hw6_1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Домашнее задание 6: классификация текстов"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "В этом домашнем задании вам предстоит построить классификатор текстов и поучаствовать в соревновании на Kaggle!\n",
    "Первым делом вам предстоит построить хороший бейзлайн, а дальше пытаться улучшать, соревнуясь с другими участниками.\n",
    "\n",
    "\n",
    "Ссылка на соревнование: https://www.kaggle.com/c/avito-category-prediction/overview\n",
    "Данные для домашнего задания можно скачать на странице соревнования.\n",
    "Чтобы ваше участие было засчитано, убедитесь, что имя в Leaderboard имеет вид: «Имя Фамилия номер_группы».\n",
    "\n",
    "Оценивание:\n",
    "\n",
    "Домашнее задание оценивается как обычно. Баллы указаны напротив заданий.\n",
    "\n",
    "За соревнование даются бонусные баллы следующим образом.\n",
    "Если вы пересекли baseline_2 на приватном лидерборде, ваша оценка равна \n",
    "\n",
    "10 - 10 * (i - 1) / M\n",
    "\n",
    "где M — количество студентов, принявших участие в соревновании;\n",
    "\n",
    "i — место (начиная с 1) студента в приватном лидерборде среди всех таких студентов.\n",
    "\n",
    "Правила:\n",
    "\n",
    "* Домашнее задание оценивается в 10 баллов.\n",
    "\n",
    "* Плагиат не допускается. При обнаружении случаев списывания, 0 за работу выставляется всем участникам нарушения, даже если можно установить, кто у кого списал.\n",
    "\n",
    "* Старайтесь сделать код как можно более оптимальным. В частности, будет штрафоваться использование циклов в тех случаях, когда операцию можно совершить при помощи инструментов библиотек, о которых рассказывалось в курсе.  \n",
    "\n",
    "* В течение 3 суток после окончания соревнования в соответствующее задание на anytask необходимо прислать код, воспроизводящий ответы для посылки, фигурирующей в приватном лидерборде. В случае отсутствия кода, воспроизводящего результат, в установленный срок студенту выставляется 0 в качестве оценки за соревнование. Если не оговорено иное, использовать любые внешние данные в соревнованиях '''запрещено'''. Под внешними данными понимаются размеченные данные, где разметка имеет прямое отношение к решаемой задаче. Грубо говоря, сборник текстов с википедии не считается внешними данными.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Мы будем работать с датасетом объявлений Avito. Нам предстоит по заголовку и тексту объявления предсказывать категорию объявления."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import numpy as np\n",
    "from matplotlib import pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = pd.read_csv('/Users/andreydyadyunov/Downloads/train.csv')\n",
    "test = pd.read_csv('/Users/andreydyadyunov/Downloads/test.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>title</th>\n",
       "      <th>description</th>\n",
       "      <th>Category_name</th>\n",
       "      <th>Category</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Эбу Renault / Nissan 1,5 DCI Delphi 2001-2008 год</td>\n",
       "      <td>Комплект ЭБУ (мозги, компьютер мотора, двигате...</td>\n",
       "      <td>Запчасти и аксессуары</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Утюг утп 1000 ватт СССР 1987 год</td>\n",
       "      <td>Продам/\\n Фото № 1-2 /\\n /\\nУтюг УТП 1000 ватт...</td>\n",
       "      <td>Бытовая техника</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Возвму машину с выкупом</td>\n",
       "      <td>Возьму машину в аренду с последующим выкупом н...</td>\n",
       "      <td>Предложение услуг</td>\n",
       "      <td>114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Полусапожки</td>\n",
       "      <td>полусапожки в отличном состоянии, один раз оде...</td>\n",
       "      <td>Одежда, обувь, аксессуары</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Босоножки кожаные</td>\n",
       "      <td>Кожаные(натур) босоножки Karlo Pasolini, 40 рр...</td>\n",
       "      <td>Одежда, обувь, аксессуары</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                               title  \\\n",
       "0  Эбу Renault / Nissan 1,5 DCI Delphi 2001-2008 год   \n",
       "1                   Утюг утп 1000 ватт СССР 1987 год   \n",
       "2                            Возвму машину с выкупом   \n",
       "3                                        Полусапожки   \n",
       "4                                  Босоножки кожаные   \n",
       "\n",
       "                                         description  \\\n",
       "0  Комплект ЭБУ (мозги, компьютер мотора, двигате...   \n",
       "1  Продам/\\n Фото № 1-2 /\\n /\\nУтюг УТП 1000 ватт...   \n",
       "2  Возьму машину в аренду с последующим выкупом н...   \n",
       "3  полусапожки в отличном состоянии, один раз оде...   \n",
       "4  Кожаные(натур) босоножки Karlo Pasolini, 40 рр...   \n",
       "\n",
       "               Category_name  Category  \n",
       "0      Запчасти и аксессуары        10  \n",
       "1            Бытовая техника        21  \n",
       "2          Предложение услуг       114  \n",
       "3  Одежда, обувь, аксессуары        27  \n",
       "4  Одежда, обувь, аксессуары        27  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Задание 1 (0.5 балла)\n",
    "\n",
    "Выведете на экран информацию о пропусках в данных. Если пропуски присутствуют заполните их пустой строкой."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Пропуски в трейне:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "title            0\n",
       "description      2\n",
       "Category_name    0\n",
       "Category         0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print('Пропуски в трейне:')\n",
    "train.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Пропуски в трейне:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "title            0\n",
       "description      0\n",
       "Category_name    0\n",
       "Category         0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train['description'] = train['description'].fillna('')\n",
    "print('Пропуски в трейне:')\n",
    "train.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Пропуски в тесте:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "title          0\n",
       "description    1\n",
       "itemid         0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print('Пропуски в тесте:')\n",
    "test.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Пропуски в тесте:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "title          0\n",
       "description    0\n",
       "itemid         0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test['description'] = test['description'].fillna('')\n",
    "print('Пропуски в тесте:')\n",
    "test.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Задание 2 (0.5 балла)\n",
    "Сконкатенируйте заголовок и описание в единую строку и поместите результат в отдельный столбец. Таким образом мы будет работать с одним текстом, а не с двумя."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>Category_name</th>\n",
       "      <th>Category</th>\n",
       "      <th>title_description</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Запчасти и аксессуары</td>\n",
       "      <td>10</td>\n",
       "      <td>Эбу Renault / Nissan 1,5 DCI Delphi 2001-2008 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Бытовая техника</td>\n",
       "      <td>21</td>\n",
       "      <td>Утюг утп 1000 ватт СССР 1987 год Продам/\\n Фот...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Предложение услуг</td>\n",
       "      <td>114</td>\n",
       "      <td>Возвму машину с выкупом Возьму машину в аренду...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Одежда, обувь, аксессуары</td>\n",
       "      <td>27</td>\n",
       "      <td>Полусапожки полусапожки в отличном состоянии, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Одежда, обувь, аксессуары</td>\n",
       "      <td>27</td>\n",
       "      <td>Босоножки кожаные Кожаные(натур) босоножки Kar...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Category_name  Category  \\\n",
       "0      Запчасти и аксессуары        10   \n",
       "1            Бытовая техника        21   \n",
       "2          Предложение услуг       114   \n",
       "3  Одежда, обувь, аксессуары        27   \n",
       "4  Одежда, обувь, аксессуары        27   \n",
       "\n",
       "                                   title_description  \n",
       "0  Эбу Renault / Nissan 1,5 DCI Delphi 2001-2008 ...  \n",
       "1  Утюг утп 1000 ватт СССР 1987 год Продам/\\n Фот...  \n",
       "2  Возвму машину с выкупом Возьму машину в аренду...  \n",
       "3  Полусапожки полусапожки в отличном состоянии, ...  \n",
       "4  Босоножки кожаные Кожаные(натур) босоножки Kar...  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train['title_description'] = train.title + ' ' + train.description\n",
    "test['title_description'] = test.title + ' ' + test.description\n",
    "train = train.drop(['title', 'description'], axis=1)\n",
    "test = test.drop(['title', 'description'], axis=1)\n",
    "train.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Задание 3 (0.5 балла)\n",
    "Давайте немного посмотрим на наши данные. Визуализируйте (где явно просят) или выведете информацию о следующем:\n",
    "\n",
    "1. Сколько всего уникальных классов необходимо предсказать?\n",
    "2. Постройте столбчатую диаграмму распределения по классам. Нормализуйте график, чтобы видеть доли, а не абсолютные значения.\n",
    "3. На сколько большой дисбаланс? Какой класс имеет самую большую долю в выборке (укажите класс и долю), какой класс наименьшую (укажите класс и долю)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Необходимо предсказать:\n",
      "50 строк\n",
      "Наибольшую долю имеет класс: Одежда, обувь, аксессуары - 0.1807067572782698\n",
      "Наименьшую долю имеет класс: Недвижимость за рубежом - 3.8969854337769914e-05\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print('Необходимо предсказать:\\n{rows} строк'.format(rows=len(train.Category.unique())))\n",
    "fig = plt.figure(figsize=(20, 10))\n",
    "plt.xticks(rotation=80)\n",
    "plt.title('Доля по категориям')\n",
    "df = pd.DataFrame(train['Category_name'].value_counts())\n",
    "df['categ'] = df.index\n",
    "df = df.reset_index(drop = True)\n",
    "plt.bar(df['categ'], df['Category_name']/df['Category_name'].sum())\n",
    "print('Наибольшую долю имеет класс: {maxi} - {part1}\\nНаименьшую долю имеет класс: {mini} - {part2}'.format(maxi=df.categ[0], mini=df.categ[49], part1=(df.Category_name[0]/df.Category_name.sum()), part2=(df.Category_name[49]/df.Category_name.sum())))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Задание 4 (0.5 балла)\n",
    "\n",
    "Далее мы будем пока работать только с train частью. Для test части нам неизвестны истинные значения, и они понадобятся только тогда, когда мы будем отправлять решение на Kaggle.\n",
    "\n",
    "1. Предобработайте данные (train часть) с помощью CountVectorizer.\n",
    "2. Какого размера получилась матрица?\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "vectorizer = CountVectorizer()\n",
    "X = vectorizer.fit_transform(train.title_description)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Размер матрицы:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(4234042, 2125805)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print('Размер матрицы:')\n",
    "X.shape\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Задание 5 (0.5 балла)\n",
    "\n",
    "В предыдущем пункте у вас должная была получится очень большая матрица. Это не дело.\n",
    "Если вы взгляните на текст, то увидете, что там есть множество специальных символов.\n",
    "\n",
    "Давайте также посмотрим на словарь, который получился в результате построения CountVectorizer, его можно найти в поле _vocabulary инстанса этого класса.\n",
    "\n",
    "1. Найдите в этом словаре все слова, которые начинаются на цифру. Сколько таких слов нашлось?\n",
    "\n",
    "2. Найдите все слова, которые начинаются на символы пунктуации. Сколько таких слов нашлось? "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-12-e1d560f19ab2>:5: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df1.numbers[j] = 1\n"
     ]
    }
   ],
   "source": [
    "df1 = pd.DataFrame(vectorizer.vocabulary_.keys(), columns=['words'])\n",
    "df1['numbers'] = 0\n",
    "for j, i in enumerate(df1.words):\n",
    "    if i[0] in '0123456789': \n",
    "        df1.numbers[j] = 1\n",
    "    else:\n",
    "        pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Элементов начинающихся на цифру:\n",
      "610871\n"
     ]
    }
   ],
   "source": [
    "print(\"Элементов начинающихся на цифру:\\n{numb}\".format(numb=df1.numbers.sum()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-14-dbaf10a797c1>:5: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df1.punctuation[j] = 1\n"
     ]
    }
   ],
   "source": [
    "from string import punctuation as punc_lst\n",
    "df1['punctuation'] = 0\n",
    "for j, i in enumerate(df1.words):\n",
    "    if i[0] in punc_lst: \n",
    "        df1.punctuation[j] = 1\n",
    "    else:\n",
    "        pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Элементов начинающихся на знак препинания:\n",
      "2043\n"
     ]
    }
   ],
   "source": [
    "print(\"Элементов начинающихся на знак препинания:\\n{punc}\".format(punc=df1.punctuation.sum()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Задание 6 (2.5 балла)\n",
    "\n",
    "В scikit-learn мы можем оценивать процесс подсчета матрицы через CountVectorizer. У CountVectorizer, как и у других наследников \\_VectorizerMixin, есть аргумент tokenizer и preprocessor. preprocessor применится в самом начале к каждой строке вашего датасета, tokenizer же должен принять строку и вернуть токены.\n",
    "Давайте напишем кастомный токенайзер, которые сделает все, что нам нужно: \n",
    "\n",
    "0. Приведет все буквы к нижнему регистру\n",
    "1. Разобьет текст на токены с помощью word_tokenizer из пакета nltk\n",
    "2. Удалит все токены содержащие числа и пунктуацию\n",
    "3. Удалит все токены, которые перечислены в nltk.corpus.stopwords('russian')\n",
    "4. Проведет стемминг с помощью SnowballStemmer\n",
    "\n",
    "Продемонстрируйте работу вашей функии на самом первом описании товара в датасете.\n",
    "\n",
    "Важно: так как функию будет необходимо запускать на большом корпусе, подумайте об эффективной реализации. Постарайтесь решить задачу в минимальное количество циклов и копирований.\n",
    "\n",
    "Референсная реализация на компьютере автора работала 230 секунд на 100к строчек."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: nltk in /Users/andreydyadyunov/Downloads/anaconda3/lib/python3.8/site-packages (3.5)\n",
      "Requirement already satisfied: click in /Users/andreydyadyunov/Downloads/anaconda3/lib/python3.8/site-packages (from nltk) (7.1.2)\n",
      "Requirement already satisfied: tqdm in /Users/andreydyadyunov/Downloads/anaconda3/lib/python3.8/site-packages (from nltk) (4.50.2)\n",
      "Requirement already satisfied: regex in /Users/andreydyadyunov/Downloads/anaconda3/lib/python3.8/site-packages (from nltk) (2020.10.15)\n",
      "Requirement already satisfied: joblib in /Users/andreydyadyunov/Downloads/anaconda3/lib/python3.8/site-packages (from nltk) (0.17.0)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[nltk_data] Downloading package punkt to\n",
      "[nltk_data]     /Users/andreydyadyunov/nltk_data...\n",
      "[nltk_data]   Package punkt is already up-to-date!\n",
      "[nltk_data] Downloading package stopwords to\n",
      "[nltk_data]     /Users/andreydyadyunov/nltk_data...\n",
      "[nltk_data]   Package stopwords is already up-to-date!\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "!pip install nltk\n",
    "import nltk\n",
    "nltk.download('punkt')\n",
    "from nltk.corpus import stopwords\n",
    "from nltk.tokenize import word_tokenize\n",
    "nltk.download('stopwords')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "import re\n",
    "from nltk.stem.snowball import SnowballStemmer\n",
    "stemmer = SnowballStemmer(\"russian\")\n",
    "def preprocessor(string):\n",
    "    string = string.lower()\n",
    "    string = word_tokenize(string)\n",
    "    df = pd.DataFrame(string ,columns = ['tokens'])\n",
    "    string = df.loc[~df['tokens'].isin(list(stopwords.words('russian'))) == True].values.tolist()\n",
    "    description_new = []\n",
    "    for i in string:\n",
    "        if re.search(r'[\\d\\W]',i[0]) == None:\n",
    "            description_new.append(i[0])\n",
    "\n",
    "    return description_new\n",
    "\n",
    "def tokenizer (string: str):\n",
    "    string = [stemmer.stem(s) for s in string]\n",
    "    return string\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Задание 7 (1 балл)\n",
    "\n",
    "1. Возьмите случайные 500к строк из датасета. Постройте по ним CountVectorizer с применением вашего токенизатора.\n",
    "2. Разбейте полученную матрицу на train, test в отношении 4 к 1. Не забудьте про target переменную.\n",
    "3. Обучите SGDClassifier на полученной выборке.\n",
    "4. Посчитайте метрику accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.83445"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.linear_model import SGDClassifier\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import accuracy_score\n",
    "classifier = SGDClassifier()\n",
    "df3 = train[['Category', 'title_description']].sample(n=100000, random_state=4)\n",
    "X = df3.title_description\n",
    "Y = df3.Category\n",
    "vectorizer = CountVectorizer(tokenizer=lambda x: tokenizer(x), preprocessor=lambda y: preprocessor(y))\n",
    "X = vectorizer.fit_transform(X)\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size=0.2, random_state=4)\n",
    "classifier.fit(X_train, y_train)\n",
    "y_test_pred = classifier.predict(X_test)\n",
    "accuracy_score(y_test_pred, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Задание 8 (1 балл)\n",
    "\n",
    "1. Повторите 7 задание, но с tf-idf векторизатором. Как изменилось качество.\n",
    "2. Мы можем еще сильнее уменьшите размер нашей матрицы, если отбросим значения df близкие к единице. Скорее всего такие слова не несут много информации о категории, так как встречаются достаточно часто. Ограничьте максимальный df в параметрах TfIdfVectorizer, поставьте верхнюю границу равную 0.9. Как изменился размер матрицы, как изменилось качество?\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.83315"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "vectorizer = TfidfVectorizer(tokenizer=lambda x: tokenizer(x), preprocessor=lambda y: preprocessor(y), max_df=0.9)\n",
    "X = df3.title_description\n",
    "Y = df3.Category\n",
    "X_tfid = vectorizer.fit_transform(X)\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_tfid, Y, test_size=0.2, random_state=4)\n",
    "classifier.fit(X_train, y_train)\n",
    "y_test_pred = classifier.predict(X_test)\n",
    "accuracy_score(y_test_pred, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Качество улучшилось"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Задание 9 (1 балл)\n",
    "Еще один популяпный трюк, который позволит уменьшить количество признаков называется hashing trick. Его суть в том, то мы случайно группируем признаки ииии  ..... складываем их! А потом удаляем исходные признаки. В итоге все наши признаки это просто суммы исходных. Звучит странно, но это отлично работает. Давайте проверим этот трюк в нашем сеттинге.\n",
    "Также при таком подходе вам не нужно хранить словарь token->index, что тоже иногда полезно.\n",
    "\n",
    "1. Повторите задание 7 с HashingVectorizer, укажите количество фичей равное 30000.\n",
    "2. Какой из трех подходов показал самый высокий результат?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/andreydyadyunov/Downloads/anaconda3/lib/python3.8/site-packages/sklearn/feature_extraction/text.py:484: UserWarning: The parameter 'token_pattern' will not be used since 'tokenizer' is not None'\n",
      "  warnings.warn(\"The parameter 'token_pattern' will not be used\"\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.8161"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.feature_extraction.text import HashingVectorizer\n",
    "vectorizer = HashingVectorizer(tokenizer=lambda x: tokenizer(x), preprocessor=lambda y: preprocessor(y), n_features=30000)\n",
    "X = df3.title_description\n",
    "Y = df3.Category\n",
    "X_hash = vectorizer.fit_transform(X)\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_hash, Y, test_size=0.2, random_state=4)\n",
    "classifier.fit(X_train, y_train)\n",
    "y_test_pred = classifier.predict(X_test)\n",
    "accuracy_score(y_test_pred, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Задание 10 (1.5 балла)\n",
    "\n",
    "Пришло время выйти в мир. Отправляйтесь на Kaggle Inclass и сделайте первую посылку. Там вы найдете инструкцию как сформировать файл с предсказаниями и отправить его."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Не помню точно, какой вариант дал лучший результат\n",
    "Но мне разрешили отправить несколько\n",
    "ВАРИАНТ 1:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "vectorizer = TfidfVectorizer(tokenizer=lambda x: tokenizer(x), preprocessor=lambda y: preprocessor(y), max_df=0.8)\n",
    "df4 = train[['title_description', 'Category']]\n",
    "X = df4.title_description\n",
    "Y = df4.Category\n",
    "X_tfid = vectorizer.fit_transform(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test = test['title_description']\n",
    "X_test = vectorizer.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "classifier = SGDClassifier(max_iter=4000, alpha=0.0000001, penalty='elasticnet', l1_ratio=0.05)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SGDClassifier(alpha=1e-07, l1_ratio=0.03, max_iter=4000, penalty='elasticnet')"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "classifier.fit(X_tfid, Y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9205078740362046"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train_pred = classifier.predict(X_tfid)\n",
    "accuracy_score(y_train_pred, Y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_test_pred = classifier.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions1=pd.DataFrame(y_test_pred, columns=['Category'])\n",
    "predictions1['id'] = test.itemid\n",
    "predictions1.to_csv('predictions12.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "ВАРИАНТ 2:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "vectorizer = TfidfVectorizer(tokenizer=lambda x: tokenizer(x), preprocessor=lambda y: preprocessor(y), max_df=0.85)\n",
    "df4 = train[['title_description', 'Category']]\n",
    "X = df4.title_description\n",
    "Y = df4.Category\n",
    "X_tfid = vectorizer.fit_transform(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test = test['title_description']\n",
    "X_test = vectorizer.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "classifier = SGDClassifier(max_iter=4000, alpha=0.0000001)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "classifier.fit(X_tfid, Y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_train_pred = classifier.predict(X_tfid)\n",
    "accuracy_score(y_train_pred, Y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_test_pred = classifier.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions1=pd.DataFrame(y_test_pred, columns=['Category'])\n",
    "predictions1['id'] = test.itemid\n",
    "predictions1.to_csv('predictions_new.csv', index=False)"
   ]
  }
 ],
 "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}