time": "2019-12-07T06:19:35.270922Z"
    }
   },
   "outputs": [],
   "source": [
    "\"\"\"用均值填补缺失值\"\"\"\n",
    "inputer_mean = SimpleImputer(missing_values=np.nan, strategy='mean')\n",
    "X_missing_fill_with_mean = inputer_mean.fit_transform(X_missing)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 262,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T06:19:36.059064Z",
     "start_time": "2019-12-07T06:19:36.040216Z"
    }
   },
   "outputs": [],
   "source": [
    "import pandas as pd \n",
    "import numpy as np\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.impute import SimpleImputer\n",
    "\n",
    "\n",
    "\n",
    "def rfr_fill_missing_value(X, y, strategy='mean'):\n",
    "    \"\"\"\n",
    "       用随机森林填补缺失值。\n",
    "       基本策略是：\n",
    "       假设特征数据存在缺失，而target并未缺失。\n",
    "       这时候我们可以把target作为特征数据，把那些数据缺失的特征作为target，\n",
    "       通过随机森林模型训练，我们可以预测出缺失值。\n",
    "\n",
    "       基本步骤是：\n",
    "       1. 先找出数据缺失的特征，按从小到大排列其索引值。\n",
    "       2. 从缺失值最少的特征开始，我们把其他含有缺失值的特征先用0或者均值填充。\n",
    "       3. 把target作为特征的一部分进行训练\n",
    "       4. 预测(predict)缺失值\n",
    "       \n",
    "       Parameters\n",
    "       ----------\n",
    "       X : array_like\n",
    "           data containing missing values\n",
    "           \n",
    "       y : array_like\n",
    "           target, it will be as parts of training data\n",
    "           \n",
    "       Returns\n",
    "       -------\n",
    "       X_missing_in_rfr : pd.DataFrame\n",
    "    \"\"\"\n",
    "    # 将X转换成DataFrame\n",
    "    X_copy = X.copy()\n",
    "    X_missing_reg = pd.DataFrame(X_copy)\n",
    "    \n",
    "    # 1. 先找出数据缺失的特征，按从小到大排列其索引值。\n",
    "    sorted_missing_index = np.argsort(X_missing_reg.notnull().sum(axis=0))\n",
    "    \n",
    "    for i in sorted_missing_index:\n",
    "        # 2. 从缺失值最少的特征开始，我们把其他含有缺失值的特征先用0或者均值填充。\n",
    "        # 2.1 取出要填补特征的列\n",
    "        column_of_missing_values = X_missing_reg.iloc[:, i]\n",
    "\n",
    "        # 2.2 取出不含填补特征的列\n",
    "        data = pd.concat([X_missing_reg, pd.DataFrame(y)], axis=1)\n",
    "        temp_df = data.iloc[:, X_missing_reg.columns!=i]\n",
    "\n",
    "        # 2.3 填补数据，根据不同的策略填充不同的数据\n",
    "        if strategy == \"mean\":\n",
    "#             temp_df.fillna(temp_df.mean(), axis=0, inplace=True)\n",
    "            imputer_mean = SimpleImputer(missing_values=np.nan, strategy=strategy)\n",
    "            temp_arr = imputer_mean.fit_transform(temp_df)\n",
    "        elif strategy == 'fill_with_zero':\n",
    "            imputer_zero = SimpleImputer(missing_values=np.nan, \n",
    "                                         strategy='constant', fill_value=0)\n",
    "            temp_arr = imputer_zero.fit_transform(temp_df)\n",
    "            \n",
    "        # 3. 划分训练集和测试集\n",
    "        # 3.1 获取缺失值的索引\n",
    "        null_value_index = column_of_missing_values.isnull()\n",
    "\n",
    "        # 3.2 将不含缺失值的数据作为训练集\n",
    "#         X_train = temp_df.loc[~null_value_index, :]\n",
    "        y_train = column_of_missing_values.loc[~null_value_index]\n",
    "        X_train = temp_arr[y_train.index, :]\n",
    "        \n",
    "        \n",
    "        # 3.3 将含有缺失值的数据作为测试集\n",
    "        y_test = temp_df.loc[null_value_index]\n",
    "        X_test = temp_arr[y_test.index, :]\n",
    "\n",
    "        # 4. 创建模型，并预测缺失值\n",
    "        rfr = RandomForestRegressor(n_estimators=100, random_state=1)\n",
    "        rfr = rfr.fit(X_train, y_train)\n",
    "        y_predict = rfr.predict(X_test)\n",
    "\n",
    "        # 5. 将训练得到的数据填充对应列缺失值\n",
    "        X_missing_reg.loc[null_value_index, i] = y_predict\n",
    "#         break\n",
    "    return X_missing_reg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 274,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T06:27:36.952595Z",
     "start_time": "2019-12-07T06:27:32.446850Z"
    }
   },
   "outputs": [],
   "source": [
    "X_missing_reg = rfr_fill_missing_value(X_missing.values, y_full, 'fill_with_zero')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 对填好的数据进行建模"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 284,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T06:33:19.553668Z",
     "start_time": "2019-12-07T06:33:03.069048Z"
    }
   },
   "outputs": [],
   "source": [
    "X = [X_full, X_missing_fill_with_mean, X_missing_fill_with_zero, X_missing_reg]\n",
    "\n",
    "# 定义空列表接收返回的均方误差\n",
    "mse = []\n",
    "for data in X:\n",
    "    rfr = RandomForestRegressor(n_estimators=100, random_state=1)\n",
    "    score = cross_val_score(rfr, data, y_full, cv=8,\n",
    "                            scoring='neg_mean_squared_error').mean()\n",
    "    mse.append(score * (-1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 285,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T06:33:20.508365Z",
     "start_time": "2019-12-07T06:33:20.499437Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[21.997858562468984,\n",
       " 36.452253607669896,\n",
       " 39.388518943390366,\n",
       " 28.766275797433025]"
      ]
     },
     "execution_count": 285,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mse"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 画图"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 279,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T06:30:24.025184Z",
     "start_time": "2019-12-07T06:30:23.466189Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABpwAAAMeCAYAAAAAnadAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAAXEQAAFxEByibzPwAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzs3Xn8bdd8P/7Xm4gIIYaKIFyJGJIiCEq1Qqu+1FBTKcptxVBUwhctOqQaRVFp8SNtEUPzVfNUNUupKSUxJcYQMSQRZCISkrx/f+x9ck8+PTf3fj77c+8n9+b5fDzO43zO2uvstfY++5wk+5W1VnV3AAAAAAAAYKUut9YdAAAAAAAAYNsmcAIAAAAAAGASgRMAAAAAAACTCJwAAAAAAACYROAEAAAAAADAJAInAAAAAAAAJhE4AQAAAAAAMInACQAAAAAAgEkETgAAAAAAAEwicAIAAAAAAGASgRMAAAAAAACTCJwAAAAAAACYROAEAAAAAADAJAInAAAAAAAAJhE4AQDApUhV9fg4YK37sr2pqgPGc3vUWvdlueaui0t8bOU+bfXzuS1/htuCqjpiPL/rV3GfW/wzq6r1G/lOnFFVH6+qP9xSbV8WVNVRc+f0wqr6SVUdW1Uvqqp1a90/AODSY4e17gAAALB9qapDkqS7D9mW27iU+cLc3zsmufn491eS/GLrd4fLujFoWJ/kxO4+Yi37MueXSY6fe33dJHdOcueq2r+7D1qLTlXVrkkOTnJGdx+2Fn1YJacm+WGSayfZb3w8uaoeu5rXQFUdnGTXJId19xmrtV8AYMszwgkAAFhtfz0+Lm1tnJPka0lOWv3ubFndvd/skeRec5vutWTb9m6b/Qy3ESdnOL9nbkbddRm+g+u3YH+W6wfz34fuvnaSR4/bnlxVd1yjfu2a4VwdvEbtr5ZXdvctu/s6SW6V5GNJrpDk1VX1m6vYzsEZzteuq7hPAGArEDgBAACXCd19dHffrLsfudZ9YWV8hltWdz9zPL9vX+u+rJbufnWSj48vf28t+7I96e4vJrlHhtGXleTv17ZHAMClgcAJAAAA2J6dNj5fZU17sZ3p7nOTHDq+vENV7bWW/QEA1p7ACQAAANguVdUVktx6fPnNtezLdup9SXr8+w5r2REAYO0JnAAAYBtRVeuqqqvqxKrar6o+WVXnVNV/VtUuVfXkqjq5qn5cVX8/974Tx/ftU1X/UFU/qKrzqur4qnp6Ve2wsXY20o9Dxu2HjK8PGF93VfVcvV5UvmRfV6uqv6qqL1XVT8fH0VX1x0vqrbiNBfs46pLqjXX3rqoj5s7VN6vqBVV19QV1Z+d3XVXtWVVHVtUpVXVWVX26qu6+qfa2lKrabfzMvzEex+nj9XLXS3jPI6rqv6vq7Kr6eVV9saqeWVVX3ERbD6iqz4zv+35V/VtV7bakzhHjuVpfVVcez+m3x+v4+Ko6qKrqEtpYzmf46Ko6ZjyG06rq1VV1zUXXyvxnuJw2l3t+q+qK47n84njMZ1XV58fj3nFTx3QJx3rVqrpg7MMVxrJHjv3+WVVdfixbP5b9v43s56LPZyPbZ9/9TvLRsfguS76HR22ir5u8TlZDDfZK8rokN0ryoySvXVDvtlX1lvEaObeqjquqZ1XVlTay3+tU1SvG6/YXVXVqVR1VVfdZUm/93Ln69lh8wyXn6sSNtHG1qnpuVX19/ExPqao3VNXNFtSd9J2aqrt/mg1rql2sf1W1Uw3/XPpcVZ0x93vytJr7505t+GfO7HzdcNz07SXna93S9je3DQBg6/APXwAA2PZcLcmHkvw8w7/T/5/x9e0y3Pi7TpKnV9Vbuvvoufe9Lsltk/wwyVlJbp5h3Y0Dquq+3X3BCvvz0wzreMzcanz+woK6F6mqXZN8JslNkvwiyYlJdhyP43ZVtVt3P29KGytRVfdI8tYkV05yXpLvZ7hh/Ywkv19Vv93dJyx4680z3NC+RpJvZfic7pDkP6vq9t19zGr39ZJU1S2TfCDJbklOzzC641oZrpd7VNWB4/o2s/qXS3JkkoeMRT9M8rMktxgf96qq3+ruXyxo62lJXpjkjAw39tcleViSm1bVHRZcWzsn+UiS22f43M/NcP4OG7f/48RjPzzJY8eXZ2b4HP9oPI5VsdzzO3p7knsmuTDJd5KcP/bpsCR3SfKAlfSlu8+qquPGfd0syZeS3GbcvPNYdlySW45ln1pJO0lOyYbv3FWS7JXhGpkfObTRUUQruE6W64a1OHj+rySP6e6fLOnP+iT/kuF39GcZjm+fJM9N8qCqunt3/3iu/rWSHJ1kjwzX7Lcy/E7cJUPwdlB3/9NY/SfZcK52zHB9/zLJ8XNd+MHSjlbVDTP8nt84w8ihE5NcN8nDk9y/qh7S3e9ZcIxb9Du1CT/KEBJdbVYwBqgfSvLrSS7IhtBt3wzXwN5JHjeW/SIX/y3fJ8kVknxl3Ja5ehdZZhsAwFZghBMAAGx7dk1yRHfvkeGmWjLcZLxvd69L8smx7CZL3neLJA/r7t26+1eS3CfDTdZ7JXn0SjvT3Z/t7v1mj7ny/RaVz3n62Mfjktygu2/a3TdKcu9x+5/NRmZMaGNZqur6Sd6c4SbyK5Jco7v3THKDDDdz1yV5x6xfSxyR4Ybnjbv7JhluwH45yeWzlW96jv9n/1uTXDVDiHGt7t63u3fLcH47yT9U1S5zb/vzDGHTaUnuNV4ne2a4efv9JHdO8vgFzf1qkhdkWMtlt7nPsDMEnLdf8J6/znAu7zrW3y3Jv43bnrDS404uCgwfmyHUeXR379rd10/y2xluQE+2kvNbVbfKEDadmeQW3b3neJ3snSGAuX9NWwNnFiLNQqXbZAilZ38nG4LaFQVO3f3Kue/agWPxZ5d8Dw/cyNtXcp0s1y8zBBdfyBBWnDOW75HkYr8PVXXbbAib/irJruPv583H9986yeuX7P/h476OTrJ7d99s/B2eBYVPn1Xs7nfNnat7jcU/WHKu7jW/83Ek0lszhE0fT7Ju/A7umiEw2jnJv1fVjRYc+xb7Tm2G2XV25bmyR2UIgk5OcvPu3ru7906yf4bg6DFVde0k6e6LnZdsCOLuteR8LQ3oNrsNAGDrEDgBAMC25/wMN0iT5Bvj86fm/q/3WdnSGQ1e0d0XTaU11n/Z+PIRW6Kjm3BUhlEnf9Ddp86VfyDDTfmrZRittTX9aZJdkvx3kid29zlJ0t3fT/LgDGHMr2YI65Y6L8k9uvvE8T0/TPKicduUIGEl7pPhpvUvM9yIPqaGqds+n2H0xvkZzu+vJxeNbnrq+N7Hdvd/znbU3ccn+dsMo0R2XdDWNZO8vLv/cjb6qbv/Y6yfLD72aya5T3cfNdb/ZZJnX0L95Zhdy2+ZH2HU3R9OcsjEfc8s6/yOZuHH98dzOuvXt5LcLcldM4wUWamLAqcxuNgvQwh6Ti4eOJ2b5PMT2lmplVwnyzUfXOyT5OoZwt7dk7ypqv5gru4zMvxGHtndf9vd5499+mqG7/p5Se5ZVfNB1ewz/Gx3nzFX/o4Mn9/U39H/kyF8Oz3Jg7r7pLFP53b3wUk+nCF0OmjBe7fkd2pTdhqffzZX9uUMv+8P7O5vzJV/PsMouMr0AHhrtAEALIMp9QAAYNtz8iwImbPRaazmvGNB2fuT/FmGEGWr6u4PJsm47sidxz7cNsk9smFqpits5W7dZXx+c3dfbGqu7v5JVX04yUPHekvP54uW3IRONqxtsrX/2+uO4/NVs2FUyyKztVJumuGG9S+TvHdppe4+PMnhG9nH+Un+ZkH5JR37u5ZM9zhff9HoseWYXcsfXLDtnUleMnH/yfLPbzJ8Rz+a5K5V9cYM0+sdl+Sr3X3sKvRpFjjdKsPIwV2SfDbDSJ3bVtX1MnzGnxjDiK1tJdfJJGOw9c/jyMW/zDDCaha6z77r/77gfd+oqmOT/NpYbxbQvSNDoHhgVf00yccyTJF34izomWjWp6PGwHqpNyf5rbl687bkd2pTrjE+nz4r6O5PJflUDeuW3T7D9/LWSe6e4fcmmfj7vjXaAACWxwgnAADY9ly4mWVLnbKgbHZTc5cF2zZmOXU3qqquVVWvz7DWyceTvDTJnZL8x1i2Fq45Pn93I9u/u6TevI8tKFu0nszWcPXx+SXdXZfwmIVIsxvGZ/WCNZo24bieW+dmziUd+/86V0sDvglm1+dpC7adPGF/85Z7fmfH93sZppTbN8N0Z19KcmZVvbWqpoa+X8/wvbllNoxoOnZ87JfhRnyy8vWbplrJdbJaZtOM7lFVs2t92d/17j4tw6i112cInt+TYR2nU6rqpXP7XqlV/f1Zxe/URlXVTtkQrB43V75zVf1Tkh9nWKvvXzKM4Pp0kkVr4K2k7S3eBgCwPAInAAC47LjegrLdx+czl7Gf203tyDjl11syTEF1ZIYb4juP6+D8SZKzp7axQrMb4ntsZPseS+rNWzq6aS3NRhosPI6qunpVXb+qrjoWzQK+q1bVjgvq71ZVh1TVXy7Y3UqOe0ueq9m1vNuCbddfwf4WXe/LPb9Jku4+a5xS7hYZ1ru5bYZ12O6T5Kiquu4K+jfbd2e40b57kt/JMFrtuAyB0y5JHjhWXavAaS2/H4tGuazou97d3+juA7v7hhmmmLxbhtF0T0ryrnF6ypXaFn9/fjvDCKrZ9TfzygxTlB6VYaTYlcc1ltYn+d4qtb012gAAlkHgBAAAlx0PWlA2W7R+fk2XC8bna43B0EWq6o5JfmMT7Vww1r2kqYz2zDAt1DlJDuzuL8ym+aqqvZL8yiq0sRKz9WR+f8GxXyPDzdVk8WimS5NZqPDbVbVohM5HMoyWeNj4+msZbmJfIRuuiXl3zrBW0R+vcj+3hNm1fM8F2x64oCzZcM1f7Lqrql2TPGZB/eWe31TVH1bVv1bV3ZOku8/r7mO6+5Akr8owauV+G+nf5pr1a32GEUW/yBA4zcqSi4cCU5w/Pm8LU5bdeXw+PRvCwtl3/SFLK1fV3tkwIuxjc+WHjp/hDZKku8/s7o929yOSfCXD6KebLt1fNv9czfp016q69oLtv7+0T2upqnZI8hfjyw9093fH8ismma2X9Zju/kx3nztuu0Y2va7SJs/XKrQBAGwBAicAALjsOLCq/nD2oqrul+Tx48vXztX7QZJzM4zAeMRc/X0zrCHy8020M5vK6KKb+1V1hXGdppldx+crZRjdlKq6XFXdK8mHkuy8Cm2sxMuS/DTDjeOXV9XO476vl+HYr5Vh1Mi7Jrazpb0nw5pBuyZ5+9j/2Tl+VoZz/rMM6wiluy9M8g/je/+5qi4Ka6rqZkmeN76crX9zaTa7lu9bVY+bBYdVdY8kz97Ie741Pj9uVjCGTW/LhvXE5i3r/I5+Jcmjkzy3qq4z1841M4zOSKb/N/r86KVZ0PSlbLiBf1J3/2BiGzOzc7ZfVe05K6yqXatqv1VqY5LxN+HAJE8ei942N83cCzMEjQ+rqr8Yw5PZ9f7mJFdM8v4l62vdMsNn+KzZb8P4nv2yYfTRos/w1Ay/K7uPof3sfTtX1R3m6r0/w+e2a5K3zIKtqtqpqg7LMJrq50kOW+apWHVVdfMM05/eIUOfnjq3+SrZsCbXr829504ZwthNjeRb9PteVTW/dtXUNgCALUDgBAAAlx3/leR1VXVqVZ2W5B0ZAp+3JnnDrFJ3X5Dk1ePLI6rq/VX1niTHJPlOhrWWLskLxucjq+qbVfWVDKMKXjZX5/NJjk9SST5dVV9L8qMMNzArG9bfWTQt2ua2sWzj/6H/kAwjr/4kyU+q6oQkJ2W42XtSkvuP5+hSaxwt9sAMN7p/K8l3q+pbGc7rczMEEAd296lzb3tBkn/PEIy8d7xOTszwOe2d5BMZ1h+6VOvujyV5SYb/3n1lktOr6ntJ3pfkfzbytn8enx9dVZ+rqjdlCJT2SXLQgjZWcn7/Nck3MkzR9/2q+lZVfTXJ9zMEVKdkCDqmODob1nM7duzreRlG3ySrOJ3eGFy9PkM4fHxVfaWqvpnhHKxfrXaW6bpV9fnx8fUkZ2VY12enDOfgosCxuz+bIWC8IMnfJjmjqr491rtVki8m+cMl+39OhiDxcUl+XFVfHT/3YzIEIB/p7uOWvGd2vbwkw2/bx6rqa+Nv3k+SPHOu3oUZrqtvZxhJ+u3x9+f0DNfhuUn+oLu/PeEcTfH4qvpCVX0/w+/C72T43b53dx8/qzSu1fX+8eVbq+obVXVqht+QG2RDWLmx3/cXZbiOn1dV36mq4zKcg3euYhsAwBYgcAIAgMuOx2e4sfqLJFfNcGP1qUkesmBx+f+b5PlJTkzym0lunmFEwG9lEyOcuvvVGdak+USGacJunOGm5Lvm6lyQIcB5WYYb7usyrL3zjxluyH9yrLpwCrfNaWOluvu9SW6T4Wb6GRnW/Tkxw03Q23T3N6a2sTV09xczBBmHZbj5er0MN7zfmeRO3f3GJfUv6O6HJnlkhvN/lQw3a7+Q5OlJ7tbd52y9I1i57n5qkkdlCAJ2SrJjkpcnue9G6r85Q0jypQzX+m8keW+Ga3Hh572C83tWhtEgz8tws/46SW6YIcR8SZL9u/uHKzzkWRtnJ/ny+HJ+ZM7s79WaTm/m0UmekuTrGY7lehl+V/57ldvZXFfIEBbdKsNvwnkZruX/m+S2SwLAdPerktwxw0i0czOsf/WVJH+Z5I7dfdqS+p9Nsn+SIzIEhHtm+P35fJKnJbn3JfTtrzNck1/IMPpmXYYA/4NL2vh2hrW9np8heNojQ3B25HgM78za2S3JLTL8jwqfyhDg7d3dH1lQ98EZwtdvZTiG8zP8jwy3SfLusc7Gft8/mOE7+IEMgeZNMwR9SwPZFbcBAGwZ9b//uxIAANiejKNUbpjkRt194tr2BtZWVXWSdHdtqi4AALD5jHACAAAAAABgEoETAAAAAAAAkwicAAAAAAAAmMQaTgAAAAAAAExihBMAAAAAAACTCJwAAAAAAACYROAEAAAAAADAJAInAAAAAAAAJhE4AQAAAAAAMInACQAAAAAAgEkETgAAAAAAAEyyw1p3ANj+VNUpSXZO8t217gsAAAAAABu1R5Jzuvs6U3dU3b0K/QHYoKrOuuIVr7jLXnvttdZdAQAAAABgI0444YScd955Z3f3VafuywgnYEv47l577bXPcccdt9b9AAAAAABgI/bdd98cf/zxqzJTlTWcAAAAAAAAmETgBAAAAAAAwCQCJwAAAAAAACYROAEAAAAAADCJwAkAAAAAAIBJBE4AAAAAAABMInACAAAAAABgEoETAAAAAAAAkwicAAAAAAAAmETgBAAAAAAAwCQCJwAAAAAAACYROAEAAAAAADCJwAkAAAAAAIBJBE4AAAAAAABMInACAAAAAABgEoETAAAAAAAAkwicAAAAAAAAmETgBAAAAAAAwCQCJwAAAAAAACYROAEAAAAAADCJwAkAAAAAAIBJBE4AAAAAAABMInACAAAAAABgEoETAAAAAAAAkwicAAAAAAAAmETgBAAAAAAAwCQCJwAAAAAAACYROAEAAAAAADDJDmvdAWA79fOTk6Mft9a9AAC2Zbc/fK17AAAAwGYywgkAAAAAAIBJBE4AAAAAAABMInACAAAAAABgEoETAAAAAAAAkwicAAAAAAAAmETgBAAAAAAAwCQCJwAAAAAAACYROAEAAAAAADCJwAkAAAAAAIBJBE4AAAAAAABMInACAAAAAABgEoETAAAAAAAAkwicAAAAAAAAmETgBAAAAAAAwCQCJwAAAAAAACYROAEAAAAAADCJwAkAAAAAAIBJBE4AAAAAAABMInACAAAAAABgEoETAAAAAAAAkwicAAAAAAAAmETgBAAAAAAAwCQCJwAAAAAAACYROAEAAAAAADCJwAkAAAAAAIBJBE4AAAAAAABMInACAAAAAABgEoETAAAAAAAAkwicAAAAAAAAmETgBAAAAAAAwCQCJwAAAAAAACYROAEAAAAAADCJwAkAAAAAAIBJBE4AAAAAAABMInACAAAAAABgEoETAAAAAAAAkwicAAAAAAAAmETgBAAAAAAAwCQCJwAAAAAAACYROAEAAAAAADCJwAkAAAAAAIBJBE4AAAAAAABMInACAAAAAABgEoETAAAAAAAAkwicAAAAAAAAmETgBAAAAAAAwCQCJwAAAAAAACYROAEAAAAAADCJwAkAAAAAAIBJBE4AAAAAAABMInACAAAAAABgEoETAAAAAAAAkwicAAAAAAAAmETgBAAAAAAAwCQCJwAAAAAAACYROAEAAAAAADCJwAkAAAAAAIBJBE4AAAAAAABMInAiSVJV16yqs6vqpKraYcm2Y6vq/Kracyv045Cq6qpat6XbWtD2+rHtA7Z22wAAAAAAsC0TOJEk6e4fJ3llkj2SPGBWXlV3SbJfkiO7+1tr1D0AAAAAAOBSTODEvBcnOTfJQXNlByXpJM/bSn14WZKbJ/n+Vmpv3tvHto9eg7YBAAAAAGCbtcOmq3BZ0d2nVNWrkzyhqvZP8qMk903ytu7+ylbqw4/Gdre67j4zyZlr0TYAAAAAAGzLjHBiqb9P8ssMI5uelOTySf5uTXsEAAAAAABcqgmcuJju/k6SNyT5/SQHJnlvdx+z3P1U1YlVdVRVHVxVp42v96+qQ6vqrKr6alX96oL3HVJVXVXrNrLfqqo/raovVdXPquoHVfWGqrr+grq3q6r3VdUPxzY/XVW/ewl9Xj+2fcCCbevGbYdW1Z5V9a6qOnNs/yVVteOS+jtV1WHj9lOr6uVV9QdVdVxVfXszTuGi/p1YVUeNf/92Vb2/qn5cVTdccI6eUFVfrKpzq+qUqnppVV1lwT6vX1XvrKozquqEqnpSVT1rfM8Rm9Gn4xY9kuy1kmMEAAAAAGDbJHBikednmG7xakmeO2E/+2cYJfWeJDdM8v4kD03y1iQ3TfLnK9jnnyf5pyQXJnllkvcluX+S/6yqy88qVdWNk3w0yR2TvCPJvyS5epJ3VtUdV3g8SbJ7kv8en1+TYc2rg5M8bUm9w5P8aZIPJ/lAkscnOSzJO5O8eUL7qaonjvtcl+TzSc5bUuUVSV6e5Ozx+dgMn8O7qqrm9nPFsX93SfL6JN9J8tIkD0zy6iSfnNJPAAAAAAAuO6zhxCI/yRCk7JTk1An72THJ3br7pKq6X5Jdktx67vX1VrDPhyQ5K8mvdffPk6SqPpfk6UlulOSbY717J7lykod395FjvRcm+XSGgOVTKzymP0pyZJJHdveFVfWCJCckuV/GqQer6lpJHpnkFd39hLHs7Ayh0+HjKLKV2jvJC5I8rLvfuHRjVf1mkscleW13r58r/+ckj0ly1yQfGYvvneQmSR7S3W8aw6gvJ9m1u5+1OZ3p7n0XlY+jnPbZ3IMCAAAAAGDbZoQTizwlyc7j38+csJ8vdfdJ499nLXhdi992iU5LcqUkN5sVdPfLu3tdd39zSb0kuVVVXW6sd8pY7/kraHfmrCQHdfeF4z5PzhDSXHuuzrrx+QtzZV/McLy7T2g7Sa6b5DmLwqbRg2d/jNP/HVpVh2YY3ZUkvzFX90bz/ezuTvKlrCwIBAAAAADgMswIJy6mqnZN8sQkH8swuumRVfWcuaBoOc7exOuVeE6Gqfk+V1VfSfI/ST6Y5C3dPT+13FuSPDXJM5Ksr6qjM0wRd+TEEUYf6u4fLyk7JxcPz05I0kluOVd2i/F5PhRbiVMzTM23MTcenx+1ke3zgdesL7dM8rVxhNOvJvnGpB4CAAAAAHCZI3BiqSdnWLvpxRnCjQcn+bMMIdSa6+6PV9VeSX4vyW2T/GaGcOVpVXXH7j53rHdeVd0hye8m+fUkt8sQVv1VVd2zu49aYRe+tRl9PL2qXpPk8VV1lQxh1MOTvK67f7TCdme+2t2/2Ix61+zun2yizn8k+WqSw6vqN5LsOz7+eGIfAQAAAAC4jDGlHhcZw5GDknw9ybu7+zNJPpHkj6tq6lRwk1XVjlW1f5LLdfcruvvA7r5Jkhcm2S/D+k6zujdJsk93v7O7n9Hdd80QOu2Q5K8ndGOTYU9V7Z3kgCTvTnLPJHdP8vIkfzKh3c01G7V00yV9ulpVPamq7j5XfM8M62p9Jsn6JDdI8oTufs1W6CcAAAAAANsRgRPznpDkGkn+YVzPJ0lelGSnJE9bs15tcMUM4cjLl5QfNz7/ylzZoRmm3bvWXNk3kpy7pN6WcFCG83j/7r52d+/e3U/u7nO2cLvJMJVgkjyjqq4wV35gkpcm2Weu7O+SfK6779ndV+3uvbv7FVuhjwAAAAAAbGdMqUeSpKqulGHNox8led3cpndlGDXzuKp63ipMCbdi3X12Vb0pyUOr6qNJPpdk1wwjm36RYYq4mVcneVCSY6rqPUl+nuQeSa6S5G1buKufyzAF4dFV9ckkZyX5aYbz+IHuXo21rBbq7v+qqn/NEDB9vqo+lOS6Se6f5EtJXrWkn39YVe9O8rUkP0tyRpIvJ/lwd1+4pfoJAAAAAMD2ReDEzGOS7JbkOd3981lhd19YVS/JMKroKUmevUb9m/mjJMcmeUSSxyb5ZYZp//6mu78yq9Td76uq30nyjCQPSHLVDOsvHZxhpM+W9KYkf5pkjyQPTXLNJJcft31vXGvqe1uw/ccm+UKSxyd5XJJTkvx/SQ7p7p/O1fu7DOfmFknunGHtrhq3fbiq7tHdF2zBfgIAAAAAsJ2oDTOnAauhqj6d5OQkD5yNEqqqHZPcKclHkzyluw9bwy6mqnbLMKrpb7r7JXPlO2cIFg9Ncuvu/vwK93/cPje6+j7HvfHBq9JfAOAy6vaHr3UPAAAAtmv77rtvjj/++OO7e9+p+zLCic1WVQ9IcpvNqHpMd2/paesuza6T5JZJ/r2qTsow3d/VktwtSSc5tqpuneSBm7Gv07v7xVugj1dJskuSp1fVLZOcluTCse+/m+THSU7cAu0CAAAAALAdEjixHPdN8qjNqPfabPl1ki7N7pHkr5L8eoZzVklOzTD135+M6yytz+ZNT/idJKseOHX3CVX1uxmmGLxXkmtkCMa+n+SNSQ7r7jNWu10AAAAAALZPAic2W3evT7J+jbtxqdfdX0sjpCyKAAAgAElEQVTy8E3UOSLJEVujP5fQh/cled9a9gEAAAAAgO3D5da6AwAAAAAAAGzbBE4AAAAAAABMInACAAAAAABgEoETAAAAAAAAkwicAAAAAAAAmETgBAAAAAAAwCQCJwAAAAAAACYROAEAAAAAADCJwAkAAAAAAIBJBE4AAAAAAABMInACAAAAAABgEoETAAAAAAAAkwicAAAAAAAAmETgBAAAAAAAwCQCJwAAAAAAACYROAEAAAAAADCJwAkAAAAAAIBJBE4AAAAAAABMInACAAAAAABgEoETAAAAAAAAkwicAAAAAAAAmETgBAAAAAAAwCQCJwAAAAAAACYROAEAAAAAADCJwAkAAAAAAIBJBE4AAAAAAABMInACAAAAAABgEoETAAAAAAAAkwicAAAAAAAAmETgBAAAAAAAwCQCJwAAAAAAACYROAEAAAAAADCJwAkAAAAAAIBJBE4AAAAAAABMInACAAAAAABgEoETAAAAAAAAkwicAAAAAAAAmETgBAAAAAAAwCQCJwAAAAAAACYROAEAAAAAADCJwAkAAAAAAIBJBE4AAAAAAABMInACAAAAAABgEoETAAAAAAAAkwicAAAAAAAAmETgBAAAAAAAwCQCJwAAAAAAACYROAEAAAAAADCJwAkAAAAAAIBJBE4AAAAAAABMInACAAAAAABgEoETAAAAAAAAkwicAAAAAAAAmGSHte4AsJ260u7J7Q9f614AAAAAALAVGOEEAAAAAADAJAInAAAAAAAAJhE4AQAAAAAAMInACQAAAAAAgEkETgAAAAAAAEwicAIAAAAAAGASgRMAAAAAAACTCJwAAAAAAACYROAEAAAAAADAJAInAAAAAAAAJhE4AQAAAAAAMInACQAAAAAAgEkETgAAAAAAAEwicAIAAAAAAGASgRMAAAAAAACTCJwAAAAAAACYROAEAAAAAADAJAInAAAAAAAAJhE4AQAAAAAAMInACQAAAAAAgEkETgAAAAAAAEwicAIAAAAAAGASgRMAAAAAAACTCJwAAAAAAACYROAEAAAAAADAJAInAAAAAAAAJhE4AQAAAAAAMInACQAAAAAAgEkETgAAAAAAAEwicAIAAAAAAGASgRMAAAAAAACT7LDWHQC2TyefnDzucWvdCwAAAADW2uGHr3UPgK3BCCcAAAAAAAAmETgBAAAAAAAwicAJAAAAAACASQROAAAAAAAATCJwAgAAAAAAYBKBEwAAAAAAAJMInAAAAAAAAJhE4AQAAAAAAMAkAicAAAAAAAAmETgBAAAAAAAwicAJAAAAAACASQROAAAAAAAATCJwAgAAAAAAYBKBEwAAAAAAAJMInAAAAAAAAJhE4AQAAAAAAMAkAicAAAAAAAAmETgBAAAAAAAwicAJAAAAAACASQROAAAAAAAATCJwAgAAAAAAYBKBEwAAAAAAAJMInAAAAAAAAJhE4AQAAAAAAMAkAicAAAAAAAAmETgBAAAAAAAwicAJAAAAAACASQROAAAAAAAATCJwAgAAAAAAYBKBEwAAAAAAAJMInAAAAAAAAJhE4AQAAAAAAMAkAicAAAAAAAAmETgBAAAAAAAwicAJAAAAAACASQROAAAAAAAATCJwAgAAAAAAYBKBEwAAAAAAAJMInAAAAAAAAJhE4AQAAAAAAMAkAicAAAAAAAAmETgBAAAAAAAwicAJAAAAAACASQROAAAAAAAATCJwAgAAAAAAYBKBEwAAAAAAAJMInAAAAAAAAJhE4AQAAAAAAMAkAicAAAAAAAAmETgBAAAAAAAwicAJAAAAAACASQROAAAAAAAATCJwAgAAAAAAYBKBEwAAAAAAAJMInAAAAAAAAJhE4AQAAAAAAMAkAicAAAAAAAAmmRQ4VdU1q+rsqjqpqnZYsu3Yqjq/qvac1sXN6schVdVVtW5Lt7Wg7fVj2wds7baXo6puXlXvrarTq+rcqjqmqm66oN5mH89Y74jNqHdEVfXKej7Nal0bVXVUVZ24Kp0CAAAAAIDtzKTAqbt/nOSVSfZI8oBZeVXdJcl+SY7s7m9N6iGTVdVVkrw/ye2TvDbJK5Kcl+Q6a9kvAAAAAABg+7AaU+q9OMm5SQ6aKzsoSSd53irsf3O8LMnNk3x/K7U37+1j20evQdub604ZQsGDu/vg7n5Kd98xyccW1N0Wjmc5Nuva2NzRWgAAAAAAwP+2w6arXLLuPqWqXp3kCVW1f5IfJblvkrd191em7n8z+/Cjsd2trrvPTHLmWrS9DLuPz9+bL+zu/zXN3TZyPJttLa8NAAAAAAC4rFiNEU5J8vdJfplhZNOTklw+yd+t0r6Zrta6AwAAAAAAwPZrVQKn7v5Okjck+f0kByZ5b3cfs9z9VNWJVXVUVR1cVaeNr/evqkOr6qyq+mpV/eqC9x0yTom2biP7rar606r6UlX9rKp+UFVvqKrrL6h7u6p6X1X9cGzz01X1u5fQ5/Vj2wcs2LZu3HZoVe1ZVe+qqjPH9l9SVTsuqb9TVR02bj+1ql5eVX9QVcdV1bc34xQuaruTvGYs/uisbGPTx13S8ayV8fM7s6o+Mr7eZezj9+bqnFJVn1nw3o1eG+O1NjtHSfKoufPzv0Z/je/Zoar+tqq+W1VnV9VHq+pmKzyuI+bbW/I4ZEndK1TVX1TV16vqvKr6TlU9p6p2WFJv9rkfMp63x1fVZ6vqywvav3pVvbSqTq6qc6vqf6rq91ZyLAAAAAAAXLZNnlJvzvOTPCrJ1ZI8d8J+9k9y/STvSbI+yfuTnJ7krePrP0/yiGXu888zjLj6YpJXJrl6kockuVVV7dfdFyRJVd04yUeTXJDk35OcneTeSd5ZVb/R3Z9a4THtnuS/M6wj9JoMUw4enOS0XHwk2OEZju3I8fXjkzwoyauy/M/qjGz4HG6Z5D4ZQsHvjGXLDgTXSnd3VR2T5KZj0T7j8/Wq6qoZ1gvbLcM1shyvz/C5JMmzM1wf797Ee16T5F5J/l+SPZPcM8nbquoWs+toGd6VJdMcZrg2bpHklFlBVV0uyTvGdt8/9vEOSf4yyQ0yfC+WqiRvzHD9HJPh2DZsrLpakk9kWN/qg0m+Obb99qo6qLv/aXMOoKqO28imvTbn/QAAAAAAbB9WM3D6SZJzk+yU5NQJ+9kxyd26+6Squl+SXZLceu719Vawz4ckOSvJr3X3z5Okqj6X5OlJbpThZnsyhEtXTvLw7j5yrPfCJJ9OcpckKw2c/ihDiPTI7r6wql6Q5IQk98sYOFXVtZI8MskruvsJY9nZGUKnw8dRZJutu89I8hfjftZnCJxe1d1HrfAY1tpnkxxQVVfJEDgdmyHw2SfJ+WOd/1nODrv7VbO/q+rZSY7t7r+4hLfcIMndMl6P4/vekeFzvEmSZa1Z1t1vS/K2uT78ZpJnJHlfkn+eq/qIDGHT33T3IXP135/kkVX1t919wpLdPzLD9/s23f2FBc0fmiFsenZ3z67Bv8xwrb+4qt7T3d9azvEAAAAAAHDZtVprOCXJU5LsPP79zAn7+dLsZn6GkGjp65WsR3RakisluWjqs+5+eXev6+5vLqmXDCOfLjfWO2Ws9/wVtDtzVpKDuvvCcZ8nJ/lykmvP1Vk3Ps+HA1/McLy7T2h7e/HZ8XnvDEHJ5zMEPDfPEPbM19lSKsmz5q7HJPnA+HztBfU3f8fD9I5vTvLdJA+bXSujB4/Pu47TMx5aVYeO/akkd16wyxsmefCisKmqKsnDM4yi+vtZeXf/OMOouB2SPHRz+t3d+y56ZAhUAQAAAAC4jFiVEU5VtWuSJyb5WIbRTY+squcsuTG/uc7exOuVeE6Gqcg+V1VfyTAS5oNJ3tLd583Ve0uSp2YYZbK+qo5O8skkRy53hNESHxpv5s87JxcPz07IMDXcLefKbjE+z4dil1WzMOmmGUY1fTTJhePf5yT5WZY5wmgFOkMoNO+c8XklQejwxqorZhjpdJUkv9Pdpy+pcuPx+aCN7GJRIPmO7v70RupfO8O0kh/p7vOXbJsFVCtalwoAAAAAgMum1Rrh9OQMaze9eHxcIcmfrdK+J+vuj2dYU+aJGabFu1OG9Yw+XVU7zdU7L8PaOL+X5LUZAoDnJPlqVR0woQubnJpsDBlek+TxVfXaqnpdhun0XtfdP5rQ9nZhnDLuJxlGM+2T5Ljxsc9YduwK1lBarlO6+5xNV1u2VyS5XZLHbmT6uyQ5u7trI49Fo+++uKBsqV7hNgAAAAAAuJjJgdO4ps5BSb6e5N3d/Zkkn0jyx1W15lPBVdWOVbV/kst19yu6+8DuvkmSFybZL8P6TrO6N0myT3e/s7uf0d13zRAE7JDkryd04xeb0c+9kxyQ5N1J7pnk7klenuRPJrS7vTkmya0yTBe3NHCaOp3e5gQsm/wcl6uqnphhja9/7O5/20i1bybZpaquu+S9e1TVk6rqDsts9rQkpyfZp6qWjnKcjar72jL3CQAAAADAZdhqjHB6QpJrJPmH7p7dtH9Rkp2SPG0V9j/VFZN8JkN4M++48flX5soOzTDt3rXmyr6R5Nwl9baEgzKcx/t397W7e/fufvIWGlGzrfpskgck+Wl3fzfDZ7guyW0yTJM4xUlJbjRfsCCMWVVVdeckL8kwFeUlfVfeMj4/c1x/aeYZSV6a5DrLaXdcH+rfMkzF93/n+nONJM9Ocn6SNy5nnwAAAAAAXLZNuqFeVVfKsObRj5K8bm7TuzKMynhcVT1vLaeE6+6zq+pNSR5aVR9N8rkku2YY2fSLJP8xV/3VSR6U5Jiqek+Snye5R4ap9d62hbv6uQxT/h1dVZ/M/8/e3cdrXtd1Hn9/FJVIMcIVJ7VFKW9mNrzJm8RUpEe5amBCrLYVjK1KeYe0rdniFqGllqRtmc3uomBmamol5ipaoC3lDULezHiTZKCGJoiACnjDZ//4XUeOwxnmnPmec65h5vl8POZxzfldv/P7fa9zrrm4OK/z/f6Sq5J8JdPX8ezuXo1rWd3cLcxi2pYk3f25qvpypu/n6Ayn1yQ5uar+PNMSiIckuTzJCYPHvSlvzLT85CeTnPKdLSlXdPdps7//SZKfSfKMJA+uqvMyXcvq0ZmuRfbWXTj385L8WJIXVdUjM11D7HFJ7pzkxO7e6TKQAAAAAACwYHQGx1OSHJTk1O6+ZmFjd19fVS/NNKvopEyzJubpSUkuTPJzSZ6a5BuZlv37ze7+2MJO3f32qvqJTDNHjk6yf6b48OxMM0nW0huSPDPJXZM8McmBSW45u++zVfWQ7v7sGo9hd7cwi2nbom3bkmzKNBNtxKmz259P8vgkX0jywsFj7sxBs9snL3HfxZmuh7bw7+moJM9NclymZRY/k2nML96Va1d195VV9dAkz88UWQ9P8pEkz+zuv1jp8QAAAAAA2LvVDavgMU9V9d4klyY5ZrbkWarq1kkOS3JOkpO6+2VzHCIsW1VtPeCAjRuPPXbrzncGAAAAYI+2Zcu8RwDsyKZNm7Jt27Zt3b1p9Fhreo2aJKmqozNdY2dnLujutV62bnd2pySHJnl9VV2Sabm/2yc5IkknubCq7pfkmGUca/FybLutqjoi0+PbmU939+lrPZ5Re9rjAQAAAACA5Vrz4JTkqCTHL2O/M7P210nanT0qya8neWimr1llWtbtvCS/1N3vrqrNWd7yhN9ejm039/As7/G8O8nNIdDsaY8HAAAAAACWxZJ6wKqzpB4AAAAACyypB7uv1VxS7xarMSAAAAAAAAD2XoITAAAAAAAAQwQnAAAAAAAAhghOAAAAAAAADBGcAAAAAAAAGCI4AQAAAAAAMERwAgAAAAAAYIjgBAAAAAAAwBDBCQAAAAAAgCGCEwAAAAAAAEMEJwAAAAAAAIYITgAAAAAAAAwRnAAAAAAAABgiOAEAAAAAADBEcAIAAAAAAGCI4AQAAAAAAMAQwQkAAAAAAIAhghMAAAAAAABDBCcAAAAAAACGCE4AAAAAAAAMEZwAAAAAAAAYIjgBAAAAAAAwRHACAAAAAABgiOAEAAAAAADAEMEJAAAAAACAIYITAAAAAAAAQwQnAAAAAAAAhghOAAAAAAAADBGcAAAAAAAAGCI4AQAAAAAAMERwAgAAAAAAYIjgBAAAAAAAwBDBCQAAAAAAgCGCEwAAAAAAAEMEJwAAAAAAAIYITgAAAAAAAAwRnAAAAAAAABgiOAEAAAAAADBEcAIAAAAAAGCI4AQAAAAAAMAQwQkAAAAAAIAhghMAAAAAAABDBCcAAAAAAACGCE4AAAAAAAAMEZwAAAAAAAAYIjgBAAAAAAAwRHACAAAAAABgiOAEAAAAAADAEMEJAAAAAACAIYITAAAAAAAAQwQnAAAAAAAAhghOAAAAAAAADBGcAAAAAAAAGLLPvAcA7Jk2bEi2bJn3KAAAAAAAWA9mOAEAAAAAADBEcAIAAAAAAGCI4AQAAAAAAMAQwQkAAAAAAIAhghMAAAAAAABDBCcAAAAAAACGCE4AAAAAAAAMEZwAAAAAAAAYIjgBAAAAAAAwRHACAAAAAABgiOAEAAAAAADAEMEJAAAAAACAIYITAAAAAAAAQwQnAAAAAAAAhghOAAAAAAAADBGcAAAAAAAAGCI4AQAAAAAAMERwAgAAAAAAYIjgBAAAAAAAwBDBCQAAAAAAgCGCEwAAAAAAAEMEJwAAAAAAAIYITgAAAAAAAAwRnAAAAAAAABgiOAEAAAAAADBEcAIAAAAAAGCI4AQAAAAAAMAQwQkAAAAAAIAhghMAAAAAAABDBCcAAAAAAACGCE4AAAAAAAAMEZwAAAAAAAAYss+8BwDsmS69+tKccNYJ8x4GAAAA7FG2HLll3kMAgCWZ4QQAAAAAAMAQwQkAAAAAAIAhghMAAAAAAABDBCcAAAAAAACGCE4AAAAAAAAMEZwAAAAAAAAYIjgBAAAAAAAwRHACAAAAAABgiOAEAAAAAADAEMEJAAAAAACAIYITAAAAAAAAQwQnAAAAAAAAhghOAAAAAAAADBGcAAAAAAAAGCI4AQAAAAAAMERwAgAAAAAAYIjgBAAAAAAAwBDBCQAAAAAAgCGCEwAAAAAAAEMEJwAAAAAAAIYITgAAAAAAAAwRnAAAAAAAABgiOAEAAAAAADBEcAIAAAAAAGCI4AQAAAAAAMAQwQkAAAAAAIAhghMAAAAAAABDBCcAAAAAAACGCE4AAAAAAAAMEZwAAAAAAAAYIjgBAAAAAAAwRHACAAAAAABgiOAEAAAAAADAEMEJAAAAAACAIYITAAAAAAAAQwQnAAAAAAAAhghOAAAAAAAADBGcAAAAAAAAGCI4AQAAAAAAMERwAgAAAAAAYIjgBAAAAAAAwBDBCQAAAAAAgCGCEwAAAAAAAEMEJwAAAAAAAIYITgAAAAAAAAwRnAAAAAAAABgiOAEAAAAAADBEcAIAAAAAAGCI4AQAAAAAAMAQwQkAAAAAAIAhghMAAAAAAABDBCcAAAAAAACGCE4AAAAAAAAMEZwAAAAAAAAYIjgBAAAAAAAwRHACAAAAAABgiOAEAAAAAADAkKHgVFUHVtXVVXVJVe2z3X0XVtU3q+ruY0Nc1jhOqaquqoPX+lxLnHvz7NyHr/e5V6Kq7l1Vb6uqK6rq2qq6oKruucR+y348s/3OWMZ+Z1RV79rIx6zWc6Oqzq2qf1mVQQEAAAAAwB5mKDh19+VJ/jjJXZMcvbC9qh6R5L5JXtvd/zw0QoZV1W2TvCPJg5KcmeQVSa5Lcqd5jgsAAAAAANgzrMaSeqcluTbJiYu2nZikk7xwFY6/HH+Y5N5JPrdO51vsL2bnfv8czr1ch2WKgs/u7md390nd/ZAk71li35vD41mJZT03ljtbCwAAAAAAuLF9dr7LTevuz1fVK5M8raoekOSyJEcleXN3f2z0+Mscw2Wz86677r4yyZXzOPcKbJjdfnbxxu6+0TJ3N5PHs2zzfG4AAAAAAMDeYjVmOCXJ7yT5RqaZTc9Icsskv71Kx2ZczXsAAAAAAADAnmtVglN3X5zkNUn+U5InJ3lbd1+w0uNU1b9U1blV9eyq+uLs4wdU1Quq6qqq+nhV/YclPu+U2ZJoB+/guFVVz6yqj1TVV6vqX6vqNVV1lyX2fWBVvb2q/m12zvdW1WNvYsybZ+c+fIn7Dp7d94KquntVvaWqrpyd/6VVdevt9t+3ql42u/8LVfXyqvqZqtpaVZ9expdwqXN3klfNNp+zsG1Hy8fd1OOZl9n378qq+tvZx7ebjfGzi/b5fFW9b4nP3eFzY/ZcW/gaJcnxi74+N5r9Nfucfarq+VX1maq6uqrOqap77cJjOnf2/P6Z2bH+raoeU1VPr6rLquqSqnrkdl+Dp1XVh6vq2tnj/YPZ9bm2P/YRs+NfMft39PaqOnS7fc6YPc59Zv82PjX7t/GBqvqRlT4eAAAAAAD2bsNL6i3yoiTHJ7l9kt8aOM4DktwlyVuTbE7yjiRXJHnT7OPnJvm5FR7zuZlmXH04yR8nOSDJE5Lcp6ru293fSpKq+oEk5yT5VpLXJ7k6yU8m+auqelh3/8MuPqYNSf5fpusIvSrTkoPPTvLFfOdMsC2ZHttrZx//YpKfTnJ6Vv69+nJu+D4cmuTITFHw4tm2FQfBeenurqoLktxztmnj7PbOVbV/puuFHZTpObISf5Lp+5IkJ2d6fpy1k895VZLHJPmzJHdP8ugkb66qH1p4Hq3AQUl+d3bOp2T6/lyX5M9nHz8/yY/O9n1FkhOS/H2Sl2f6Gjwjyaaq+rGF5RFnofAdSS5P8qeZZrcdk+RdVXXv7r58uzGckuSkTM+5/TNF47Oq6pDuvmpnD6Cqtu7grkN29rkAAAAAAOw5VjM4fSnJtUn2TfKFgePcOskR3X1JVT0uye2S3G/Rx3fehWM+IclVSX6ku69Jkqr6YJL/luRuST412+8nk3x3kp/t7tfO9vvdJO9N8ogkuxqcnpTpB/rHdff1VfXiJBcleVxmwamq7pDkuCSv6O6nzbZdnSk6bZnNIlu27v5ykufNjrM5U3A6vbvP3cXHMG/nJzl8NqNnY5ILMwWfjUm+OdvnAys5YHefvvD3qjo5yYXd/byb+JTvT3JEZs/H2ef9Zabv4z2SrPSaZfsmOba7/6GqHpzkfkkOW/TxnWfneHim2HRmd29eNOb/lSlMPTLJ3842nzi7/bHu3jrb74wk788Uyv5kuzGcmORhCzMSq+rS2baHJvm/K3w8AAAAAADspVYzOJ2UZL8k1yf5tUxL6+2Kjyz8MD9TJPr0dh/vyvWIvpgpTNwrU6hId78800yR7fdLpplPr+vu67v780kO3oVzLnZVkhO7+/rZuS+tqo8mueOifRbO8aFF2z6c6fFuyA0zk/ZW589ufzDJvZP8Y6bZQPee3S7eZ61Ukv++6PmYJGdnCk53zMqD05cWzZq7aomPv3f292O/PYCqFyz6/ANmtw/LLDh19+O/Y8BVB2YKUsnSz+OXbbf85dmZgtMdl9j3Rrp701LbZzOfNi51HwAAAAAAe55VCU5V9T1Jnp7kPZlmNx1XVadu94P55bp6Jx/vilMzLTP2war6WKaZMO9M8sbuvm7Rfm9M8stJnpNkc1W9P9MSZq9d6Qyj7bxriaXMvpbvjGcXZVoabvG1dn5odvupsBCT7pkpZJyTKW5uzPS1/GpWHnxWqjMtd7fY12a3uxJCl/tc/4HZ7fE7uH/D4g+q6qhMkeqhmSLT9bO7brnE575+u49HHg8AAAAAAHupW6zScZ6V6dpNp83+3CrJr67SsYd1999luqbM0zMti3dYpuvlvLeq9l2033VJHpzkp5KcmeS2mWLVx2fXxtlV/7yMMV6R6fpAv1hVZ1bVqzMtp/fq7r5s4Nx7hO6+KNOyjffIFJm2zv5snG27cBeuobRSn+/ur+18tzVzYHfXEn9+cWGHqvrDJH+VaebX6zJdA+xuN3HMnT43AQAAAABgZ4ZnOM2uqXNikk8mOau7u6rOS/ILVfWC7r509ByD47t1pllDl3b3KxZt/51M13B6Qqa4lKq6R5J9u/uvMv3QPlV130wzon4jybm7OIyvL2OcP5jk8CRnJXl0km9lWvLvubt4zj3RBUnuk+TfZ4pN12eKnV/KNLtuRC9jn51+H9fIwgy3e2bRdcSq6vZJfj7JJ7r7nVV150xR9V3d/eOL9rup5fHm9ZgAAAAAANiDrMYMp6dlutbM73X3wg/tX5Jk3yS/sgrHH3WbJO/Lja/XtHV2++8WbXtBpmX37rBo2z8luXa7/dbCiZm+jo/v7jt294buftacZ9Tsbs5PcnSSr3T3ZzJ9Dw9Ocv9MUXDEJdluJlBVreY1zka8cXb7nKq61aLtT07yB7nhWkkHzW6/HXmrar8kW9Z8hAAAAAAA7NWGfqBeVd+V6ZpHlyV59aK73pJpVsYJVfXCeS4J191XV9Ubkjyxqs5J8sEk35NpZtPXk/z1ot1fmWkJsguq6q1JrknyqExL6715jYf6wUyzU95fVX+f5KokX8n0dTy7u1fjWlY3dwvXcdqWJN39uar6cqbv5/k7/KzleU2Sk6vqzzMtM3dIksuTnDB43GHd/e6q+j+ZAtM/VtW7knxfkscn+UiS02e7bssUm/5zVV2f6fl7VJL9Zvfvv64DBwAAAABgrzE6w+kpmWZV/FF3X7OwsbuvT/LSJN+d5KTBc6yGJ2W6ptSBSZ6a6Qf15yU5vLs/trBTd789yU8k+XimmTS/NLvr2UlOWeMxviHJhUnumuSJSX4tyYsyzW7ZVlV3WePz3xwszGLatmjbtiRXZpqJNuLUJL+V5EFJ/muSh2SKObuLpyZ5Zqal/05I8sAkf5TpOfyVJOnuazMtx3hOpufvEzI9zx+UKQo/drsZUgAAAAAAsCrqhlXwmKeqem+m2SnHzILdwvWnDssUEE7q7pfNcYiwbFW19YC7HrDx2JcfO+fGI2IAABlnSURBVO+hAAAAwB5ly5FWzgdg9WzatCnbtm3b1t2bRo+15teoqaqjM11jZ2cu6O61XrZud3anJIcmeX1VXZJpub/bJzki06yWC6vqfkmOWcaxruju09ZspKukqo7I9Ph25tPdffrOd5uvPe3xAAAAAADAcq15cMp0DZnjl7HfmVn76yTtzh6V5NeTPDTT16ySfCHTkmi/NLuOz+YkJy/jWBcn2e2DU5KHZ3mP59254TpFu7M97fEAAAAAAMCyrHlw6u7NSTav9Xlu7rr7E0l+dif7nJHkjPUYz3ro7lOy9tfGWjd72uMBAAAAAIDlusW8BwAAAAAAAMDNm+AEAAAAAADAEMEJAAAAAACAIYITAAAAAAAAQwQnAAAAAAAAhghOAAAAAAAADBGcAAAAAAAAGCI4AQAAAAAAMERwAgAAAAAAYIjgBAAAAAAAwBDBCQAAAAAAgCGCEwAAAAAAAEMEJwAAAAAAAIYITgAAAAAAAAwRnAAAAAAAABgiOAEAAAAAADBEcAIAAAAAAGCI4AQAAAAAAMAQwQkAAAAAAIAhghMAAAAAAABDBCcAAAAAAACGCE4AAAAAAAAMEZwAAAAAAAAYIjgBAAAAAAAwRHACAAAAAABgiOAEAAAAAADAEMEJAAAAAACAIYITAAAAAAAAQwQnAAAAAAAAhghOAAAAAAAADBGcAAAAAAAAGCI4AQAAAAAAMERwAgAAAAAAYIjgBAAAAAAAwBDBCQAAAAAAgCGCEwAAAAAAAEMEJwAAAAAAAIYITgAAAAAAAAwRnAAAAAAAABgiOAEAAAAAADBEcAIAAAAAAGCI4AQAAAAAAMAQwQkAAAAAAIAhghMAAAAAAABDBCcAAAAAAACGCE4AAAAAAAAMEZwAAAAAAAAYIjgBAAAAAAAwRHACAAAAAABgiOAEAAAAAADAEMEJAAAAAACAIYITAAAAAAAAQwQnAAAAAAAAhuwz7wEAe6YNt9uQLUdumfcwAAAAAABYB2Y4AQAAAAAAMERwAgAAAAAAYIjgBAAAAAAAwBDBCQAAAAAAgCGCEwAAAAAAAEMEJwAAAAAAAIYITgAAAAAAAAwRnAAAAAAAABgiOAEAAAAAADBEcAIAAAAAAGCI4AQAAAAAAMAQwQkAAAAAAIAhghMAAAAAAABDBCcAAAAAAACGCE4AAAAAAAAMEZwAAAAAAAAYIjgBAAAAAAAwRHACAAAAAABgiOAEAAAAAADAEMEJAAAAAACAIYITAAAAAAAAQwQnAAAAAAAAhghOAAAAAAAADBGcAAAAAAAAGCI4AQAAAAAAMERwAgAAAAAAYIjgBAAAAAAAwBDBCQAAAAAAgCGCEwAAAAAAAEMEJwAAAAAAAIYITgAAAAAAAAzZZ94DAPZQl16anHDCvEfB7mDLlnmPAAAAAABYY2Y4AQAAAAAAMERwAgAAAAAAYIjgBAAAAAAAwBDBCQAAAAAAgCGCEwAAAAAAAEMEJwAAAAAAAIYITgAAAAAAAAwRnAAAAAAAABgiOAEAAAAAADBEcAIAAAAAAGCI4AQAAAAAAMAQwQkAAAAAAIAhghMAAAAAAABDBCcAAAAAAACGCE4AAAAAAAAMEZwAAAAAAAAYIjgBAAAAAAAwRHACAAAAAABgiOAEAAAAAADAEMEJAAAAAACAIYITAAAAAAAAQwQnAAAAAAAAhghOAAAAAAAADBGcAAAAAAAAGCI4AQAAAAAAMERwAgAAAAAAYIjgBAAAAAAAwBDBCQAAAAAAgCGCEwAAAAAAAEMEJwAAAAAAAIYITgAAAAAAAAwRnAAAAAAAABgiOAEAAAAAADBEcAIAAAAAAGCI4AQAAAAAAMAQwQkAAAAAAIAhghMAAAAAAABDBCcAAAAAAACGCE4AAAAAAAAMEZwAAAAAAAAYIjgBAAAAAAAwRHACAAAAAABgiOAEAAAAAADAEMEJAAAAAACAIYITAAAAAAAAQwQnAAAAAAAAhghOAAAAAAAADBGcAAAAAAAAGCI4AQAAAAAAMERwAgAAAAAAYIjgBAAAAAAAwBDBCQAAAAAAgCGCEwAAAAAAAEMEJwAAAAAAAIYITgAAAAAAAAwRnAAAAAAAABgiOAEAAAAAADBEcAIAAAAAAGCI4ATLVFUHVtXVVXVJVe2z3X0XVtU3q+ruA8d/bFWdX1Vfq6qrquqdg+M9vKq6qjYvY99/qapzR84HAAAAAMDeS3CCZeruy5P8cZK7Jjl6YXtVPSLJfZO8trv/eVeOXVU/nOQvk9wmycuTvCbJLscrAAAAAABYT/vsfBdgkdOSPCPJiUneMNt2YpJO8sKB4x6T6d/jo7v7s0lSVYIwAAAAAAA3C36gDSvQ3Z9P8sokh1XVA6rq4CRHJXlzd39s4NAbZsf/7KJzXT9wPAAAAAAAWDeCE6zc7yT5RqaZTc9Icsskvz14zBodFAAAAAAAzIsl9WCFuvviqnpNkp9Nck2St3X3BSs9TlVtTvKq7bb1og+f1N1nzLafm+Tg7j54B8d4ZHefu9IxjKqqrTu465B1HQgAAAAAAHMlOMGueVGS45PcPslv7eIxPrToc49Mcuh2x/rQLo8OAAAAAADWkeAEu+ZLSa5Nsm+SL+zKAbr7wiQXJklV3SXJod39vFUb4Tro7k1LbZ/NfNq4zsMBAAAAAGBOXMMJds1JSfab/f3X5jkQAAAAAACYN8EJVqiqvifJ05O8J8mbkhxXVd8/p+Hcck7nBQAAAACAbxOcYOWelenaTafN/twqya/OaSzzCl0AAAAAAPBtghOsQFXdNsmJST6Z5Kzufl+S85L8QlVtWMNTd5a+5trD1/CcAAAAAACwLIITrMzTknxvkt/r7p5te0mSfZP8yhqe94tJ7lRVByxsqKqfSHL4Gp4TAAAAAACWRXCCZaqq70ryy0kuS/LqRXe9JcmnkpxQVXdYo9O/JdP1ms6oqodU1VOSvD7JO9bofAAAAAAAsGyCEyzfU5IclOSPuvuahY3dfX2Slyb57iQnrdG5/zTJrye5X5K/SfLUJD+X5HVrdD4AAAAAAFi2umFVMIDVUVVbNx5wwMatxx4776GwO9iyZd4jAAAAAACWsGnTpmzbtm1bd28aPdY+qzEgYFJVRye5/zJ2vaC737zW4wEAAAAAgPUgOMHqOirJ8cvY78wkghMAAAAAAHsEwQlWUXdvTrJ5zsMAAAAAAIB1dYt5DwAAAAAAAICbN8EJAAAAAACAIYITAAAAAAAAQwQnAAAAAAAAhghOAAAAAAAADBGcAAAAAAAAGCI4AQAAAAAAMERwAgAAAAAAYIjgBAAAAAAAwBDBCQAAAAAAgCGCEwAAAAAAAEMEJwAAAAAAAIYITgAAAAAAAAwRnAAAAAAAABgiOAEAAAAAADBEcAIAAAAAAGCI4AQAAAAAAMAQwQkAAAAAAIAhghMAAAAAAABDBCcAAAAAAACGCE4AAAAAAAAMEZwAAAAAAAAYIjgBAAAAAAAwRHACAAAAAABgiOAEAAAAAADAEMEJAAAAAACAIYITAAAAAAAAQwQnAAAAAAAAhghOAAAAAAAADBGcAAAAAAAAGCI4AQAAAAAAMERwAgAAAAAAYIjgBAAAAAAAwBDBCQAAAAAAgCGCEwAAAAAAAEMEJwAAAAAAAIYITgAAAAAAAAwRnAAAAAAAABgiOAEAAAAAADBEcAIAAAAAAGCI4AQAAAAAAMAQwQkAAAAAAIAhghMAAAAAAABDBCcAAAAAAACGCE4AAAAAAAAMEZwAAAAAAAAYIjgBAAAAAAAwRHACAAAAAABgiOAEAAAAAADAEMEJAAAAAACAIYITAAAAAAAAQwQnAAAAAAAAhuwz7wEAe6gNG5ItW+Y9CgAAAAAA1oEZTgAAAAAAAAwRnAAAAAAAABgiOAEAAAAAADBEcAIAAAAAAGCI4AQAAAAAAMAQwQkAAAAAAIAhghMAAAAAAABDBCcAAAAAAACGCE4AAAAAAAAMEZwAAAAAAAAYIjgBAAAAAAAwRHACAAAAAABgiOAEAAAAAADAEMEJAAAAAACAIYITAAAAAAAAQwQnAAAAAAAAhghOAAAAAAAADBGcAAAAAAAAGCI4AQAAAAAAMERwAgAAAAAAYIjgBAAAAAAAwBDBCQAAAAAAgCGCEwAAAAAAAEMEJwAAAAAAAIYITgAAAAAAAAwRnAAAAAAAABgiOAEAAAAAADBEcAIAAAAAAGCI4AQAAAAAAMCQ6u55jwHYw1TVVbe5zW1ud8ghh8x7KAAAAAAA7MBFF12U66677uru3n/0WIITsOqq6vNJ9kvymWV+ykKZumhtRgQwxGsUsLvy+gTsrrw+Abszr1Hwne6a5GvdfafRAwlOwNxV1dYk6e5N8x4LwPa8RgG7K69PwO7K6xOwO/MaBWvHNZwAAAAAAAAYIjgBAAAAAAAwRHACAAAAAABgiOAEAAAAAADAEMEJAAAAAACAIdXd8x4DAAAAAAAAN2NmOAEAAAAAADBEcAIAAAAAAGCI4AQAAAAAAMAQwQkAAAAAAIAhghMAAAAAAABDBCcAAAAAAACGCE4AAAAAAAAMEZwAAAAAAAAYIjgB66Kq7lNVZ1fVVVV1XVV9sqqeuMR+j6mq91fVV6tqW1UdOY/xAnuX5bxGVdVRVdU7+HP4nIYO7AWq6ser6ver6pVV9ZyqutMS+xxfVR+tqmuq6vyqesg8xgrsfXb2GlVVz7qJ91AHz2fUwN6mqn6zqnqJ7d5DwSqq7hv9OwNYVVV13yTnJdkvyfuSfDXJ4Zmi909395tm+x2b5PVJLkvyZ0k2Jnlkkkd299+t/8iBvcEKXqNOSfK8JC9a4jCnd/en12O8wN6lqn4/ybOSXJ7k6iQHJ/lipvdHW2f7/EqS301ycZI3J/nRJJuS3Le7/2kOwwb2Est8jTojyZFJXrHEIV7S3V9el8ECe62qumeSDyW5TXfXou3eQ8EqE5yANVdV5yW5T5LHdfffzLb9eJKzk3y4u+9TVfsnuSjTD3jv390Xz/Z7a5K7dfem+Ywe2NMt5zVqtu2sJD/Y3fea22CBvUpVHZPkjUl+I8nzu7ur6qgkb0pydnc/djY74BNJPpPkgd19RVXdKskHkvxrdz9mPqMH9nTLeY2a7feRJJ/r7v84v9ECe7OqOifTLxVmITh5DwVrw5J6wJqqqnsleVCS5y38IDdJuvudSbYmObSqbpnkp5LcIcnvL8SmmRcn2TibgQCwqlbwGpUk90/yj+s/SmAvdnKSD3T3qT37TcHufkuS8zP9Bm6SHJfk1klO7e4rZvt8I8lpSR5VVQeu/7CBvcROX6Oq6ruS3CveQwFzUlVPyhSbPrPdXd5DwRrYZ94DAPZs3f3xqrptkuuXuPtWSb45+/ths9uzttvn75Ncm+nNgf9JAVbVcl+jZtci+L5Mv+0GsF6OS/L1JbbfKtPyn8n0HqqTvHW7ff420y8YPizJX67VAIG92nJeo+6T6WdP3kMB666q7pBpybz3J3lHkv+x6G7voWANmOEErLnuvm72WyLfVlV3S3KPTL8R960kB2X6ge/W7T73W5nW0r37Og0X2Mss8zXqh2d3PaqqPl1V11bVJ6rq5Kq69XqPGdg7dPdHu/uTi7dV1SMyzbh822zTQZmWqvrSdp/7uSTXxXsoYI0s8zVq4T3U8VX1uaq6pqo+VFVPqyo/kwLW2u8l2T/Jf8mNf8nQeyhYA/7jDszLr85uFy4cu3+SK7t7qd+QuzLTzAKA9bL9a9QDZrcPTvKe2favJ3lBktet79CAvVFVPbqq/mem68t9JMlzZnftn+SLO/g076GAdXETr1GL30P9dZL/nel16+VJXrLe4wT2HlV1RJKfT/Li7v7oErt4DwVrwJJ6wLqrqgcmeXKSDyV57WzzN5Ncs4NP+XqS/dZhaAA7eo36eqYlGDZ39/9v715DLSvLOID/H9PsZuqkkVSgSTVdnHJA0yRLS7CILnQhizCoPnX5UBaVMklNWHaPIPwQGFEiplEUYUVaoZZOUTNZSdLtQxeytCzFyebpw1qHNptzpoF99j4zs38/2Ky13/WufZ7z5eXd+7/Wu/409ntAkmuSvLSqtnb3jzeiXmBpfCTJk8f9zye5c9w3hwL2B2uNUXcnuTrDHOqfSVJV70pyfZI3V9WlK3MrgPVSVYdnuEjwlxkuElyNORTMgTucgIUaHxr7uQzr5L5+XKoqGdb4ftAapx2W4UGOAHO11hjV3Zd097mTP4iMxy4a35638GKBZXNShjsEvp3hWQQrzyAwhwL2B6uOUd391u5++UrYNLbdk+T9Gcaol21ArcDB78Ikj0/yhu6+b40+5lAwBwInYNE+leRJSbZ1948m2v+QZFNVPXyVcx6Z5K5FFAcsvbXGqLWsPLfgxPmVBJB0957uvjnJCzOMPe+oqodkmEMdP91/vAtzU8yhgAXYyxi1FnMoYC6qanOGJdI/09037KWrORTMgcAJWJiqOj/DMlXXJvnQ1OGfjtutU+cck+SEJH+ee4HAUtvbGFVVR1TVI1Y5bdO4XevKOICZVFVNvh+v0r0uycMyPFvgpxku2jl+6tSTM1ydaw4FzM3/G6Oq6piqOmKVU82hgHn5QIa7k+6vqu0rryRnJ8nEe3MomAOBE7AQVXVGksuS/D7Ja7p7z1SXbyTZk+HH3kkry1R9b74VAstsb2PU+CPJX5K8e5VTnz9ufz73IoGlUlVHV9UfMywJM+2ocfuvJF8f982hgIXZxzHq2Aw/2J6/Sh9zKGBeVi4UfGuGMWrl9ayxfeW9ORTMQXX3RtcAHOSq6ilJbkhyeJIzuvvHa/S7KsMa3m/K8MPv6Um+Mh5+XHf/YwHlAktmX8aoqvpehivdTu3uX4xtm5N8J8mjkmzt7p8srmpgGVTVzUk2Zxibdo1tT8hwRe6vunvL2HZLhuennNfdX66qlyS5IkOIvrl96QPm4P+NUUmenuTXSR6QYQ71x7HPM5N8c2w/YfIZmQDzUlUXJ3lvd9dEmzkUrDOBEzB3VXVdkuckuSXDF4tpn+3u31TVsUm+m+H5KfcmeXCS+5O8qruvXlC5wJLZlzEqyXEZlof5T5IfJKkkZ2RYamFbd79/IcUCS6WqTkny/Qx3gX8pw/zo5RnmSM/r7hvHfk/IEIA/Ov+bQ92T5JyVPgDrbV/GqKp6RZIrk/wjw1zrIUlOS9JJ3tDdl29A6cASWiNwMoeCdSZwAuauqnZn+FF2LWd19/Vj3wcneUeS52ZYwuqD3b1j7kUCS2tfx6iqOjXJJUmekWR3kpuTfLy7r11AmcCSqqqtSbZnGHv+nWF5lwu7+1dT/TYlec/Y77dJ3jfdB2C97csYVVXnJrk4ydOS3J0hpLq0u3+48IKBpbVa4DS2m0PBOhI4AQAAAAAAMJNDNroAAAAAAAAADmwCJwAAAAAAAGYicAIAAAAAAGAmAicAAAAAAABmInACAAAAAABgJgInAAAAAAAAZiJwAgAAAAAAYCYCJwAAAAAAAGYicAIAAAAAAGAmAicAAAAAAABmInACAABgqVTVc6qqx9erJ9ofPdF+8dh2SFW9s6pur6rdVfX7qvpIVT1o6jN7L6/fLvY/BACAxTt0owsAAACADbQlyRcn9qddnuS1SW5I8rUkz07y9iTHJHndVN+/J/n0Kp9x5zrUCQAA+zWBEwAAAMvspDX2U1WnZAibvtrdLx7bHpjkxiSvrapLuvu2iVPu6u6L5l0wAADsjyypBwAAwDLbssZ+kpw5bq9daeju3UmuzPB9+vT5lgYAAAcOgRMAAADL6vtJHlNVR4/vtyT57sTx+ybaJ30yydFJrphveQAAcOAQOAEAALCsdo7bk6rqsCRPTLJr4vi3ktyf5I1Vtb2qjkiGu5y6+67uvi8AAEASz3ACAABged2aZE+GO5juTPLATARO3X1bVV2Q5GNJLkzypqr6VJIPd/c/V/m8o6pq+yrtX+/um9a9egAA2I8InAAAAFhW9ya5PclJSf42tu2c7NDdn6yqm5N8IMlZSbZluOPpRd29Y+rzjswQTE27I4nACQCAg5ol9QAAAFhmuzLc4bQlQzD0p+kO3X1Td5+d5LQMwdFxSa6vqhOnuv6uu2uV1yfm/D8AAMCGEzgBAACwzHYmeWqSp2Xq7qZp3f3DJGcmuSrJQ5O8d+7VAQDAAULgBAAAwDLbleRhSc7NxPObkqSqrqmqH1RVrbR19/1J3ja+3bqwKgEAYD8ncAIAAGCZ7VxjP0lOSPKMJI+dat89bu+ZV1EAAHCgETgBAACwzH6d5F/j/q6pY98atxdMtb953H57XkUBAMCB5tCNLgAAAAA2Snd3Vf0sySlJbk3yyInDlyZ5ZZK3VNXJSXYkeUqSc5L8JslHpz7uqKravsaf+kJ3/2JdiwcAgP2IwAkAAIBltyvJpu6+Z+JxTenuO6rq9CTbk7wgyWlJ/pzksiTbuvuvU59zZJIL1/gbO5IInAAAOGhVd290DQAAAAAAABzAPMMJAAAAAACAmQicAAAAAAAAmInACQAAAAAAgJkInAAAAAAAAJiJwAkAAAAAAICZCJwAAAAAAACYicAJAAAAAACAmQicAAAAAAAAmInACQAAAAAAgJkInAAAAAAAAJiJwAkAAAAAAICZCJwAAAAAAACYicAJAAAAAACAmQicAAAAAAAAmInACQAAAAAAgJkInAAAAAAAAJjJfwFPf26t5flFuwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 1800x900 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "x_labels = ['X_full', 'X_missing_fill_with_mean', 'X_missing_fill_with_zero', 'X_missing_reg']\n",
    "\n",
    "colors = ['r', 'g', 'b', 'orange']\n",
    "\n",
    "fig = plt.figure(figsize=(12, 6))\n",
    "ax = plt.subplot(111)\n",
    "for i in range(len(mse)):\n",
    "    ax.barh(i, mse[i], color=colors[i], alpha=0.6, align='center')\n",
    "ax.set_title('Imputation Techniques with Boston Data')\n",
    "ax.set_xlim(left=np.min(mse) * 0.9,\n",
    "right=np.max(mse) * 1.1)\n",
    "ax.set_yticks(np.arange(len(mse)))\n",
    "ax.set_xlabel('MSE')\n",
    "ax.set_yticklabels(x_labels)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 282,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T06:32:08.202708Z",
     "start_time": "2019-12-07T06:32:08.186343Z"
    }
   },
   "outputs": [],
   "source": [
    "X_missing_copy = X_missing.copy()\n",
    "data = X_missing_copy.apply(lambda x: x.fillna(x.mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 283,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T06:32:16.568221Z",
     "start_time": "2019-12-07T06:32:16.561069Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestRegressor(bootstrap=True, criterion='mse', max_depth=None,\n",
       "                      max_features='auto', max_leaf_nodes=None,\n",
       "                      min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                      min_samples_leaf=1, min_samples_split=2,\n",
       "                      min_weight_fraction_leaf=0.0, n_estimators=100,\n",
       "                      n_jobs=None, oob_score=False, random_state=1, verbose=0,\n",
       "                      warm_start=False)"
      ]
     },
     "execution_count": 283,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 机器学习的调参思想"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 机器学习的调参方法\n",
    "1. 画学习曲线\n",
    "2. 网格搜索\n",
    " \n",
    "在数据集很小的情况下，以上两种方法都非常有效。但是，当数据集很大时，这种调参的方法非常浪费时间。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 机器学习的调参思想\n",
    "找准目标：  \n",
    "一般在机器学习中我们都有一个评估模型的指标，我们要做的就是尽可能的让这个指标变得更好。在机器学习中，我们把**衡量一个模型在未知数据上的准确程度，称之为泛化误差(Generalization error)。**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    ">泛化误差。  \n",
    "当模型在未知数据（测试集或者袋外数据）上表现糟糕时，我们说模型的泛化程度不够，泛化误差大，模型的效果\n",
    "不好。泛化误差受到模型的结构（复杂度）影响。看下面这张图，它准确地描绘了泛化误差与模型复杂度的关系，\n",
    "当模型太复杂，模型就会过拟合，泛化能力就不够，所以泛化误差大。当模型太简单，模型就会欠拟合，拟合能力\n",
    "就不够，所以误差也会大。只有当模型的复杂度刚刚好的才能够达到泛化误差最小的目标"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![](images/2_3.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "模型的复杂度与参数的关系：  \n",
    "对于树模型来说，树越茂盛，深度越深，枝叶越多，模型就越复杂。所\n",
    "以树模型是天生位于图的右上角的模型，随机森林是以树模型为基础，所以随机森林也是天生复杂度高的模型。随\n",
    "机森林的参数，都是向着一个目标去：减少模型的复杂度，把模型往图像的左边移动，防止过拟合。当然了，调参\n",
    "没有绝对，也有天生处于图像左边的随机森林，所以调参之前，我们要先判断，模型现在究竟处于图像的哪一边。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. 模型过于简单或者过于复杂，会让泛化误差变大，而我们的目标是让泛化误差达到最低。\n",
    "2. 模型太复杂很容易过拟合， 模型太简单很容易欠拟合\n",
    "3. 对树模型和树的集成模型来说，树的深度越深，枝叶越多，模型越复杂\n",
    "4. 树模型和树的集成模型的目标，都是减少模型复杂度，把模型往图像的左边移动\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "树模型中参数的影响：  \n",
    "![](images/2_4.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 示例：随机森林在乳腺癌数据上的调参"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 288,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T10:30:13.444561Z",
     "start_time": "2019-12-07T10:30:13.438155Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_breast_cancer\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "import numpy as np \n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 289,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T10:30:25.837490Z",
     "start_time": "2019-12-07T10:30:25.802766Z"
    }
   },
   "outputs": [],
   "source": [
    "data = load_breast_cancer()\n",
    "X = data.data\n",
    "y = data.target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 293,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T10:32:34.843195Z",
     "start_time": "2019-12-07T10:32:32.612691Z"
    }
   },
   "outputs": [],
   "source": [
    "# 创建模型并评估模型\n",
    "rfc = RandomForestClassifier(n_estimators=100, random_state=1)\n",
    "score_pre = cross_val_score(rfc, X, y, cv=8).mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 295,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T10:32:39.011520Z",
     "start_time": "2019-12-07T10:32:39.004653Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.956250698636262"
      ]
     },
     "execution_count": 295,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "score_pre"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 第一步：调整n_estimators"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 309,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T10:46:45.715816Z",
     "start_time": "2019-12-07T10:46:01.820819Z"
    }
   },
   "outputs": [],
   "source": [
    "# 定义一个空列表用于接收返回的分数\n",
    "scores = []\n",
    "for i in range(0, 200, 10):\n",
    "    rfc = RandomForestClassifier(n_estimators=i+1, random_state=1)\n",
    "    score = cross_val_score(rfc, X, y, cv=8).mean()\n",
    "    scores.append(score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 310,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T10:46:46.533668Z",
     "start_time": "2019-12-07T10:46:46.525196Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "最大分数为：0.9614590319695954, 对应的n_estimators为：161\n"
     ]
    }
   ],
   "source": [
    "print('最大分数为：{}, 对应的n_estimators为：{}'.format(max(scores), \n",
    "                                             scores.index(max(scores))*10 +1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 311,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T10:46:50.419456Z",
     "start_time": "2019-12-07T10:46:49.927874Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1800x750 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12, 5))\n",
    "plt.plot(range(1, 201, 10), scores)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 305,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T10:42:17.115717Z",
     "start_time": "2019-12-07T10:41:37.084465Z"
    }
   },
   "outputs": [],
   "source": [
    "scores = []\n",
    "for i in range(155, 166):\n",
    "    rfc = RandomForestClassifier(n_estimators=i, random_state=1)\n",
    "    score = cross_val_score(rfc, X, y, cv=8).mean()\n",
    "    scores.append(score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 307,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T10:45:26.635380Z",
     "start_time": "2019-12-07T10:45:26.629348Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "最大分数为：0.9614590319695954\n"
     ]
    }
   ],
   "source": [
    "print('最大分数为：{}'.format(max(scores)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 308,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T10:45:37.993659Z",
     "start_time": "2019-12-07T10:45:37.369113Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1a7a05ddef0>]"
      ]
     },
     "execution_count": 308,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1800x750 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12, 5))\n",
    "plt.plot(range(155, 166), scores)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "由上图我们可以看出n_estimators取155时，模型达到最优"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 第二步：调整max_depth"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 312,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T10:52:12.066616Z",
     "start_time": "2019-12-07T10:51:08.456308Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=8, error_score='raise-deprecating',\n",
       "             estimator=RandomForestClassifier(bootstrap=True, class_weight=None,\n",
       "                                              criterion='gini', max_depth=None,\n",
       "                                              max_features='auto',\n",
       "                                              max_leaf_nodes=None,\n",
       "                                              min_impurity_decrease=0.0,\n",
       "                                              min_impurity_split=None,\n",
       "                                              min_samples_leaf=1,\n",
       "                                              min_samples_split=2,\n",
       "                                              min_weight_fraction_leaf=0.0,\n",
       "                                              n_estimators=155, n_jobs=None,\n",
       "                                              oob_score=False, random_state=1,\n",
       "                                              verbose=0, warm_start=False),\n",
       "             iid='warn', n_jobs=None,\n",
       "             param_grid={'max_depth': array([ 1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16, 17,\n",
       "       18, 19])},\n",
       "             pre_dispatch='2*n_jobs', refit=True, return_train_score=False,\n",
       "             scoring=None, verbose=0)"
      ]
     },
     "execution_count": 312,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "parameters = {\n",
    "    'max_depth' : np.arange(1, 20)\n",
    "}\n",
    "\n",
    "rfc = RandomForestClassifier(n_estimators=155, random_state=1)\n",
    "GS = GridSearchCV(rfc, parameters, cv=8)\n",
    "GS.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 315,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T10:52:28.014859Z",
     "start_time": "2019-12-07T10:52:28.007423Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'max_depth': 9}"
      ]
     },
     "execution_count": 315,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 316,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T10:52:28.930974Z",
     "start_time": "2019-12-07T10:52:28.924522Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.961335676625659"
      ]
     },
     "execution_count": 316,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.best_score_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 第三步：调整max_features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 319,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T10:56:38.119810Z",
     "start_time": "2019-12-07T10:53:33.346832Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=8, error_score='raise-deprecating',\n",
       "             estimator=RandomForestClassifier(bootstrap=True, class_weight=None,\n",
       "                                              criterion='gini', max_depth=None,\n",
       "                                              max_features='auto',\n",
       "                                              max_leaf_nodes=None,\n",
       "                                              min_impurity_decrease=0.0,\n",
       "                                              min_impurity_split=None,\n",
       "                                              min_samples_leaf=1,\n",
       "                                              min_samples_split=2,\n",
       "                                              min_weight_fraction_leaf=0.0,\n",
       "                                              n_estimators=155, n_jobs=None,\n",
       "                                              oob_score=False, random_state=1,\n",
       "                                              verbose=0, warm_start=False),\n",
       "             iid='warn', n_jobs=None,\n",
       "             param_grid={'max_features': array([ 1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16, 17,\n",
       "       18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29])},\n",
       "             pre_dispatch='2*n_jobs', refit=True, return_train_score=False,\n",
       "             scoring=None, verbose=0)"
      ]
     },
     "execution_count": 319,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "parameters = {\n",
    "    'max_features' : np.arange(1, 30)\n",
    "}\n",
    "\n",
    "rfc = RandomForestClassifier(n_estimators=155, random_state=1)\n",
    "GS = GridSearchCV(rfc, parameters, cv=8)\n",
    "GS.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 320,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T10:56:38.851184Z",
     "start_time": "2019-12-07T10:56:38.841760Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'max_features': 18}"
      ]
     },
     "execution_count": 320,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 321,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T10:56:39.645278Z",
     "start_time": "2019-12-07T10:56:39.637343Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9666080843585237"
      ]
     },
     "execution_count": 321,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.best_score_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 第四步：调整min_samples_leaf\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 323,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T10:59:26.022814Z",
     "start_time": "2019-12-07T10:57:30.024077Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=8, error_score='raise-deprecating',\n",
       "             estimator=RandomForestClassifier(bootstrap=True, class_weight=None,\n",
       "                                              criterion='gini', max_depth=None,\n",
       "                                              max_features=18,\n",
       "                                              max_leaf_nodes=None,\n",
       "                                              min_impurity_decrease=0.0,\n",
       "                                              min_impurity_split=None,\n",
       "                                              min_samples_leaf=1,\n",
       "                                              min_samples_split=2,\n",
       "                                              min_weight_fraction_leaf=0.0,\n",
       "                                              n_estimators=155, n_jobs=None,\n",
       "                                              oob_score=False, random_state=1,\n",
       "                                              verbose=0, warm_start=False),\n",
       "             iid='warn', n_jobs=None,\n",
       "             param_grid={'min_samples_leaf': array([ 1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16, 17,\n",
       "       18, 19])},\n",
       "             pre_dispatch='2*n_jobs', refit=True, return_train_score=False,\n",
       "             scoring=None, verbose=0)"
      ]
     },
     "execution_count": 323,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "parameters = {\n",
    "    'min_samples_leaf' : np.arange(1, 20)\n",
    "}\n",
    "\n",
    "rfc = RandomForestClassifier(n_estimators=155, max_features=18, random_state=1)\n",
    "GS = GridSearchCV(rfc, parameters, cv=8)\n",
    "GS.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 324,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T10:59:26.855140Z",
     "start_time": "2019-12-07T10:59:26.848652Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'min_samples_leaf': 1}"
      ]
     },
     "execution_count": 324,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 325,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T10:59:27.643333Z",
     "start_time": "2019-12-07T10:59:27.635437Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9666080843585237"
      ]
     },
     "execution_count": 325,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.best_score_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 第五步：调整min_samples_split\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 326,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T11:01:51.541006Z",
     "start_time": "2019-12-07T10:59:28.461734Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=8, error_score='raise-deprecating',\n",
       "             estimator=RandomForestClassifier(bootstrap=True, class_weight=None,\n",
       "                                              criterion='gini', max_depth=None,\n",
       "                                              max_features=18,\n",
       "                                              max_leaf_nodes=None,\n",
       "                                              min_impurity_decrease=0.0,\n",
       "                                              min_impurity_split=None,\n",
       "                                              min_samples_leaf=1,\n",
       "                                              min_samples_split=2,\n",
       "                                              min_weight_fraction_leaf=0.0,\n",
       "                                              n_estimators=155, n_jobs=None,\n",
       "                                              oob_score=False, random_state=1,\n",
       "                                              verbose=0, warm_start=False),\n",
       "             iid='warn', n_jobs=None,\n",
       "             param_grid={'min_samples_split': array([ 2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16, 17, 18,\n",
       "       19, 20, 21])},\n",
       "             pre_dispatch='2*n_jobs', refit=True, return_train_score=False,\n",
       "             scoring=None, verbose=0)"
      ]
     },
     "execution_count": 326,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "parameters = {\n",
    "    'min_samples_split' : np.arange(2, 22)\n",
    "}\n",
    "\n",
    "rfc = RandomForestClassifier(n_estimators=155, max_features=18, random_state=1)\n",
    "GS = GridSearchCV(rfc, parameters, cv=8)\n",
    "GS.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 327,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T11:02:22.800200Z",
     "start_time": "2019-12-07T11:02:22.794210Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'min_samples_split': 10}"
      ]
     },
     "execution_count": 327,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 328,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T11:02:31.946585Z",
     "start_time": "2019-12-07T11:02:31.939565Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.968365553602812"
      ]
     },
     "execution_count": 328,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.best_score_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 最后一步：调整criterion\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 330,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T11:04:13.158864Z",
     "start_time": "2019-12-07T11:03:55.385190Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=8, error_score='raise-deprecating',\n",
       "             estimator=RandomForestClassifier(bootstrap=True, class_weight=None,\n",
       "                                              criterion='gini', max_depth=None,\n",
       "                                              max_features=18,\n",
       "                                              max_leaf_nodes=None,\n",
       "                                              min_impurity_decrease=0.0,\n",
       "                                              min_impurity_split=None,\n",
       "                                              min_samples_leaf=1,\n",
       "                                              min_samples_split=10,\n",
       "                                              min_weight_fraction_leaf=0.0,\n",
       "                                              n_estimators=155, n_jobs=None,\n",
       "                                              oob_score=False, random_state=1,\n",
       "                                              verbose=0, warm_start=False),\n",
       "             iid='warn', n_jobs=None,\n",
       "             param_grid={'criterion': ['gini', 'entropy']},\n",
       "             pre_dispatch='2*n_jobs', refit=True, return_train_score=False,\n",
       "             scoring=None, verbose=0)"
      ]
     },
     "execution_count": 330,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "parameters = {\n",
    "    'criterion' : ['gini', 'entropy']\n",
    "}\n",
    "\n",
    "rfc = RandomForestClassifier(n_estimators=155, min_samples_split=10,\n",
    "                             max_features=18, random_state=1)\n",
    "GS = GridSearchCV(rfc, parameters, cv=8)\n",
    "GS.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 331,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T11:04:13.994072Z",
     "start_time": "2019-12-07T11:04:13.985146Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'criterion': 'gini'}"
      ]
     },
     "execution_count": 331,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 332,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T11:04:14.757413Z",
     "start_time": "2019-12-07T11:04:14.752948Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.968365553602812"
      ]
     },
     "execution_count": 332,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.best_score_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 示例：随机森林在titanic数据集上的表现"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 433,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T11:47:19.191183Z",
     "start_time": "2019-12-07T11:47:19.170385Z"
    }
   },
   "outputs": [],
   "source": [
    "import seaborn as sns\n",
    "titanic = sns.load_dataset('titanic')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 434,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T11:47:20.206602Z",
     "start_time": "2019-12-07T11:47:20.188251Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 891 entries, 0 to 890\n",
      "Data columns (total 15 columns):\n",
      "survived       891 non-null int64\n",
      "pclass         891 non-null int64\n",
      "sex            891 non-null object\n",
      "age            714 non-null float64\n",
      "sibsp          891 non-null int64\n",
      "parch          891 non-null int64\n",
      "fare           891 non-null float64\n",
      "embarked       889 non-null object\n",
      "class          891 non-null category\n",
      "who            891 non-null object\n",
      "adult_male     891 non-null bool\n",
      "deck           203 non-null category\n",
      "embark_town    889 non-null object\n",
      "alive          891 non-null object\n",
      "alone          891 non-null bool\n",
      "dtypes: bool(2), category(2), float64(2), int64(4), object(5)\n",
      "memory usage: 80.6+ KB\n"
     ]
    }
   ],
   "source": [
    "titanic.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 435,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T11:47:21.198938Z",
     "start_time": "2019-12-07T11:47:21.191780Z"
    }
   },
   "outputs": [],
   "source": [
    "titanic.drop(['alive', 'adult_male','alone', 'who', 'deck', 'class', 'embark_town'], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 438,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T11:47:42.968733Z",
     "start_time": "2019-12-07T11:47:42.960612Z"
    }
   },
   "outputs": [],
   "source": [
    "sex_labels = titanic['sex'].unique().tolist()\n",
    "titanic['sex'] = titanic['sex'].apply(lambda x:sex_labels.index(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 454,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T11:53:35.921634Z",
     "start_time": "2019-12-07T11:53:35.914658Z"
    }
   },
   "outputs": [],
   "source": [
    "# 删除embarked中含有nan的整条数据\n",
    "titanic_data = titanic[titanic['embarked'].notnull()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 457,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T11:54:18.323138Z",
     "start_time": "2019-12-07T11:54:16.784049Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "D:\\anaconda\\lib\\site-packages\\ipykernel_launcher.py:2: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  \n"
     ]
    }
   ],
   "source": [
    "embarked_labels = titanic['embarked'].unique().tolist()\n",
    "titanic_data['embarked'] = titanic_data['embarked'].apply(lambda x: embarked_labels.index(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 461,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T11:54:55.844539Z",
     "start_time": "2019-12-07T11:54:55.823659Z"
    }
   },
   "outputs": [],
   "source": [
    "def fill_na(x):\n",
    "    x['age'] = x['age'].fillna(x['age'].mean())\n",
    "    return x\n",
    "titanic_data = titanic_data.groupby('sex').apply(fill_na)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 487,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T12:03:01.281043Z",
     "start_time": "2019-12-07T12:03:01.272643Z"
    }
   },
   "outputs": [],
   "source": [
    "X = titanic_data.loc[:, titanic_data.columns!='survived']\n",
    "# 将数据扁平化\n",
    "y = titanic_data.loc[:, titanic_data.columns=='survived'].values.ravel()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 构建模型\n",
    "### 第一步：按默认创建模型"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 513,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T12:17:38.997005Z",
     "start_time": "2019-12-07T12:17:38.991584Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 489,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T12:03:07.948732Z",
     "start_time": "2019-12-07T12:03:05.444561Z"
    }
   },
   "outputs": [],
   "source": [
    "rfc = RandomForestClassifier(n_estimators=100, random_state=1)\n",
    "score = cross_val_score(rfc, X, y, cv=10).mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T12:03:08.926968Z",
     "start_time": "2019-12-07T12:03:08.918853Z"
    }
   },
   "source": [
    "## 第二步：调整n_estimators"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 492,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T12:06:15.348861Z",
     "start_time": "2019-12-07T12:05:26.733403Z"
    }
   },
   "outputs": [],
   "source": [
    "scores = []\n",
    "for i in range(0, 200, 10):\n",
    "    rfc = RandomForestClassifier(n_estimators=i+1, random_state=1)\n",
    "    score = cross_val_score(rfc, X, y, cv=10).mean()\n",
    "    scores.append(score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 493,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T12:06:16.268448Z",
     "start_time": "2019-12-07T12:06:16.262580Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "最大分数为：0.8121807967313586, 对应的n_estimators为：81\n"
     ]
    }
   ],
   "source": [
    "print('最大分数为：{}, 对应的n_estimators为：{}'.format(max(scores), \n",
    "                                             scores.index(max(scores))*10 +1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 495,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T12:06:41.109706Z",
     "start_time": "2019-12-07T12:06:40.721838Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1a7a604a9e8>]"
      ]
     },
     "execution_count": 495,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAx0AAAIHCAYAAAD6ozHaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAAXEQAAFxEByibzPwAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzs3Xl4lNXd//HPyb5AEkjCEnZIwi4ICIqAqHUBW21r1bpr3avVp26tWts+T2trXat1q1sVl7ZWq60/wH0BAUEW2ZewLwmQhOwh65zfHzOZzIQkJGS2ZN6v6/Ia577v3PehUpjPnO/5HmOtFQAAAAD4S0SwBwAAAACgayN0AAAAAPArQgcAAAAAvyJ0AAAAAPArQgcAAAAAvyJ0AAAAAPArQgcAAAAAvyJ0AAAAAPArQgcAAAAAvyJ0AAAAAPArQgcAAAAAvyJ0AAAAAPArQgcAAAAAvyJ0AAAAAPArQgcAAAAAvyJ0AAAAAPCrqGAPIFQZY/ZLSpC0J9hjAQAAAHxogKRKa22fQD3QWGsD9axOxRhTGhsb233YsGHBHgoAAADgM9u2bVN1dXWZtTYpUM9kpqNle4YNGzZq/fr1wR4HAAAA4DOjR4/Whg0bAlrNw5oOAAAAAH5F6AAAAADgV4QOAAAAAH5F6AAAAADgV4QOAAAAAH5F6AAAAADgV4QOAAAAAH5F6AAAAADgV4QOAAAAAH5F6AAAAADgV4QOAAAAAH5F6AAAAADgV4QOAAAAAH5F6AAAAADgV4QOAAAAAH5F6AAAAADgV4QOAAAAAH4VFewBAACOTW29Q/PW5qm23uq88RmKjuR7JABAaCJ0AEAndKiiRje9vkJLdxySJH226YD+cvEERUaYII8MAIAj8bUYAHQym/aX6tynvnIHDkmat3a/fv2fdbLWBnFkAAA0j9ABAJ3Ih+v364fPLNbeosNHnHtj6W79+ZOcIIwKAIDWEToAoBOw1uovn+bohtdWqLKmXpJkjHTXWcN1/MAU93VPfJqjOUt2BmeQAAC0gNABACHucE29bvn7Kj368Rb3se6xUXr5yhN086mZ+ttVJyirVzf3ud/8d73eX50bjKECANAsQgcAhLDc4sO64K+LNXdNnvvY4NQEvXvzVJ06opckKSUhRnOumayM5DhJkrXS7W99q4U5+UEZMwAATRE6ACBErdh1SOc+tUjr9pW6j03LTNN7N5+szF7dva7tmxyvOddMUY+EaElSbb3VDa+t0Oo9xQEdMwAAzSF0AEAIemv5Hl38/FIVlFe7j1198mC9cvUJSkmIafZnMnt109+unqyEmEhJUmVNva5+5Rttyy8PyJgBAGgJoQMAQkhdvUO/+38bdPfba1RT75AkRUca/en8sfrN90Yr6igbAI4fkKLnLpuo6Ejnfh2HKmp0xUvLtL+kyu9jR/A5HFZr95bo7RV7daCU/+Yb80r17qq92llQEeyhAGGPzQEBIESUVNbqlr+v1MKcAvex1MQYPXf5RJ0wuGeb7zMjO12PXDBO//PPb2WttK/4sK54eaneuuGkFmdJ0HnllRzWwpwCLcwp0Fc5+SqqrJUkJcVF6elLJ2h6VnqQRxh41lq9vGinHpi7QQ7X1jUDeyZoelaapmel66RhqUqOjw7uIIEwY9hIqnnGmPWjRo0atX79+mAPBUAY2JZfruteXa7tHt/IjuqbpBeunKR+KfHHdM9XFu3Qb9/f4H4/cVAPvX7NFMW7yq/QOVXW1GnpjkNauKVAC3PylXOw5fK5yAijX50zUldNHSxjwmO3+uq6ev3q3XX614q9LV4TGWE0fkCKO4SM65981FlEoCsZPXq0NmzYsMFaOzpQz/RZ6DDG9JD0gKTzJCVJekfSrdba0lZ/sOX73SbpWklDJZVK+lrSPdbaTS1c30fSCkn3WWtfOZZnNrkfoQNAQHy++aBu/fsqlVXVuY/NHttHj1wwTgkxHZuQfvSjzfrLZ1vd708b0Ut/vXyiovmA1Wk4HFYb95dqgStkLN9Z5C69a05khFGkMV7XXDRpgP7v+6MVG9W1A2d+WbVufH2FVuwqch+Lj47U4dr6Vn+ue1yUTh6WpunZaZqRla4BPRP8PVQgqIIROnxSXmWMiZf0oaQTXK9bJF0mKUPSmcdwv/sl/Z+kD1z3GyTpe5JmGmMmWWu3Nbk+RtK/XM8DgE7BWqsXF+7QH+dvdJeASNLtZ2TrZ6dl+uSb6dvPyFZBeY3+vmy3JOmzTQf1i7fX6JELxikiIjy++e6MDpZWuUqm8vXV1gIVlNe0ev2gVO/Sod2FlbpuznLludby/HP5Hm3LL9dzl09UWrfYQPwSAm7dvhKvX7MkTRrUQ89eNlH1DquFOfnOErStBTpU4f2/Z1lVnT5Yv18frN8vydmWenpWuqZnpemkYanqHkcpFtBRPpnpMMbcK+csx2PW2jtcxyZLWirpImvtW+24V5Kkg5JetNbe4nH8dEmfSHrSWnubx/FkSf+WdJrr0NXMdAAIdVW19brv3XV6Z2VjCUh8dKQev2iczh7T16fPqndY3fzGSvcHKkm6dtoQ3XfOyLApuQl1VbX1WrbjkPuD8ab9Za1e3z02SlMzU90fjAelJh5xTXPf+mckx+mFKydpdEayz38NwfT/1uTqzn+tVlXt0Wd3HA6rDXmlWpCTr4VbCrR81yHV1rf8WSgywmjCwBRNz0rXjOx0je2XrEgCOzq5TlteZYzZLildUh9rbYXH8U8llVlrv9+Oe50oaYmkc6y18zyOG0nVkj6x1s72OL5C0jBJj0j6nQgdAELcwdIq3fD6Cq3a3biHRr+UeL1wxSSNykjyyzOraut19d++0ZLthe5jv5w1QjeeMswvz0PrrLXatL/MHTKW7jikmrqWS6YijDRugPOD7ynZaRrXP6VNaxCaW98QHx2pRy4Yp3OO8224DQaHw+rxT7Z4lRBGGOn+745q8zqWiuo6Ld1R6C5f25bfeqer5PhoTctM0/SsNE3LSlP/HpRiofPplOVVrrUUQyT91zNwuHwm6a523rLS9TpS0jyP49mSoiXtaXJ9saQpkjr/n54Aurw1e4t1/ZwV2u/RznTykJ569tIJSvVj2UtcdKSev2Kifvz811qf61xq9+D8TeqZEKMLTxjgt+eiUX5ZtRZtLdCCLflauLVA+WXVrV7fLyVeM7LTNSMrTVOHpSk5of0lPrFRkXroR8dpRN8kdyenw7X1uvnNldp8IEv/c3pWpy2zq6iu08//+a0+2nDAfexYOnYlxkbptBG9ddqI3pKc3d6+ysnXgpwCLdpaoGJXN7AGJYdrNXdtnuauzZMkDU1P1AzXjNOJQ1OVGEtjUKA5vvh/Rm/X65pmzm2TlGyM6WmtPdTG+62Tc03IvcaYjZK+lHMx+YuSHJLmNLn+TGttvTGG0AEgpP3n2326++01qvb4RvviyQP1v+eOVkyU/xd2d4+L1itXT9aPnlusXYXO73d++e816pEYozNG9T7KT6O9qmrrtWJXkbuMZ0Ne631VEmMiddKwNM3Idq7NGJya4JPyN2OMrpk2RFm9uumWN1eq1NWw4MlPc7Rlf5kevXBcp/ugvOeQc82KZxnasPREvXjlCRqSdmSpWXv0S4nXRScM1EUnDFS9w2rdvhItdIWQlbuKVOfwrhDZnl+h7fkVemXxTkVHGk0Y2EMzsp0hZExGcqcNdYCvdbi8yhgzXdICSbdZa59scm6WnLMVY62169pxz2Fyrt8Y7HG4RNI11tp3WviZmZI+F+VVXU69wyrCiNpzdFoOh9WjH2/W05839sCIjDD6zfdG6fITBwX89/buwkqd/9xi9zftsVEReu2aKZo8pO17geBI1lrlHCx3zmTkFGjpjkKvNQZNGSMd1y/ZvVbg+IEpfu8qtj2/XNfOWa7tHiVEI/p01wtXTOo0HZu+3l6om15f4d6PRJJOHZ6uJy4+Xkl+XvBdXl2nJdsK3WVxO46y6WCPhGidnOnsiHXGqN7qkcg+OQgNnXJNhzHmJEmLJV1vrX2hybmGxd9TrLXL2nHPuyU9KMlIypXUU1KcnN2sLrPWFjbzMzN1DKHDGNNSqhg2atSoWEJHcH2wbr/ueOtbDevVTc9eNvGY9ysAgqWsqlY//+e3+mTjQfexlIRoPXPJBE3NTAvauDbmlerCvy5xt+ntHhelt244SSP7+mdNSVdVWF6tr7YWuDtNHShtvWQqIznOufg7O00nD0sLyofQ0qpa/ezNVfpyS777WM/EGD176QRNGZoa8PG0xxtLd+k3/1nvNdtwwylDdfdZI4KyuHvPoUqvLmOeba+b6hYbpRtPGaprpg1lrxwEXWcNHcdJWi3nnhx/aXLubEnzJU231n7VxvuNcd1vhZydr3YYY7rJ2UL355LmWmu/28zPzRSho0ux1mr6Q59rb9FhSc662bdvnKqefFOETmJXYYWum7NcWw40bt6W1aubXrxyUrPdhgJt2Y5Duvylpe5yr/Tusfr3TVM7zTfewVBT59CKXUWucpt8rdvXeslUfHSkThqW6m5nOyw9MSRmbesdVn/6YJOeX7DdfSwqwuh33x+jiycPDOLImldb79D/vb9Br329y30sJipCfzp/rH5wfP8gjqxRXb1Da/aVuDdtXLWnWPWOIz9j9UmK0+1nZuv8Cf3pgoWg6ayhI01Svjza5Xqcu0LSq2pHeZUx5jeSfitpnLV2TZNzX0maKqmntba4ybmZoryqS1mzt1jnPrXI69i4/sl647oT1a2T1R8j/CzeWqCfvrnSaxHqd0b20uMXjQ+pnv8fbzigG15b7t4nZHBqgv5141Sld++aezm0l7VW2/Ir3OU0X28vVGVN6xvNjemX5FpYnK4Jg1JCekO+d1bs1T3/Xuu1keCVJw3Sr747KmQ2kCyqqNFP31jp1XmtV/dY/fXyiTp+YI8gjqx1pVW17lKs91fnqeSw94L0EX26657ZI3VKdtsXvQO+0im7V1lrC4wx+yRNbOb0ZNfrgWbOtaRhQfjmZs5tkHSy65riZs6jC2noDOJp9d4S3fjaCr101aSQ/osc4ctaq9e/3qXfvr/B61vOn84cpjvPHB5yi0rPGNVbD/7wON39jvM7np2Flbrqb8v0j+tPDKlwFEhFFTVatK3A/Y11rsdmc83pnRTr3i9jWmaaX7uQ+dr5E/trSHqibnhthXuNz6tLdinnYLmevmRC0NcgbN5fpmvnfKM9hw67jx3XP1nPXz5JfZLjgjiyo0uKi9ZZo/vorNF9dNeZI/T0F1v1yqKd7oC3aX+Zrnx5maZnpemXs0Z0ub1TgKZ8tU/H85KukTTKWrvZdSxW0nZJpdbake241+8k/UrSaGvthibnlsm563kfa+2BJudmipmOLsNaqxkPf+7+i2ZoeqLXwsdzxvbVkxcfz9Q0QkpNnUO/fX+93ly6230sNipCD/3oOJ03vl8QR3Z0z36xTX/6YJP7/UlDU/W3q09QXHTXD/c1dQ6t2l3krs1fs69Erf3VGBcdoSlDnCVTM7LTldWrW0iUTHVEXslhXT9nhdbuK3EfG9gzQS9eOUnZvbsHZUwfbzig//nHKlV4zCydNz5Dfzr/uE77+3LPoUo98tFm/efbXK/jxkg/OL6f7jxzuDJYu4gA6JTlVZJkjBknabmcbXPPk1Qg6VlJV0n6lbX2gXbc6yJJ/5D0npwBotgYEyfpbkn/K2mVtXZCMz83U4SOLmPt3hJ976nGZUCLf3maHpy/Sf9d3fgH9aVTBur33x/T6f+yR9dQWF6tm95YqWU7GruD90mK0/NXTNRx/VOCOLK2sdbqgbkb9eJXO9zHZo3po6cumdDlwr21VjsKKtwhY8m2Qq8Pts0Z1TdJ07OdXYgmDurRaT/0tqaqtl53v73G68/ZbrFR+vNF4/WdALZUttbqmS+26ZGPNrvDnzHS3WeN0I2nDO0Sf+av2VusP8zbqK+3e+8mEBsVoZ9MG6KbZg7zeycuhLdOGzokyRhzq6THXW9rJcVKWibnIvIa1zW/l1RkrX20lfvEyLkj+QRJdXIGmFQ5NwZ0yLkvx6fN/NxMETq6jAfnb9JzXzrbi04a1ENv3zRVNXUOXTtnuRZ4dFy59bRM3X7m8GANE5Dk7AR17avLta+4sQRk/IAUPX/5RPVKCu0SEE8Oh9Ud/1qtd1ftcx+7ZMpAPdAFwn1JZa0WbyvQAlfQaGhQ0ZL07rHOmYysdJ2cmRY2a1ystXr2y216+EPvD/x3nTVcN50yzO+/Dw7X1Ovud9bo/SbB54kfj9fpI7vWXjLWWn2++aD+OG+Tcg6We53rmRijW0/L1CVTBgVkDx+En04dOiTJGHOypNsk9ZH0kaSHGgKH67yVtMtaO/go90mSdL+kH0gaKKla0kpJv7fWftzCz8wUoaNLsNbqlIe/0O5Dzs3Lfv3dUfrJtCGSpMqaOl364lKt2t24pOc33xulq08eEpSxAh+s26/b3/rWa3HxDyf00x9+MLZTfhteW+/QdXOW64vNnTvc19Y7tHpPsTtkrN5TrGYaCbnFREVoypCe7i5TI/p07/RBqyM+2XBAtzUpbTp3XIYe+pH/SpuaK/EalJqgF6+YpKwglXgFQl29Q/9asVePfbzliF3qB6cm6Bdnj9DZY/qE9e9H+F6nDx1dCaEjeNbtK9F3/9JYWrXkntPUN7mxxrW4skYX/nWJVxvSP180Xt8/PrRr5uE7dfUO7S06LEeQ//x6f3WeHv9ki/t9hJHumTVS104f0qk/IFTW1OmyF5dqpUe4/+33RumqEA/3uwornCFji7Nkqqy65T0TJGf3oIaQMXlIz04ZEv1py4EyXfvqcvcXQJI0tl+yXrjC94u4V+wq0g2vrVBBeeOH7pMzU/X0JROUkhAebdIrquv0wsLten7B9iM6pE0c1EP3zh6hiYPYwBO+QegIIYSO4Hnog0165gtnadXEQT30zk1Tj7hmf0mVzn92sbucJSrC6IUrJ+nU4b0COlYEVl29Q28t36vHPznyG8Fg6x4bpScvOb7L/B4srqzRBc8tcZd9GCM98ePjde64jCCPrFFpVa0Wb23cHdrzw3FzUhNj3CFjWlaaenei0rdgaa5dbXr3WD3vw3a1b6/Yq3ubtO29aupg3XfOyJBp2xtIB0ur9PgnOfrnN7uPmJ2bNaaP7j57hIakBX+fH3RuhI4QQugIDmutTn3kC+0sdH54uP+7o3TNtOa/Xd2eX64LnluiwgpnBV9cdITeuHYK3wR1QdZafbLxoB6cv1HbPLqYhYohaYl64YpJyuzVLdhD8am8ksP60bNL3OE+OtLopStP0Iwg7Svgufnagpx8fdvC5msNYiIjdMKQHu52tiP7JIVcy+LOoLbeod//vw16dYn3xnx//MFYnT/x2Dfmq3dYPTh/o15Y2Ni8IDrS6P/OC80NCgMt50CZHpy/SZ9uOuh1PCrC6NIpA3Xr6Vmdqj0zQguhI4QQOoJjfW6JznnSu2tVa+0D1+4t0cUvfK1yVxlFUlyU/nXjVA3v03Xrf8PNt3ucXV48u0JJzm/eo4L8ATIqIkJnje6t/z13jJITumanma0Hy3XBc4tV5NrkMCEmUm9ed6LGDwhMR649hyrdXaYWbS1QaVXrJVOZvbo5N+bLTtOUIT2VEMNGor7y5tLd+vV/1qnOI+hdP2OofnH2iHZ3OCs5XKtb/75KX3o0BumZGKPnLpuoyUP44sjTkm2F+sO8jV5rXSTn7OqNM4fpmmlDKA1EuxE6QgihIzge/nCTnv7cWVp1/MAUvfvTk4/6M4u3Feiql79xT833TorV2zdO1YCeCX4dK/xrV2GFHvpws+au8d4k0hjpRxP66/Yzs73W+sB/vt1TrEte+NpdZ94jIVr/unGqX2Z2yqvr3Ls4L8wp0I6C1me2eiREa5prJmN6Vhq/J/xs6fZC3fTGSh2qcPeI0czh6Xry4uPb3OJ1e365rp2z3GvvpRF9uuvFKyepfw/+3G6Ow2H1/ppcPfTBZq8ueZLUNzlOd5w5XD84vl+Xa28N/yF0hBBCR+BZa3Xao1+6P2T86pyRunb60Db97Afr8vTTN1a6618Hpybo7ZumKo2p507nUEWNnvw0R28s3aXaeu8/n2YOT9cvzh6hkX2TgjS68LUwJ18/eeUb93+TjOQ4vfPTqR3+kF/vsFq3r0QLc/K1YEuBVu4u8vomvanoSKOJg5wlUzOy0jU6g5KpQNtzqFLXzVmuTfvL3MeGpifqxSsmaWh660H0yy35uuXNlSrzmLE6e3QfPXrhOCXGMit1NFW19ZqzZKee+mzrEbN+I/sm6d7ZIzQ9Kzjlj+hcCB0hhNAReBtySzX7yYXu94t+eZr6tWNn1n8s261f/nut+/3ojCT94/oT1Z0NljqFqtp6vbxoh579fNsRXYdGZyTp3tkjdXJmWpBGB0l6f3Wubv3HKvf+DVm9uulfN57U7u5CucWHnSEjp0CLthao2FW61ZJh6YnOkJGdpilDUvlwGgIqqut0+1vf6sP1B9zHkuKi9NQlE5pd82Ot1Utf7dAf5m30Whx92+lZuu30LIJjOxVX1uipz7ZqzpJdXgvwJWlGdrrumcWXM2gdoSOEEDoC75EPN+upz7dKcm6s9t7NRy+tauqZL7bqoQ82u99PGdJTr/5kMvWuIazeYfXvlc4e9XklVV7n+qXE666zhuvccRl8KAkRry7eqd/8t/HPxQkDU/T6tVNaXTtRUV2npTsKtWCLc23G0ZoBJMdHa1pmmmZkp2laVnq7vnxA4DgcVn/+NEdPfprjPhZhpPvOGaWfnDzY3Ta6uq5e9727Tm+v2Ou+Lj46Uo9eOE6zx/YN+Li7kj2HKvXQh5u9NlOUGstQ7zhzuM/bG6NrIHSEEEJHYFlrdfqjX2q7q7Tqvtkjdd2MtpVWNb3PA3M36sWvGruhnDmqt565dIKiwrD1Yiiz1urLLfl6cP4mrzINyfmN6c9Oy9LlJw0iMIagxz7arCc/2+p+P3N4ul64YpK7vanDYbU+t1QLcvK1MCdfK3YVHVEq5ykqwmjCwB7OdRnZ6RrbL5na9E5k7po83fGvb1VV2/iN+4WT+ut33x+jksO1uvG1FV57vvRLidfzV0zU6IzkYAy3S2qp4UZcdISumTZEN54yjFl/eCF0hBBCR2BtzCvVrCcaS6u++sWpx7yg0OGwuvPt1fr3yn3uYxdNGqAHzx/bqTds60rW7SvRH+dv1KKthV7HYyIjdOXUQbr51Myw2RCsM7LW6r731unNpbvdx84bn6FpmWlamFOgr7YWeC00bs6QtET3nhknDu3JB6JObt2+El0/Z7lyPWYrxw9I0YHSKq8ZzEmDeui5yyey3s4PWmstnpoYo9u+k6WLJw8My71PcCRCRwghdATWox9t1l9c35yOG5Ci/xxDaZWn2nqHbnxthVd/8xtPGaZfzhrRofuiY/YWVerRj7bo3VX7jjj3/fEZuuPM4XQd6yTqHVa3vLlS89ftb9P13eOiNC0zzb1nBv+du578smrd9PoKLd9V1Oz5iyYN0O++P0YxUXzo9ae6eof+uXyPHv84x2uHd0kampaou88eobNG9+ZLuDBH6AghhI7Asdbq9Me+dLdPvHf2CF0/Y1iH71tVW68rXlqmZTsbp5uPtWwLHVNSWaunv9iqVxbtPGLR49Rhqbp39kiN6UepRWdTXVevq//2jRZvKzziXGSE0fgBKZqelaYZ2ek6rl8yJY5hoLquXve/t05vLW9cvxEZYXT/OSN15dTBfNANoPLqOj2/YLteWLBdh2vrvc5NHNRD984ewWa6YYzQEUIIHYGzaX+pzv5zY2nVwrtP9dm3oCWHa/Xj57/WxrxS97FHLhinH3VgF120XXVdveYs3qWnPt+qksPeHYpG9OmuX84aoVOy0/kg0omVVdXqujnL9fX2QxrQM965MV9Wuk4alqrkeEqmwpG1Vq8s3qnHPtqibnFRevhH4zQti85zwXKgtEqPf7xFby3fo6bdqGeN6aO7zhp+1FbH6HoIHSGE0BE4notSx/VP1n9umebT+x8sq9KPnl2i3YcqJTm/dXvusok6Y1Rvnz4HjRo2snr4w83aW+S9kVWfpDjdfma2zp/Qn8XCXYTDYVVZW69utLKFh6raesVERtB5LkRs3l+mP32wSZ95lB1LzkYOl0wZqFtPz2KtTRgJRuhgrhtBZa3V3LWNO077o31ir+5xeu2aye4/TOsdVje/uVJLtx9ZEoKOW7y1QOc+/ZVu+8e3XoGjW2yU7jpruD6/c6YunDSAwNGFREQYAgeOEBcdSeAIIcP7dNfLV52gN6+borEe5ax1Dqs5S3bplIc+118+zVFlTV0rdwGOHaEDQZVzsNyry4a/erYPSk3UnJ9MVvc45wejmjqHrn11udbnlvjleeFo0/5SXfW3ZbrkxaVat6+xnC0qwuiqqYP15V0zdfOpmYqPoQUuAATL1GFp+s/NJ+vJi4/XgJ6Ne+BU1NTr0Y+36NRHvtA/v9mt+qa1WEAHEToQVHPXNM5yHNc/2a8dbUZlJOmlK09QrKtzSll1na58+RvtKmx9ozK0bn9Jle5+e7VmP7FQX2zO9zp3zti++uT2U/Tbc0crlWl7AAgJERFG547L0Ce3n6JfnTPSa/3VgdJq/eKdtZr1xAJ9tumAKMOHrxA6EFTz/Fxa1dTkIT31zKUT3KU9BeXVuuylpTpYWnWUn0RTpVW1evjDTZr5yOd6a/lerwWKkwf31Ls/naqnL52gwWmJwRskAKBFsVGRunb6UC2461TdcMpQr3bGWw6U6yevLNfFL3ytNXuLW7kL0DaEDgRNzoEy5Rwsd7+fPcb/oUOSTh/ZWw+df5z7/Z5Dh3XFy8uO6K6E5tXUOfTKoh2a+fAXevrzbV67EA9LT9QLV0zSP284UccP7BHEUQIA2io5IVr3zBqpz++cqR9O6CfPhoJfbz+kc59apFv/vkp7XA1ZgGPByj8EjecC8jH9kjQwNXCbhZ0/sb+KKmv0+7kbJUmb9pfp2le/0ZyfTGHNQQustZq/br8e+mCTdhZ6/8WT1i1Wt5+RrQsn9WcvBgDopPqlxOuxC8frmmlD9OD8TVqYU+A+99/VuZq/Lk9XnDRYt5yaqR6Gp69oAAAgAElEQVSJMUEcafvVO6zW7C3WwpwCfbunWD0SYjQ9K00nZ6YpvTvlv4FA6EDQBLq0qqlrpw9VYUWNnv1imyTpm51FuuXNlXru8omK5oOzlxW7DumBuRu1crf3FHtCTKSunzFU100fqkS6FwFAlzA6I1mvXTNFC7bk64/zN7n3uqqtt3rpqx16a/ke3Xxqpq6aOlhx0aH7Rd3eokotzCnQwpx8fZVToNIq785c76x0bmI5qm+SpmenaUZWuiYO6hHSv6bOjH06WsA+Hf619WCZvvPYAvf7L++aqUGpga/9t9bqnn+v1T++2eM+9sPj++mRC8bR6lHSjoIK/Wn+Jn2wfr/X8cgIox+fMEC3fSdLvbrHBWl0AAB/q3dYvbdqnx79aLNyS7zXP2Ykx+nOs4br++P7hcTfmeXVdfp6W6EW5uRrYU6Bthe0v1FMXHSEpgxJ1fSsNJ2Sna7MXt265Aa2bA4YQggd/vXEJzl6/JMtkqTRGUmae+v0oI2lrt6hW95c5fXB+pppQ/Src0Z2yT9o2qKwvFpPfpqjN5buVl2TtonfGdlbv5w1XJm9ugdpdACAQKuqrdcri3fq6c+3qqzJjMGovkm6Z/YITc9KD+iY6h1W6/aVaGFOvhbkFGjlrqIj/s7yFB1pNHFQD500NE25xYe1ICdfeSWtN5LpkxSn6Vlpmp6drmmZaerZycrKWkLoCCGEDv866/EF2nygTJJ011nDdfOpmUEdT1Vtva7+2zda4rFhYCiMK9AO19Tr5UU79OwX21Re7f2Xyrj+ybp39khNGZoapNEBAIKtqKJGT32+VXOW7FRtvfdnyOlZabpn1kiNykjy2/Nziw/rq5wCLcjJ16KtBSqqbL0JzND0RM3ISteM7DRNGZLqVQpsrdW2/Ar3zMiSbYU6XFvf4r2MkcZkJDtDiKsUy7PjV2dC6AghhA7/2XqwXN957Ev3+y/unBkSbVXLqmp18Qtfe21s98cfjtXFkwcGcVSBUe+w+vfKvXrs4y1HfOszoGe87jprhL47tm9ITJ8DAIJvd2GlHvlos/67OtfruDHSD4/vrzvOzFZGSnwLP912lTV1Wrr9kBa4gsFWj66XzUmOj9a0zDRNz0rTtKw09e/R9iY11XX1Wrmr2PWsfK/PA81JiInUiUNT3SFkWHpip6mQIHSEEEKH/zz5aY4e+9hZWjWqb5Lm3Ra80qqmCsqrdeFzS9x1oBFGeubSCTo7QO18g6HpQsEGyfHR+tlpmbr8pEGKjWJRHQDgSKv3FOsP8zZq6Y5DXsdjoiL0k5OH6KaZw7w2Hzwah8NqQ16p84P/lgIt33XoiBkVT1ERRhMG9nCXQI3tl+zei6ujCsur9dXWAvdi9AOl1a1en5Ecp+lZ6ZqRna6TM1OVkhC6pViEjhBC6PCfs/+8QJv2h05pVVN7iyp1/rOL3X+4xERG6JWrT9DUzLQgj8y3NuaV6g/zNnq1RJScv96rTh6sm2dmKjmh7X9RAADCk7VWn28+qD/O2+S1/5YkpSRE62enZemyEwe2+AXWgdIqLdjinMn4amuBDlXUtPq8wakJmpGdrulZ6TpxaE91j/P/31XWWuUcLHePc+mOQq99qpoyRjquf4pmuGZBjh+YElKdMQkdIYTQ4R/b8st1+qONpVWf3zlTQ0KgtKqpLQfKdMFzS9wbBibGROrv15+o4/qnBHlkHZdXcliPfrRF76zcq6b/9z9vfIbuPHO4BvQM3J4pAICuoa7eoXdW7tWjH23RwTLvWYEBPeN191kjdM7Yvqquc2jZzkNa6PoA37DGsyXd46J08rA0Tc9O0/TM9IDu69WSqtp6rdhVpAVbnIvYm1YLNJUYE6mThqVpRrYzhAxOTQhqKRahI4QQOvzjqc9y9MhHztKqkX2TND+ESquaWrGrSJe9uNS9qKxnYoxuOTVTM7I7V91mg7KqWj335Ta99NWOI76dOXFoT907e2SXCFUAgOCqrKnTSwt36Lkvt6mixnth9sCeCdpfWqWaupZnCSKMdHxDyVRWusb1Tw75jWcPllVp0dYCLdxSoAU5BSoob70Uq3+PeGcpVlaazhrdJ+BrJgkdIYTQ4R+znljo/jbgzjOzdctpWUEeUeu+2HxQ1766/IgWfP1S4jXDtZHQ1My0dtWrBlptvUN/X7ZbT3ySo8ImU9aZvbrpnlkjdNqIXp0uRAEAQluBq/36m820X29qQM94zchylkydNCw1pP9ePRprrTbtL3N3xVq641CLIWtgzwQtuPvUAI+Q0BFSCB2+t6OgQqc+8oX7/Wd3nKKh6d2CN6A2+s+3+/Tzf36rlv68jIwwGj8gRdOz0jQjO13j+qf4bBFbR1hr9eH6A3rog01HbJCU1i1Wt5+RrQsn9Q/5b48AAJ3b9vxyPfzhZs1f17gfVrfYKJ00LFUzXH93BmOD4ECpqq3Xsh2H3OtBPMvJLp0yUA/8YGzAxxSM0BF19EsA35i3Ns/97yP6dO8UgUOSzhvfT9m9u+u9b/dpwZYj6zbrHVYrdhVpxa4i/fmTHHe7vhnZzj9I+yZ3vGVge63cXaQ/zN2o5buKvI7HR0fq+hlDdf2MoV69ygEA8Jeh6d307GUTtW5fidbnlmhoejeNHxBaC6v9KS46UjOynV2tJOfC+YaOWGeM6h3k0QUOMx0tYKbD92Y/sVAbXB/Ybz8jW7eeHtqlVS05WFbl3JjI9Y1F05KlprJ6dXO10EvTiUNTFRftv/azOwsq9NCHmzRv7X6v4xFGuuiEAfr5d7LVKynOb88HAAChj5kOdFk7CyrcgUOSZo/tvPte9Ooepx9O6K8fTujv7if+5ZZ8LdiSrxW7io6oW805WK6cg+V6edEOxURFaMqQnq7dUdOV3bubT9ZSHKqo0ZOf5uiNpbuO6Gd++ohe+sWsEcru3b3DzwEAADgWhA4ExFyP0qrhvbsrs1fnKK06mogIozH9kjWmX7JuPjVT5dV1WrKt0NVCL1+7Ciu9rq+pc7imVAv0wLyN6p0U695IaHpmmnoktm8joaraev1t0U4988VWlVXVeZ0b2y9Z98weoanDutb+IgAAoPMhdCAgPNdzdOZZjqPpFhulM0b1dtdo7iqs0AJXKdaSbYUqr/YOBgdKq/X2ir16e8Ve50ZC/ZLddZ+t1bs6HFbvfbtPj3y4WbklVV7n+qXE6+6zh+t7x2UEvAUfAABAcwgd8LtdhRVan9tYWnXOcX2COJrAGpSaqMtTE3X5iYNUW+/Qyl1FWpCTrwVbCrR2X4nXtdZKq/eWaPXeEv3ls63q3tDZIztdp2SnuzfsW7S1QH+Yt9Hrf1NJSoqL0i2nZeqKkwb7dd0IAABAexE64HeepVXZvbsps1d4ri2IjozQlKGpmjI0VXedJRWWV+urrQX60rUgPb/J7q1l1XX6aMMBfbThgCRpSFqi0rvFatnOQ03ua3TFSYN1y6mZ7S7PAgAACARCB/wuXEqr2iu1W6zOG99P543v595IqGEtyDc7ilRT772R0I6CCu1ost/Gd4/rq7vPGqGBqQmBHDoAAEC7EDrgV7sLK7Vun0dpFaGjWcYYjeybpJF9k3TDKcNUWVOnpdsPObti5eRre7532Jg8uKfuPWekxg9ICdKIAQAA2o7QAb/yLK3K6tVNWbRtbZOEmCidOqKXTh3RS5K0t6hSC3MKtLeoUpMG9dTM4ek+abULAAAQCIQO+BWlVb7Rv0eCLp48MNjDAAAAOCbhsf88gmLPoUqvDk3nHEfoAAAACEeEDviN5yxHZq9u7IgNAAAQpggd8BtKqwAAACAROuAnew5VavVej9IqQgcAAEDYInTAL+ava5zlGJqeqOze3YI4GgAAAAQToQN+MXftfve/nzO2L+1dAQAAwhihAz63t6hSq/cUu9+zngMAACC8ETrgc/M9ZjmGpiVqRB+6VgEAAIQzQgd8bm6TrlWUVgEAAIQ3Qgd8al/xYX1LaRUAAAA8EDrgU/M9ZjmGpCVqZF9KqwAAAMIdoQM+5V1a1YfSKgAAABA64Du5xYe1ajelVQAAAPBG6IDPzF/X2LVqcGqCRvVNCuJoAAAAECoIHfCZeXStAgAAQDMIHfCJvJLDWrGryP2e0ioAAAA0IHTAJzw3BByUmqDRGZRWAQAAwInQAZ+gtAoAAAAtIXSgw/aXVGm5Z2nVGEqrAAAA0IjQgQ6bv65xlmNAz3iN6UdpFQAAABoROtBhlFYBAACgNYQOdMiBUu/SqnPoWgUAAIAmCB3okPlr82St89/794jX2H7JwR0QAAAAQg6hAx0yz6NV7jmUVgEAAKAZhA4cs4OlVfpm1yH3ezYEBAAAQHN8FjqMMT2MMc8YY/YZY8qMMa8YY465jZEx5jZjzFpjTIUxJs8Y864xZkQz10UYY+4wxuQYYw4bYz5v7jr43vx1+71Kq47rT2kVAAAAjuST0GGMiZf0oaSbJK2V9DdJ50p6+xjvd7+kP0vaK+lZSV9JmiVpiTFmWJPLn5T0iKQqSU9L6iPpC2NMz2N5NtpuLl2rAAAA0Aa+mun4uaQTJD1mrT3bWnurpLMlnWGMubA9N3LNjtwn6Wlr7Sxr7Z3W2gsknSMpRdKtHtdOlXSzpCWSJllr75R0oiSHpN/74NeFFhwsq9I3OymtAgAAwNH5KnRcK6lc0q8bDlhrl0n6TNIl7bzXKEmxkuY1Of6ZpFpJWU2eK0m/tNZWu55bIufsyEXGmMh2Phtt9KFHaVW/lHiNo7QKAAAALehw6DDG9JE0RNJn1tqKJqc/kzSznbesdL2ObHI8W1K0pD0ex6ZKKpaz/Krpc3tKGtvOZ6ONvEur+lBaBQAAgBb5Yqajt+t1TTPntklKbuf6inWStki61xgz2xiTaIwZK2mOnGVTc5o8e5211tHMcyVpaDueizbKL6vWsh2UVgEAAKBtonxwj4YOVfnNnCtxvWZIOtTM+SNYax3GmNmSPpE0t8m9LrTWLmry7KM9t1XGmPUtnGq6YB0uH6zfL4ertCojOU7jB6QEd0AAAAAIab6Y6ahzvR5u5lyN6zWhnfc8X9Ig17/nytmZKlnStcaY1CbP9uVz0Qbz1jSWVs2iaxUAAACOwheho2EdR1wz56JdrzFtvZkxZoykP0paLmmotbafpHRJj8vZEevVJs/u0HOttaOb+0eNJVrwUFBeraU7Ct3vKa0CAADA0fgidOS6Xgc3c66X67W4Hfc7X85xXWut3SFJ1tpya+3tkhZJmm2MaajnyfXhc9EGH6xrLK3qmxyn4ymtAgAAwFF0OHRYawsk7ZM0sZnTk12vB9pxy4avzjc3c26DJONxzWpJo1ybE3b0uWiDeR5dq2aN6auICEqrAAAA0Dpf7dMxT9J0Y8zwhgPGmFhJP5C0yVrb3GLvljRc29xC7vGu14ZF6XPlLK+6tMl1F0uykha247k4ioLyan29vbG06pzj+gRxNAAAAOgsfBU6npazne2bxpj+xpg4Sc/J2T3q9Xbea53r9YGGMipjTJwx5tdy7nq+ylrbMIPxb0l7JT1kjJlmnG6UM+x8aq3d37FfFjx96NG1qk9SnI4f0CO4AwIAAECn4IuWubLWrjbG3CHnYu9dcu4cHitpmaSHG64zxvxeUpG19tFWbveupJWSvi/pu8aYAkmpci4Od0i6y+O5VcaYH0v6f3LOahyWFC+pUNKNvvi1oZFXadXYPpRWAQAAoE18NdMha+2TkmZIekfOsHG/pOnW2hqPy+6T9LOj3KdG0qmSHpEzwKRKqpa0QNLZ1tpPm1y/SNI4OWdWVkr6q6Qx1lq6T/lQYXm1lmzzKK2iaxUAAADayCczHQ1cAWBRK+fb9NW4tbZUzhmNu452rev63ZJuasu1ODYfbTjgVVo1YSClVQAAAGgbn810oGvzLK06ewylVQAAAGg7QgeO6lBFjRZ7llYdR2kVAAAA2o7QgaP6aP1+1btqq3onxWoipVUAAABoB0IHjmouGwICAACgAwgdaFVRk9Kq2XStAgAAQDsROtCqjzY0llald4/VxEGUVgEAAKB9CB1o1dy1jZu6zxrTR5GUVgEAAKCdCB1oUXFljRZvLXC/p7QKAAAAx4LQgRZ9tP6A6lylVWndYnXC4J5BHhEAAAA6I0IHWuTdtYrSKgAAABwbQgeaVVpVq0WUVgEAAMAHCB1o1vb8CndpVbfYKE0eQmkVAAAAjg2hA83KLT7s/vf+PeIprQIAAMAxI3SgWZ6ho29yXBBHAgAAgM6O0IFm5ZVUuf+9b0p8EEcCAACAzo7QgWbllTTOdGQw0wEAAIAOIHSgWbnFHjMdycx0AAAA4NgROtAsr5kOyqsAAADQAYQOHKGmzqGDZdXu9xkplFcBAADg2BE6cIQDpVWytvF9H9Z0AAAAoAMIHTiCZ+eqtG4xio2KDOJoAAAA0NkROnAEz/UcLCIHAABARxE6cATPzlWs5wAAAEBHETpwBO/dyJnpAAAAQMcQOnAE73a5zHQAAACgYwgdOAIbAwIAAMCXCB04AjMdAAAA8CVCB7wcrqlXUWWt+z0zHQAAAOgoQge8eM5yREYY9eoeG8TRAAAAoCsgdMCL53qO3t1jFRXJbxEAAAB0DJ8o4SXXc2PAFEqrAAAA0HGEDnjJ8+pcxSJyAAAAdByhA168O1cx0wEAAICOI3TAS25J40xHBjMdAAAA8AFCB7zkFrOmAwAAAL5F6ICbtVZ5HqEjgz06AAAA4AOEDriVVtWpoqbe/b4vu5EDAADABwgdcPNcRB4TFaHUxJggjgYAAABdBaEDbk3b5RpjgjgaAAAAdBWEDrjtYz0HAAAA/IDQAbc8r93IWc8BAAAA3yB0wM2zvIqZDgAAAPgKoQNuucx0AAAAwA8IHXDLK2GmAwAAAL5H6IAkyeGw3qGD3cgBAADgI4QOSJIKK2pUU+dwv6e8CgAAAL5C6IAk785V3WKjlBQXHcTRAAAAoCshdECSlNtkY0AAAADAVwgdkNR0jw7WcwAAAMB3CB2Q1LRzFTMdAAAA8B1CByRJ+4obZzroXAUAAABfInRAkpTnETpY0wEAAABfInRAktijAwAAAH5D6IDq6h06UEr3KgAAAPgHoQM6WFYth2183zeZmQ4AAAD4DqEDXu1yeybGKD4mMoijAQAAQFdD6ID2sTEgAAAA/IjQgSadqyitAgAAgG8ROtCkcxUzHQAAAPAtQgeUy0wHAAAA/IjQAWY6AAAA4FeEDnjNdLAxIAAAAHyN0BHmqmrrVVhR435P9yoAAAD4GqEjzO33KK0yRuqdROgAAACAbxE6wlyux8aAvbrHKjqS3xIAAADwLZ99wjTG9DDGPGOM2WeMKTPGvGKMSTqG+1xljLFH+ecqj+sHG2NecD33sDFmiTHmXF/9urq6PK+NAVnPAQAAAN+L8sVNjDHxkj6UdILrdYukyyRlSDqznbdbLemBFs6dJOk0SQddz82WtEhSN0n/lFQi6XxJ/zHG/I+19ol2Pjvs5HnMdPRjETkAAAD8wCehQ9LP5Qwcj1lr75AkY8zrkpYaYy601r7V1htZa1dJWtX0uDEmQtJaOUPJfNfhv0jqIWmatfZr13W/lfS1pEeNMf+11u445l9VGNjnNdPBeg4AAAD4nq/Kq66VVC7p1w0HrLXLJH0m6RIfPePHkkZJut9aa40xfSSdIem9hsDhem6RpKclRbrOoxWeMx19mekAAACAH3Q4dLg+/A+R9Jm1tqLJ6c8kzfTBMyLkDDTLrLXvuw6PlGQkrWjmRwpcr2kdfXZX57mmI4OZDgAAAPiBL8qrerte1zRzbpukZGNMT2vtoQ4841xJwyVd4HGszvXao5nrR7he93TgmWEhl5kOAAAA+JkvQkdDh6r8Zs6VuF4zJHUkdNwiaZ+k9zyO5Uiyki42xvzRVVYlY0xvSTdKqpf0ydFubIxZ38KpYR0Yb6dQVlWrsqo69/uMFGY6AAAA4Hu+WNPR8Kn1cDPnGra6TjjWmxtjRko6XdJz1lr3J2Rr7X5J/5DUX9JiY8zNxpg7JC2WlC7pA2tt3rE+NxzkeWwMGB1plJYYG8TRAAAAoKvyxUxHwzqO5r4mj3a9xnTg/tfLOaPxcjPnrpPkkHOx+lNNzj3Ylptba0c3d9w1AzKq7cPsfHKLG3Nin+Q4RUSYII4GAAAAXZUvZjpyXa+DmznXy/VafCw3di0gv1DSYmttbtPz1toKa+1lcs52nCHpd65TH1prvzqWZ4YTz5kONgYEAACAv3Q4dFhrC+RcbzGxmdOTXa8HjvH20+RcD/LOUcaQa639RM4yLCvp3mN8XljJ85jpoHMVAAAA/MVX+3TMkzTdGDO84YAxJlbSDyRtstY2t8i8LRq6Vb17tAuNMadImirp79balcf4vLCS6znTQecqAAAA+ImvQsfTcq6teNMY098YEyfpOTlnKV7vwH3PkrTbWruzDdf+n6RqSfd14HlhxXNNRwahAwAAAH7ik9BhrV0t6Q5J4yXtknMNx1WSlkl6uOE6Y8zvXR2mjsoY019SlpzdqI527WxJMyQ92caAAnmv6aC8CgAAAP7iq5kOWWuflPOD/ztyho37JU231tZ4XHafpJ+18Zanu15bDR2uxeZ/lHMX8gfaM+ZwZq31mulgITkAAAD8xRctc92stYskLWrlfJt7slprX5X0ahuuc0ga19b7wqmoslbVdQ73ezYGBAAAgL/4bKYDnYvnLEd8dKSS46NbuRoAAAA4doSOMOW1niMlTsawMSAAAAD8g9ARpuhcBQAAgEAhdISp3BLPReSs5wAAAID/EDrCVF6xx8aAdK4CAACAHxE6wlReiWd5FTMdAAAA8B9CR5jKZaYDAAAAAULoCEP1Dqv9pZ7dqwgdAAAA8B9CRxjKL6tWvcO631NeBQAAAH8idIQhz85VyfHRSojx6cb0AAAAgBdCRxjy7lzFLAcAAAD8i9ARhrw7V7GeAwAAAP5F6AhDnp2rWM8BAAAAfyN0hKHcYs/dyJnpAAAAgH8ROsIQGwMCAAAgkAgdYSi3hI0BAQAAEDiEjjBTU+dQQXm1+30GoQMAAAB+RugIMwdKq2Qb9wVU7+TY4A0GAAAAYYHQEWb2eSwiT+8eq9ioyCCOBgAAAOGA0BFmvBaRszEgAAAAAoDQEWZyi1lEDgAAgMAidIQZz5mOvrTLBQAAQAAQOsJMnudu5Mx0AAAAIAAIHWHGcyF5RgqhAwAAAP5H6AgzeZ4bA1JeBQAAgAAgdISRypo6lRyudb+nvAoAAACBQOgII56dq6IijNK7szEgAAAA/I/QEUY8O1f1TopTZIQJ4mgAAAAQLggdYSTPa48O1nMAAAAgMAgdYYTOVQAAAAgGQkcYYWNAAAAABAOhI4x4tsulcxUAAAAChdARRnI9yqtY0wEAAIBAIXSECWut90wHazoAAAAQIISOMFFyuFaVNfXu94QOAAAABAqhI0x4bgwYGxWhHgnRQRwNAAAAwgmhI0x4dq7KSImXMWwMCAAAgMAgdISJ3BI2BgQAAEBwEDrCRJ5X5yrWcwAAACBwCB1hwrNzVT82BgQAAEAAETrCxD7PmQ46VwEAACCACB1hwnMhOWs6AAAAEEiEjjDgcFjtZ2NAAAAABAmhIwwUVFSrtt663zPTAQAAgEAidISBPI+NAbvHRql7HBsDAgAAIHAIHWEgt9h7Y0AAAAAgkAgdYcBrY0Da5QIAACDACB1hgI0BAQAAEEyEjjDguTFgBovIAQAAEGCEjjCQW8LGgAAAAAgeQkcY8OxelcGaDgAAAAQYoaOLq6136ECZZ3kVMx0AAAAILEJHF3egtEq2cV9A9WFNBwAAAAKM0NHFeS4iT02MUVx0ZBBHAwAAgHBE6OjiPDcGZI8OAAAABAOho4vznOlgjw4AAAAEA6Gji/Oc6ehHu1wAAAAEAaGji8st9pzpoLwKAAAAgUfo6OLy2BgQAAAAQUbo6OI813RkMNMBAACAICB0dGFVtfU6VFHjfs9MBwAAAIKB0NGFeS4ijzBS7+6xQRwNAAAAwhWhowvzLK3qnRSnqEj+cwMAACDw+BTahXltDMh6DgAAAAQJoaML89oYkPUcAAAACBKfhQ5jTA9jzDPGmH3GmDJjzCvGmKRjuM9Vxhh7lH+u8rg+0xjznjEm1xhTYYxZZ4z5uTEm7AOVZ7tcOlcBAAAgWKJ8cRNjTLykDyWd4HrdIukySRmSzmzn7VZLeqCFcydJOk3SQddze0la4Dr3niSH6/xjkoZK+lk7n92leG8MyEwHAAAAgsMnoUPSz+UMHI9Za++QJGPM65KWGmMutNa+1dYbWWtXSVrV9Lhr5mKtnKFkvuvw/0hKkTTcWrvHdV2kpE8l3WSM+V9rbcGx/7I6N881HRmUVwEAACBIfFWCdK2kckm/bjhgrV0m6TNJl/joGT+WNErS/dZa6zo2QdK6hsDhem69pI8lRco52xG2vDYGTKG8CgAAAMHR4dBhjOkjaYikz6y1FU1OfyZppg+eESFnoFlmrX3f41SlpCHGmKafqMe4XvcoTJVW1aq8us79nvIqAAAABIsvZjp6u17XNHNum6RkY0zPDj7jXEnDJT3c5Pi/JaVJesEY09+1mP1mSRdK+sham9fB53ZaeR7rOWIiI5SaGBPE0QAAACCc+WJNR0OHqvxmzpW4XjMkHerAM26RtE/OxeJu1trXjTHDJf1KzoXrDd6XdEUHntfp5Xp0ruqTHKeICBPE0QAAACCc+SJ0NNTwHG7mXI3rNeFYb26MGSnpdDnXctQ1OZch56yG5FxTclhSupxdrs6T9Gob7r++hVPDjnXMocB7ETnrOQAAABA8viivaljH0dwn22jXa0dqe66XZCW93My5xyVlSbpZUoq1tpecgaNS0t+MMZM78NxOzbO8KoP1HAAAAAgiX8x05LpeBzdzrpfrtfhYbuxaQH6hpMXW2twm52IkfV/Se9baZxqOW2u/NsZcI2cHq1CVCeEAAB62SURBVIslLWvtGdba0S08e72c3bI6Jc/yqr7MdAAAACCIOjzT4doHY5+kic2cbphpOHCMt58m53qQd5o511POGZTNzZzb4HrNOMbndnp5bAwIAACAEOGrfTrmSZruWtQtSTLGxEr6gaRN1trmFpm3xQWu13ebOVcoZ9lVc2svxntcE5bySljTAQAAgNDgq9DxtCSHpDddrWvjJD0n50zD6x2471mSdltrdzY9Ya2tlXOW41xjzFkNx40xY+Vc6yFJH3Tg2Z2WtVa5Jcx0AAAAIDT4JHRYa1dLukPOGYZdcq7huErO9RTuvTWMMb83xtzRlnsaY/rLuUh8cSuX/UbOEqsPjDFFxphCOfcLyZb0hZytc8NOYUWNauoc7vcZKYQOAAAABI+vZjpkrX1S0gw5118sk3S/pOnW2hqPy+6T9LM23vJ012uLocNa+5acsyGfSDJy7hmyW9ITks611tr2/Bq6Cs/1HIkxkUqK80W/AAAAAODY+PTTqLV2kaRFrZxv8w511tpX1YZ9Nqy1H8vZqQou3p2r4mUMGwMCAAAgeHw204HQkeexMWDfZBaRAwAAILgIHV1QXgkbAwIAACB0EDq6oH3Fnu1yCR0AAAAILkJHF+Q508Fu5AAAAAg2QkcX5Lmmg/IqAAAABBuho4upd1gdKKt2v2emAwAAAMFG6OhiDpZVqd7RuD0JMx0AAAAINkJHF5PrsTFgj4RoxcdEBnE0AAAAAKGjy8n12qODWQ4AAAAEH6Gji8kr8WyXy3oOAAAABB+ho4vxLK9ipgMAAAChgNDRxXjOdNC5CgAAAKGA0NHFeG4MSOcqAAAAhAJCRxfjuZA8I4XQAQAAgOAjdHQh1XX1Kiivcb/vm0x5FQAAAIKP0NGF7PcorTJG6kPoAAAAQAggdHQhnp2r0rvFKjqS/7wAAAAIPj6VdiHenatYzwEAAIDQQOjoQjwXkfejXS4AAABCBKGjC8ktYWNAAAAAhB5CRxeS5zHTQecq/P/27j7arrI+8Pj3lxeSIAmQAoGEaqgjCKxlFRTfwFIZXzpap9SCyLgEZxitjtplsavjC0xbX8bRwVY6Wp3pWFxS2jJQmKJQqEaqlWq0KjI4QoeK2NzLuwlJCHn9zR/PPsm+l3PuPck955697/1+1jpr5+znufvss/OcvfdvP2+SJElNYdAxh0yYGNA+HZIkSWoIg445ZMyaDkmSJDWQQcccsWX7Lh57Ytfe99Z0SJIkqSkMOuaIen+OxQuDIw9ZMsK9kSRJkvYx6Jgj6iNXrVqxlAULYoR7I0mSJO1j0DFH1Gs6VjtcriRJkhrEoGOOmDBHhxMDSpIkqUEMOuaIiXN0WNMhSZKk5jDomCPGNu0LOtZY0yFJkqQGMeiYI8Y31ppXWdMhSZKkBjHomAMyc0JNh306JEmS1CQGHXPAxsd38sTOPXvfO3qVJEmSmsSgYw6o13IsXbyAww5ePMK9kSRJkiYy6JgD6v05Vh+6jAgnBpQkSVJzGHTMAfWajtWH2bRKkiRJzWLQMQeMTRi5yk7kkiRJahaDjjlgfMLIVdZ0SJIkqVkMOuaAiX06rOmQJElSsxh0zAFj1nRIkiSpwQw6Wm73nuT+TftqOtY4MaAkSZIaxqCj5R7esp1de3Lv+2OcGFCSJEkNY9DRcmMb9zWtWrF0EU9ZsmiEeyNJkiQ9mUFHy43XmlY5R4ckSZKayKCj5eo1Hc7RIUmSpCYy6Gi5ek2HI1dJkiSpiQw6Wq5e07HGoEOSJEkNZNDRcmP1mg6bV0mSJKmBDDpabnxCnw5rOiRJktQ8Bh0ttmPXHh7asn3v+9VODChJkqQGMuhosQcee4LcNy8gR9u8SpIkSQ1k0NFi9U7kRxyyhCWLFo5wbyRJkqTuDDpabOLEgNZySJIkqZkMOlpsbJMTA0qSJKn5DDpabHxjfbhcR66SJElSMxl0tNh4rabD5lWSJElqKoOOFtuwsd6nw5oOSZIkNZNBR4uNb3JiQEmSJDWfQUdLbduxm42P79z73uZVkiRJaiqDjpaqj1y1cEFw1HKDDkmSJDWTQUdL1UeuWrV8CQsXxAj3RpIkSerNoKOlJszRYSdySZIkNZhBR0uNbawPl2vQIUmSpOYy6GipevOq1c5GLkmSpAYz6GipCc2rDDokSZLUYAYdLTW+aV9Nh306JEmS1GQDCzoi4vCI+FREbIiIzRFxRUSsOIDtXBgROc3rwirvvdPku3dQ369JMpPxep8OJwaUJElSgy0axEYiYhlwM/C8ank38AZgNfDy/dzc7cCHeqS9EHgp8GD1/g+Bw7vkWwBcXMs3pzy2bRdbd+ze+96JASVJktRkAwk6gHdRAo6PZ+bFABFxJfDNiDg3M6/ud0OZ+V3gu5PXR8QC4A5KUHJTlfeybtuIiPOAg4CP7Of3aIV6f44lixaw8ikHjXBvJEmSpKkNqnnVRcAW4NLOisxcD6wDzh/QZ5wHnARckpnZK1MVnFwK/ANw3YA+u1HGJ3Uij3BiQEmSJDXXjIOOiDgaOA5Yl5lbJyWvA84cwGd0Aon1mXnDNNlfD5zINMFJm43Vhss9xv4ckiRJarhBNK9aVS2/3yXtHuDQiFiZmY/O4DNeA5wAnNNH3ncDd2bmTTP4vEabUNNhfw5JkiQ13CCCjs4IVQ91SdtULVcDMwk63g5sAK6fKlNEnA48G3hrvxuOiDt7JD29772bZRMnBrSmQ5IkSc02iD4du6rlti5pO6rlwQe68Yg4ETgL+HRm7pom+zsogc7nD/Tz2mBDfbhc5+iQJElSww2ipqPTj6NbO5/F1XImwyu9GUjgs1NliogjgLOBz3bpW9JTZp7cY3t3UjquN87EiQFtXiVJkqRmG0RNx1i1XNsl7ahqufFANlx1ID8XuC0zx6bJ/lpKkHPNgXxWW+zZk9y/yeZVkiRJao8ZBx2Z+TClv8WpXZJPq5YPHODmT6f0B7m2j7yvo/QbufUAP6sVHtm6gx279+x9b02HJEmSmm5Q83TcCJwRESd0VkTEEkpzpx9mZrdO5v3ojFY15XwbEXEU8AvADX30+2i1+shVhyxZxIqli6fILUmSJI3eoIKOTwJ7gKsi4tiIWAp8mlJLceUMtvsK4L7MvHeafP+S8l3+dgaf1QpjEzqRW8shSZKk5htI0JGZtwMXU4ar/TGlD8eFwHrgY518EfHBiLi4n21GxLHAM4Db+sh+VrXsJ2+rOTGgJEmS2mZQNR1k5uXASyj9L9YDlwBnZOaOWrb3UYa17cf+BBIvBR7JzLv63HZr1ZtXWdMhSZKkNhjEkLl7ZebXga9PkR77sa3PAZ/rM+9x/W637cY2WdMhSZKkdhlYTYdmx3itT8cxh1rTIUmSpOYz6GiZep8OZyOXJElSGxh0tMiu3Xt4cLNBhyRJktrFoKNFHti8nT25773NqyRJktQGBh0tUu/PsfIpB7F08cIR7o0kSZLUH4OOFpk4cpW1HJIkSWoHg44WmThylf05JEmS1A4GHS0yVgs61jgxoCRJklrCoKNFJjSvcuQqSZIktYRBR4uMb3JiQEmSJLWPQUeLjDsxoCRJklrIoKMlnti5m0e27tj73poOSZIktYVBR0uM1/pzRMCqFQYdkiRJageDjpaoD5e7avlSFi/0v06SJEnt4J1rS0wcucpaDkmSJLWHQUdL1Gs6VjsxoCRJklrEoKMlJtR02IlckiRJLWLQ0RIT5uhwuFxJkiS1iEFHS4zVmletsU+HJEmSWsSgoyXqEwMeY58OSZIktYhBRwtsfmInm7fv2vve0askSZLUJgYdLVCfGHDxwuCIpywZ4d5IkiRJ+8egowXq/TmOPnQpCxbECPdGkiRJ2j8GHS0wVuvP4RwdkiRJahuDjhaoD5e72uFyJUmS1DIGHS0wttGJASVJktReBh0t4MSAkiRJajODjhaoj1612poOSZIktYxBR8Nl5oTRq5wYUJIkSW1j0NFwj27dwfZde/a+X2PzKkmSJLWMQUfD1ZtWHXzQQlYsWzTCvZEkSZL2n0FHw01sWrWUCCcGlCRJUrsYdDTchE7kNq2SJElSCxl0NNzYpok1HZIkSVLbGHQ0XH1iQGs6JEmS1EYGHQ03XuvTsdrhciVJktRCBh0NV+/TccxhNq+SJElS+xh0NNjuPcn9j9WCDms6JEmS1EIGHQ320Obt7N6Te9+vtqZDkiRJLWTQ0WAbav05Dl22mIMPcmJASZIktY9BR4ON14bLdeQqSZIktZVBR4ON14fLdY4OSZIktZRBR4NNmBjQ/hySJElqKYOOBqvXdDhylSRJktrKoKPBJvbpsKZDkiRJ7WTQ0WAbJvTpsKZDkiRJ7WTQ0VDbd+3m4S3b97539CpJkiS1lUFHQz2waV/AEQGrVti8SpIkSe1k0NFQ9ZGrjjhkCQct8r9KkiRJ7eSdbENN6ETuHB2SJElqMYOOhhpzuFxJkiTNEQYdDTW2sT5crkGHJEmS2sugo6HGN9WGy3WODkmSJLWYQUdD1Ws6bF4lSZKkNjPoaKh6Tccx1nRIkiSpxQw6Gmjr9l1s2rZz73tnI5ckSVKbGXQ0UH243EULgiOXLxnh3kiSJEkzY9DRQPXhcletWMrCBTHCvZEkSZJmxqCjgSZMDGh/DkmSJLWcQUcDOTGgJEmS5hKDjgaq13Q4cpUkSZLabtGod0BPdtEZP8cZzziS8U3bePbPHj7q3ZEkSZJmxKCjgY5ftZzjVy0f9W5IkiRJA2HzKkmSJElDZdAhSZIkaagGFnRExOER8amI2BARmyPiiohYcQDbuTAicprXhT3+9teq9A/O+AtJkiRJGoiB9OmIiGXAzcDzquXdwBuA1cDL93NztwMf6pH2QuClwINd9mE58AfAT4AP7+dnSpIkSRqSQXUkfxcl4Ph4Zl4MEBFXAt+MiHMz8+p+N5SZ3wW+O3l9RCwA7qAEJTd1+dMPAGuA12Xm4/v/FSRJkiQNw6CaV10EbAEu7azIzPXAOuD8AX3GecBJwCWZmfWEiHgO8Hbg1v0JcCRJkiQN34yDjog4GjgOWJeZWyclrwPOHMBnLKAENOsz84YuaZ+p3r5jpp8lSZIkabAG0bxqVbX8fpe0e4BDI2JlZj46g894DXACcE6XtLdQmnZ9BzgvIpYCtwHXTa4RkSRJkjT7BhF0dEaoeqhL2qZquRqYSdDxdmADcH19ZdV5/Hert6dUL4CLgS9W/Umm7N8REXf2SHr6ge+uJEmSpI5B9OnYVS23dUnbUS0PPtCNR8SJwFnApzNz16TktwBHUvqTvBF4KvAiyghar6L3KFiSJEmSZskgajo6/TiWdklbXC0PmsH23wwk8Nkuaa+ulhdn5uerf/8kIv418E/AhRFxcWbu6bXxzDy52/qqBuSkA99tSZIkSTCYmo6xarm2S9pR1XLjgWy46iR+LnBbZo51ybIa2ANcVV+Zmdsp/ToOo9SESJIkSRqRGQcdmfkwpb/FqV2ST6uWDxzg5k+nBBbX9kjfDDyUmVu6pHVqWXZ0SZMkSZI0SwY1T8eNwBkRcUJnRUQsAc4GfpiZ3TqZ96MzWtV1PdLvAH6mGrFqr4gIStOo8cz86QF+tiRJkqQBGFTQ8UmqZk4RcWwVBHyaUktx5Qy2+wrgvsy8t0f61ZR+KZdMWv9W4BlVuiRJkqQRGkRHcjLz9oi4GPh94MfATmAJsB74WCdfRHwQ+GlmXjbdNiPiWErg8OdTfO6NEXEN8N6IOJ0yV8fJwMuAfwY+fMBfSpIkSdJAxCDnz4uIFwO/ARwN3AJ8NDN31NIT+HFmru1jWxcAVwDvzMw/nCLfAuBtwJsoQcoTwNeAd2fmj2bwXR5bsmTJ8qc/3ek6JEmSNHfcc889bN++fXNmrpg+92AMNOiYSyLifsr8Ij+ZxY/tRDj3zOJnzlUey8HyeA6Ox3JwPJaD5fEcHI/lYHk8B6dzLHcAj2fm0bP1wQYdDdKZHb3X3CHqn8dysDyeg+OxHByP5WB5PAfHYzlYHs/BGeWxHFRHckmSJEnqyqBDkiRJ0lAZdEiSJEkaKoMOSZIkSUNl0CFJkiRpqBy9SpIkSdJQWdMhSZIkaagMOiRJkiQNlUGHJEmSpKEy6JAkSZI0VAYdkiRJkobKoEOSJEnSUBl0SJIkSRoqg44GiIjDI+JTEbEhIjZHxBURsWLU+9V0EfHzEXFLRDwWEdsj4u6IOG9SntdERPZ4nTmiXW+ciPhOj2N066R8/yoi1kfE1oj4QUT88oh2uZEiYu0U5a3zuqLKa9mcQkT8YkTsioi1PdIviIj/ExHbIuLbEfHCHvlOjIgvRMRPI2I8It47zP1uoqmOZUQcEhGXRcR9VZ5Hq+vRwV3yPjJVmZ4PpjmW75ziN712Ut5jI+KqiHiwOq4fj4jFs/MtmqPX8YyIW6c7l07KP2/LZj/3QlW+vq7fwyybiwaxER24iFgG3Aw8r1reDbwBWA28fIS71mgR8Wzg68DBwDeBrcCZwJ9FxM7MvLbKegqwG/hIl838eBZ2tfEi4iDgZOArwG2Tkn9Uy3cO8BfAw8AfAycB10XEL2bm12Zpd5tuI/ChHmmrgTcBD1bvLZs9RMRTKWVtYY/0dwMfoxynPwJOB74UEc/OzH+s5Tse+DtgOfCnQAAfiAgy88PD/RbNMNWxjIiFwI3AGcAGyjXoucBbgWOB19TyrgVWAtcDd07a1HcGv+fNM125pPymH6WUyck21rZzJPBVYC1wTfU3vw4cArx5cHvcbNMcz89Tfrvd/DqwN+iYz2Wz33uhfq/fQy+bmelrhC/gvZQfz2W1dadV684d9f419VX9yLYAZ9XWvaw6brfX1t0A/HDU+9vkF3BqddzOmyLPCuAh4BHgabX1XwDuHPV3aMML+DSwDVhTvbdsdj9OJ1OCiaxeayelrwW2A/8POLxatxj4HnDjpLy3VNt4TW3db1X/D08d9XdtwLH899X6y4HF1bpDKTduCTynlve11boXjPp7NfFYVnnuAP66j23992ob76itOwfYAzx/1N+1Kcezx9+9oMr/m7V187Zs9nMvtD/X72GXTZtXjd5FlAJzaWdFZq4H1gHnj2qnmiwinkkJzN6fmV/urM/Mv6FcLJ9VPcGD8uTpe7O/l61ySrWc6jj9CnAE8InMrD+F/y/ASdXTFvUQEU8D/i3wqczcUK22bE4SEcdSLqKPUS5+3bwROAj4vcz8KUBm7gQuA14RET9TbetnKRffr2TmX9X+/lPADuBXh/IlGmI/juX3gN+ojiGZuQn4kyr9lFreUyg3HncMZYcbrJ9jWbVaeCbT/KarmuU3AP8E/Lda0jXVutcPYJcbrc+y2cvvAWOU33HHvCyb+3Ev1Nf1ezbKpkHHCEXE0cBxwLrM3DopeR2likyTZOYPKVV9n+ySvBjYBXuP72rgW7O3d610KuXkf/cUeV5ULW+YtP424Aksq9N5P+Xp/EfAsjmFIyjNfU4HxnvkeRHlSdwXJq1fR7mmnVG9f3G1nFBmq3PteuZ+mZ32WGbmGcDpWT3OrOm0395VW3cq8IMu16r5oJ9y+fOUJuvT/aZPAZYBX6gf9+rfX2Hul0vo73g+SUS8mPIg4YOZ+UQtaV6WzX7vhej/+j30smnQMVqrquX3u6TdAxwaEStncX9aIzO3d57MdUTEccDxwLcyczflRATl6eePIuKJiLgrIt5XRfQqTqXUtn256oj204i4ISJOreVZRXmSNKG9bHWcfwz83KztbctExFGUJ8qfy8yHqtWWze7uyMzzq6ftvawCNmTmo/WVVQ3SdvaVxenOr3O9zPZzLOlxo/aKavn3tXWnAIsj4hsRsSUiHoqIP4uIfzGoHW6wfo5l5zd9QZRBYbZFxO0R8baIqN9rzfdyCX2WzS5+i9JE6E8mrZ+3ZbPPe6F+r99DL5sGHaPVGaHqoS5pnR/j6lnal7ngt6tlpxPfc6vl8ykdo/6I0qzig8Cfz+6uNVN1g/ssSjlbRulgdhNlEIPbIuJ5VdYVwKbM3NFlM5uwnE7lzZTmQPWnUZbNLqqL4HQ67ZO7qZfF6c6vc7rM9nksnyQiXgD8AqUG/u5q3dOAI4ETKA8oPkNpHnMusL5Kn7P6PJb13/QXgf9BKYOfBP5rLd905XJ5RBxygLvaCgdSNqsy9mrgj+u1HPO9bPYw+V6o3+v30Mumo1eNVqfqa1uXtE7heNKwhXqy6ub4IuB24Kpq9Q7KaCwXZub9Vb6FwF8CZ0fEKZk5p0e26MPRlBP09Zl5eWdllOFHvwpcQhnBZhfdyymU42w57SIiFlFG/vhyZv7fWpJl88D1WxanO79aZiepyusnKU9F31NLOhL4GnB5Zl5Ty38OcDVwMfDOWdzVJtoMXEv5TW8BiIj/CNwKvD0iPlr91vu97m8Z7u62ztsoo89NHhnMslnT415oUOdMmGHZtKZjtDrV2ku7pHXa1M7nphZ9qTrwfY7SzvvfdZ6iZOZ/zsxXdm7qqnW7Ke3rYR502JtOZt6XmS+tBxzV+r8H/gp4VZQ5Y7bSvZxCKauW0+5eBayh1CDtZdmckX7L4nTnV8vsk11CaapyeTWgCQCZ+e3MfEn9pq5a/78oQ5LO+/Kame/MzF/rBBzVuseBD1DK22ur1V7391MVDL8J+JtJHaEtmzW97oUY3DkTZlg2DTpGa6xaru2SdlS13NglTRNdDpwIXJqZ/9BH/k6H6acPb5fmhLsp54i1lLK6MrpPWnkUltNezqN01Jvc6bkXy+b0xuhyzqxqilayryxOd361zNZExEuA91Hac79nmux1dwNH9Dg36Mm/aa/7++8sSo3GNdNlnGS+lc1e90L9Xr+HXjYNOkYoMx+mTMh0apfk06rlA7O3R+0TERdQqhJvpgz/Vk9b3hk+c5JO5/xekf+8UnU666Z+nG6v/l0fQpOIOIIyApvldJIoMzr/MnBL/elnlWbZPHC3Uy6gayetfw7ladwDtXzQ+/xqma1ExDGUvkSPU+aHemJS+oIp2sZbZinnwohY3iVp8vH5AbCT3uXy8cnnC/E6ykSq/3tygmWzmOpeiP6v30MvmwYdo3cjcEZEnNBZERFLgLMpE4f16jA571XD530GuA/4N5m5p5a2nNIZqtsTu1+qlj8Y+k42XER8GLhr8ihpVXX2yygn+rsoncv3UE5qdZ2q668OeVfb6JeApwDX1VdaNmfsi9VyyrKYZWbyfwTeWB8RLCKeRZmN1zLL3uD4euAYSpOMu7pkuxL4ZnVeqP/toZThOB+md+f+Oa+aL+EB4IIuyRN+01WTq1uBX62fdyNiFfBSLJcTVGXubODvetwPzfuyOdW9UKWv6/dslE2DjtHrdNq7KiKOjYillJmLV1N+TOoiIk6m3HwkcHZmPlJPz8zNlLH43xIRJ9b+7pmUEYISjy/AlyhPhz/aWRERQXlSchxwQ2Zuysx/pnRyPj8i3lo9XXox8J8oQxjeNPu73nidYUf/tr7SsjkzmXkb8G3g3RFxNkBE/Aqlo+ndlGPb8QlKOf5MRBwcEWuAz1I6pP7prO54A1VN0q6lPMX8aNUOvpsvUYbTfE/tbxdTJnY7BLiyy1wf88ndwE+A365qjQCIiBcBv0lpYllvGvQJYDnlur8yIg4DrqCMIOhvf6LTgMOYdB6tmddlc7p7IYD9vH4PtWzGHP6/aI2IeCfw+9XbncASyoXzjB5DnM17EdGZqOZbwC1dsvxPypO7r1Ce1n+DcqPxYspN9qWZ+YFZ2dmGi4gbKEMR3kWZdfQk4GmU8btPr05YRMSRlBP/iZTRLZZRRrs4LzOvHcGuN1pE3AMcnJnHdEl7EZbNKUXE71Auisdl5r2T0o6nTAa4hn1l8XHgZVVQ0skXlGZD51JGX1lYvf4gM981/G/RDL2OZdUk4wrKMfwEpTzWfScz/7KqKfoGpQnb7ZS236dS2nl/FzgzMx8b7rdohimO5TnAX1AmWv0WZZSfF1BuBi/KzCsmbedjwLsp59Ck/Pavz8yzh/4lGmSq33mVfgllFvJXZubNXdLnddns514oM3+0P9fvoZbNzPTVgBflhuNqSvXV+4GDRr1PTX5RbiByiteZVb7TgC9Thnh7FPhr4BWj3v8mvSjtXX8XuJfyRO4uyszZK7vkXQZcWp28rgGeO+r9b+ILeGpVDq+dIo9lc+pj+DvVMVzbI30lZf6DrwGfB54xxbbeSGnrfDPw+lF/t6YcS8rT4KnOo1fU8h5G6ag6Rgnwvg+8F1g26u/XhGNZpb2ScgO8DXiQUov0/Cm29WrKLNHrgP8ALBj192vS8azSOw9nDp1iG/O2bPZ7L1Tl7fv6PayyaU2HJEmSpKGyT4ckSZKkoTLokCRJkjRUBh2SJEmShsqgQ5IkSdJQGXRIkiRJGiqDDkmSJElDZdAhSZIkaagMOiRJkiQNlUGHJEmSpKEy6JAkSZI0VAYdkiRJkobKoEOSJEnSUBl0SJIkSRoqgw5JkiRJQ2XQIUmSJGmoDDokSZIkDZVBhyRJkqSh+v87PLuJzjxQ8wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 900x600 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(range(1, 201, 10), scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 500,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T12:08:34.592995Z",
     "start_time": "2019-12-07T12:08:14.999293Z"
    }
   },
   "outputs": [],
   "source": [
    "scores = []\n",
    "for i in range(75, 85):\n",
    "    rfc = RandomForestClassifier(n_estimators=i+1, random_state=1)\n",
    "    score = cross_val_score(rfc, X, y, cv=10).mean()\n",
    "    scores.append(score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 501,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T12:08:35.579275Z",
     "start_time": "2019-12-07T12:08:35.571339Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "最大分数为： 0.8133043922369765\n"
     ]
    }
   ],
   "source": [
    "print('最大分数为：', max(scores))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 502,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T12:08:36.834051Z",
     "start_time": "2019-12-07T12:08:36.473445Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1a79cda99b0>]"
      ]
     },
     "execution_count": 502,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 900x600 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(range(75, 85), scores)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "由上图我们可以确定n_estimators选择为79"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 第二步：调整max_depth"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 503,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T12:11:49.541850Z",
     "start_time": "2019-12-07T12:11:15.821805Z"
    }
   },
   "outputs": [],
   "source": [
    "parameters = {\n",
    "    'max_depth':np.arange(1, 20)\n",
    "}\n",
    "\n",
    "rfc = RandomForestClassifier(n_estimators=79, random_state=1)\n",
    "\n",
    "GS = GridSearchCV(rfc, parameters, cv=10)\n",
    "\n",
    "GS = GS.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 504,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T12:11:50.476290Z",
     "start_time": "2019-12-07T12:11:50.469794Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'max_depth': 5}"
      ]
     },
     "execution_count": 504,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 506,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T12:11:55.509797Z",
     "start_time": "2019-12-07T12:11:55.502142Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8357705286839145"
      ]
     },
     "execution_count": 506,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.best_score_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 第三步：调整max_features\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 507,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T12:13:31.988499Z",
     "start_time": "2019-12-07T12:13:16.264133Z"
    }
   },
   "outputs": [],
   "source": [
    "parameters = {\n",
    "    'max_features':np.arange(1, len(X.columns)+1)\n",
    "}\n",
    "\n",
    "rfc = RandomForestClassifier(n_estimators=79, max_features=5,random_state=1)\n",
    "\n",
    "GS = GridSearchCV(rfc, parameters, cv=10)\n",
    "\n",
    "GS = GS.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 508,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T12:13:32.936105Z",
     "start_time": "2019-12-07T12:13:32.930153Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'max_features': 4}"
      ]
     },
     "execution_count": 508,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 509,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T12:13:33.817495Z",
     "start_time": "2019-12-07T12:13:33.810551Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8245219347581553"
      ]
     },
     "execution_count": 509,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.best_score_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 第四步：调整min_samples_leaf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 510,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T12:16:19.581219Z",
     "start_time": "2019-12-07T12:15:23.670478Z"
    }
   },
   "outputs": [],
   "source": [
    "parameters = {\n",
    "    'min_samples_leaf':np.arange(1, 30)\n",
    "}\n",
    "\n",
    "rfc = RandomForestClassifier(n_estimators=79, max_features=5,random_state=1)\n",
    "\n",
    "GS = GridSearchCV(rfc, parameters, cv=10)\n",
    "\n",
    "GS = GS.fit(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 511,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T12:16:20.520864Z",
     "start_time": "2019-12-07T12:16:20.513920Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'min_samples_leaf': 4}"
      ]
     },
     "execution_count": 511,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 512,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-07T12:16:21.305538Z",
     "start_time": "2019-12-07T12:16:21.297598Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8346456692913385"
      ]
     },
     "execution_count": 512,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GS.best_score_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 使用多参数网格搜索"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "start_time": "2019-12-07T13:51:56.172Z"
    }
   },
   "outputs": [],
   "source": [
    "parameters = {\n",
    "    'n_estimators' : np.arange(50, 150),\n",
    "    'max_depth' : np.arange(1, 20),\n",
    "    'max_features': np.arange(1, len(X.columns)),\n",
    "    'min_samples_leaf': np.arange(1, 10),\n",
    "    'min_samples_split': np.arange(2, 10)\n",
    "}\n",
    "rfc = RandomForestClassifier(random_state=1)\n",
    "GS = GridSearchCV(rfc, parameters, n_jobs=-1)\n",
    "GS = GS.fit(X, y)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "start_time": "2019-12-07T14:26:59.649Z"
    }
   },
   "outputs": [],
   "source": [
    "GS.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "start_time": "2019-12-07T14:27:03.977Z"
    }
   },
   "outputs": [],
   "source": [
    "GS.best_score_"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": true
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}