{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Пробное программирование"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.model_selection import GridSearchCV, cross_val_score, train_test_split\n",
    "from sklearn.utils import shuffle \n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Загружаем и \"обрабатываем\" данные. Как видно, все признаки имеют тип float, а тип стекла - int."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RI               float64\n",
       "Na               float64\n",
       "Mg               float64\n",
       "Al               float64\n",
       "Si               float64\n",
       "K                float64\n",
       "Ca               float64\n",
       "Ba               float64\n",
       "Fe               float64\n",
       "Type_of_glass      int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('glass.csv')\n",
    "\n",
    "df.drop(['Id'], axis = 1, inplace = True)\n",
    "\n",
    "df.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 406,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>RI</th>\n",
       "      <th>Na</th>\n",
       "      <th>Mg</th>\n",
       "      <th>Al</th>\n",
       "      <th>Si</th>\n",
       "      <th>K</th>\n",
       "      <th>Ca</th>\n",
       "      <th>Ba</th>\n",
       "      <th>Fe</th>\n",
       "      <th>Type_of_glass</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.52101</td>\n",
       "      <td>13.64</td>\n",
       "      <td>4.49</td>\n",
       "      <td>1.10</td>\n",
       "      <td>71.78</td>\n",
       "      <td>0.06</td>\n",
       "      <td>8.75</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.51761</td>\n",
       "      <td>13.89</td>\n",
       "      <td>3.60</td>\n",
       "      <td>1.36</td>\n",
       "      <td>72.73</td>\n",
       "      <td>0.48</td>\n",
       "      <td>7.83</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.51618</td>\n",
       "      <td>13.53</td>\n",
       "      <td>3.55</td>\n",
       "      <td>1.54</td>\n",
       "      <td>72.99</td>\n",
       "      <td>0.39</td>\n",
       "      <td>7.78</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.51766</td>\n",
       "      <td>13.21</td>\n",
       "      <td>3.69</td>\n",
       "      <td>1.29</td>\n",
       "      <td>72.61</td>\n",
       "      <td>0.57</td>\n",
       "      <td>8.22</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.51742</td>\n",
       "      <td>13.27</td>\n",
       "      <td>3.62</td>\n",
       "      <td>1.24</td>\n",
       "      <td>73.08</td>\n",
       "      <td>0.55</td>\n",
       "      <td>8.07</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        RI     Na    Mg    Al     Si     K    Ca   Ba   Fe  Type_of_glass\n",
       "0  1.52101  13.64  4.49  1.10  71.78  0.06  8.75  0.0  0.0              1\n",
       "1  1.51761  13.89  3.60  1.36  72.73  0.48  7.83  0.0  0.0              1\n",
       "2  1.51618  13.53  3.55  1.54  72.99  0.39  7.78  0.0  0.0              1\n",
       "3  1.51766  13.21  3.69  1.29  72.61  0.57  8.22  0.0  0.0              1\n",
       "4  1.51742  13.27  3.62  1.24  73.08  0.55  8.07  0.0  0.0              1"
      ]
     },
     "execution_count": 406,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 407,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>RI</th>\n",
       "      <th>Na</th>\n",
       "      <th>Mg</th>\n",
       "      <th>Al</th>\n",
       "      <th>Si</th>\n",
       "      <th>K</th>\n",
       "      <th>Ca</th>\n",
       "      <th>Ba</th>\n",
       "      <th>Fe</th>\n",
       "      <th>Type_of_glass</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>209</th>\n",
       "      <td>1.51623</td>\n",
       "      <td>14.14</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.88</td>\n",
       "      <td>72.61</td>\n",
       "      <td>0.08</td>\n",
       "      <td>9.18</td>\n",
       "      <td>1.06</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>210</th>\n",
       "      <td>1.51685</td>\n",
       "      <td>14.92</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.99</td>\n",
       "      <td>73.06</td>\n",
       "      <td>0.00</td>\n",
       "      <td>8.40</td>\n",
       "      <td>1.59</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>211</th>\n",
       "      <td>1.52065</td>\n",
       "      <td>14.36</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.02</td>\n",
       "      <td>73.42</td>\n",
       "      <td>0.00</td>\n",
       "      <td>8.44</td>\n",
       "      <td>1.64</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>212</th>\n",
       "      <td>1.51651</td>\n",
       "      <td>14.38</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.94</td>\n",
       "      <td>73.61</td>\n",
       "      <td>0.00</td>\n",
       "      <td>8.48</td>\n",
       "      <td>1.57</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>213</th>\n",
       "      <td>1.51711</td>\n",
       "      <td>14.23</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.08</td>\n",
       "      <td>73.36</td>\n",
       "      <td>0.00</td>\n",
       "      <td>8.62</td>\n",
       "      <td>1.67</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          RI     Na   Mg    Al     Si     K    Ca    Ba   Fe  Type_of_glass\n",
       "209  1.51623  14.14  0.0  2.88  72.61  0.08  9.18  1.06  0.0              7\n",
       "210  1.51685  14.92  0.0  1.99  73.06  0.00  8.40  1.59  0.0              7\n",
       "211  1.52065  14.36  0.0  2.02  73.42  0.00  8.44  1.64  0.0              7\n",
       "212  1.51651  14.38  0.0  1.94  73.61  0.00  8.48  1.57  0.0              7\n",
       "213  1.51711  14.23  0.0  2.08  73.36  0.00  8.62  1.67  0.0              7"
      ]
     },
     "execution_count": 407,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.tail()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Т.к. записи упорядочены по типу стекла, перемешаем строки."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>RI</th>\n",
       "      <th>Na</th>\n",
       "      <th>Mg</th>\n",
       "      <th>Al</th>\n",
       "      <th>Si</th>\n",
       "      <th>K</th>\n",
       "      <th>Ca</th>\n",
       "      <th>Ba</th>\n",
       "      <th>Fe</th>\n",
       "      <th>Type_of_glass</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>169</th>\n",
       "      <td>1.51994</td>\n",
       "      <td>13.27</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.76</td>\n",
       "      <td>73.03</td>\n",
       "      <td>0.47</td>\n",
       "      <td>11.32</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>1.51926</td>\n",
       "      <td>13.20</td>\n",
       "      <td>3.33</td>\n",
       "      <td>1.28</td>\n",
       "      <td>72.36</td>\n",
       "      <td>0.60</td>\n",
       "      <td>9.14</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.11</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>150</th>\n",
       "      <td>1.51665</td>\n",
       "      <td>13.14</td>\n",
       "      <td>3.45</td>\n",
       "      <td>1.76</td>\n",
       "      <td>72.48</td>\n",
       "      <td>0.60</td>\n",
       "      <td>8.38</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.17</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>1.51761</td>\n",
       "      <td>12.81</td>\n",
       "      <td>3.54</td>\n",
       "      <td>1.23</td>\n",
       "      <td>73.24</td>\n",
       "      <td>0.58</td>\n",
       "      <td>8.39</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>200</th>\n",
       "      <td>1.51508</td>\n",
       "      <td>15.15</td>\n",
       "      <td>0.00</td>\n",
       "      <td>2.25</td>\n",
       "      <td>73.50</td>\n",
       "      <td>0.00</td>\n",
       "      <td>8.34</td>\n",
       "      <td>0.63</td>\n",
       "      <td>0.00</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          RI     Na    Mg    Al     Si     K     Ca    Ba    Fe  Type_of_glass\n",
       "169  1.51994  13.27  0.00  1.76  73.03  0.47  11.32  0.00  0.00              5\n",
       "51   1.51926  13.20  3.33  1.28  72.36  0.60   9.14  0.00  0.11              1\n",
       "150  1.51665  13.14  3.45  1.76  72.48  0.60   8.38  0.00  0.17              3\n",
       "15   1.51761  12.81  3.54  1.23  73.24  0.58   8.39  0.00  0.00              1\n",
       "200  1.51508  15.15  0.00  2.25  73.50  0.00   8.34  0.63  0.00              7"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = shuffle(df) \n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Разделим датасет на обучающую выборку и тест. Оставим на тест 30%."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = df['Type_of_glass']\n",
    "X = df.copy()\n",
    "X.drop('Type_of_glass', axis = 1, inplace = True)\n",
    "\n",
    "X_train, X_holdout, y_train, y_holdout = train_test_split(X.values, y, test_size = .3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Решающее дерево"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 410,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0"
      ]
     },
     "execution_count": 410,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dec_tree = DecisionTreeClassifier(criterion='entropy')\n",
    "\n",
    "dec_tree.fit(X_train, y_train)\n",
    "\n",
    "dec_tree.score(X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Если строить дерево без ограничения глубины, то оно идеально подстроится под обучающую выборку."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 411,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6153846153846154"
      ]
     },
     "execution_count": 411,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dec_tree.score(X_holdout, y_holdout)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Точность на отложеных данных - 61.5%."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Теперь будем использовать кросс-валидацию. Оптимизируем максимальную глубину дерева."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 412,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 4 folds for each of 10 candidates, totalling 40 fits\n",
      "Best params: {'max_depth': 7}\n",
      "Best cross validaton score 0.6577181208053692\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  33 out of  40 | elapsed:    8.2s remaining:    1.7s\n",
      "[Parallel(n_jobs=4)]: Done  40 out of  40 | elapsed:    8.2s finished\n"
     ]
    }
   ],
   "source": [
    "tree_params = {'max_depth': range(1, 11, 1)}\n",
    "\n",
    "cv_best_tree = GridSearchCV(dec_tree, tree_params, cv = 4, n_jobs = 4, verbose = True)\n",
    "\n",
    "cv_best_tree.fit(X_train, y_train)\n",
    "\n",
    "print(\"Best params:\", cv_best_tree.best_params_)\n",
    "print(\"Best cross validaton score\", cv_best_tree.best_score_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Оптимальная глубина - 7."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 413,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6307692307692307"
      ]
     },
     "execution_count": 413,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_best_tree.score(X_holdout, y_holdout)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Точность на отложенных данных - 63.1%."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Метод ближайших соседей"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 414,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6778523489932886"
      ]
     },
     "execution_count": 414,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "knn = KNeighborsClassifier(n_neighbors = 10, n_jobs = 4) #10 ближайших соседей\n",
    "\n",
    "knn.fit(X_train, y_train)\n",
    "\n",
    "knn.score(X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Метод ближайших соседей на обучающей выборке дает точность 67.8%."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 415,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6461538461538462"
      ]
     },
     "execution_count": 415,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "knn.score(X_holdout, y_holdout)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "На отложенных данных - 64.6%."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Теперь будем оптимизировать количество ближайших соседей и их веса (одинаковые, либо то расстояния)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 416,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 4 folds for each of 40 candidates, totalling 160 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  42 tasks      | elapsed:   10.2s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best params: {'n_neighbors': 3, 'weights': 'distance'}\n",
      "Best cross validaton score 0.6845637583892618\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done 160 out of 160 | elapsed:   16.4s finished\n"
     ]
    }
   ],
   "source": [
    "weights = ['uniform', 'distance']\n",
    "knn_params = {'n_neighbors': range(1, 21, 1), 'weights': weights}\n",
    "\n",
    "cv_best_knn = GridSearchCV(knn, knn_params, cv = 4, n_jobs = 4, verbose = True)\n",
    "\n",
    "cv_best_knn.fit(X_train, y_train)\n",
    "\n",
    "print(\"Best params:\", cv_best_knn.best_params_)\n",
    "print(\"Best cross validaton score\", cv_best_knn.best_score_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Модель определила, что лучше всего использовать 3х соседей и веса, зависящие от расстояния."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 417,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6923076923076923"
      ]
     },
     "execution_count": 417,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_best_knn.score(X_holdout, y_holdout)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Точность на тестовых данных - 69.2%."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Пробный анализ ошибки (метод ближайших соседей)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 453,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.metrics import accuracy_score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Мы хотим построить графики зависимости значений ошибки и ее стандартного отклонения от объема выборки на обучении и контроле."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 419,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(214, 10)"
      ]
     },
     "execution_count": 419,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "В датасете 214 записей."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 425,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 4 folds for each of 40 candidates, totalling 160 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  42 tasks      | elapsed:   10.6s\n",
      "[Parallel(n_jobs=4)]: Done 160 out of 160 | elapsed:   16.8s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 4 folds for each of 40 candidates, totalling 160 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  42 tasks      | elapsed:   10.1s\n",
      "[Parallel(n_jobs=4)]: Done 160 out of 160 | elapsed:   16.3s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 4 folds for each of 40 candidates, totalling 160 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  42 tasks      | elapsed:   10.9s\n",
      "[Parallel(n_jobs=4)]: Done 160 out of 160 | elapsed:   17.1s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 4 folds for each of 40 candidates, totalling 160 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  42 tasks      | elapsed:   11.4s\n",
      "[Parallel(n_jobs=4)]: Done 160 out of 160 | elapsed:   17.6s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 4 folds for each of 40 candidates, totalling 160 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  42 tasks      | elapsed:   10.6s\n",
      "[Parallel(n_jobs=4)]: Done 160 out of 160 | elapsed:   16.9s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 4 folds for each of 40 candidates, totalling 160 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  42 tasks      | elapsed:   10.6s\n",
      "[Parallel(n_jobs=4)]: Done 160 out of 160 | elapsed:   16.8s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 4 folds for each of 40 candidates, totalling 160 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_split.py:605: Warning: The least populated class in y has only 3 members, which is too few. The minimum number of members in any class cannot be less than n_splits=4.\n",
      "  % (min_groups, self.n_splits)), Warning)\n",
      "[Parallel(n_jobs=4)]: Done  42 tasks      | elapsed:   11.0s\n",
      "[Parallel(n_jobs=4)]: Done 160 out of 160 | elapsed:   17.2s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 4 folds for each of 40 candidates, totalling 160 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  42 tasks      | elapsed:   10.7s\n",
      "[Parallel(n_jobs=4)]: Done 160 out of 160 | elapsed:   16.9s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 4 folds for each of 40 candidates, totalling 160 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Done  42 tasks      | elapsed:   11.4s\n",
      "[Parallel(n_jobs=4)]: Done 160 out of 160 | elapsed:   17.6s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 4 folds for each of 40 candidates, totalling 160 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_split.py:605: Warning: The least populated class in y has only 3 members, which is too few. The minimum number of members in any class cannot be less than n_splits=4.\n",
      "  % (min_groups, self.n_splits)), Warning)\n",
      "[Parallel(n_jobs=4)]: Done  42 tasks      | elapsed:   10.0s\n",
      "[Parallel(n_jobs=4)]: Done 160 out of 160 | elapsed:   16.3s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 4 folds for each of 40 candidates, totalling 160 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_split.py:605: Warning: The least populated class in y has only 3 members, which is too few. The minimum number of members in any class cannot be less than n_splits=4.\n",
      "  % (min_groups, self.n_splits)), Warning)\n",
      "[Parallel(n_jobs=4)]: Done  42 tasks      | elapsed:   10.0s\n",
      "[Parallel(n_jobs=4)]: Done 160 out of 160 | elapsed:   16.2s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 4 folds for each of 40 candidates, totalling 160 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_split.py:605: Warning: The least populated class in y has only 3 members, which is too few. The minimum number of members in any class cannot be less than n_splits=4.\n",
      "  % (min_groups, self.n_splits)), Warning)\n",
      "[Parallel(n_jobs=4)]: Done  42 tasks      | elapsed:   10.7s\n",
      "[Parallel(n_jobs=4)]: Done 160 out of 160 | elapsed:   16.9s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 4 folds for each of 40 candidates, totalling 160 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_split.py:605: Warning: The least populated class in y has only 2 members, which is too few. The minimum number of members in any class cannot be less than n_splits=4.\n",
      "  % (min_groups, self.n_splits)), Warning)\n",
      "[Parallel(n_jobs=4)]: Done  42 tasks      | elapsed:   10.8s\n",
      "[Parallel(n_jobs=4)]: Done 153 out of 160 | elapsed:   16.0s remaining:    0.6s\n",
      "[Parallel(n_jobs=4)]: Done 160 out of 160 | elapsed:   16.2s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 4 folds for each of 40 candidates, totalling 160 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_split.py:605: Warning: The least populated class in y has only 2 members, which is too few. The minimum number of members in any class cannot be less than n_splits=4.\n",
      "  % (min_groups, self.n_splits)), Warning)\n",
      "[Parallel(n_jobs=4)]: Done  42 tasks      | elapsed:   10.4s\n",
      "[Parallel(n_jobs=4)]: Done 160 out of 160 | elapsed:   14.4s finished\n"
     ]
    }
   ],
   "source": [
    "test_size = np.linspace(0.2, 0.8, 14)\n",
    "\n",
    "test_acc_error = []\n",
    "test_squ_error = []\n",
    "\n",
    "train_acc_error = []\n",
    "train_squ_error = []\n",
    "\n",
    "for i in test_size:\n",
    "    X_train, X_holdout, y_train, y_holdout = train_test_split(X.values, y, test_size = i)\n",
    "    cv_best_knn.fit(X_train, y_train)\n",
    "\n",
    "    cv_best_knn_prediction_train = cv_best_knn.predict(X_train)\n",
    "    cv_best_knn_prediction_test = cv_best_knn.predict(X_holdout)\n",
    "\n",
    "    train_acc_error.append(1 - accuracy_score(cv_best_knn_prediction_train, y_train))\n",
    "    train_squ_error.append(mean_squared_error(cv_best_knn_prediction_train, y_train))\n",
    "\n",
    "    test_acc_error.append(1 - accuracy_score(cv_best_knn_prediction_test, y_holdout))\n",
    "    test_squ_error.append(mean_squared_error(cv_best_knn_prediction_test, y_holdout))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 452,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_acc_err = pd.DataFrame(data=[test_size, test_acc_error, train_acc_error]).transpose()\n",
    "data_acc_err.columns = ['объем тестовых данных', 'средняя ошибка на тестовых данных', 'средняя ошибка на обучающей выборке']\n",
    "\n",
    "data_squ_err = pd.DataFrame(data=[test_size, test_squ_error, train_squ_error]).transpose()\n",
    "data_squ_err.columns = ['объем тестовых данных', 'среднеквадратичная ошибка на тестовых данных', 'среднеквадратичная ошибка на обучающей выборке']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 454,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x2710e626cf8>"
      ]
     },
     "execution_count": 454,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 842.4x595.44 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot( 'объем тестовых данных', 'средняя ошибка на тестовых данных', data=data_acc_err)\n",
    "plt.plot( 'объем тестовых данных', 'средняя ошибка на обучающей выборке', data=data_acc_err)\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 455,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x271106bf710>"
      ]
     },
     "execution_count": 455,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 842.4x595.44 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot( 'объем тестовых данных', 'среднеквадратичная ошибка на тестовых данных', data=data_squ_err)\n",
    "plt.plot( 'объем тестовых данных', 'среднеквадратичная ошибка на обучающей выборке', data=data_squ_err)\n",
    "plt.legend()s"
   ]
  }
 ],
 "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": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}