{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "1cc23995",
   "metadata": {},
   "source": [
    "\n",
    "# Регрессия и поиск аномалий по данным Student Performance\n",
    "\n",
    "Задача:\n",
    "- определить ключевую зависимую переменную\n",
    "- построить регрессионную модель\n",
    "- оценить её точность, значимость и достоверность\n",
    "- найти наиболее аномальные объекты\n",
    "- удалить их\n",
    "- построить регрессию повторно по сокращённой выборке\n",
    "- сравнить модели до и после очистки\n",
    "\n",
    "Ноутбук специально сделан **без итогового содержательного вывода**.  \n",
    "После запуска можно будет добавить короткий вывод по реальным результатам.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d50423b8",
   "metadata": {},
   "source": [
    "## 1. Импорт библиотек"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "3f947a46",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import mean_absolute_error, mean_squared_error, r2_score\n",
    "from sklearn.ensemble import IsolationForest\n",
    "\n",
    "import statsmodels.api as sm\n",
    "from statsmodels.stats.outliers_influence import OLSInfluence\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d34506a2",
   "metadata": {},
   "source": [
    "## 2. Загрузка данных и проверка столбцов"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "5b8b8442",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Размер датасета: (395, 33)\n",
      "\n",
      "Столбцы:\n",
      "- school\n",
      "- sex\n",
      "- age\n",
      "- address\n",
      "- famsize\n",
      "- Pstatus\n",
      "- Medu\n",
      "- Fedu\n",
      "- Mjob\n",
      "- Fjob\n",
      "- reason\n",
      "- guardian\n",
      "- traveltime\n",
      "- studytime\n",
      "- failures\n",
      "- schoolsup\n",
      "- famsup\n",
      "- paid\n",
      "- activities\n",
      "- nursery\n",
      "- higher\n",
      "- internet\n",
      "- romantic\n",
      "- famrel\n",
      "- freetime\n",
      "- goout\n",
      "- Dalc\n",
      "- Walc\n",
      "- health\n",
      "- absences\n",
      "- G1\n",
      "- G2\n",
      "- G3\n"
     ]
    },
    {
     "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>school</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>address</th>\n",
       "      <th>famsize</th>\n",
       "      <th>Pstatus</th>\n",
       "      <th>Medu</th>\n",
       "      <th>Fedu</th>\n",
       "      <th>Mjob</th>\n",
       "      <th>Fjob</th>\n",
       "      <th>...</th>\n",
       "      <th>famrel</th>\n",
       "      <th>freetime</th>\n",
       "      <th>goout</th>\n",
       "      <th>Dalc</th>\n",
       "      <th>Walc</th>\n",
       "      <th>health</th>\n",
       "      <th>absences</th>\n",
       "      <th>G1</th>\n",
       "      <th>G2</th>\n",
       "      <th>G3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>18</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>A</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>at_home</td>\n",
       "      <td>teacher</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>17</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>T</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>at_home</td>\n",
       "      <td>other</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>15</td>\n",
       "      <td>U</td>\n",
       "      <td>LE3</td>\n",
       "      <td>T</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>at_home</td>\n",
       "      <td>other</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>15</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>T</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>health</td>\n",
       "      <td>services</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>15</td>\n",
       "      <td>14</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>GP</td>\n",
       "      <td>F</td>\n",
       "      <td>16</td>\n",
       "      <td>U</td>\n",
       "      <td>GT3</td>\n",
       "      <td>T</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>other</td>\n",
       "      <td>other</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>6</td>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 33 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "  school sex  age address famsize Pstatus  Medu  Fedu     Mjob      Fjob  ...  \\\n",
       "0     GP   F   18       U     GT3       A     4     4  at_home   teacher  ...   \n",
       "1     GP   F   17       U     GT3       T     1     1  at_home     other  ...   \n",
       "2     GP   F   15       U     LE3       T     1     1  at_home     other  ...   \n",
       "3     GP   F   15       U     GT3       T     4     2   health  services  ...   \n",
       "4     GP   F   16       U     GT3       T     3     3    other     other  ...   \n",
       "\n",
       "  famrel freetime  goout  Dalc  Walc health absences  G1  G2  G3  \n",
       "0      4        3      4     1     1      3        6   5   6   6  \n",
       "1      5        3      3     1     1      3        4   5   5   6  \n",
       "2      4        3      2     2     3      3       10   7   8  10  \n",
       "3      3        2      2     1     1      5        2  15  14  15  \n",
       "4      4        3      2     1     2      5        4   6  10  10  \n",
       "\n",
       "[5 rows x 33 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "df = pd.read_csv(\"student_data.csv\")\n",
    "df.columns = df.columns.str.strip()\n",
    "\n",
    "print(\"Размер датасета:\", df.shape)\n",
    "print(\"\\nСтолбцы:\")\n",
    "for col in df.columns:\n",
    "    print(\"-\", col)\n",
    "\n",
    "display(df.head())\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "523b806a",
   "metadata": {},
   "source": [
    "\n",
    "## 3. Выбор зависимой переменной\n",
    "\n",
    "В датасете есть три оценки:\n",
    "- `G1` — первая\n",
    "- `G2` — вторая\n",
    "- `G3` — итоговая\n",
    "\n",
    "В качестве ключевой зависимой переменной выбираем **`G3`**, так как это итоговый результат обучения.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "bf6b653e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Размер рабочей выборки: (395, 16)\n"
     ]
    },
    {
     "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>age</th>\n",
       "      <th>Medu</th>\n",
       "      <th>Fedu</th>\n",
       "      <th>traveltime</th>\n",
       "      <th>studytime</th>\n",
       "      <th>failures</th>\n",
       "      <th>famrel</th>\n",
       "      <th>freetime</th>\n",
       "      <th>goout</th>\n",
       "      <th>Dalc</th>\n",
       "      <th>Walc</th>\n",
       "      <th>health</th>\n",
       "      <th>absences</th>\n",
       "      <th>G1</th>\n",
       "      <th>G2</th>\n",
       "      <th>G3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>18</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>17</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>15</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>15</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>15</td>\n",
       "      <td>14</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>16</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>6</td>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   age  Medu  Fedu  traveltime  studytime  failures  famrel  freetime  goout  \\\n",
       "0   18     4     4           2          2         0       4         3      4   \n",
       "1   17     1     1           1          2         0       5         3      3   \n",
       "2   15     1     1           1          2         3       4         3      2   \n",
       "3   15     4     2           1          3         0       3         2      2   \n",
       "4   16     3     3           1          2         0       4         3      2   \n",
       "\n",
       "   Dalc  Walc  health  absences  G1  G2  G3  \n",
       "0     1     1       3         6   5   6   6  \n",
       "1     1     1       3         4   5   5   6  \n",
       "2     2     3       3        10   7   8  10  \n",
       "3     1     1       5         2  15  14  15  \n",
       "4     1     2       5         4   6  10  10  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "target = \"G3\"\n",
    "\n",
    "numeric_features = [\n",
    "    \"age\", \"Medu\", \"Fedu\", \"traveltime\", \"studytime\", \"failures\",\n",
    "    \"famrel\", \"freetime\", \"goout\", \"Dalc\", \"Walc\", \"health\",\n",
    "    \"absences\", \"G1\", \"G2\"\n",
    "]\n",
    "\n",
    "data = df[numeric_features + [target]].dropna().copy()\n",
    "print(\"Размер рабочей выборки:\", data.shape)\n",
    "display(data.head())\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a2fb9922",
   "metadata": {},
   "source": [
    "## 4. Первичный анализ связи с `G3`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "34b7640f",
   "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>corr_with_G3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>G3</th>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>G2</th>\n",
       "      <td>0.904868</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>G1</th>\n",
       "      <td>0.801468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Medu</th>\n",
       "      <td>0.217147</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fedu</th>\n",
       "      <td>0.152457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>studytime</th>\n",
       "      <td>0.097820</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>famrel</th>\n",
       "      <td>0.051363</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>absences</th>\n",
       "      <td>0.034247</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freetime</th>\n",
       "      <td>0.011307</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Walc</th>\n",
       "      <td>-0.051939</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Dalc</th>\n",
       "      <td>-0.054660</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>health</th>\n",
       "      <td>-0.061335</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>traveltime</th>\n",
       "      <td>-0.117142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>goout</th>\n",
       "      <td>-0.132791</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>age</th>\n",
       "      <td>-0.161579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>failures</th>\n",
       "      <td>-0.360415</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            corr_with_G3\n",
       "G3              1.000000\n",
       "G2              0.904868\n",
       "G1              0.801468\n",
       "Medu            0.217147\n",
       "Fedu            0.152457\n",
       "studytime       0.097820\n",
       "famrel          0.051363\n",
       "absences        0.034247\n",
       "freetime        0.011307\n",
       "Walc           -0.051939\n",
       "Dalc           -0.054660\n",
       "health         -0.061335\n",
       "traveltime     -0.117142\n",
       "goout          -0.132791\n",
       "age            -0.161579\n",
       "failures       -0.360415"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "corr_with_target = data.corr(numeric_only=True)[target].sort_values(ascending=False)\n",
    "display(corr_with_target.to_frame(\"corr_with_G3\"))\n",
    "\n",
    "plt.figure(figsize=(8, 5))\n",
    "corr_with_target.drop(target).sort_values().plot(kind=\"barh\")\n",
    "plt.title(\"Корреляции признаков с G3\")\n",
    "plt.xlabel(\"Корреляция\")\n",
    "plt.grid(axis=\"x\", alpha=0.3)\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2145e908",
   "metadata": {},
   "source": [
    "## 5. Обучение базовой регрессионной модели"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "90b0d915",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                     G3   R-squared:                       0.849\n",
      "Model:                            OLS   Adj. R-squared:                  0.841\n",
      "Method:                 Least Squares   F-statistic:                     112.1\n",
      "Date:                Fri, 13 Mar 2026   Prob (F-statistic):          3.04e-113\n",
      "Time:                        17:10:24   Log-Likelihood:                -631.13\n",
      "No. Observations:                 316   AIC:                             1294.\n",
      "Df Residuals:                     300   BIC:                             1354.\n",
      "Df Model:                          15                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "==============================================================================\n",
      "                 coef    std err          t      P>|t|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "const         -0.4684      1.701     -0.275      0.783      -3.815       2.878\n",
      "age           -0.1980      0.088     -2.246      0.025      -0.372      -0.024\n",
      "Medu           0.0942      0.126      0.746      0.456      -0.154       0.343\n",
      "Fedu          -0.1883      0.125     -1.504      0.134      -0.435       0.058\n",
      "traveltime     0.1310      0.158      0.827      0.409      -0.181       0.443\n",
      "studytime     -0.0660      0.134     -0.493      0.623      -0.330       0.198\n",
      "failures      -0.4163      0.159     -2.613      0.009      -0.730      -0.103\n",
      "famrel         0.3346      0.121      2.760      0.006       0.096       0.573\n",
      "freetime       0.0106      0.113      0.094      0.925      -0.212       0.233\n",
      "goout          0.1376      0.113      1.219      0.224      -0.085       0.360\n",
      "Dalc          -0.1050      0.151     -0.694      0.488      -0.403       0.193\n",
      "Walc           0.0618      0.115      0.537      0.592      -0.165       0.288\n",
      "health         0.0591      0.075      0.787      0.432      -0.089       0.207\n",
      "absences       0.0449      0.013      3.505      0.001       0.020       0.070\n",
      "G1             0.1607      0.061      2.624      0.009       0.040       0.281\n",
      "G2             0.9775      0.052     18.787      0.000       0.875       1.080\n",
      "==============================================================================\n",
      "Omnibus:                      159.409   Durbin-Watson:                   2.084\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):              837.588\n",
      "Skew:                          -2.107   Prob(JB):                    1.32e-182\n",
      "Kurtosis:                       9.772   Cond. No.                         418.\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    }
   ],
   "source": [
    "\n",
    "X = data[numeric_features].copy()\n",
    "y = data[target].copy()\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(\n",
    "    X, y, test_size=0.2, random_state=42\n",
    ")\n",
    "\n",
    "X_train_sm = sm.add_constant(X_train)\n",
    "X_test_sm = sm.add_constant(X_test)\n",
    "\n",
    "model_full = sm.OLS(y_train, X_train_sm).fit()\n",
    "print(model_full.summary())\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6694d084",
   "metadata": {},
   "source": [
    "## 6. Уравнение регрессии"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "8824f666",
   "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>feature</th>\n",
       "      <th>coef</th>\n",
       "      <th>p_value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>const</td>\n",
       "      <td>-0.468444</td>\n",
       "      <td>7.831579e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>age</td>\n",
       "      <td>-0.198011</td>\n",
       "      <td>2.544634e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Medu</td>\n",
       "      <td>0.094194</td>\n",
       "      <td>4.559590e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Fedu</td>\n",
       "      <td>-0.188338</td>\n",
       "      <td>1.335327e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>traveltime</td>\n",
       "      <td>0.131009</td>\n",
       "      <td>4.088209e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>studytime</td>\n",
       "      <td>-0.066050</td>\n",
       "      <td>6.225769e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>failures</td>\n",
       "      <td>-0.416258</td>\n",
       "      <td>9.425501e-03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>famrel</td>\n",
       "      <td>0.334612</td>\n",
       "      <td>6.135543e-03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>freetime</td>\n",
       "      <td>0.010623</td>\n",
       "      <td>9.250999e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>goout</td>\n",
       "      <td>0.137633</td>\n",
       "      <td>2.237088e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Dalc</td>\n",
       "      <td>-0.105011</td>\n",
       "      <td>4.882054e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Walc</td>\n",
       "      <td>0.061756</td>\n",
       "      <td>5.916848e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>health</td>\n",
       "      <td>0.059078</td>\n",
       "      <td>4.319745e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>absences</td>\n",
       "      <td>0.044864</td>\n",
       "      <td>5.270368e-04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>G1</td>\n",
       "      <td>0.160748</td>\n",
       "      <td>9.146999e-03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>G2</td>\n",
       "      <td>0.977534</td>\n",
       "      <td>1.359277e-52</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       feature      coef       p_value\n",
       "0        const -0.468444  7.831579e-01\n",
       "1          age -0.198011  2.544634e-02\n",
       "2         Medu  0.094194  4.559590e-01\n",
       "3         Fedu -0.188338  1.335327e-01\n",
       "4   traveltime  0.131009  4.088209e-01\n",
       "5    studytime -0.066050  6.225769e-01\n",
       "6     failures -0.416258  9.425501e-03\n",
       "7       famrel  0.334612  6.135543e-03\n",
       "8     freetime  0.010623  9.250999e-01\n",
       "9        goout  0.137633  2.237088e-01\n",
       "10        Dalc -0.105011  4.882054e-01\n",
       "11        Walc  0.061756  5.916848e-01\n",
       "12      health  0.059078  4.319745e-01\n",
       "13    absences  0.044864  5.270368e-04\n",
       "14          G1  0.160748  9.146999e-03\n",
       "15          G2  0.977534  1.359277e-52"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "G3 = -0.4684 - 0.1980*age + 0.0942*Medu - 0.1883*Fedu + 0.1310*traveltime - 0.0660*studytime - 0.4163*failures + 0.3346*famrel + 0.0106*freetime + 0.1376*goout - 0.1050*Dalc + 0.0618*Walc + 0.0591*health + 0.0449*absences + 0.1607*G1 + 0.9775*G2\n"
     ]
    }
   ],
   "source": [
    "\n",
    "coef_table = pd.DataFrame({\n",
    "    \"feature\": model_full.params.index,\n",
    "    \"coef\": model_full.params.values,\n",
    "    \"p_value\": model_full.pvalues.values\n",
    "})\n",
    "\n",
    "display(coef_table)\n",
    "\n",
    "equation_parts = [f\"{model_full.params.iloc[0]:.4f}\"]\n",
    "for feature, coef in model_full.params.iloc[1:].items():\n",
    "    sign = \"+\" if coef >= 0 else \"-\"\n",
    "    equation_parts.append(f\" {sign} {abs(coef):.4f}*{feature}\")\n",
    "\n",
    "equation = \"G3 = \" + \"\".join(equation_parts)\n",
    "print(equation)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5e97e93b",
   "metadata": {},
   "source": [
    "## 7. Точность и качество базовой модели"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "313f4891",
   "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>Метрика</th>\n",
       "      <th>Значение</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>MAE_train</td>\n",
       "      <td>1.149863e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>RMSE_train</td>\n",
       "      <td>1.783041e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>R2_train</td>\n",
       "      <td>8.486404e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>MAE_test</td>\n",
       "      <td>1.348164e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>RMSE_test</td>\n",
       "      <td>2.122209e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>R2_test</td>\n",
       "      <td>7.803580e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>F_stat_pvalue</td>\n",
       "      <td>3.044478e-113</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         Метрика       Значение\n",
       "0      MAE_train   1.149863e+00\n",
       "1     RMSE_train   1.783041e+00\n",
       "2       R2_train   8.486404e-01\n",
       "3       MAE_test   1.348164e+00\n",
       "4      RMSE_test   2.122209e+00\n",
       "5        R2_test   7.803580e-01\n",
       "6  F_stat_pvalue  3.044478e-113"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "y_pred_train = model_full.predict(X_train_sm)\n",
    "y_pred_test = model_full.predict(X_test_sm)\n",
    "\n",
    "metrics_full = pd.DataFrame({\n",
    "    \"Метрика\": [\"MAE_train\", \"RMSE_train\", \"R2_train\", \"MAE_test\", \"RMSE_test\", \"R2_test\", \"F_stat_pvalue\"],\n",
    "    \"Значение\": [\n",
    "        mean_absolute_error(y_train, y_pred_train),\n",
    "        mean_squared_error(y_train, y_pred_train) ** 0.5,\n",
    "        r2_score(y_train, y_pred_train),\n",
    "        mean_absolute_error(y_test, y_pred_test),\n",
    "        mean_squared_error(y_test, y_pred_test) ** 0.5,\n",
    "        r2_score(y_test, y_pred_test),\n",
    "        model_full.f_pvalue\n",
    "    ]\n",
    "})\n",
    "\n",
    "display(metrics_full)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "16ba2d1a",
   "metadata": {},
   "source": [
    "## 8. Диагностика остатков"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "a3bc2608",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "residuals = y_train - y_pred_train\n",
    "\n",
    "plt.figure(figsize=(7, 5))\n",
    "plt.scatter(y_pred_train, residuals, alpha=0.7)\n",
    "plt.axhline(0, linestyle=\"--\")\n",
    "plt.xlabel(\"Предсказанные значения\")\n",
    "plt.ylabel(\"Остатки\")\n",
    "plt.title(\"Остатки базовой модели\")\n",
    "plt.grid(alpha=0.3)\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n",
    "sm.qqplot(residuals, line=\"45\", fit=True)\n",
    "plt.title(\"Q-Q plot остатков\")\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4ea33626",
   "metadata": {},
   "source": [
    "\n",
    "## 9. Поиск аномалий\n",
    "\n",
    "Для устойчивого отбора используем три подхода:\n",
    "1. **Стандартизованные остатки**\n",
    "2. **Cook's distance**\n",
    "3. **Isolation Forest**\n",
    "\n",
    "Объект считается наиболее аномальным, если он отмечен **минимум 2 из 3 методов**.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "380b5b30",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Число аномалий: 12\n"
     ]
    },
    {
     "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>age</th>\n",
       "      <th>Medu</th>\n",
       "      <th>Fedu</th>\n",
       "      <th>traveltime</th>\n",
       "      <th>studytime</th>\n",
       "      <th>failures</th>\n",
       "      <th>famrel</th>\n",
       "      <th>freetime</th>\n",
       "      <th>goout</th>\n",
       "      <th>Dalc</th>\n",
       "      <th>...</th>\n",
       "      <th>G1</th>\n",
       "      <th>G2</th>\n",
       "      <th>G3</th>\n",
       "      <th>student_resid</th>\n",
       "      <th>cooks_d</th>\n",
       "      <th>flag_resid</th>\n",
       "      <th>flag_cook</th>\n",
       "      <th>flag_iso</th>\n",
       "      <th>votes</th>\n",
       "      <th>anomaly</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>259</th>\n",
       "      <td>17</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>10</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>-4.350386</td>\n",
       "      <td>0.068008</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>296</th>\n",
       "      <td>19</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>10</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>-4.202122</td>\n",
       "      <td>0.061774</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>310</th>\n",
       "      <td>19</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>9</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>-4.196503</td>\n",
       "      <td>0.046214</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>341</th>\n",
       "      <td>18</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>-4.814491</td>\n",
       "      <td>0.042420</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>239</th>\n",
       "      <td>18</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>...</td>\n",
       "      <td>7</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>-3.262131</td>\n",
       "      <td>0.036015</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>264</th>\n",
       "      <td>18</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>-5.091042</td>\n",
       "      <td>0.034292</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>337</th>\n",
       "      <td>17</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>-3.839841</td>\n",
       "      <td>0.030653</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>343</th>\n",
       "      <td>17</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>9</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>-3.536514</td>\n",
       "      <td>0.023329</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>333</th>\n",
       "      <td>18</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>-3.811963</td>\n",
       "      <td>0.021069</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>247</th>\n",
       "      <td>22</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>...</td>\n",
       "      <td>6</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>1.290461</td>\n",
       "      <td>0.019418</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>170</th>\n",
       "      <td>16</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>-1.783054</td>\n",
       "      <td>0.015719</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>153</th>\n",
       "      <td>19</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.477533</td>\n",
       "      <td>0.013972</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>173</th>\n",
       "      <td>16</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>8</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>-2.895338</td>\n",
       "      <td>0.057248</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>276</th>\n",
       "      <td>18</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>10</td>\n",
       "      <td>9</td>\n",
       "      <td>9</td>\n",
       "      <td>-1.595195</td>\n",
       "      <td>0.054333</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>367</th>\n",
       "      <td>17</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>7</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>-2.797750</td>\n",
       "      <td>0.036048</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>15 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     age  Medu  Fedu  traveltime  studytime  failures  famrel  freetime  \\\n",
       "259   17     2     2           1          4         0       3         4   \n",
       "296   19     4     4           2          2         0       2         3   \n",
       "310   19     1     2           1          2         1       4         2   \n",
       "341   18     4     4           1          2         1       4         3   \n",
       "239   18     2     2           1          2         1       5         5   \n",
       "264   18     2     2           1          3         0       4         3   \n",
       "337   17     3     2           1          2         0       4         3   \n",
       "343   17     2     2           1          2         1       3         3   \n",
       "333   18     2     2           1          2         0       4         3   \n",
       "247   22     3     1           1          1         3       5         4   \n",
       "170   16     3     4           3          1         2       3         4   \n",
       "153   19     3     2           1          1         3       4         5   \n",
       "173   16     1     3           1          2         3       4         3   \n",
       "276   18     3     2           2          2         0       4         1   \n",
       "367   17     1     1           3          1         1       5         2   \n",
       "\n",
       "     goout  Dalc  ...  G1  G2  G3  student_resid   cooks_d  flag_resid  \\\n",
       "259      1     1  ...  10   9   0      -4.350386  0.068008           1   \n",
       "296      4     2  ...  10   9   0      -4.202122  0.061774           1   \n",
       "310      4     2  ...   9   9   0      -4.196503  0.046214           1   \n",
       "341      3     2  ...  10  10   0      -4.814491  0.042420           1   \n",
       "239      4     3  ...   7   7   0      -3.262131  0.036015           1   \n",
       "264      3     1  ...   9  10   0      -5.091042  0.034292           1   \n",
       "337      2     2  ...   7   8   0      -3.839841  0.030653           1   \n",
       "343      1     1  ...   9   8   0      -3.536514  0.023329           1   \n",
       "333      3     1  ...   8   8   0      -3.811963  0.021069           1   \n",
       "247      5     5  ...   6   8   8       1.290461  0.019418           0   \n",
       "170      5     2  ...   6   5   0      -1.783054  0.015719           0   \n",
       "153      4     1  ...   5   0   0       1.477533  0.013972           0   \n",
       "173      5     1  ...   8   7   0      -2.895338  0.057248           0   \n",
       "276      1     1  ...  10   9   9      -1.595195  0.054333           0   \n",
       "367      1     1  ...   7   6   0      -2.797750  0.036048           0   \n",
       "\n",
       "     flag_cook  flag_iso  votes  anomaly  \n",
       "259          1         0      2        1  \n",
       "296          1         0      2        1  \n",
       "310          1         0      2        1  \n",
       "341          1         0      2        1  \n",
       "239          1         0      2        1  \n",
       "264          1         0      2        1  \n",
       "337          1         0      2        1  \n",
       "343          1         0      2        1  \n",
       "333          1         0      2        1  \n",
       "247          1         1      2        1  \n",
       "170          1         1      2        1  \n",
       "153          1         1      2        1  \n",
       "173          1         0      1        0  \n",
       "276          1         0      1        0  \n",
       "367          1         0      1        0  \n",
       "\n",
       "[15 rows x 23 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "influence = OLSInfluence(model_full)\n",
    "\n",
    "student_resid = influence.resid_studentized_external\n",
    "cooks_d = influence.cooks_distance[0]\n",
    "\n",
    "resid_flag = (np.abs(student_resid) > 3).astype(int)\n",
    "cook_flag = (cooks_d > 4 / len(X_train)).astype(int)\n",
    "\n",
    "iso = IsolationForest(contamination=0.05, random_state=42)\n",
    "iso_flag = (iso.fit_predict(X_train) == -1).astype(int)\n",
    "\n",
    "anomaly_df = X_train.copy()\n",
    "anomaly_df[\"G3\"] = y_train.values\n",
    "anomaly_df[\"student_resid\"] = student_resid\n",
    "anomaly_df[\"cooks_d\"] = cooks_d\n",
    "anomaly_df[\"flag_resid\"] = resid_flag\n",
    "anomaly_df[\"flag_cook\"] = cook_flag\n",
    "anomaly_df[\"flag_iso\"] = iso_flag\n",
    "anomaly_df[\"votes\"] = anomaly_df[[\"flag_resid\", \"flag_cook\", \"flag_iso\"]].sum(axis=1)\n",
    "anomaly_df[\"anomaly\"] = (anomaly_df[\"votes\"] >= 2).astype(int)\n",
    "\n",
    "print(\"Число аномалий:\", int(anomaly_df[\"anomaly\"].sum()))\n",
    "display(anomaly_df.sort_values([\"votes\", \"cooks_d\", \"student_resid\"], ascending=[False, False, False]).head(15))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d1e9a0bc",
   "metadata": {},
   "source": [
    "## 10. Наиболее аномальные объекты"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "2ec81ab3",
   "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>age</th>\n",
       "      <th>Medu</th>\n",
       "      <th>Fedu</th>\n",
       "      <th>traveltime</th>\n",
       "      <th>studytime</th>\n",
       "      <th>failures</th>\n",
       "      <th>famrel</th>\n",
       "      <th>freetime</th>\n",
       "      <th>goout</th>\n",
       "      <th>Dalc</th>\n",
       "      <th>...</th>\n",
       "      <th>G1</th>\n",
       "      <th>G2</th>\n",
       "      <th>G3</th>\n",
       "      <th>student_resid</th>\n",
       "      <th>cooks_d</th>\n",
       "      <th>flag_resid</th>\n",
       "      <th>flag_cook</th>\n",
       "      <th>flag_iso</th>\n",
       "      <th>votes</th>\n",
       "      <th>anomaly</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>259</th>\n",
       "      <td>17</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>10</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>-4.350386</td>\n",
       "      <td>0.068008</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>296</th>\n",
       "      <td>19</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>10</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>-4.202122</td>\n",
       "      <td>0.061774</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>310</th>\n",
       "      <td>19</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>9</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>-4.196503</td>\n",
       "      <td>0.046214</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>341</th>\n",
       "      <td>18</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>-4.814491</td>\n",
       "      <td>0.042420</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>239</th>\n",
       "      <td>18</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>...</td>\n",
       "      <td>7</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>-3.262131</td>\n",
       "      <td>0.036015</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>264</th>\n",
       "      <td>18</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>-5.091042</td>\n",
       "      <td>0.034292</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>337</th>\n",
       "      <td>17</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>-3.839841</td>\n",
       "      <td>0.030653</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>343</th>\n",
       "      <td>17</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>9</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>-3.536514</td>\n",
       "      <td>0.023329</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>333</th>\n",
       "      <td>18</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>-3.811963</td>\n",
       "      <td>0.021069</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>247</th>\n",
       "      <td>22</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>...</td>\n",
       "      <td>6</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>1.290461</td>\n",
       "      <td>0.019418</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     age  Medu  Fedu  traveltime  studytime  failures  famrel  freetime  \\\n",
       "259   17     2     2           1          4         0       3         4   \n",
       "296   19     4     4           2          2         0       2         3   \n",
       "310   19     1     2           1          2         1       4         2   \n",
       "341   18     4     4           1          2         1       4         3   \n",
       "239   18     2     2           1          2         1       5         5   \n",
       "264   18     2     2           1          3         0       4         3   \n",
       "337   17     3     2           1          2         0       4         3   \n",
       "343   17     2     2           1          2         1       3         3   \n",
       "333   18     2     2           1          2         0       4         3   \n",
       "247   22     3     1           1          1         3       5         4   \n",
       "\n",
       "     goout  Dalc  ...  G1  G2  G3  student_resid   cooks_d  flag_resid  \\\n",
       "259      1     1  ...  10   9   0      -4.350386  0.068008           1   \n",
       "296      4     2  ...  10   9   0      -4.202122  0.061774           1   \n",
       "310      4     2  ...   9   9   0      -4.196503  0.046214           1   \n",
       "341      3     2  ...  10  10   0      -4.814491  0.042420           1   \n",
       "239      4     3  ...   7   7   0      -3.262131  0.036015           1   \n",
       "264      3     1  ...   9  10   0      -5.091042  0.034292           1   \n",
       "337      2     2  ...   7   8   0      -3.839841  0.030653           1   \n",
       "343      1     1  ...   9   8   0      -3.536514  0.023329           1   \n",
       "333      3     1  ...   8   8   0      -3.811963  0.021069           1   \n",
       "247      5     5  ...   6   8   8       1.290461  0.019418           0   \n",
       "\n",
       "     flag_cook  flag_iso  votes  anomaly  \n",
       "259          1         0      2        1  \n",
       "296          1         0      2        1  \n",
       "310          1         0      2        1  \n",
       "341          1         0      2        1  \n",
       "239          1         0      2        1  \n",
       "264          1         0      2        1  \n",
       "337          1         0      2        1  \n",
       "343          1         0      2        1  \n",
       "333          1         0      2        1  \n",
       "247          1         1      2        1  \n",
       "\n",
       "[10 rows x 23 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "most_anomalous = anomaly_df[anomaly_df[\"anomaly\"] == 1].copy()\n",
    "most_anomalous = most_anomalous.sort_values(\n",
    "    [\"votes\", \"cooks_d\", \"student_resid\"], ascending=[False, False, False]\n",
    ")\n",
    "\n",
    "display(most_anomalous.head(10))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3ade059b",
   "metadata": {},
   "source": [
    "## 11. Визуализация аномалий"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "925ca5bb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "plt.figure(figsize=(8, 5))\n",
    "plt.scatter(y_pred_train, student_resid, alpha=0.7, label=\"Объекты\")\n",
    "if len(most_anomalous) > 0:\n",
    "    anomalous_pred = model_full.predict(sm.add_constant(most_anomalous[numeric_features], has_constant='add'))\n",
    "    plt.scatter(anomalous_pred, most_anomalous[\"student_resid\"], s=80, marker=\"x\", label=\"Аномалии\")\n",
    "plt.axhline(3, linestyle=\"--\")\n",
    "plt.axhline(-3, linestyle=\"--\")\n",
    "plt.xlabel(\"Предсказанные значения\")\n",
    "plt.ylabel(\"Стандартизованные остатки\")\n",
    "plt.title(\"Аномалии по регрессионной модели\")\n",
    "plt.legend()\n",
    "plt.grid(alpha=0.3)\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d63283b2",
   "metadata": {},
   "source": [
    "## 12. Удаление аномалий и повторная регрессия"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "90608b89",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Размер train до очистки: 316\n",
      "Размер train после очистки: 304\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                     G3   R-squared:                       0.918\n",
      "Model:                            OLS   Adj. R-squared:                  0.914\n",
      "Method:                 Least Squares   F-statistic:                     215.2\n",
      "Date:                Fri, 13 Mar 2026   Prob (F-statistic):          1.27e-146\n",
      "Time:                        17:10:25   Log-Likelihood:                -488.76\n",
      "No. Observations:                 304   AIC:                             1010.\n",
      "Df Residuals:                     288   BIC:                             1069.\n",
      "Df Model:                          15                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "==============================================================================\n",
      "                 coef    std err          t      P>|t|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "const         -1.7908      1.184     -1.513      0.131      -4.121       0.539\n",
      "age           -0.0143      0.062     -0.229      0.819      -0.137       0.109\n",
      "Medu           0.0440      0.087      0.506      0.613      -0.127       0.215\n",
      "Fedu          -0.0801      0.086     -0.930      0.353      -0.250       0.089\n",
      "traveltime     0.0506      0.110      0.461      0.645      -0.166       0.267\n",
      "studytime      0.0860      0.093      0.930      0.353      -0.096       0.268\n",
      "failures      -0.5027      0.113     -4.462      0.000      -0.724      -0.281\n",
      "famrel         0.2423      0.084      2.877      0.004       0.077       0.408\n",
      "freetime       0.0593      0.078      0.759      0.448      -0.094       0.213\n",
      "goout          0.0101      0.079      0.129      0.898      -0.145       0.165\n",
      "Dalc          -0.0034      0.105     -0.033      0.974      -0.210       0.203\n",
      "Walc           0.0826      0.079      1.042      0.298      -0.073       0.239\n",
      "health        -0.0767      0.052     -1.468      0.143      -0.179       0.026\n",
      "absences       0.0182      0.009      2.067      0.040       0.001       0.036\n",
      "G1             0.0779      0.042      1.851      0.065      -0.005       0.161\n",
      "G2             0.9896      0.036     27.741      0.000       0.919       1.060\n",
      "==============================================================================\n",
      "Omnibus:                       96.315   Durbin-Watson:                   2.061\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):              390.354\n",
      "Skew:                          -1.293   Prob(JB):                     1.72e-85\n",
      "Kurtosis:                       7.912   Cond. No.                         423.\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    }
   ],
   "source": [
    "\n",
    "clean_train = anomaly_df[anomaly_df[\"anomaly\"] == 0].copy()\n",
    "\n",
    "X_train_clean = clean_train[numeric_features]\n",
    "y_train_clean = clean_train[\"G3\"]\n",
    "\n",
    "X_train_clean_sm = sm.add_constant(X_train_clean)\n",
    "model_clean = sm.OLS(y_train_clean, X_train_clean_sm).fit()\n",
    "\n",
    "print(\"Размер train до очистки:\", len(X_train))\n",
    "print(\"Размер train после очистки:\", len(X_train_clean))\n",
    "print(model_clean.summary())\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b79df031",
   "metadata": {},
   "source": [
    "## 13. Метрики модели после очистки"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "29c0a7b3",
   "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>Метрика</th>\n",
       "      <th>Значение</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>MAE_train_clean</td>\n",
       "      <td>8.202420e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>RMSE_train_clean</td>\n",
       "      <td>1.207813e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>R2_train_clean</td>\n",
       "      <td>9.181009e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>MAE_test_clean</td>\n",
       "      <td>1.195890e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>RMSE_test_clean</td>\n",
       "      <td>2.126471e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>R2_test_clean</td>\n",
       "      <td>7.794749e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>F_stat_pvalue_clean</td>\n",
       "      <td>1.265592e-146</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Метрика       Значение\n",
       "0      MAE_train_clean   8.202420e-01\n",
       "1     RMSE_train_clean   1.207813e+00\n",
       "2       R2_train_clean   9.181009e-01\n",
       "3       MAE_test_clean   1.195890e+00\n",
       "4      RMSE_test_clean   2.126471e+00\n",
       "5        R2_test_clean   7.794749e-01\n",
       "6  F_stat_pvalue_clean  1.265592e-146"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "y_pred_train_clean = model_clean.predict(X_train_clean_sm)\n",
    "y_pred_test_clean = model_clean.predict(X_test_sm)\n",
    "\n",
    "metrics_clean = pd.DataFrame({\n",
    "    \"Метрика\": [\"MAE_train_clean\", \"RMSE_train_clean\", \"R2_train_clean\", \"MAE_test_clean\", \"RMSE_test_clean\", \"R2_test_clean\", \"F_stat_pvalue_clean\"],\n",
    "    \"Значение\": [\n",
    "        mean_absolute_error(y_train_clean, y_pred_train_clean),\n",
    "        mean_squared_error(y_train_clean, y_pred_train_clean) ** 0.5,\n",
    "        r2_score(y_train_clean, y_pred_train_clean),\n",
    "        mean_absolute_error(y_test, y_pred_test_clean),\n",
    "        mean_squared_error(y_test, y_pred_test_clean) ** 0.5,\n",
    "        r2_score(y_test, y_pred_test_clean),\n",
    "        model_clean.f_pvalue\n",
    "    ]\n",
    "})\n",
    "\n",
    "display(metrics_clean)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bea133fa",
   "metadata": {},
   "source": [
    "## 14. Сравнение моделей до и после очистки"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "66bdbb80",
   "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>Показатель</th>\n",
       "      <th>До очистки</th>\n",
       "      <th>После очистки</th>\n",
       "      <th>Изменение</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>R2_train</td>\n",
       "      <td>0.848640</td>\n",
       "      <td>0.918101</td>\n",
       "      <td>0.069461</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>R2_test</td>\n",
       "      <td>0.780358</td>\n",
       "      <td>0.779475</td>\n",
       "      <td>-0.000883</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>MAE_test</td>\n",
       "      <td>1.348164</td>\n",
       "      <td>1.195890</td>\n",
       "      <td>-0.152273</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>RMSE_test</td>\n",
       "      <td>2.122209</td>\n",
       "      <td>2.126471</td>\n",
       "      <td>0.004262</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Adj_R2_train</td>\n",
       "      <td>0.841072</td>\n",
       "      <td>0.913835</td>\n",
       "      <td>0.072763</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Показатель  До очистки  После очистки  Изменение\n",
       "0      R2_train    0.848640       0.918101   0.069461\n",
       "1       R2_test    0.780358       0.779475  -0.000883\n",
       "2      MAE_test    1.348164       1.195890  -0.152273\n",
       "3     RMSE_test    2.122209       2.126471   0.004262\n",
       "4  Adj_R2_train    0.841072       0.913835   0.072763"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "compare = pd.DataFrame({\n",
    "    \"Показатель\": [\"R2_train\", \"R2_test\", \"MAE_test\", \"RMSE_test\", \"Adj_R2_train\"],\n",
    "    \"До очистки\": [\n",
    "        r2_score(y_train, y_pred_train),\n",
    "        r2_score(y_test, y_pred_test),\n",
    "        mean_absolute_error(y_test, y_pred_test),\n",
    "        mean_squared_error(y_test, y_pred_test) ** 0.5,\n",
    "        model_full.rsquared_adj\n",
    "    ],\n",
    "    \"После очистки\": [\n",
    "        r2_score(y_train_clean, y_pred_train_clean),\n",
    "        r2_score(y_test, y_pred_test_clean),\n",
    "        mean_absolute_error(y_test, y_pred_test_clean),\n",
    "        mean_squared_error(y_test, y_pred_test_clean) ** 0.5,\n",
    "        model_clean.rsquared_adj\n",
    "    ]\n",
    "})\n",
    "\n",
    "compare[\"Изменение\"] = compare[\"После очистки\"] - compare[\"До очистки\"]\n",
    "display(compare)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "21dee706",
   "metadata": {},
   "source": [
    "## 15. Сравнение коэффициентов"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "efcfb5cf",
   "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>coef_before</th>\n",
       "      <th>coef_after</th>\n",
       "      <th>abs_change</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>const</th>\n",
       "      <td>-0.468444</td>\n",
       "      <td>-1.790753</td>\n",
       "      <td>1.322309</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>age</th>\n",
       "      <td>-0.198011</td>\n",
       "      <td>-0.014311</td>\n",
       "      <td>0.183700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>studytime</th>\n",
       "      <td>-0.066050</td>\n",
       "      <td>0.086027</td>\n",
       "      <td>0.152077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>health</th>\n",
       "      <td>0.059078</td>\n",
       "      <td>-0.076683</td>\n",
       "      <td>0.135761</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>goout</th>\n",
       "      <td>0.137633</td>\n",
       "      <td>0.010121</td>\n",
       "      <td>0.127511</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fedu</th>\n",
       "      <td>-0.188338</td>\n",
       "      <td>-0.080146</td>\n",
       "      <td>0.108192</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Dalc</th>\n",
       "      <td>-0.105011</td>\n",
       "      <td>-0.003420</td>\n",
       "      <td>0.101591</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>famrel</th>\n",
       "      <td>0.334612</td>\n",
       "      <td>0.242309</td>\n",
       "      <td>0.092303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>failures</th>\n",
       "      <td>-0.416258</td>\n",
       "      <td>-0.502701</td>\n",
       "      <td>0.086443</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>G1</th>\n",
       "      <td>0.160748</td>\n",
       "      <td>0.077858</td>\n",
       "      <td>0.082890</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>traveltime</th>\n",
       "      <td>0.131009</td>\n",
       "      <td>0.050644</td>\n",
       "      <td>0.080365</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Medu</th>\n",
       "      <td>0.094194</td>\n",
       "      <td>0.044032</td>\n",
       "      <td>0.050162</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freetime</th>\n",
       "      <td>0.010623</td>\n",
       "      <td>0.059264</td>\n",
       "      <td>0.048640</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>absences</th>\n",
       "      <td>0.044864</td>\n",
       "      <td>0.018236</td>\n",
       "      <td>0.026629</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Walc</th>\n",
       "      <td>0.061756</td>\n",
       "      <td>0.082644</td>\n",
       "      <td>0.020887</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>G2</th>\n",
       "      <td>0.977534</td>\n",
       "      <td>0.989621</td>\n",
       "      <td>0.012087</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            coef_before  coef_after  abs_change\n",
       "const         -0.468444   -1.790753    1.322309\n",
       "age           -0.198011   -0.014311    0.183700\n",
       "studytime     -0.066050    0.086027    0.152077\n",
       "health         0.059078   -0.076683    0.135761\n",
       "goout          0.137633    0.010121    0.127511\n",
       "Fedu          -0.188338   -0.080146    0.108192\n",
       "Dalc          -0.105011   -0.003420    0.101591\n",
       "famrel         0.334612    0.242309    0.092303\n",
       "failures      -0.416258   -0.502701    0.086443\n",
       "G1             0.160748    0.077858    0.082890\n",
       "traveltime     0.131009    0.050644    0.080365\n",
       "Medu           0.094194    0.044032    0.050162\n",
       "freetime       0.010623    0.059264    0.048640\n",
       "absences       0.044864    0.018236    0.026629\n",
       "Walc           0.061756    0.082644    0.020887\n",
       "G2             0.977534    0.989621    0.012087"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "coef_compare = pd.DataFrame({\n",
    "    \"coef_before\": model_full.params,\n",
    "    \"coef_after\": model_clean.params\n",
    "}).fillna(0)\n",
    "\n",
    "coef_compare[\"abs_change\"] = (coef_compare[\"coef_after\"] - coef_compare[\"coef_before\"]).abs()\n",
    "display(coef_compare.sort_values(\"abs_change\", ascending=False))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ce897091",
   "metadata": {},
   "source": [
    "## 16. Итог для заполнения после запуска"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e6df354a",
   "metadata": {},
   "source": [
    "\n",
    "## 16. Итог\n",
    "\n",
    "В качестве зависимой переменной выбрана **`G3`** — итоговая оценка студента.\n",
    "\n",
    "### Базовая модель\n",
    "Исходная регрессия показала хорошее качество:\n",
    "- **R²_train = 0.849**\n",
    "- **R²_test = 0.780**\n",
    "- **MAE_test = 1.348**\n",
    "- **RMSE_test = 2.122**\n",
    "- модель в целом значима: **p-value F-критерия = 3.04e-113**\n",
    "\n",
    "Наиболее значимым предиктором оказался **`G2`** (`coef = 0.978`, `p < 0.001`).  \n",
    "Также значимы **age**, **failures**, **famrel**, **absences** и **G1**.\n",
    "\n",
    "Уравнение регрессии:\n",
    "**G3 = -0.468 - 0.198·age + 0.094·Medu - 0.188·Fedu + 0.131·traveltime - 0.066·studytime - 0.416·failures + 0.335·famrel + 0.011·freetime + 0.138·goout - 0.105·Dalc + 0.062·Walc + 0.059·health + 0.045·absences + 0.161·G1 + 0.978·G2**\n",
    "\n",
    "### Аномалии\n",
    "Консенсусом трёх методов было найдено **12 аномальных объектов**.  \n",
    "Наиболее аномальные наблюдения — студенты с **G1 ≈ 7–10, G2 ≈ 7–10, но G3 = 0**.  \n",
    "Именно они сильнее всего искажают регрессию, потому что итоговая оценка резко не согласуется с промежуточными.\n",
    "\n",
    "### Модель после удаления аномалий\n",
    "После очистки обучающей выборки качество на train заметно выросло:\n",
    "- **R²_train: 0.849 → 0.918**\n",
    "- **Adj. R²_train: 0.841 → 0.914**\n",
    "\n",
    "На тестовой выборке изменения небольшие:\n",
    "- **R²_test: 0.780 → 0.779**\n",
    "- **MAE_test: 1.348 → 1.196**\n",
    "- **RMSE_test: 2.122 → 2.126**\n",
    "\n",
    "Модель после очистки тоже значима:\n",
    "- **p-value F-критерия = 1.27e-146**\n",
    "\n",
    "### Сравнение моделей\n",
    "После удаления аномалий модель стала лучше описывать обучающую выборку и дала меньшую среднюю абсолютную ошибку на тесте, но **R²_test практически не изменился**, а **RMSE_test слегка ухудшился**.  \n",
    "Это означает, что очистка убрала сильные выбросы и сделала оценки коэффициентов более устойчивыми, но не дала существенного прироста обобщающей способности.\n",
    "\n",
    "### Вывод\n",
    "Для прогноза **G3** основную роль играет **G2**, а также влияют **failures**, **famrel**, **absences**, **age** и частично **G1**.  \n",
    "Аномалии в данных действительно есть, и их удаление делает модель более стабильной, однако итоговое качество на тестовой выборке улучшается лишь незначительно.  \n",
    "Следовательно, очистка выборки в данном случае **полезна для интерпретации модели**, но не даёт сильного выигрыша в прогнозной точности.\n"
   ]
  }
 ],
 "metadata": {
  "language_info": {
   "name": "python"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}