Land_data_classification_article_Part1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Перший погляд на дані"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:11: FutureWarning: Passing a negative integer is deprecated in version 1.0 and will not be supported in future version. Instead, use None to not limit the column width.\n",
      "  # This is added back by InteractiveShellApp.init_path()\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import nltk\n",
    "import re\n",
    "import random\n",
    "from sklearn.model_selection import train_test_split, GridSearchCV, RandomizedSearchCV\n",
    "from ukrainian_stemmer import UkrainianStemmer\n",
    "from data_print import data_class_display\n",
    "\n",
    "pd.set_option('display.max_rows', None)\n",
    "pd.set_option('display.max_colwidth', -1)   \n",
    "\n",
    "land_data=pd.read_csv('land_real_estate.csv').replace(np.nan, \"\")\n",
    "land_data['text']=land_data['description']+' '+land_data['land_types_source']+' '+land_data['land_types_PcmU']\n",
    "def land_types_download():    \n",
    "    data=pd.read_csv('land_types.csv')\n",
    "    data=data[['id','land_types']]\n",
    "    data=data.set_index('id')\n",
    "    data=data.T\n",
    "    data=data.to_dict('records')\n",
    "    return (data[0])\n",
    "\n",
    "land_types=land_types_download()\n",
    "built_up={0:'Не забудована',1:'Забудована'}\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Типи земельних ділянок:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{1: 'ОЖБ',\n",
       " 2: 'ОСГ',\n",
       " 3: 'Cадова',\n",
       " 4: 'Комерційна',\n",
       " 5: 'Багатоповерхова',\n",
       " 6: 'Товарна',\n",
       " 7: 'ОЖБ та ОСГ'}"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print ('Типи земельних ділянок:')\n",
    "land_types"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Забудованість:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{0: 'Не забудована', 1: 'Забудована'}"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print ('Забудованість:')\n",
    "built_up"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Кількість помічених оглошень: 4256 екземплярів\n",
      "Перших 10 оголошень:\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>land_types_source</th>\n",
       "      <th>land_types_PcmU</th>\n",
       "      <th>description</th>\n",
       "      <th>built_up</th>\n",
       "      <th>land_types</th>\n",
       "      <th>land_squer</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>комерційного</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>5000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>під забудову</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>садівництво</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>470</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>сільськогосподарського</td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>10000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>земельних ділянок с пустомитівський р н км від львова є земельна ділянка сот приватизованої є старенький будиночок з виведеним газом та світлом вода біля будинку є кадастровий номер та ще сот в користуванні в центрі села торг</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>під забудову</td>\n",
       "      <td></td>\n",
       "      <td>увага дуже терміново сот вислобокихороше розташування поряд обжитий район поряд луцька трассаодразу за рестораном семільйонпо за соткукомунікації порядідеально під котеджі за усе</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>під забудову</td>\n",
       "      <td></td>\n",
       "      <td>за давидовом земельна ділянка розміром м прямокутної форми рівна зручний асфальтовий доїзд комунікації світло і газ поруч з ділянкою неподалік від ділянки ліс і ставок</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>зем діл вул шевченка кінець вулиці сот приватиз вздовж лісу метрів на ділянці також дерева вигляд на озеро біля ділянки світло і газ поруч збудований буд вже освітлюється у вечері вулиця розбудований р можна докупити діл поруч уо торг</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td></td>\n",
       "      <td>для будівництва і обслуговування житлового будинку господарських будівель і споруд присадибна ділянка для будівництва і обслуговування житлового будинку господарських будівель і споруд присадибна ділянка</td>\n",
       "      <td>га кадастровий номер га кадастровий номер м ужгород вул запорізька</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>ділянка під житлову забудову</td>\n",
       "      <td>для ведення особистого селянського господарства для ведення особистого селянського господарства для ведення особистого селянського господарства</td>\n",
       "      <td>га для ведення особистого селянського господарства біля річки в двадцяти метрах лісу та рибних ставків розкажу детально по телефону кадастровий номер можлива оренда</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>102</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              land_types_source  \\\n",
       "0  комерційного                   \n",
       "1  під забудову                   \n",
       "2  садівництво                    \n",
       "3  сільськогосподарського         \n",
       "4                                 \n",
       "5  під забудову                   \n",
       "6  під забудову                   \n",
       "7                                 \n",
       "8                                 \n",
       "9  ділянка під житлову забудову   \n",
       "\n",
       "                                                                                                                                                                                               land_types_PcmU  \\\n",
       "0                                                                                                                                                                                                                \n",
       "1                                                                                                                                                                                                                \n",
       "2                                                                                                                                                                                                                \n",
       "3                                                                                                                                                                                                                \n",
       "4                                                                                                                                                                                                                \n",
       "5                                                                                                                                                                                                                \n",
       "6                                                                                                                                                                                                                \n",
       "7                                                                                                                                                                                                                \n",
       "8  для будівництва і обслуговування житлового будинку господарських будівель і споруд присадибна ділянка для будівництва і обслуговування житлового будинку господарських будівель і споруд присадибна ділянка   \n",
       "9  для ведення особистого селянського господарства для ведення особистого селянського господарства для ведення особистого селянського господарства                                                               \n",
       "\n",
       "                                                                                                                                                                                                                                  description  \\\n",
       "0                                                                                                                                                                                                                                               \n",
       "1                                                                                                                                                                                                                                               \n",
       "2                                                                                                                                                                                                                                               \n",
       "3                                                                                                                                                                                                                                               \n",
       "4  земельних ділянок с пустомитівський р н км від львова є земельна ділянка сот приватизованої є старенький будиночок з виведеним газом та світлом вода біля будинку є кадастровий номер та ще сот в користуванні в центрі села торг            \n",
       "5  увага дуже терміново сот вислобокихороше розташування поряд обжитий район поряд луцька трассаодразу за рестораном семільйонпо за соткукомунікації порядідеально під котеджі за усе                                                           \n",
       "6  за давидовом земельна ділянка розміром м прямокутної форми рівна зручний асфальтовий доїзд комунікації світло і газ поруч з ділянкою неподалік від ділянки ліс і ставок                                                                      \n",
       "7  зем діл вул шевченка кінець вулиці сот приватиз вздовж лісу метрів на ділянці також дерева вигляд на озеро біля ділянки світло і газ поруч збудований буд вже освітлюється у вечері вулиця розбудований р можна докупити діл поруч уо торг   \n",
       "8  га кадастровий номер га кадастровий номер м ужгород вул запорізька                                                                                                                                                                           \n",
       "9  га для ведення особистого селянського господарства біля річки в двадцяти метрах лісу та рибних ставків розкажу детально по телефону кадастровий номер можлива оренда                                                                         \n",
       "\n",
       "   built_up  land_types land_squer  \n",
       "0  0         4           5000       \n",
       "1  0         1           1200       \n",
       "2  0         3           470        \n",
       "3  0         2           10000      \n",
       "4  1         1           1300       \n",
       "5  0         1           2100       \n",
       "6  0         1           1000       \n",
       "7  0         1           1300       \n",
       "8  0         1           1200       \n",
       "9  0         2           102        "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print ('Кількість помічених оглошень: {} екземплярів'.format(land_data.shape[0]))\n",
    "print ('Перших 10 оголошень:')\n",
    "land_data.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Розподіл вибірки за типами земельних ділянок"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data_class_display(dataframe=land_data,class_column='land_types',expl_lables=land_types)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Розподіл вибірки за забудованістю земельних ділянок"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data_class_display(dataframe=land_data,class_column='built_up',expl_lables=built_up)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Токенізатор та функція ngrams_info"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1 - grams\n",
      "ua_stemmer: True\n",
      "Кількість символів:  1100031\n",
      "Кількість токенів:  151318\n",
      "Кількість унікальних токенів:  6638\n",
      "Найбільш уживані токени:  [(('ділянк',), 6916), (('під',), 3472), (('і',), 3248), (('для',), 3032), (('будівництв',), 2584), (('на',), 2532), (('в',), 2458), (('будинк',), 2336), (('забудов',), 2310), (('земельн',), 2205), (('житлов',), 2034), (('поруч',), 1620), (('обслуговуванн',), 1614), (('та',), 1580), (('м',), 1532), (('комунікаці',), 1428), (('сот',), 1416), (('є',), 1406), (('с',), 1397), (('від',), 1362)]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3 - grams\n",
      "ua_stemmer: True\n",
      "Кількість символів:  1100031\n",
      "Кількість токенів:  151316\n",
      "Кількість унікальних токенів:  70599\n",
      "Найбільш уживані токени:  [(('обслуговуванн', 'житлов', 'будинк'), 1447), (('будівництв', 'і', 'обслуговуванн'), 1278), (('і', 'обслуговуванн', 'житлов'), 1240), (('для', 'будівництв', 'і'), 1205), (('будинк', 'господарськ', 'будівел'), 1186), (('будівел', 'і', 'споруд'), 1185), (('господарськ', 'будівел', 'і'), 1174), (('житлов', 'будинк', 'господарськ'), 1167), (('споруд', 'присадибн', 'ділянк'), 1079), (('і', 'споруд', 'присадибн'), 1075), (('продаж', 'земельн', 'ділянк'), 485), (('продаєт', 'земельн', 'ділянк'), 374), (('веденн', 'особист', 'селянськ'), 329), (('особист', 'селянськ', 'господарств'), 326), (('для', 'веденн', 'особист'), 310), (('земельн', 'ділянк', 'в'), 300), (('під', 'забудов', 'продаж'), 300), (('для', 'індивідуальн', 'садівництв'), 281), (('земельн', 'ділянк', 'під'), 260), (('будівництв', 'та', 'обслуговуванн'), 258)]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def ua_tokenizer(text,ua_stemmer=True,stop_words=[]):\n",
    "    \"\"\" Tokenizer for Ukrainian language, returns only alphabetic tokens. \n",
    "    \n",
    "    Keyword arguments:\n",
    "    text -- text for tokenize \n",
    "    ua_stemmer -- if True use UkrainianStemmer for stemming words (default True)\n",
    "    stop_words -- list of stop words (default [])\n",
    "        \n",
    "    \"\"\"\n",
    "    tokenized_list=[]\n",
    "    text=re.sub(r\"\"\"['’\"`�]\"\"\", '', text)\n",
    "    text=re.sub(r\"\"\"([0-9])([\\u0400-\\u04FF]|[A-z])\"\"\", r\"\\1 \\2\", text)\n",
    "    text=re.sub(r\"\"\"([\\u0400-\\u04FF]|[A-z])([0-9])\"\"\", r\"\\1 \\2\", text)\n",
    "    text=re.sub(r\"\"\"[\\-.,:+*/_]\"\"\", ' ', text)\n",
    "    for word in nltk.word_tokenize(text): \n",
    "        if word.isalpha():\n",
    "            word=word.lower() \n",
    "            if ua_stemmer is True:      \n",
    "                word=UkrainianStemmer(word).stem_word()\n",
    "            if word not in stop_words:\n",
    "                tokenized_list.append(word) \n",
    "    return tokenized_list\n",
    "\n",
    "\n",
    "\n",
    "def ngrams_info(series,n=1,most_common=20,ua_stemmer=True,stop_words=[]):\n",
    "    \"\"\" ngrams_info - Show detailed information about string pandas.Series column. \n",
    "    \n",
    "    Keyword arguments:\n",
    "    series -- pandas.Series object\n",
    "    most_common -- show most common words(default 50)\n",
    "    ua_stemmer -- if True use UkrainianStemmer for stemming words (default True)\n",
    "    stop_words -- list of stop words (default [])\n",
    "        \n",
    "    \"\"\"\n",
    "    print (n,'- grams')\n",
    "    print ('ua_stemmer:',ua_stemmer)\n",
    "    words=series.str.cat(sep=' ')\n",
    "    print ('Кількість символів: ',len(words))\n",
    "    words=nltk.ngrams(ua_tokenizer(words,ua_stemmer=ua_stemmer,stop_words=stop_words),n)\n",
    "    words=nltk.FreqDist(words)\n",
    "    print ('Кількість токенів: ',words.N())\n",
    "    print ('Кількість унікальних токенів: ',words.B())\n",
    "    print ('Найбільш уживані токени: ',words.most_common(most_common))\n",
    "    words.plot (most_common, cumulative = True)\n",
    "for n in (1,3):    \n",
    "    ngrams_info(land_data['text'],n=n)\n",
    "\n",
    "    \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Реалізація моделі мішка слів"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Класифікатор - <class 'nltk.classify.naivebayes.NaiveBayesClassifier'>\n",
      "Порядок n - 1\n",
      "Класифікатор за колонкою - land_types\n",
      "Точність класифікатора на навчальних даних: 0.8796913363732024\n",
      "Точність класифікатора на тестових даних: 0.8327402135231317\n",
      "  |   1   2   3   4   5   6   7 |\n",
      "--+-----------------------------+\n",
      "1 |<735>  4   5   7   8   .   9 |\n",
      "2 |   5 <86>  7   8   .   2   3 |\n",
      "3 |  76  15<271>  5   1   .   . |\n",
      "4 |  14   5   2 <61>  1   .   . |\n",
      "5 |   7   .   .   2  <3>  .   . |\n",
      "6 |   .   .   .   .   .  <.>  . |\n",
      "7 |  36   7   5   1   .   . <14>|\n",
      "--+-----------------------------+\n",
      "(row = reference; col = test)\n",
      "\n",
      "Найважливіші токени для класифікації за колонкою - land_types\n",
      "Most Informative Features\n",
      "contains(('виробництв',)) = True                6 : 1      =    620.9 : 1.0\n",
      "contains(('сільськогосподарськ',)) = True                6 : 1      =    443.5 : 1.0\n",
      " contains(('селянськ',)) = True                2 : 1      =    309.9 : 1.0\n",
      "   contains(('веденн',)) = True                6 : 1      =    282.2 : 1.0\n",
      "       contains(('me',)) = True                6 : 1      =    266.1 : 1.0\n",
      "    contains(('канал',)) = True                6 : 1      =    266.1 : 1.0\n",
      "contains(('sitalozemlyaзагальн',)) = True                6 : 1      =    266.1 : 1.0\n",
      "contains(('господарств',)) = True                2 : 3      =    257.8 : 1.0\n",
      "contains(('багатоквартирн',)) = True                5 : 1      =    234.7 : 1.0\n",
      "contains(('колективн',)) = True                3 : 1      =    226.3 : 1.0\n",
      "Класифікатор - <class 'nltk.classify.naivebayes.NaiveBayesClassifier'>\n",
      "Порядок n - 3\n",
      "Класифікатор за колонкою - land_types\n",
      "Точність класифікатора на навчальних даних: 0.5142055419151175\n",
      "Точність класифікатора на тестових даних: 0.5138790035587188\n",
      "  |   1   2   3   4   5   6   7 |\n",
      "--+-----------------------------+\n",
      "1 |<399>  4  14   1   2   .   5 |\n",
      "2 |  29 <45> 23   6   .   .   . |\n",
      "3 | 143  27<204> 11   3   .   . |\n",
      "4 |  54  16  11 <52>  .   .   5 |\n",
      "5 | 132   9  10   9  <4>  .   . |\n",
      "6 |   6   6  11   2   .  <2>  . |\n",
      "7 | 110  10  17   3   4   . <16>|\n",
      "--+-----------------------------+\n",
      "(row = reference; col = test)\n",
      "\n",
      "Найважливіші токени для класифікації за колонкою - land_types\n",
      "Most Informative Features\n",
      "contains(('сільськогосподарськ', 'земельн', 'ділянк')) = True                6 : 1      =    620.9 : 1.0\n",
      "contains(('сільськогосподарськ', 'продаж', 'земельн')) = True                6 : 1      =    620.9 : 1.0\n",
      "contains(('сільськогосподарськ', 'продаєт', 'земельн')) = True                6 : 1      =    620.9 : 1.0\n",
      "contains(('сільськогосподарськ', 'для', 'веденн')) = True                6 : 1      =    620.9 : 1.0\n",
      "contains(('селянськ', 'господарств', 'продаж')) = True                7 : 1      =    333.3 : 1.0\n",
      "contains(('селянськ', 'господарств', 'діл')) = True                7 : 1      =    333.3 : 1.0\n",
      "contains(('селянськ', 'господарств', 'кадастров')) = True                7 : 1      =    333.3 : 1.0\n",
      "contains(('селянськ', 'господарств', 'земельн')) = True                7 : 1      =    333.3 : 1.0\n",
      "contains(('селянськ', 'господарств', 'для')) = True                7 : 1      =    333.3 : 1.0\n",
      "contains(('селянськ', 'господарств', 'продаєт')) = True                7 : 1      =    333.3 : 1.0\n",
      "Класифікатор - <class 'nltk.classify.maxent.MaxentClassifier'>\n",
      "Порядок n - 1\n",
      "Класифікатор за колонкою - land_types\n",
      "  ==> Training (5 iterations)\n",
      "\n",
      "      Iteration    Log Likelihood    Accuracy\n",
      "      ---------------------------------------\n",
      "             1          -1.94591        0.019\n",
      "             2          -1.07311        0.697\n",
      "             3          -0.83080        0.697\n",
      "             4          -0.71131        0.697\n",
      "         Final          -0.64350        0.697\n",
      "Точність класифікатора на навчальних даних: 0.6969484391441599\n",
      "Точність класифікатора на тестових даних: 0.6832740213523132\n",
      "  |   1   2   3   4   5   6   7 |\n",
      "--+-----------------------------+\n",
      "1 |<869>113 216  70  13   2  26 |\n",
      "2 |   .  <4>  .   1   .   .   . |\n",
      "3 |   3   . <74>  .   .   .   . |\n",
      "4 |   1   .   . <13>  .   .   . |\n",
      "5 |   .   .   .   .  <.>  .   . |\n",
      "6 |   .   .   .   .   .  <.>  . |\n",
      "7 |   .   .   .   .   .   .  <.>|\n",
      "--+-----------------------------+\n",
      "(row = reference; col = test)\n",
      "\n",
      "Класифікатор - <class 'nltk.classify.maxent.MaxentClassifier'>\n",
      "Порядок n - 3\n",
      "Класифікатор за колонкою - land_types\n",
      "  ==> Training (5 iterations)\n",
      "\n",
      "      Iteration    Log Likelihood    Accuracy\n",
      "      ---------------------------------------\n",
      "             1          -1.94591        0.019\n",
      "             2          -1.50637        0.634\n",
      "             3          -1.44725        0.634\n",
      "             4          -1.41986        0.634\n",
      "         Final          -1.40445        0.634\n",
      "Точність класифікатора на навчальних даних: 0.6338126972991933\n",
      "Точність класифікатора на тестових даних: 0.6334519572953736\n",
      "  |   1   2   3   4   5   6   7 |\n",
      "--+-----------------------------+\n",
      "1 |<871>103 280  77  13   .  26 |\n",
      "2 |   1  <2>  .   .   .   .   . |\n",
      "3 |   1  12 <10>  .   .   1   . |\n",
      "4 |   .   .   .  <7>  .   .   . |\n",
      "5 |   .   .   .   .  <.>  .   . |\n",
      "6 |   .   .   .   .   .  <.>  . |\n",
      "7 |   .   .   .   .   .   1  <.>|\n",
      "--+-----------------------------+\n",
      "(row = reference; col = test)\n",
      "\n",
      "Класифікатор - <class 'nltk.classify.decisiontree.DecisionTreeClassifier'>\n",
      "Порядок n - 1\n",
      "Класифікатор за колонкою - land_types\n",
      "Точність класифікатора на навчальних даних: 0.9786039985969835\n",
      "Точність класифікатора на тестових даних: 0.9138790035587189\n",
      "  |   1   2   3   4   5   6   7 |\n",
      "--+-----------------------------+\n",
      "1 |<827>  4  17  22   2   .   2 |\n",
      "2 |   7<106>  2   3   .   .  10 |\n",
      "3 |  14   2<271>  .   .   .   3 |\n",
      "4 |  18   2   . <58>  1   .   1 |\n",
      "5 |   6   .   .   1 <10>  .   . |\n",
      "6 |   .   .   .   .   .  <2>  . |\n",
      "7 |   1   3   .   .   .   . <10>|\n",
      "--+-----------------------------+\n",
      "(row = reference; col = test)\n",
      "\n",
      "Класифікатор - <class 'nltk.classify.decisiontree.DecisionTreeClassifier'>\n",
      "Порядок n - 3\n",
      "Класифікатор за колонкою - land_types\n",
      "Точність класифікатора на навчальних даних: 0.839003858295335\n",
      "Точність класифікатора на тестових даних: 0.8021352313167259\n",
      "  |   1   2   3   4   5   6   7 |\n",
      "--+-----------------------------+\n",
      "1 |<843> 32 129  45   8   .  10 |\n",
      "2 |   6 <73>  2   2   .   2   2 |\n",
      "3 |  16   4<159>  .   .   .   1 |\n",
      "4 |   2   5   . <34>  .   .   . |\n",
      "5 |   2   .   .   1  <5>  .   . |\n",
      "6 |   .   .   .   .   .  <.>  . |\n",
      "7 |   4   3   .   2   .   . <13>|\n",
      "--+-----------------------------+\n",
      "(row = reference; col = test)\n",
      "\n",
      "Класифікатор - <class 'nltk.classify.naivebayes.NaiveBayesClassifier'>\n",
      "Порядок n - 1\n",
      "Класифікатор за колонкою - built_up\n",
      "Точність класифікатора на навчальних даних: 0.9214310768151526\n",
      "Точність класифікатора на тестових даних: 0.892526690391459\n",
      "  |    0    1 |\n",
      "--+-----------+\n",
      "0 |<1096>  45 |\n",
      "1 |  106 <158>|\n",
      "--+-----------+\n",
      "(row = reference; col = test)\n",
      "\n",
      "Найважливіші токени для класифікації за колонкою - built_up\n",
      "Most Informative Features\n",
      "     contains(('цегл',)) = True                1 : 0      =    119.8 : 1.0\n",
      "    contains(('стоїт',)) = True                1 : 0      =     60.1 : 1.0\n",
      "     contains(('кухн',)) = True                1 : 0      =     60.1 : 1.0\n",
      "   contains(('цеглян',)) = True                1 : 0      =     58.3 : 1.0\n",
      "     contains(('літн',)) = True                1 : 0      =     53.0 : 1.0\n",
      "contains(('одноповерхов',)) = True                1 : 0      =     49.1 : 1.0\n",
      "  contains(('опаленн',)) = True                1 : 0      =     45.2 : 1.0\n",
      "   contains(('підвал',)) = True                1 : 0      =     43.6 : 1.0\n",
      "  contains(('деревян',)) = True                1 : 0      =     38.6 : 1.0\n",
      "    contains(('горищ',)) = True                1 : 0      =     37.3 : 1.0\n",
      "Класифікатор - <class 'nltk.classify.naivebayes.NaiveBayesClassifier'>\n",
      "Порядок n - 3\n",
      "Класифікатор за колонкою - built_up\n",
      "Точність класифікатора на навчальних даних: 0.7572781480182392\n",
      "Точність класифікатора на тестових даних: 0.7544483985765125\n",
      "  |   0   1 |\n",
      "--+---------+\n",
      "0 |<903> 46 |\n",
      "1 | 299<157>|\n",
      "--+---------+\n",
      "(row = reference; col = test)\n",
      "\n",
      "Найважливіші токени для класифікації за колонкою - built_up\n",
      "Most Informative Features\n",
      "contains(('будинок', 'під', 'знос')) = True                1 : 0      =     18.4 : 1.0\n",
      "contains(('будинок', 'кв', 'м')) = True                1 : 0      =     18.4 : 1.0\n",
      "contains(('стар', 'будинок', 'під')) = True                1 : 0      =     13.4 : 1.0\n",
      "contains(('земел', 'природн', 'заповідн')) = True                1 : 0      =     10.2 : 1.0\n",
      "contains(('мож', 'бут', 'під')) = True                1 : 0      =      9.0 : 1.0\n",
      "contains(('фрукт', 'дерев', 'кущ')) = True                1 : 0      =      8.3 : 1.0\n",
      "contains(('природн', 'заповідн', 'фонд')) = True                1 : 0      =      5.9 : 1.0\n",
      "contains(('ділянц', 'є', 'стар')) = True                1 : 0      =      5.0 : 1.0\n",
      "contains(('ділянц', 'є', 'світл')) = True                1 : 0      =      5.0 : 1.0\n",
      "contains(('ділянц', 'є', 'будинок')) = True                1 : 0      =      5.0 : 1.0\n",
      "Класифікатор - <class 'nltk.classify.maxent.MaxentClassifier'>\n",
      "Порядок n - 1\n",
      "Класифікатор за колонкою - built_up\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  ==> Training (5 iterations)\n",
      "\n",
      "      Iteration    Log Likelihood    Accuracy\n",
      "      ---------------------------------------\n",
      "             1          -0.69315        0.145\n",
      "             2          -0.57463        0.884\n",
      "             3          -0.50751        0.884\n",
      "             4          -0.46603        0.884\n",
      "         Final          -0.43991        0.884\n",
      "Точність класифікатора на навчальних даних: 0.8842511399508944\n",
      "Точність класифікатора на тестових даних: 0.8725978647686833\n",
      "  |    0    1 |\n",
      "--+-----------+\n",
      "0 |<1196> 173 |\n",
      "1 |    6  <30>|\n",
      "--+-----------+\n",
      "(row = reference; col = test)\n",
      "\n",
      "Класифікатор - <class 'nltk.classify.maxent.MaxentClassifier'>\n",
      "Порядок n - 3\n",
      "Класифікатор за колонкою - built_up\n",
      "  ==> Training (5 iterations)\n",
      "\n",
      "      Iteration    Log Likelihood    Accuracy\n",
      "      ---------------------------------------\n",
      "             1          -0.69315        0.145\n",
      "             2          -0.65883        0.857\n",
      "             3          -0.64805        0.857\n",
      "             4          -0.64184        0.857\n",
      "         Final          -0.63789        0.857\n",
      "Точність класифікатора на навчальних даних: 0.8572430726061031\n",
      "Точність класифікатора на тестових даних: 0.8562277580071175\n",
      "  |    0    1 |\n",
      "--+-----------+\n",
      "0 |<1202> 202 |\n",
      "1 |    .   <1>|\n",
      "--+-----------+\n",
      "(row = reference; col = test)\n",
      "\n",
      "Класифікатор - <class 'nltk.classify.decisiontree.DecisionTreeClassifier'>\n",
      "Порядок n - 1\n",
      "Класифікатор за колонкою - built_up\n",
      "Точність класифікатора на навчальних даних: 0.9705366538056822\n",
      "Точність класифікатора на тестових даних: 0.9373665480427046\n",
      "  |    0    1 |\n",
      "--+-----------+\n",
      "0 |<1155>  41 |\n",
      "1 |   47 <162>|\n",
      "--+-----------+\n",
      "(row = reference; col = test)\n",
      "\n",
      "Класифікатор - <class 'nltk.classify.decisiontree.DecisionTreeClassifier'>\n",
      "Порядок n - 3\n",
      "Класифікатор за колонкою - built_up\n",
      "Точність класифікатора на навчальних даних: 0.9182742897229043\n",
      "Точність класифікатора на тестових даних: 0.9081850533807829\n",
      "  |    0    1 |\n",
      "--+-----------+\n",
      "0 |<1178> 105 |\n",
      "1 |   24  <98>|\n",
      "--+-----------+\n",
      "(row = reference; col = test)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "\n",
    "def bag_of_words(document_tokens,word_features):\n",
    "        \"\"\" Return the dict of bag_of_words. \n",
    "\n",
    "        Keyword arguments:\n",
    "        document_tokens -- list of tokens\n",
    "        word_features -- list of features\n",
    "\n",
    "        \"\"\"\n",
    "        \n",
    "        features={}\n",
    "        for word in word_features:\n",
    "            \n",
    "            features['contains({})'.format(word)]=(word[0] in document_tokens)\n",
    "        \n",
    "        return features\n",
    "\n",
    "\n",
    "def nltk_classifiers(dataframe,X_column,y_column,classifier=nltk.NaiveBayesClassifier,n=1,stop_words=[],ua_stemmer=True,most_common=1000): \n",
    "    \n",
    "    \n",
    "    \n",
    "    words=dataframe[X_column].str.cat(sep=' ')\n",
    "    words=nltk.ngrams(ua_tokenizer(words,ua_stemmer=ua_stemmer,stop_words=stop_words),n=n)\n",
    "    words=nltk.FreqDist(words)\n",
    "    word_features=words.most_common(most_common)\n",
    "    word_features=[words[0] for words in word_features]\n",
    "    \n",
    "    \n",
    "    labeled_featuresets=[]\n",
    "    for _,row in dataframe.iterrows():\n",
    "        \n",
    "        row[X_column]=nltk.ngrams(ua_tokenizer(row[X_column],ua_stemmer=ua_stemmer,stop_words=stop_words),n=n)\n",
    "        row[X_column]=[words[0] for words in nltk.FreqDist(row[X_column])]        \n",
    "        labeled_featuresets.append((bag_of_words(row[X_column],word_features=word_features), row[y_column]))  \n",
    "        \n",
    "    \n",
    "    train_set,test_set,_,_=train_test_split(labeled_featuresets,dataframe[y_column],stratify=dataframe[y_column],test_size=0.33)\n",
    "    \n",
    "    \n",
    "    \n",
    "    if classifier==nltk.MaxentClassifier:\n",
    "        classifier=classifier.train(train_set, max_iter=5)\n",
    "    else:\n",
    "        classifier=classifier.train(train_set)         \n",
    "    accuracy_train=nltk.classify.accuracy(classifier, train_set)\n",
    "    accuracy=nltk.classify.accuracy(classifier, test_set)\n",
    "    print('Точність класифікатора на навчальних даних:',accuracy_train)\n",
    "    print('Точність класифікатора на тестових даних:',accuracy)\n",
    "    y_true=[]\n",
    "    y_pred=[]\n",
    "    for test in test_set:\n",
    "        y_true.append(test[1])\n",
    "        y_pred.append(classifier.classify(test[0]))\n",
    "    confmat=nltk.ConfusionMatrix(y_pred,y_true)\n",
    "    print(confmat)\n",
    "    return classifier   \n",
    "\n",
    "\n",
    "classifiers=[nltk.NaiveBayesClassifier,nltk.MaxentClassifier,nltk.DecisionTreeClassifier]\n",
    "for y_column in ('land_types','built_up'):\n",
    "    for classifier in classifiers:    \n",
    "        for n in (1,3):      \n",
    "            print ('Класифікатор -',classifier)\n",
    "            print ('Порядок n -',n)           \n",
    "            print ('Класифікатор за колонкою -',y_column) \n",
    "            model=nltk_classifiers(land_data,X_column='text',y_column=y_column,classifier=classifier, n=n)\n",
    "            if classifier==nltk.NaiveBayesClassifier:\n",
    "                print ('Найважливіші токени для класифікації за колонкою -',y_column)\n",
    "                model.show_most_informative_features(10) \n",
    "                    \n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}