1_Pogled_na_podatke.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Sveučilište u Zagrebu<br>\n",
    "Fakultet elektrotehnike i računarstva\n",
    "\n",
    "## Uvod u znanost o podacima\n",
    "\n",
    "\n",
    "# Priprema i vizualizacija podataka"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.ensemble import RandomForestClassifier, ExtraTreesClassifier\n",
    "from sklearn.metrics import accuracy_score, confusion_matrix\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import OneHotEncoder, LabelEncoder\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.feature_selection import SelectKBest, mutual_info_classif\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Prikaz podataka o datotekama i člancima"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import re\n",
    "\n",
    "files = os.listdir('reuters21578')\n",
    "\n",
    "columns = ['FILENAME', 'NEWID', 'OLDID', 'LEWISSPLIT', 'CGISPLIT', 'TOPICS_ENUM', 'TEXT TYPE', 'DATE', 'TITLE', 'AUTHOR', 'DATELINE', 'TOPICS', 'PLACES', 'PEOPLE', 'ORGS', 'EXCHANGES', 'COMPANIES',  'BODY']\n",
    "rows = []\n",
    "\n",
    "for filename in os.listdir('reuters21578'):\n",
    "    if filename.endswith(\".sgm\"):\n",
    "        file_content = open(os.path.join(\"reuters21578\", filename), mode = 'r', encoding='ascii', errors='ignore').read()\n",
    "        articles = file_content.split(\"</REUTERS>\")\n",
    "        articles[0] = articles[0][articles[0].find(\"\\n\")::] ## izbacimo prvi red, onaj <!DOCSTYLE ...\n",
    "        #print(articles[0])\n",
    "        \n",
    "        for article in articles:\n",
    "            if \"<REUTERS\" in article:\n",
    "                red = {}\n",
    "                red['FILENAME'] = filename\n",
    "                red['NEWID'] = re.findall('NEWID=\"([^\"]*)\"', article)[0]\n",
    "                red['OLDID'] = re.findall('OLDID=\"([^\"]*)\"', article)[0]\n",
    "                red['TOPICS_ENUM'] = re.findall('TOPICS=\"([^\"]*)\"', article)[0]\n",
    "                red['TEXT TYPE'] = \"\".join(re.findall('TEXT TYPE=\"([^\"]*)\"', article))\n",
    "                red['LEWISSPLIT'] = re.findall('LEWISSPLIT=\"([^\"]*)\"', article)[0]\n",
    "                red['CGISPLIT'] = re.findall('CGISPLIT=\"([^\"]*)\"', article)[0]\n",
    "                for znacajka in columns[7:]:\n",
    "                    if znacajka == \"DATE\":\n",
    "                        regex = '<' + znacajka + '>' + '([^&#$]*)' + '</' + znacajka + '>'\n",
    "                    else:\n",
    "                        regex = '<' + znacajka + '>' + '([\\s\\S]*)' + '</' + znacajka + '>'\n",
    "                    match = re.findall(regex, article)\n",
    "                    if(len(match)==0):\n",
    "                        continue\n",
    "                    elif match[0] == '':\n",
    "                        continue\n",
    "                    else:\n",
    "                        match = match[0]\n",
    "                        if '<D>' in match:\n",
    "                            match = match.replace('<D>', '')\n",
    "                            match = match.split('</D>')\n",
    "                            match = match[:-1]\n",
    "                        red[znacajka] = match\n",
    "                #print(red[0:-1])\n",
    "                rows.append(red.copy())\n",
    "\n",
    "dataframe = pd.DataFrame(rows, columns=['NEWID', 'TOPICS', 'PLACES', 'PEOPLE', 'ORGS', 'EXCHANGES', 'TITLE', 'AUTHOR', 'DATELINE', 'OLDID', 'TOPICS_ENUM', 'FILENAME', 'DATE', 'COMPANIES', 'BODY', 'TEXT TYPE', 'LEWISSPLIT', 'CGISPLIT'])\n",
    "dataframe = dataframe.astype('string')\n",
    "dataframe = dataframe.astype({'NEWID' : 'int64', 'OLDID' : 'int64'})\n",
    "#dataframe = dataframe.set_index(\"NEWID\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Prikaz prvih 10 redova"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": true
   },
   "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>NEWID</th>\n",
       "      <th>TOPICS</th>\n",
       "      <th>PLACES</th>\n",
       "      <th>PEOPLE</th>\n",
       "      <th>ORGS</th>\n",
       "      <th>EXCHANGES</th>\n",
       "      <th>TITLE</th>\n",
       "      <th>AUTHOR</th>\n",
       "      <th>DATELINE</th>\n",
       "      <th>OLDID</th>\n",
       "      <th>TOPICS_ENUM</th>\n",
       "      <th>FILENAME</th>\n",
       "      <th>DATE</th>\n",
       "      <th>COMPANIES</th>\n",
       "      <th>BODY</th>\n",
       "      <th>TEXT TYPE</th>\n",
       "      <th>LEWISSPLIT</th>\n",
       "      <th>CGISPLIT</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>['cocoa']</td>\n",
       "      <td>['el-salvador', 'usa', 'uruguay']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>BAHIA COCOA REVIEW</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>SALVADOR, Feb 26 -</td>\n",
       "      <td>5544</td>\n",
       "      <td>YES</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:01:01.79</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>Showers continued throughout the week in\n",
       "the B...</td>\n",
       "      <td></td>\n",
       "      <td>TRAIN</td>\n",
       "      <td>TRAINING-SET</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>STANDARD OIL &amp;lt;SRD&gt; TO FORM FINANCIAL UNIT</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>CLEVELAND, Feb 26 -</td>\n",
       "      <td>5545</td>\n",
       "      <td>NO</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:02:20.00</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>Standard Oil Co and BP North America\n",
       "Inc said ...</td>\n",
       "      <td></td>\n",
       "      <td>TRAIN</td>\n",
       "      <td>TRAINING-SET</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>TEXAS COMMERCE BANCSHARES &amp;lt;TCB&gt; FILES PLAN</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>HOUSTON, Feb 26 -</td>\n",
       "      <td>5546</td>\n",
       "      <td>NO</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:03:27.51</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>Texas Commerce Bancshares Inc's Texas\n",
       "Commerce...</td>\n",
       "      <td></td>\n",
       "      <td>TRAIN</td>\n",
       "      <td>TRAINING-SET</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>['usa', 'brazil']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>TALKING POINT/BANKAMERICA &amp;lt;BAC&gt; EQUITY OFFER</td>\n",
       "      <td>by Janie Gabbett, Reuters</td>\n",
       "      <td>LOS ANGELES, Feb 26 -</td>\n",
       "      <td>5547</td>\n",
       "      <td>NO</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:07:13.72</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>BankAmerica Corp is not under\n",
       "pressure to act ...</td>\n",
       "      <td></td>\n",
       "      <td>TRAIN</td>\n",
       "      <td>TRAINING-SET</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>['grain', 'wheat', 'corn', 'barley', 'oat', 's...</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>NATIONAL AVERAGE PRICES FOR FARMER-OWNED RESERVE</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>WASHINGTON, Feb 26 -</td>\n",
       "      <td>5548</td>\n",
       "      <td>YES</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:10:44.60</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>The U.S. Agriculture Department\n",
       "reported the f...</td>\n",
       "      <td></td>\n",
       "      <td>TRAIN</td>\n",
       "      <td>TRAINING-SET</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6</td>\n",
       "      <td>['veg-oil', 'linseed', 'lin-oil', 'soy-oil', '...</td>\n",
       "      <td>['argentina']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>ARGENTINE 1986/87 GRAIN/OILSEED REGISTRATIONS</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>BUENOS AIRES, Feb 26 -</td>\n",
       "      <td>5549</td>\n",
       "      <td>YES</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:14:36.41</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>Argentine grain board figures show\n",
       "crop regist...</td>\n",
       "      <td></td>\n",
       "      <td>TRAIN</td>\n",
       "      <td>TRAINING-SET</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>RED LION INNS FILES PLANS OFFERING</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>PORTLAND, Ore., Feb 26 -</td>\n",
       "      <td>5550</td>\n",
       "      <td>NO</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:14:42.83</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>Red Lion Inns Limited Partnership\n",
       "said it file...</td>\n",
       "      <td></td>\n",
       "      <td>TRAIN</td>\n",
       "      <td>TRAINING-SET</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>USX &amp;lt;X&gt; DEBT DOWGRADED BY MOODY'S</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>NEW YORK, Feb 26 -</td>\n",
       "      <td>5551</td>\n",
       "      <td>YES</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:15:40.12</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>Moody's Investors Service Inc said it\n",
       "lowered ...</td>\n",
       "      <td></td>\n",
       "      <td>TRAIN</td>\n",
       "      <td>TRAINING-SET</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9</td>\n",
       "      <td>['earn']</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>CHAMPION PRODUCTS &amp;lt;CH&gt; APPROVES STOCK SPLIT</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>ROCHESTER, N.Y., Feb 26 -</td>\n",
       "      <td>5552</td>\n",
       "      <td>YES</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:17:11.20</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>Champion Products Inc said its\n",
       "board of direct...</td>\n",
       "      <td></td>\n",
       "      <td>TRAIN</td>\n",
       "      <td>TRAINING-SET</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10</td>\n",
       "      <td>['acq']</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>COMPUTER TERMINAL SYSTEMS &amp;lt;CPML&gt; COMPLETES ...</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>COMMACK, N.Y., Feb 26 -</td>\n",
       "      <td>5553</td>\n",
       "      <td>YES</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:18:06.67</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>Computer Terminal Systems Inc said\n",
       "it has comp...</td>\n",
       "      <td></td>\n",
       "      <td>TRAIN</td>\n",
       "      <td>TRAINING-SET</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   NEWID                                             TOPICS  \\\n",
       "0      1                                          ['cocoa']   \n",
       "1      2                                               <NA>   \n",
       "2      3                                               <NA>   \n",
       "3      4                                               <NA>   \n",
       "4      5  ['grain', 'wheat', 'corn', 'barley', 'oat', 's...   \n",
       "5      6  ['veg-oil', 'linseed', 'lin-oil', 'soy-oil', '...   \n",
       "6      7                                               <NA>   \n",
       "7      8                                               <NA>   \n",
       "8      9                                           ['earn']   \n",
       "9     10                                            ['acq']   \n",
       "\n",
       "                              PLACES PEOPLE  ORGS EXCHANGES  \\\n",
       "0  ['el-salvador', 'usa', 'uruguay']   <NA>  <NA>      <NA>   \n",
       "1                            ['usa']   <NA>  <NA>      <NA>   \n",
       "2                            ['usa']   <NA>  <NA>      <NA>   \n",
       "3                  ['usa', 'brazil']   <NA>  <NA>      <NA>   \n",
       "4                            ['usa']   <NA>  <NA>      <NA>   \n",
       "5                      ['argentina']   <NA>  <NA>      <NA>   \n",
       "6                            ['usa']   <NA>  <NA>      <NA>   \n",
       "7                            ['usa']   <NA>  <NA>      <NA>   \n",
       "8                            ['usa']   <NA>  <NA>      <NA>   \n",
       "9                            ['usa']   <NA>  <NA>      <NA>   \n",
       "\n",
       "                                               TITLE  \\\n",
       "0                                 BAHIA COCOA REVIEW   \n",
       "1       STANDARD OIL &lt;SRD> TO FORM FINANCIAL UNIT   \n",
       "2      TEXAS COMMERCE BANCSHARES &lt;TCB> FILES PLAN   \n",
       "3    TALKING POINT/BANKAMERICA &lt;BAC> EQUITY OFFER   \n",
       "4   NATIONAL AVERAGE PRICES FOR FARMER-OWNED RESERVE   \n",
       "5      ARGENTINE 1986/87 GRAIN/OILSEED REGISTRATIONS   \n",
       "6                 RED LION INNS FILES PLANS OFFERING   \n",
       "7               USX &lt;X> DEBT DOWGRADED BY MOODY'S   \n",
       "8     CHAMPION PRODUCTS &lt;CH> APPROVES STOCK SPLIT   \n",
       "9  COMPUTER TERMINAL SYSTEMS &lt;CPML> COMPLETES ...   \n",
       "\n",
       "                          AUTHOR                        DATELINE  OLDID  \\\n",
       "0                           <NA>             SALVADOR, Feb 26 -    5544   \n",
       "1                           <NA>            CLEVELAND, Feb 26 -    5545   \n",
       "2                           <NA>              HOUSTON, Feb 26 -    5546   \n",
       "3      by Janie Gabbett, Reuters          LOS ANGELES, Feb 26 -    5547   \n",
       "4                           <NA>           WASHINGTON, Feb 26 -    5548   \n",
       "5                           <NA>         BUENOS AIRES, Feb 26 -    5549   \n",
       "6                           <NA>       PORTLAND, Ore., Feb 26 -    5550   \n",
       "7                           <NA>             NEW YORK, Feb 26 -    5551   \n",
       "8                           <NA>      ROCHESTER, N.Y., Feb 26 -    5552   \n",
       "9                           <NA>        COMMACK, N.Y., Feb 26 -    5553   \n",
       "\n",
       "  TOPICS_ENUM       FILENAME                     DATE COMPANIES  \\\n",
       "0         YES  reut2-000.sgm  26-FEB-1987 15:01:01.79      <NA>   \n",
       "1          NO  reut2-000.sgm  26-FEB-1987 15:02:20.00      <NA>   \n",
       "2          NO  reut2-000.sgm  26-FEB-1987 15:03:27.51      <NA>   \n",
       "3          NO  reut2-000.sgm  26-FEB-1987 15:07:13.72      <NA>   \n",
       "4         YES  reut2-000.sgm  26-FEB-1987 15:10:44.60      <NA>   \n",
       "5         YES  reut2-000.sgm  26-FEB-1987 15:14:36.41      <NA>   \n",
       "6          NO  reut2-000.sgm  26-FEB-1987 15:14:42.83      <NA>   \n",
       "7         YES  reut2-000.sgm  26-FEB-1987 15:15:40.12      <NA>   \n",
       "8         YES  reut2-000.sgm  26-FEB-1987 15:17:11.20      <NA>   \n",
       "9         YES  reut2-000.sgm  26-FEB-1987 15:18:06.67      <NA>   \n",
       "\n",
       "                                                BODY TEXT TYPE LEWISSPLIT  \\\n",
       "0  Showers continued throughout the week in\n",
       "the B...                TRAIN   \n",
       "1  Standard Oil Co and BP North America\n",
       "Inc said ...                TRAIN   \n",
       "2  Texas Commerce Bancshares Inc's Texas\n",
       "Commerce...                TRAIN   \n",
       "3  BankAmerica Corp is not under\n",
       "pressure to act ...                TRAIN   \n",
       "4  The U.S. Agriculture Department\n",
       "reported the f...                TRAIN   \n",
       "5  Argentine grain board figures show\n",
       "crop regist...                TRAIN   \n",
       "6  Red Lion Inns Limited Partnership\n",
       "said it file...                TRAIN   \n",
       "7  Moody's Investors Service Inc said it\n",
       "lowered ...                TRAIN   \n",
       "8  Champion Products Inc said its\n",
       "board of direct...                TRAIN   \n",
       "9  Computer Terminal Systems Inc said\n",
       "it has comp...                TRAIN   \n",
       "\n",
       "       CGISPLIT  \n",
       "0  TRAINING-SET  \n",
       "1  TRAINING-SET  \n",
       "2  TRAINING-SET  \n",
       "3  TRAINING-SET  \n",
       "4  TRAINING-SET  \n",
       "5  TRAINING-SET  \n",
       "6  TRAINING-SET  \n",
       "7  TRAINING-SET  \n",
       "8  TRAINING-SET  \n",
       "9  TRAINING-SET  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataframe.head(n=10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Pohrana Dataframe-a"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "filename = \"my_data/data/clanci.csv\"\n",
    "dataframe.to_csv(filename)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>NEWID</th>\n",
       "      <th>TOPICS</th>\n",
       "      <th>PLACES</th>\n",
       "      <th>PEOPLE</th>\n",
       "      <th>ORGS</th>\n",
       "      <th>EXCHANGES</th>\n",
       "      <th>TITLE</th>\n",
       "      <th>AUTHOR</th>\n",
       "      <th>DATELINE</th>\n",
       "      <th>OLDID</th>\n",
       "      <th>TOPICS_ENUM</th>\n",
       "      <th>FILENAME</th>\n",
       "      <th>DATE</th>\n",
       "      <th>COMPANIES</th>\n",
       "      <th>BODY</th>\n",
       "      <th>TEXT TYPE</th>\n",
       "      <th>LEWISSPLIT</th>\n",
       "      <th>CGISPLIT</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>['cocoa']</td>\n",
       "      <td>['el-salvador', 'usa', 'uruguay']</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>BAHIA COCOA REVIEW</td>\n",
       "      <td>NaN</td>\n",
       "      <td>SALVADOR, Feb 26 -</td>\n",
       "      <td>5544</td>\n",
       "      <td>YES</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:01:01.79</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Showers continued throughout the week in\\nthe ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>TRAIN</td>\n",
       "      <td>TRAINING-SET</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>STANDARD OIL &amp;lt;SRD&gt; TO FORM FINANCIAL UNIT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>CLEVELAND, Feb 26 -</td>\n",
       "      <td>5545</td>\n",
       "      <td>NO</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:02:20.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Standard Oil Co and BP North America\\nInc said...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>TRAIN</td>\n",
       "      <td>TRAINING-SET</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>NaN</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>TEXAS COMMERCE BANCSHARES &amp;lt;TCB&gt; FILES PLAN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>HOUSTON, Feb 26 -</td>\n",
       "      <td>5546</td>\n",
       "      <td>NO</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:03:27.51</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Texas Commerce Bancshares Inc's Texas\\nCommerc...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>TRAIN</td>\n",
       "      <td>TRAINING-SET</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>['usa', 'brazil']</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>TALKING POINT/BANKAMERICA &amp;lt;BAC&gt; EQUITY OFFER</td>\n",
       "      <td>by Janie Gabbett, Reuters</td>\n",
       "      <td>LOS ANGELES, Feb 26 -</td>\n",
       "      <td>5547</td>\n",
       "      <td>NO</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:07:13.72</td>\n",
       "      <td>NaN</td>\n",
       "      <td>BankAmerica Corp is not under\\npressure to act...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>TRAIN</td>\n",
       "      <td>TRAINING-SET</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>['grain', 'wheat', 'corn', 'barley', 'oat', 's...</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NATIONAL AVERAGE PRICES FOR FARMER-OWNED RESERVE</td>\n",
       "      <td>NaN</td>\n",
       "      <td>WASHINGTON, Feb 26 -</td>\n",
       "      <td>5548</td>\n",
       "      <td>YES</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:10:44.60</td>\n",
       "      <td>NaN</td>\n",
       "      <td>The U.S. Agriculture Department\\nreported the ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>TRAIN</td>\n",
       "      <td>TRAINING-SET</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6</td>\n",
       "      <td>['veg-oil', 'linseed', 'lin-oil', 'soy-oil', '...</td>\n",
       "      <td>['argentina']</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ARGENTINE 1986/87 GRAIN/OILSEED REGISTRATIONS</td>\n",
       "      <td>NaN</td>\n",
       "      <td>BUENOS AIRES, Feb 26 -</td>\n",
       "      <td>5549</td>\n",
       "      <td>YES</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:14:36.41</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Argentine grain board figures show\\ncrop regis...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>TRAIN</td>\n",
       "      <td>TRAINING-SET</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7</td>\n",
       "      <td>NaN</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RED LION INNS FILES PLANS OFFERING</td>\n",
       "      <td>NaN</td>\n",
       "      <td>PORTLAND, Ore., Feb 26 -</td>\n",
       "      <td>5550</td>\n",
       "      <td>NO</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:14:42.83</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Red Lion Inns Limited Partnership\\nsaid it fil...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>TRAIN</td>\n",
       "      <td>TRAINING-SET</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8</td>\n",
       "      <td>NaN</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>USX &amp;lt;X&gt; DEBT DOWGRADED BY MOODY'S</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NEW YORK, Feb 26 -</td>\n",
       "      <td>5551</td>\n",
       "      <td>YES</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:15:40.12</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Moody's Investors Service Inc said it\\nlowered...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>TRAIN</td>\n",
       "      <td>TRAINING-SET</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9</td>\n",
       "      <td>['earn']</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>CHAMPION PRODUCTS &amp;lt;CH&gt; APPROVES STOCK SPLIT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ROCHESTER, N.Y., Feb 26 -</td>\n",
       "      <td>5552</td>\n",
       "      <td>YES</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:17:11.20</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Champion Products Inc said its\\nboard of direc...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>TRAIN</td>\n",
       "      <td>TRAINING-SET</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10</td>\n",
       "      <td>['acq']</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>COMPUTER TERMINAL SYSTEMS &amp;lt;CPML&gt; COMPLETES ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>COMMACK, N.Y., Feb 26 -</td>\n",
       "      <td>5553</td>\n",
       "      <td>YES</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:18:06.67</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Computer Terminal Systems Inc said\\nit has com...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>TRAIN</td>\n",
       "      <td>TRAINING-SET</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   NEWID                                             TOPICS  \\\n",
       "0      1                                          ['cocoa']   \n",
       "1      2                                                NaN   \n",
       "2      3                                                NaN   \n",
       "3      4                                                NaN   \n",
       "4      5  ['grain', 'wheat', 'corn', 'barley', 'oat', 's...   \n",
       "5      6  ['veg-oil', 'linseed', 'lin-oil', 'soy-oil', '...   \n",
       "6      7                                                NaN   \n",
       "7      8                                                NaN   \n",
       "8      9                                           ['earn']   \n",
       "9     10                                            ['acq']   \n",
       "\n",
       "                              PLACES PEOPLE ORGS EXCHANGES  \\\n",
       "0  ['el-salvador', 'usa', 'uruguay']    NaN  NaN       NaN   \n",
       "1                            ['usa']    NaN  NaN       NaN   \n",
       "2                            ['usa']    NaN  NaN       NaN   \n",
       "3                  ['usa', 'brazil']    NaN  NaN       NaN   \n",
       "4                            ['usa']    NaN  NaN       NaN   \n",
       "5                      ['argentina']    NaN  NaN       NaN   \n",
       "6                            ['usa']    NaN  NaN       NaN   \n",
       "7                            ['usa']    NaN  NaN       NaN   \n",
       "8                            ['usa']    NaN  NaN       NaN   \n",
       "9                            ['usa']    NaN  NaN       NaN   \n",
       "\n",
       "                                               TITLE  \\\n",
       "0                                 BAHIA COCOA REVIEW   \n",
       "1       STANDARD OIL &lt;SRD> TO FORM FINANCIAL UNIT   \n",
       "2      TEXAS COMMERCE BANCSHARES &lt;TCB> FILES PLAN   \n",
       "3    TALKING POINT/BANKAMERICA &lt;BAC> EQUITY OFFER   \n",
       "4   NATIONAL AVERAGE PRICES FOR FARMER-OWNED RESERVE   \n",
       "5      ARGENTINE 1986/87 GRAIN/OILSEED REGISTRATIONS   \n",
       "6                 RED LION INNS FILES PLANS OFFERING   \n",
       "7               USX &lt;X> DEBT DOWGRADED BY MOODY'S   \n",
       "8     CHAMPION PRODUCTS &lt;CH> APPROVES STOCK SPLIT   \n",
       "9  COMPUTER TERMINAL SYSTEMS &lt;CPML> COMPLETES ...   \n",
       "\n",
       "                          AUTHOR                        DATELINE  OLDID  \\\n",
       "0                            NaN             SALVADOR, Feb 26 -    5544   \n",
       "1                            NaN            CLEVELAND, Feb 26 -    5545   \n",
       "2                            NaN              HOUSTON, Feb 26 -    5546   \n",
       "3      by Janie Gabbett, Reuters          LOS ANGELES, Feb 26 -    5547   \n",
       "4                            NaN           WASHINGTON, Feb 26 -    5548   \n",
       "5                            NaN         BUENOS AIRES, Feb 26 -    5549   \n",
       "6                            NaN       PORTLAND, Ore., Feb 26 -    5550   \n",
       "7                            NaN             NEW YORK, Feb 26 -    5551   \n",
       "8                            NaN      ROCHESTER, N.Y., Feb 26 -    5552   \n",
       "9                            NaN        COMMACK, N.Y., Feb 26 -    5553   \n",
       "\n",
       "  TOPICS_ENUM       FILENAME                     DATE  COMPANIES  \\\n",
       "0         YES  reut2-000.sgm  26-FEB-1987 15:01:01.79        NaN   \n",
       "1          NO  reut2-000.sgm  26-FEB-1987 15:02:20.00        NaN   \n",
       "2          NO  reut2-000.sgm  26-FEB-1987 15:03:27.51        NaN   \n",
       "3          NO  reut2-000.sgm  26-FEB-1987 15:07:13.72        NaN   \n",
       "4         YES  reut2-000.sgm  26-FEB-1987 15:10:44.60        NaN   \n",
       "5         YES  reut2-000.sgm  26-FEB-1987 15:14:36.41        NaN   \n",
       "6          NO  reut2-000.sgm  26-FEB-1987 15:14:42.83        NaN   \n",
       "7         YES  reut2-000.sgm  26-FEB-1987 15:15:40.12        NaN   \n",
       "8         YES  reut2-000.sgm  26-FEB-1987 15:17:11.20        NaN   \n",
       "9         YES  reut2-000.sgm  26-FEB-1987 15:18:06.67        NaN   \n",
       "\n",
       "                                                BODY TEXT TYPE LEWISSPLIT  \\\n",
       "0  Showers continued throughout the week in\\nthe ...       NaN      TRAIN   \n",
       "1  Standard Oil Co and BP North America\\nInc said...       NaN      TRAIN   \n",
       "2  Texas Commerce Bancshares Inc's Texas\\nCommerc...       NaN      TRAIN   \n",
       "3  BankAmerica Corp is not under\\npressure to act...       NaN      TRAIN   \n",
       "4  The U.S. Agriculture Department\\nreported the ...       NaN      TRAIN   \n",
       "5  Argentine grain board figures show\\ncrop regis...       NaN      TRAIN   \n",
       "6  Red Lion Inns Limited Partnership\\nsaid it fil...       NaN      TRAIN   \n",
       "7  Moody's Investors Service Inc said it\\nlowered...       NaN      TRAIN   \n",
       "8  Champion Products Inc said its\\nboard of direc...       NaN      TRAIN   \n",
       "9  Computer Terminal Systems Inc said\\nit has com...       NaN      TRAIN   \n",
       "\n",
       "       CGISPLIT  \n",
       "0  TRAINING-SET  \n",
       "1  TRAINING-SET  \n",
       "2  TRAINING-SET  \n",
       "3  TRAINING-SET  \n",
       "4  TRAINING-SET  \n",
       "5  TRAINING-SET  \n",
       "6  TRAINING-SET  \n",
       "7  TRAINING-SET  \n",
       "8  TRAINING-SET  \n",
       "9  TRAINING-SET  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clanci = pd.read_csv(filename, index_col = 0)\n",
    "clanci.head(n=10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Pregled podataka"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "broj primjeraka (članaka), N = 21578\n",
      "broj značajki, n = 18\n"
     ]
    }
   ],
   "source": [
    "broj_primjeraka, broj_znacajki = clanci.shape\n",
    "print(\"broj primjeraka (članaka), N = {}\".format(broj_primjeraka))\n",
    "print(\"broj značajki, n = {}\".format(broj_znacajki))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Koliko članaka ima u svakoj datoteci:\n",
      "FILENAME\n",
      "reut2-000.sgm    1000\n",
      "reut2-001.sgm    1000\n",
      "reut2-002.sgm    1000\n",
      "reut2-003.sgm    1000\n",
      "reut2-004.sgm    1000\n",
      "reut2-005.sgm    1000\n",
      "reut2-006.sgm    1000\n",
      "reut2-007.sgm    1000\n",
      "reut2-008.sgm    1000\n",
      "reut2-009.sgm    1000\n",
      "reut2-010.sgm    1000\n",
      "reut2-011.sgm    1000\n",
      "reut2-012.sgm    1000\n",
      "reut2-013.sgm    1000\n",
      "reut2-014.sgm    1000\n",
      "reut2-015.sgm    1000\n",
      "reut2-016.sgm    1000\n",
      "reut2-017.sgm    1000\n",
      "reut2-018.sgm    1000\n",
      "reut2-019.sgm    1000\n",
      "reut2-020.sgm    1000\n",
      "reut2-021.sgm     578\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print(\"Koliko članaka ima u svakoj datoteci:\")\n",
    "print(clanci.groupby('FILENAME').size())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Značajke:\n",
    "- **newid** - id koji članak ima u Reuters-21578 klasifikaciji\n",
    "- **topics** - lista kategorija tema kojima članak pripada\n",
    "- **places** - lista kategorija mjesta kojima članak pripada\n",
    "- **people** - lista kategorija ljudi kojima članak pripada\n",
    "- **orgs** - lista kategorija orgs kojima članak pripada\n",
    "- **exchange** - lista kategorija razmjena kojima članak pripada\n",
    "- **title** - naslov članka\n",
    "- **author** - autor teksta\n",
    "- **dateline** - od kod je članak potekao, i dan u godini\n",
    "- **oldid** - id koji je članak imao u Reuters-22173 klasifikaciji\n",
    "- **topics enum** - moguće vrijednosti su [YES, NO, BYPASS], je li dokument imao topics kategoriju u originalnom datasetu\n",
    "- **filename** - ime datoteke u kojoj se nalazi članak\n",
    "- **date** - datum i vrijeme dokumenta\n",
    "- **companies** - ?\n",
    "- **body** - tijelo teksta\n",
    "- **text type** - moguće vrijednosti su [NORM, BRIEF, UNPROC], norm znači da tekst ima normalnu strukturu, brief znači da je tekst kratak, unproc znači da tekst ima neobičnu strukturu\n",
    "- **lewissplit** - moguće vrijednosti su [TRAINING, TEST, NOT_USED] ovisno o tome za što se članak koristio u eksperimentima u izvješćima: LEWIS91d (poglavlja 9 i 10), LEWIS92b, LEWIS92e i LEWIS94b\n",
    "- **cgisplit** - moguće vrijednosti su [TRAINING-SET, PUBLISHED-TESTSET] ovisno o tome za što se članak koristi u eksperimentima  u izvješćima HAYES89 i HAYES90b"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 21578 entries, 0 to 21577\n",
      "Data columns (total 18 columns):\n",
      " #   Column       Non-Null Count  Dtype \n",
      "---  ------       --------------  ----- \n",
      " 0   NEWID        21578 non-null  int64 \n",
      " 1   TOPICS       11367 non-null  string\n",
      " 2   PLACES       18798 non-null  string\n",
      " 3   PEOPLE       1156 non-null   string\n",
      " 4   ORGS         881 non-null    string\n",
      " 5   EXCHANGES    482 non-null    string\n",
      " 6   TITLE        20841 non-null  string\n",
      " 7   AUTHOR       875 non-null    string\n",
      " 8   DATELINE     19043 non-null  string\n",
      " 9   OLDID        21578 non-null  int64 \n",
      " 10  TOPICS_ENUM  21578 non-null  string\n",
      " 11  FILENAME     21578 non-null  string\n",
      " 12  DATE         21250 non-null  string\n",
      " 13  COMPANIES    0 non-null      string\n",
      " 14  BODY         19043 non-null  string\n",
      " 15  TEXT TYPE    21578 non-null  string\n",
      " 16  LEWISSPLIT   21578 non-null  string\n",
      " 17  CGISPLIT     21578 non-null  string\n",
      "dtypes: int64(2), string(16)\n",
      "memory usage: 3.0 MB\n"
     ]
    }
   ],
   "source": [
    "dataframe.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Monotoni atributi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "NEWID          21578\n",
       "TOPICS           655\n",
       "PLACES          1096\n",
       "PEOPLE           217\n",
       "ORGS              67\n",
       "EXCHANGES         57\n",
       "TITLE          20029\n",
       "AUTHOR           351\n",
       "DATELINE        8822\n",
       "OLDID          21578\n",
       "TOPICS_ENUM        3\n",
       "FILENAME          22\n",
       "DATE           21250\n",
       "COMPANIES          0\n",
       "BODY           18781\n",
       "TEXT TYPE          3\n",
       "LEWISSPLIT         3\n",
       "CGISPLIT           2\n",
       "dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataframe.nunique() # gledamo ako ima monotonih značajki"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Zaključujemo da su NEWID i OLDID monotone značajke. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "original = dataframe.copy() #čuvamo original podataka, just in case\n",
    "podaci = original.copy()\n",
    "\n",
    "podaci = podaci.drop('OLDID', axis=1, inplace=False)\n",
    "podaci.drop('NEWID', axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "TOPICS         10211\n",
       "PLACES          2780\n",
       "PEOPLE         20422\n",
       "ORGS           20697\n",
       "EXCHANGES      21096\n",
       "TITLE            737\n",
       "AUTHOR         20703\n",
       "DATELINE        2535\n",
       "TOPICS_ENUM        0\n",
       "FILENAME           0\n",
       "DATE             328\n",
       "COMPANIES      21578\n",
       "BODY            2535\n",
       "TEXT TYPE          0\n",
       "LEWISSPLIT         0\n",
       "CGISPLIT           0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "podaci.isna().sum() # broji nul vrijednosti za značajke"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Vidimo da osim značajki za klasifikacijske oznake (topics, places, people, orgs, exchanges), značajke AUTHOR i COMPANIES imaju puno nul vrijednosti. Izbacujemo te značajke iz skupa podataka."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "podaci.drop(['AUTHOR', 'COMPANIES'], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Izbacujemo zapise koji nemaju body jer nam oni zapravo ništa ne govore."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "podaci = podaci.loc[podaci.BODY.notnull(), :] # zadržavamo samo one zapise kojima body nije null"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Label Encoder\n",
    "Mijenjamo format značajki koje imaju malo unique vrijednosti (poprimaju vrijednosti iz nekog diskretnog skupa) - topics_enum, text type, lewisspit, cgisplit."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>TOPICS</th>\n",
       "      <th>PLACES</th>\n",
       "      <th>PEOPLE</th>\n",
       "      <th>ORGS</th>\n",
       "      <th>EXCHANGES</th>\n",
       "      <th>TITLE</th>\n",
       "      <th>DATELINE</th>\n",
       "      <th>TOPICS_ENUM</th>\n",
       "      <th>FILENAME</th>\n",
       "      <th>DATE</th>\n",
       "      <th>BODY</th>\n",
       "      <th>TEXT TYPE</th>\n",
       "      <th>LEWISSPLIT</th>\n",
       "      <th>CGISPLIT</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>['cocoa']</td>\n",
       "      <td>['el-salvador', 'usa', 'uruguay']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>BAHIA COCOA REVIEW</td>\n",
       "      <td>SALVADOR, Feb 26 -</td>\n",
       "      <td>2</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:01:01.79</td>\n",
       "      <td>Showers continued throughout the week in\n",
       "the B...</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>STANDARD OIL &amp;lt;SRD&gt; TO FORM FINANCIAL UNIT</td>\n",
       "      <td>CLEVELAND, Feb 26 -</td>\n",
       "      <td>1</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:02:20.00</td>\n",
       "      <td>Standard Oil Co and BP North America\n",
       "Inc said ...</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>TEXAS COMMERCE BANCSHARES &amp;lt;TCB&gt; FILES PLAN</td>\n",
       "      <td>HOUSTON, Feb 26 -</td>\n",
       "      <td>1</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:03:27.51</td>\n",
       "      <td>Texas Commerce Bancshares Inc's Texas\n",
       "Commerce...</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>['usa', 'brazil']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>TALKING POINT/BANKAMERICA &amp;lt;BAC&gt; EQUITY OFFER</td>\n",
       "      <td>LOS ANGELES, Feb 26 -</td>\n",
       "      <td>1</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:07:13.72</td>\n",
       "      <td>BankAmerica Corp is not under\n",
       "pressure to act ...</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>['grain', 'wheat', 'corn', 'barley', 'oat', 's...</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>NATIONAL AVERAGE PRICES FOR FARMER-OWNED RESERVE</td>\n",
       "      <td>WASHINGTON, Feb 26 -</td>\n",
       "      <td>2</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:10:44.60</td>\n",
       "      <td>The U.S. Agriculture Department\n",
       "reported the f...</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21573</th>\n",
       "      <td>['ship']</td>\n",
       "      <td>['hong-kong', 'japan', 'india', 'pakistan', 'i...</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>JAPAN/INDIA CONFERENCE CUTS GULF WAR RISK CHARGES</td>\n",
       "      <td>HONG KONG, Oct 19 -</td>\n",
       "      <td>2</td>\n",
       "      <td>reut2-021.sgm</td>\n",
       "      <td>19-OCT-1987 00:34:08.94</td>\n",
       "      <td>The Japan/India-Pakistan-Gulf/Japan\n",
       "shipping c...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21574</th>\n",
       "      <td>['ipi']</td>\n",
       "      <td>['ussr']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>SOVIET INDUSTRIAL GROWTH/TRADE SLOWER IN 1987</td>\n",
       "      <td>MOSCOW, Oct 19 -</td>\n",
       "      <td>2</td>\n",
       "      <td>reut2-021.sgm</td>\n",
       "      <td>19-OCT-1987 00:18:22.79</td>\n",
       "      <td>The Soviet Union's industrial output is\n",
       "growin...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21575</th>\n",
       "      <td>['gold']</td>\n",
       "      <td>['south-africa']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>SIX KILLED IN SOUTH AFRICAN GOLD MINE ACCIDENT</td>\n",
       "      <td>JOHANNESBURG, Oct 19 -</td>\n",
       "      <td>2</td>\n",
       "      <td>reut2-021.sgm</td>\n",
       "      <td>19-OCT-1987 00:05:11.26</td>\n",
       "      <td>Six black miners have been killed\n",
       "and two inju...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21576</th>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>['switzerland']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>PROJECTIONS SHOW SWISS VOTERS WANT TRIED PARTIES</td>\n",
       "      <td>BERNE, Oct 19 -</td>\n",
       "      <td>1</td>\n",
       "      <td>reut2-021.sgm</td>\n",
       "      <td>19-OCT-1987 00:03:21.69</td>\n",
       "      <td>The prospect of a dominant alliance of\n",
       "sociali...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21577</th>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>['amex']</td>\n",
       "      <td>AMERICAN EXCHANGE INTRODUCES INSTITUTIONAL INDEX</td>\n",
       "      <td>NEW YORK, Oct 19 -</td>\n",
       "      <td>1</td>\n",
       "      <td>reut2-021.sgm</td>\n",
       "      <td>19-OCT-1987 13:30:41.59</td>\n",
       "      <td>The American Stock Exchange said it has\n",
       "introd...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>19043 rows × 14 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                  TOPICS  \\\n",
       "0                                              ['cocoa']   \n",
       "1                                                   <NA>   \n",
       "2                                                   <NA>   \n",
       "3                                                   <NA>   \n",
       "4      ['grain', 'wheat', 'corn', 'barley', 'oat', 's...   \n",
       "...                                                  ...   \n",
       "21573                                           ['ship']   \n",
       "21574                                            ['ipi']   \n",
       "21575                                           ['gold']   \n",
       "21576                                               <NA>   \n",
       "21577                                               <NA>   \n",
       "\n",
       "                                                  PLACES PEOPLE  ORGS  \\\n",
       "0                      ['el-salvador', 'usa', 'uruguay']   <NA>  <NA>   \n",
       "1                                                ['usa']   <NA>  <NA>   \n",
       "2                                                ['usa']   <NA>  <NA>   \n",
       "3                                      ['usa', 'brazil']   <NA>  <NA>   \n",
       "4                                                ['usa']   <NA>  <NA>   \n",
       "...                                                  ...    ...   ...   \n",
       "21573  ['hong-kong', 'japan', 'india', 'pakistan', 'i...   <NA>  <NA>   \n",
       "21574                                           ['ussr']   <NA>  <NA>   \n",
       "21575                                   ['south-africa']   <NA>  <NA>   \n",
       "21576                                    ['switzerland']   <NA>  <NA>   \n",
       "21577                                            ['usa']   <NA>  <NA>   \n",
       "\n",
       "      EXCHANGES                                              TITLE  \\\n",
       "0          <NA>                                 BAHIA COCOA REVIEW   \n",
       "1          <NA>       STANDARD OIL &lt;SRD> TO FORM FINANCIAL UNIT   \n",
       "2          <NA>      TEXAS COMMERCE BANCSHARES &lt;TCB> FILES PLAN   \n",
       "3          <NA>    TALKING POINT/BANKAMERICA &lt;BAC> EQUITY OFFER   \n",
       "4          <NA>   NATIONAL AVERAGE PRICES FOR FARMER-OWNED RESERVE   \n",
       "...         ...                                                ...   \n",
       "21573      <NA>  JAPAN/INDIA CONFERENCE CUTS GULF WAR RISK CHARGES   \n",
       "21574      <NA>      SOVIET INDUSTRIAL GROWTH/TRADE SLOWER IN 1987   \n",
       "21575      <NA>     SIX KILLED IN SOUTH AFRICAN GOLD MINE ACCIDENT   \n",
       "21576      <NA>   PROJECTIONS SHOW SWISS VOTERS WANT TRIED PARTIES   \n",
       "21577  ['amex']   AMERICAN EXCHANGE INTRODUCES INSTITUTIONAL INDEX   \n",
       "\n",
       "                          DATELINE  TOPICS_ENUM       FILENAME  \\\n",
       "0              SALVADOR, Feb 26 -             2  reut2-000.sgm   \n",
       "1             CLEVELAND, Feb 26 -             1  reut2-000.sgm   \n",
       "2               HOUSTON, Feb 26 -             1  reut2-000.sgm   \n",
       "3           LOS ANGELES, Feb 26 -             1  reut2-000.sgm   \n",
       "4            WASHINGTON, Feb 26 -             2  reut2-000.sgm   \n",
       "...                            ...          ...            ...   \n",
       "21573         HONG KONG, Oct 19 -             2  reut2-021.sgm   \n",
       "21574            MOSCOW, Oct 19 -             2  reut2-021.sgm   \n",
       "21575      JOHANNESBURG, Oct 19 -             2  reut2-021.sgm   \n",
       "21576             BERNE, Oct 19 -             1  reut2-021.sgm   \n",
       "21577          NEW YORK, Oct 19 -             1  reut2-021.sgm   \n",
       "\n",
       "                          DATE  \\\n",
       "0      26-FEB-1987 15:01:01.79   \n",
       "1      26-FEB-1987 15:02:20.00   \n",
       "2      26-FEB-1987 15:03:27.51   \n",
       "3      26-FEB-1987 15:07:13.72   \n",
       "4      26-FEB-1987 15:10:44.60   \n",
       "...                        ...   \n",
       "21573  19-OCT-1987 00:34:08.94   \n",
       "21574  19-OCT-1987 00:18:22.79   \n",
       "21575  19-OCT-1987 00:05:11.26   \n",
       "21576  19-OCT-1987 00:03:21.69   \n",
       "21577  19-OCT-1987 13:30:41.59   \n",
       "\n",
       "                                                    BODY  TEXT TYPE  \\\n",
       "0      Showers continued throughout the week in\n",
       "the B...          0   \n",
       "1      Standard Oil Co and BP North America\n",
       "Inc said ...          0   \n",
       "2      Texas Commerce Bancshares Inc's Texas\n",
       "Commerce...          0   \n",
       "3      BankAmerica Corp is not under\n",
       "pressure to act ...          0   \n",
       "4      The U.S. Agriculture Department\n",
       "reported the f...          0   \n",
       "...                                                  ...        ...   \n",
       "21573  The Japan/India-Pakistan-Gulf/Japan\n",
       "shipping c...          0   \n",
       "21574  The Soviet Union's industrial output is\n",
       "growin...          0   \n",
       "21575  Six black miners have been killed\n",
       "and two inju...          0   \n",
       "21576  The prospect of a dominant alliance of\n",
       "sociali...          0   \n",
       "21577  The American Stock Exchange said it has\n",
       "introd...          0   \n",
       "\n",
       "       LEWISSPLIT  CGISPLIT  \n",
       "0               2         1  \n",
       "1               2         1  \n",
       "2               2         1  \n",
       "3               2         1  \n",
       "4               2         1  \n",
       "...           ...       ...  \n",
       "21573           1         1  \n",
       "21574           1         1  \n",
       "21575           1         1  \n",
       "21576           1         1  \n",
       "21577           1         1  \n",
       "\n",
       "[19043 rows x 14 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "le = LabelEncoder()\n",
    "le.fit(podaci.loc[:, 'TOPICS_ENUM'])\n",
    "podaci.loc[:, 'TOPICS_ENUM'] = le.transform(podaci.loc[:, 'TOPICS_ENUM'])\n",
    "le.fit(podaci.loc[:, 'TEXT TYPE'])\n",
    "podaci.loc[:, 'TEXT TYPE'] = le.transform(podaci.loc[:, 'TEXT TYPE'])\n",
    "le.fit(podaci.loc[:, 'LEWISSPLIT'])\n",
    "podaci.loc[:, 'LEWISSPLIT'] = le.transform(podaci.loc[:, 'LEWISSPLIT'])\n",
    "le.fit(podaci.loc[:, 'CGISPLIT'])\n",
    "podaci.loc[:, 'CGISPLIT'] = le.transform(podaci.loc[:, 'CGISPLIT'])\n",
    "podaci"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>TOPICS</th>\n",
       "      <th>PLACES</th>\n",
       "      <th>PEOPLE</th>\n",
       "      <th>ORGS</th>\n",
       "      <th>EXCHANGES</th>\n",
       "      <th>TITLE</th>\n",
       "      <th>DATELINE</th>\n",
       "      <th>TOPICS_ENUM</th>\n",
       "      <th>FILENAME</th>\n",
       "      <th>DATE</th>\n",
       "      <th>BODY</th>\n",
       "      <th>TEXT TYPE</th>\n",
       "      <th>LEWISSPLIT</th>\n",
       "      <th>CGISPLIT</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>10377</td>\n",
       "      <td>18398</td>\n",
       "      <td>1012</td>\n",
       "      <td>854</td>\n",
       "      <td>402</td>\n",
       "      <td>19043</td>\n",
       "      <td>19043</td>\n",
       "      <td>19043.000000</td>\n",
       "      <td>19043</td>\n",
       "      <td>18745</td>\n",
       "      <td>19043</td>\n",
       "      <td>19043.0</td>\n",
       "      <td>19043.000000</td>\n",
       "      <td>19043.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>646</td>\n",
       "      <td>1095</td>\n",
       "      <td>214</td>\n",
       "      <td>66</td>\n",
       "      <td>57</td>\n",
       "      <td>18253</td>\n",
       "      <td>8822</td>\n",
       "      <td>NaN</td>\n",
       "      <td>22</td>\n",
       "      <td>18745</td>\n",
       "      <td>18781</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>['earn']</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>['reagan']</td>\n",
       "      <td>['ec']</td>\n",
       "      <td>['nyse']</td>\n",
       "      <td>PROPOSED OFFERINGS RECENTLY FILED WITH THE SEC</td>\n",
       "      <td>WASHINGTON, April 9 -</td>\n",
       "      <td>NaN</td>\n",
       "      <td>reut2-016.sgm</td>\n",
       "      <td>26-FEB-1987 15:01:01.79</td>\n",
       "      <td>The Bundesbank left credit policies\\nunchanged...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>3735</td>\n",
       "      <td>10685</td>\n",
       "      <td>106</td>\n",
       "      <td>296</td>\n",
       "      <td>69</td>\n",
       "      <td>25</td>\n",
       "      <td>111</td>\n",
       "      <td>NaN</td>\n",
       "      <td>947</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.626372</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.637767</td>\n",
       "      <td>0.962191</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.527006</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.553764</td>\n",
       "      <td>0.190739</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          TOPICS   PLACES      PEOPLE    ORGS EXCHANGES  \\\n",
       "count      10377    18398        1012     854       402   \n",
       "unique       646     1095         214      66        57   \n",
       "top     ['earn']  ['usa']  ['reagan']  ['ec']  ['nyse']   \n",
       "freq        3735    10685         106     296        69   \n",
       "mean         NaN      NaN         NaN     NaN       NaN   \n",
       "std          NaN      NaN         NaN     NaN       NaN   \n",
       "min          NaN      NaN         NaN     NaN       NaN   \n",
       "25%          NaN      NaN         NaN     NaN       NaN   \n",
       "50%          NaN      NaN         NaN     NaN       NaN   \n",
       "75%          NaN      NaN         NaN     NaN       NaN   \n",
       "max          NaN      NaN         NaN     NaN       NaN   \n",
       "\n",
       "                                                 TITLE  \\\n",
       "count                                            19043   \n",
       "unique                                           18253   \n",
       "top     PROPOSED OFFERINGS RECENTLY FILED WITH THE SEC   \n",
       "freq                                                25   \n",
       "mean                                               NaN   \n",
       "std                                                NaN   \n",
       "min                                                NaN   \n",
       "25%                                                NaN   \n",
       "50%                                                NaN   \n",
       "75%                                                NaN   \n",
       "max                                                NaN   \n",
       "\n",
       "                          DATELINE   TOPICS_ENUM       FILENAME  \\\n",
       "count                        19043  19043.000000          19043   \n",
       "unique                        8822           NaN             22   \n",
       "top         WASHINGTON, April 9 -            NaN  reut2-016.sgm   \n",
       "freq                           111           NaN            947   \n",
       "mean                           NaN      1.626372            NaN   \n",
       "std                            NaN      0.527006            NaN   \n",
       "min                            NaN      0.000000            NaN   \n",
       "25%                            NaN      1.000000            NaN   \n",
       "50%                            NaN      2.000000            NaN   \n",
       "75%                            NaN      2.000000            NaN   \n",
       "max                            NaN      2.000000            NaN   \n",
       "\n",
       "                           DATE  \\\n",
       "count                     18745   \n",
       "unique                    18745   \n",
       "top     26-FEB-1987 15:01:01.79   \n",
       "freq                          1   \n",
       "mean                        NaN   \n",
       "std                         NaN   \n",
       "min                         NaN   \n",
       "25%                         NaN   \n",
       "50%                         NaN   \n",
       "75%                         NaN   \n",
       "max                         NaN   \n",
       "\n",
       "                                                     BODY  TEXT TYPE  \\\n",
       "count                                               19043    19043.0   \n",
       "unique                                              18781        NaN   \n",
       "top     The Bundesbank left credit policies\\nunchanged...        NaN   \n",
       "freq                                                    4        NaN   \n",
       "mean                                                  NaN        0.0   \n",
       "std                                                   NaN        0.0   \n",
       "min                                                   NaN        0.0   \n",
       "25%                                                   NaN        0.0   \n",
       "50%                                                   NaN        0.0   \n",
       "75%                                                   NaN        0.0   \n",
       "max                                                   NaN        0.0   \n",
       "\n",
       "          LEWISSPLIT      CGISPLIT  \n",
       "count   19043.000000  19043.000000  \n",
       "unique           NaN           NaN  \n",
       "top              NaN           NaN  \n",
       "freq             NaN           NaN  \n",
       "mean        1.637767      0.962191  \n",
       "std         0.553764      0.190739  \n",
       "min         0.000000      0.000000  \n",
       "25%         1.000000      1.000000  \n",
       "50%         2.000000      1.000000  \n",
       "75%         2.000000      1.000000  \n",
       "max         2.000000      1.000000  "
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "podaci.describe(include='all')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Vizualizacija podataka"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Pie chart"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x216 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "count_per_lewissplit = clanci.groupby('LEWISSPLIT').count().loc[:, 'NEWID'].values\n",
    "labels = clanci.groupby('LEWISSPLIT').count().index.values\n",
    "percent = (count_per_lewissplit/len(clanci))*100\n",
    "\n",
    "# pokazuje postotak članaka koji su se koristi u kojim skupinama\n",
    "plt.figure(figsize=(10, 3))\n",
    "plt.pie(percent, labels=labels, autopct='%.2f')\n",
    "plt.title(\"Udio podataka za treniranje i testiranje, LEWIS SPLIT\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Inkonzistentni podaci"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# inkonzistentni podaci\n",
    "dateline_helper_dataframe = podaci.copy()\n",
    "dateline_helper_dataframe['DATELINE'] = dateline_helper_dataframe.DATELINE.str.upper()\n",
    "dateline_helper_dataframe['DATELINE'] = dateline_helper_dataframe.DATELINE.str.replace('JANUARY', 'JAN')\n",
    "dateline_helper_dataframe['DATELINE'] = dateline_helper_dataframe.DATELINE.str.upper().str.replace('FEBRUARY', 'FEB')\n",
    "dateline_helper_dataframe['DATELINE'] = dateline_helper_dataframe.DATELINE.str.replace('MARCH', 'MAR').str.replace('NARCH', 'MAR')\n",
    "dateline_helper_dataframe['DATELINE'] = dateline_helper_dataframe.DATELINE.str.replace('APRIL', 'APR')\n",
    "dateline_helper_dataframe['DATELINE'] = dateline_helper_dataframe.DATELINE.str.replace('JUNE', 'JUN')\n",
    "dateline_helper_dataframe['DATELINE'] = dateline_helper_dataframe.DATELINE.str.replace('JULY', 'JUL')\n",
    "dateline_helper_dataframe['DATELINE'] = dateline_helper_dataframe.DATELINE.str.replace('AUGUST', 'AUG')\n",
    "dateline_helper_dataframe['DATELINE'] = dateline_helper_dataframe.DATELINE.str.replace('SEPTEMBER', 'SEP')\n",
    "dateline_helper_dataframe['DATELINE'] = dateline_helper_dataframe.DATELINE.str.replace('OCTOBER', 'OCT')\n",
    "dateline_helper_dataframe['DATELINE'] = dateline_helper_dataframe.DATELINE.str.replace('NOVEMBER', 'NOV')\n",
    "dateline_helper_dataframe['DATELINE'] = dateline_helper_dataframe.DATELINE.str.replace('DECEMBER', 'DEC')\n",
    "\n",
    "\n",
    "dateline_helper_dataframe = dateline_helper_dataframe.loc[dateline_helper_dataframe.DATELINE.notnull(), :]\n",
    "dateline_helper_dataframe = dateline_helper_dataframe.DATELINE.str.extract(r',[\\s]*(\\w+)[\\s]*[\\d]*[\\s]*-')\n",
    "dateline_helper_dataframe = dateline_helper_dataframe.loc[dateline_helper_dataframe[0].notnull(), :]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "months = [\"JAN\", \"FEB\", \"MAR\", \"APR\", \"MAY\", \"JUN\", \"JUL\", \"AUG\", \"SEP\", \"OCT\", \"NOV\", \"DEC\"]\n",
    "count_months = {}\n",
    "for month in months:\n",
    "    count_months[month] = 0\n",
    "    \n",
    "nul = 0\n",
    "for month in dateline_helper_dataframe.to_numpy():\n",
    "    try:\n",
    "        count_months[month[0]] += 1\n",
    "    except:\n",
    "        nul += 1\n",
    "\n",
    "#print(count_months)\n",
    "plt.figure(figsize=(15, 5))\n",
    "plt.plot(count_months.keys(), count_months.values())\n",
    "plt.grid()\n",
    "plt.title(\"Broj objavljenih članaka po mjesecima u 1987. godini\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Podaci o kategorijama"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "exchanges = []\n",
    "orgs = []\n",
    "people = []\n",
    "places = []\n",
    "topics = []\n",
    "\n",
    "\n",
    "file_content = open(os.path.join(\"reuters21578\", \"all-exchanges-strings.lc.txt\"), mode = 'r').read()\n",
    "exchanges = file_content.split(\"\\n\")\n",
    "file_content = open(os.path.join(\"reuters21578\", \"all-orgs-strings.lc.txt\"), mode = 'r').read()\n",
    "orgs = file_content.split(\"\\n\")\n",
    "file_content = open(os.path.join(\"reuters21578\", \"all-people-strings.lc.txt\"), mode = 'r').read()\n",
    "people = file_content.split(\"\\n\")\n",
    "file_content = open(os.path.join(\"reuters21578\", \"all-places-strings.lc.txt\"), mode = 'r').read()\n",
    "places = file_content.split(\"\\n\")\n",
    "file_content = open(os.path.join(\"reuters21578\", \"all-topics-strings.lc.txt\"), mode = 'r').read()\n",
    "topics = file_content.split(\"\\n\")\n",
    "\n",
    "#print(\"Sve moguće vrijednosti klase exchanges:\\n\", exchanges)\n",
    "#print(\"Sve moguće vrijednosti klase orgs:\\n\", orgs)\n",
    "#print(\"Sve moguće vrijednosti klase people:\\n\", people)\n",
    "#print(\"Sve moguće vrijednosti klase places:\\n\", places)\n",
    "#print(\"Sve moguće vrijednosti klase topics:\\n\", topics)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x20b53368550>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "most_freq_topic = {}\n",
    "for topic in topics:\n",
    "    most_freq_topic[topic] = dataframe.TOPICS.str.contains(topic).sum()\n",
    "\n",
    "del most_freq_topic['']\n",
    "topic_map_top_10 = dict(sorted(most_freq_topic .items(), key=lambda x: -x[1])[0:10]) #sortiramo od najčešćeg topica prema manje čestima\n",
    "\n",
    "plt.figure(figsize=(15, 5))\n",
    "plt.bar(topic_map_top_10.keys(), topic_map_top_10.values())\n",
    "plt.title(\"10 ukupno najčešćih topic-a\")\n",
    "plt.grid()\n",
    "plt.show()\n",
    "\n",
    "count_topic_train = {}\n",
    "count_topic_test = {}\n",
    "for topic in topic_map_top_10.keys():\n",
    "    count_topic_train[topic] = dataframe[dataframe.LEWISSPLIT == \"TRAIN\"].TOPICS.str.contains(topic).sum()\n",
    "    count_topic_test[topic] =  dataframe[dataframe.LEWISSPLIT == \"TEST\"].TOPICS.str.contains(topic).sum()\n",
    "\n",
    "plt.figure(figsize=(15, 5))\n",
    "plt.grid()\n",
    "plt.bar(topic_map_top_10.keys(), count_topic_train.values(), color='red', label='TRAIN')\n",
    "plt.bar(topic_map_top_10.keys(), count_topic_test.values(), color='green', label='TEST', bottom=list(count_topic_train.values())) # bottom -> postavi vrijednost NA count_per_title_died\n",
    "plt.title(\"Udio podataka s određenom klasom u train/test setu\")\n",
    "plt.legend(loc='best')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "most_freq_exchanges = {}\n",
    "for exchange in exchanges:\n",
    "    most_freq_exchanges[exchange] = dataframe.EXCHANGES.str.contains(exchange).sum()\n",
    "\n",
    "del most_freq_exchanges['']\n",
    "exchanges_map_top_10 = dict(sorted(most_freq_exchanges.items(), key=lambda x: -x[1])[0:10]) #sortiramo od najčešćeg topica prema manje čestima\n",
    "\n",
    "plt.figure(figsize=(15, 5))\n",
    "plt.bar(exchanges_map_top_10.keys(), exchanges_map_top_10.values())\n",
    "plt.title(\"10 ukupno najčešćih exchange-a\")\n",
    "plt.grid()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA20AAAE/CAYAAADVKysfAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAAuGklEQVR4nO3deZhlVXnv8e9PQBmaUaDFhthq0ARoY0JJNInaHRzjgIkSMQ50ovZNrnFIUANxTJSERDEmUW+CihAxtCgOBIiCaANOKDjQDBKItDIJKkhsRLThvX/sVelDdVV116kqahf1/TxPPXXOntY679nTu9fa+6SqkCRJkiT1033mugKSJEmSpImZtEmSJElSj5m0SZIkSVKPmbRJkiRJUo+ZtEmSJElSj5m0SZIkSVKPmbRJkv5XkqVJKsnWc12X2ZDk+UnOGni/fZLPJFmS5NQkB2zhcl6fZGWSZyV55zjj75PktCQvbe8fm+SKgfHrkjxhBj7SrEny5iQnzXU9JEkmbZI0LyT50yQXJrkjyQnjjD84ybeS/CTJ55I8aA6q2XtV9aGqetLA+58ARwBfBc6tqku2cFF/A/wB8BrgL8YZfzRwTlW9t5VzflU9fFqVlyQtWPfKK6mSdC90PfBW4MnAdoMjkuwOfAx4CfAfwFuADwOPvofrOC9V1TeBB05xnruAJ00y/qjp1kuSpFG2tEnSPFBVH6uqTwA/HGf07wGXVtVHquqnwJuBX0nyS+Mta2zXvMm6wSV5dpv+gCQnJHnrwLjlSa4ds9yjklyW5JYkH0iy7eC0SY5IclOSG5L84cC8Oyf5tyTfT/Kd1v1w3GNUq+8pbfofJ7k0ycjA+COT/Hcbd1mS3x0YtzLJ5wfe75/k7CQ3J7kxyV+24fcZWM4PW3m7tXHbJjmpDf9Rkq8mWdzG7dY+9/UtBp8YL1bNI5NcnOTWJB8ejdU4n3dlki8k+ec27beSHDwmdu9vMb0uyVuTbDXwOV7fYnpTi9nObdxoV9hVrb43JDlivDq06R+d5IvtM38zyfJJpn1oks+2GP0gyYeS7DLR9JKkyZm0SdL8tz/wzdE3VXUb8N9t+NBaUvV3wBOm0G3w+XStgQ8FHga8fmDcA4CdgSXAi4F3J9m1jfvnNu4hwOOBFwF/yMSeCawGdgFOA941MO6/gce25f0VcFKSvcb5fDsCnwE+RdfS9ovAOW30K4Bntbo8ELgFeHcbd3hb9j7A/YE/Bm5v4z4IbE8X+z2Bf5jkM/w+8BTgwcAjgJWTTPvrwLeB3YE3AR8bTSKBE4ENrf6/StcC+JI2bmX7W0EX20XcPVa0cfu2+Y7MOPfaJVkCnEHX2rsb8Grg1CR7TFDfAH9LF7tfpovVmyf5fJKkSZi0SdL8twi4dcywW4Edp7HMV9Hdr7W8qq6awnzvqqprqupmuvu6njcw7ufAX1fVz6vqTGA98PDWKvRc4Kiq+nFVrQOOBV44STmfr6ozq+pOukTpV0ZHtBbH66vqrqr6MHAlcNA4y3g68L2qOraqftrKvqCN+z/A66rq2qq6gy7heE66B7T8nC5Z+8WqurOqLqqq/2mJ4VOBP66qW9rnPHeSz/BPrZ4303VrfeQk094EvLMt88PAFcDTWgvfU4FXVdVtVXUTXaJ4WJvv+cA7qurbVbUeOAo4LHd/0MxftXnXAh/g7t/ZqBcAZ7aY31VVZwMXAr8zXmWr6qqqOruq7qiq7wPvoEuAJUlD8J42SZr/1gM7jRm2E/DjaSzzNXQJ1tgufZtzzcDr73D3e8V+WFUbBt7/hC7h3B24b5t+cN4lk5TzvTHL2TbJ1lW1IcmLgD8Hlrbxo2WMtQ9dq9x4HgR8PMldA8PuBBbTJYn7AKtbl7+TgNe1YTdX1S2T1HuyzzDZfXXXVVUNvB+N7YOAbYAbkoyOuw8bv4cHsmlct26fY9TY72zZOOU/CDg0yTMGhm0DfC7JY4H/HJ2/qvZPsifwT3Qtnju2Om1pXCRJY9jSJknz36UMtDQl2YGue+KlE0x/G10XvlEPGGeaJwGvT/LsKc63z8DrX6B7gMrm/ICu9WrwiZe/AFy3BfPeTbqnZr4X+FPg/lW1C3AJXXe9sa6hi9N4rgGeWlW7DPxtW1XXtdauv6qq/YDfoGuxe1GbZ7dZundrSQayMjbG9hrgDmD3gXruVFWjXWOvZ9O4bgBuHBi2Jd/ZNcAHx8Rjh6o6pj0Zc1H7Gy33b4ECHlFVO9G11I33HUiStoBJmyTNA0m2bg+q2ArYqj0MY7S3xMeBA9I9NGRb4I3AxVX1rQkW9w26LnLbtAd4PGecaS6lu9/q3UmeOTDf77SHbTyArgvlWC9Lsne73+ov6Z5iOanWxfEU4OgkO7bE68/pWrCmage6ZOH78L/35U3022unAw9I8qok92tl/3ob9y+tPg9qy9kjySHt9Yoky1q3zv+hSzjvrKob6Fqc3pNk1xbfxw3xGcazJ/CKtsxD6e4TO7OVeRZwbJKd2oNHHppktCviycCfJXlwkkV0P1Xw4TEtnm9I93t1+9PdRzjed3YS8IwkT04yuv4tT7L3BPXdka4F+EftfrjXTDcAkrSQmbRJ0vzwerqHXRxJ12pxextGu2fo2XT3kN1C99CKw8ZfDABvoGthuoXuQR3/Pt5E7VH4Twfem+SpdN0Cvwmso0sUxju5//c27tvt763jTDOel9O15H0b+HxbzvFbOO9gnS+jux/uS3StScuAL0ww7Y+BJwLPoOuqeCXdQzkA/pHuASdnJfkx8GW6uELXwvhRuoTtcuBcNiaYL6RL4r5Fdx/aq6b6GSZwAd3DQn5A9z0/p6pGnyT6IrrupZfRfacfBUYfvHI83fd2HnA18FO6WA86F7iK7iEsb6+qs8aMp6quAQ6hS8S/T9fy9homPo/4K+DX6O6tPIPuJykkSUPK3bvIS5I0nCTrgJdU1Wfmui4TSfJHwAuq6rfnui5bKslKurj+1gwvdyldIrfNmJY3SVLP2NImSVpI9qdLVCRJmjd8eqQkaUFoP3S9L3DoHFdFkqQpsXukJEmSJPWY3SMlSZIkqcdM2iRJkiSpx3pxT9vuu+9eS5cunetq9Mptt93GDjvsMNfVmHeM23CM29QZs+EYt+EYt6kzZsMxbsMxblNnzDZ10UUX/aCq9hhvXC+StqVLl3LhhRfOdTV6Zc2aNSxfvnyuqzHvGLfhGLepM2bDMW7DMW5TZ8yGY9yGY9ymzphtKsl3Jhpn90hJkiRJ6jGTNkmSJEnqMZM2SZIkSeoxkzZJkiRJ6jGTNkmSJEnqMZM2SZIkSeoxkzZJkiRJ6jGTNkmSJEnqMZM2SZIkSeoxkzZJkiRJ6jGTNkmSJEnqsa3nugJ9tvTIM+as7COWbWDlHJW/7pinzUm5kiRJkjZlS5skSZIk9ZhJmyRJkiT1mN0jNePsVipJkiTNHFvaJEmSJKnHTNokSZIkqcdM2iRJkiSpx0zaJEmSJKnHNpu0JTk+yU1JLhkz/OVJrkhyaZK/Hxh+VJKr2rgnz0alJUmSJGmh2JKnR54AvAv4t9EBSVYAhwCPqKo7kuzZhu8HHAbsDzwQ+EySh1XVnTNdcUmSJElaCDbb0lZV5wE3jxn8J8AxVXVHm+amNvwQYHVV3VFVVwNXAQfNYH0lSZIkaUEZ9p62hwGPTXJBknOTPKoNXwJcMzDdtW2YJEmSJGkIqarNT5QsBU6vqgPa+0uAzwKvBB4FfBh4CF03yi9V1UltuvcDZ1bVqeMscxWwCmDx4sUHrl69eiY+z4xae92tc1b24u3gxtvnpuxlS3ae1vzGbf5Zv349ixYtmutqzCvGbDjGbTjGbeqM2XCM23CM29QZs02tWLHioqoaGW/cltzTNp5rgY9Vl/F9JcldwO5t+D4D0+0NXD/eAqrqOOA4gJGRkVq+fPmQVZk9K488Y87KPmLZBo5dO+zXMz3rnr98WvMbt/lnzZo19HEb7DNjNhzjNhzjNnXGbDjGbTjGbeqM2dQM2z3yE8BvAyR5GHBf4AfAacBhSe6X5MHAvsBXZqCekiRJkrQgbbZJIsnJwHJg9yTXAm8CjgeOb90kfwYc3lrdLk1yCnAZsAF4mU+OlCRJkqThbTZpq6rnTTDqBRNMfzRw9HQqJUmSJEnqDNs9UpIkSZJ0DzBpkyRJkqQeM2mTJEmSpB4zaZMkSZKkHjNpkyRJkqQeM2mTJEmSpB4zaZMkSZKkHjNpkyRJkqQeM2mTJEmSpB4zaZMkSZKkHjNpkyRJkqQeM2mTJEmSpB4zaZMkSZKkHjNpkyRJkqQeM2mTJEmSpB4zaZMkSZKkHjNpkyRJkqQeM2mTJEmSpB4zaZMkSZKkHjNpkyRJkqQeM2mTJEmSpB4zaZMkSZKkHtts0pbk+CQ3JblknHGvTlJJdh8YdlSSq5JckeTJM11hSZIkSVpItqSl7QTgKWMHJtkHeCLw3YFh+wGHAfu3ed6TZKsZqakkSZIkLUCbTdqq6jzg5nFG/QPwWqAGhh0CrK6qO6rqauAq4KCZqKgkSZIkLURD3dOW5JnAdVX1zTGjlgDXDLy/tg2TJEmSJA0hVbX5iZKlwOlVdUCS7YHPAU+qqluTrANGquoHSd4NfKmqTmrzvR84s6pOHWeZq4BVAIsXLz5w9erVM/WZZsza626ds7IXbwc33j43ZS9bsvO05jdu88/69etZtGjRXFdjXjFmwzFuwzFuU2fMhmPchmPcps6YbWrFihUXVdXIeOO2HmJ5DwUeDHwzCcDewNeSHETXsrbPwLR7A9ePt5CqOg44DmBkZKSWL18+RFVm18ojz5izso9YtoFj1w7z9Uzfuucvn9b8xm3+WbNmDX3cBvvMmA3HuA3HuE2dMRuOcRuOcZs6YzY1U+4eWVVrq2rPqlpaVUvpErVfq6rvAacBhyW5X5IHA/sCX5nRGkuSJEnSArIlj/w/GfgS8PAk1yZ58UTTVtWlwCnAZcCngJdV1Z0zVVlJkiRJWmg224+sqp63mfFLx7w/Gjh6etWSJEmSJMGQT4+UJEmSJN0zTNokSZIkqcdM2iRJkiSpx0zaJEmSJKnHTNokSZIkqcdM2iRJkiSpx0zaJEmSJKnHTNokSZIkqcdM2iRJkiSpx0zaJEmSJKnHTNokSZIkqcdM2iRJkiSpx0zaJEmSJKnHTNokSZIkqcdM2iRJkiSpx0zaJEmSJKnHTNokSZIkqcdM2iRJkiSpx0zaJEmSJKnHTNokSZIkqcdM2iRJkiSpx0zaJEmSJKnHNpu0JTk+yU1JLhkY9rYk30pycZKPJ9llYNxRSa5KckWSJ89SvSVJkiRpQdiSlrYTgKeMGXY2cEBVPQL4L+AogCT7AYcB+7d53pNkqxmrrSRJkiQtMJtN2qrqPODmMcPOqqoN7e2Xgb3b60OA1VV1R1VdDVwFHDSD9ZUkSZKkBWUm7mn7I+A/2+slwDUD465twyRJkiRJQ0hVbX6iZClwelUdMGb464AR4PeqqpK8G/hSVZ3Uxr8fOLOqTh1nmauAVQCLFy8+cPXq1dP9LDNu7XW3zlnZi7eDG2+fm7KXLdl5WvMbt/ln/fr1LFq0aK6rMa8Ys+EYt+EYt6kzZsMxbsMxblNnzDa1YsWKi6pqZLxxWw+70CSHA08HDq6Nmd+1wD4Dk+0NXD/e/FV1HHAcwMjISC1fvnzYqsyalUeeMWdlH7FsA8euHfrrmZZ1z18+rfmN2/yzZs0a+rgN9pkxG45xG45xmzpjNhzjNhzjNnXGbGqG6h6Z5CnAXwDPrKqfDIw6DTgsyf2SPBjYF/jK9KspSZIkSQvTZpskkpwMLAd2T3It8Ca6p0XeDzg7CcCXq+qPq+rSJKcAlwEbgJdV1Z2zVXlJkiRJurfbbNJWVc8bZ/D7J5n+aODo6VRKkiRJktSZiadHSpIkSZJmiUmbJEmSJPWYSZskSZIk9ZhJmyRJkiT1mEmbJEmSJPWYSZskSZIk9ZhJmyRJkiT1mEmbJEmSJPWYSZskSZIk9ZhJmyRJkiT1mEmbJEmSJPWYSZskSZIk9ZhJmyRJkiT1mEmbJEmSJPWYSZskSZIk9ZhJmyRJkiT1mEmbJEmSJPWYSZskSZIk9ZhJmyRJkiT1mEmbJEmSJPWYSZskSZIk9ZhJmyRJkiT12GaTtiTHJ7kpySUDw3ZLcnaSK9v/XQfGHZXkqiRXJHnybFVckiRJkhaCLWlpOwF4yphhRwLnVNW+wDntPUn2Aw4D9m/zvCfJVjNWW0mSJElaYDabtFXVecDNYwYfApzYXp8IPGtg+OqquqOqrgauAg6amapKkiRJ0sIz7D1ti6vqBoD2f882fAlwzcB017ZhkiRJkqQhpKo2P1GyFDi9qg5o739UVbsMjL+lqnZN8m7gS1V1Uhv+fuDMqjp1nGWuAlYBLF68+MDVq1fPwMeZWWuvu3XOyl68Hdx4+9yUvWzJztOa37jNP+vXr2fRokVzXY15xZgNx7gNx7hNnTEbjnEbjnGbOmO2qRUrVlxUVSPjjdt6yGXemGSvqrohyV7ATW34tcA+A9PtDVw/3gKq6jjgOICRkZFavnz5kFWZPSuPPGPOyj5i2QaOXTvs1zM9656/fFrzG7f5Z82aNfRxG+wzYzYc4zYc4zZ1xmw4xm04xm3qjNnUDNs98jTg8Pb6cOCTA8MPS3K/JA8G9gW+Mr0qSpIkSdLCtdkmiSQnA8uB3ZNcC7wJOAY4JcmLge8ChwJU1aVJTgEuAzYAL6uqO2ep7pIkSZJ0r7fZpK2qnjfBqIMnmP5o4OjpVEqSJEmS1Bm2e6QkSZIk6R5g0iZJkiRJPWbSJkmSJEk9ZtImSZIkST1m0iZJkiRJPWbSJkmSJEk9ZtImSZIkST1m0iZJkiRJPWbSJkmSJEk9ZtImSZIkST1m0iZJkiRJPWbSJkmSJEk9ZtImSZIkST1m0iZJkiRJPWbSJkmSJEk9ZtImSZIkST1m0iZJkiRJPWbSJkmSJEk9ZtImSZIkST1m0iZJkiRJPWbSJkmSJEk9ZtImSZIkST02raQtyZ8luTTJJUlOTrJtkt2SnJ3kyvZ/15mqrCRJkiQtNEMnbUmWAK8ARqrqAGAr4DDgSOCcqtoXOKe9lyRJkiQNYbrdI7cGtkuyNbA9cD1wCHBiG38i8KxpliFJkiRJC9bQSVtVXQe8HfgucANwa1WdBSyuqhvaNDcAe85ERSVJkiRpIUpVDTdjd6/aqcBzgR8BHwE+CryrqnYZmO6WqtrkvrYkq4BVAIsXLz5w9erVQ9VjNq297tY5K3vxdnDj7XNT9rIlO09rfuM2/6xfv55FixbNdTXmFWM2HOM2HOM2dcZsOMZtOMZt6ozZplasWHFRVY2MN27raSz3CcDVVfV9gCQfA34DuDHJXlV1Q5K9gJvGm7mqjgOOAxgZGanly5dPoyqzY+WRZ8xZ2Ucs28Cxa6fz9Qxv3fOXT2t+4zacpXMatzs59vO3zUnZ64552pyUO11r1qyhj/utvjNuwzFuU2fMhmPchmPcps6YTc107mn7LvDoJNsnCXAwcDlwGnB4m+Zw4JPTq6IkSZIkLVxDN0lU1QVJPgp8DdgAfJ2u5WwRcEqSF9MldofOREUlSZIkaSGaVj+yqnoT8KYxg++ga3WTJEmSJE3TdB/5L0mSJEmaRSZtkiRJktRjJm2SJEmS1GMmbZIkSZLUYyZtkiRJktRjJm2SJEmS1GMmbZIkSZLUYyZtkiRJktRjJm2SJEmS1GMmbZIkSZLUYyZtkiRJktRjJm2SJEmS1GMmbZIkSZLUYyZtkiRJktRjJm2SJEmS1GMmbZIkSZLUYyZtkiRJktRjJm2SJEmS1GMmbZIkSZLUYyZtkiRJktRjJm2SJEmS1GNbz3UFJGlYS488Y87KPmLZBlbOUfnrjnnanJQrSZLmxrRa2pLskuSjSb6V5PIkj0myW5Kzk1zZ/u86U5WVJEmSpIVmut0j/xH4VFX9EvArwOXAkcA5VbUvcE57L0mSJEkawtDdI5PsBDwOWAlQVT8DfpbkEGB5m+xEYA3wF9OppCRp5titVJKk+SVVNdyMySOB44DL6FrZLgJeCVxXVbsMTHdLVW3SRTLJKmAVwOLFiw9cvXr1UPWYTWuvu3XOyl68Hdx4+9yUvWzJztOa37gNx7hNnTEbjnEbjnGbX9avX8+iRYvmuhrzjnEbjnGbOmO2qRUrVlxUVSPjjZtO0jYCfBn4zaq6IMk/Av8DvHxLkrZBIyMjdeGFFw5Vj9k011ejj107N8+Jme7VaOM2HOM2dcZsOMZtOMZtflmzZg3Lly+f62rMO8ZtOMZt6ozZppJMmLRN5whwLXBtVV3Q3n+U7v61G5PsVVU3JNkLuGkaZUiSpHlqrhNdu+JKurcY+kEkVfU94JokD2+DDqbrKnkacHgbdjjwyWnVUJIkSZIWsOn2tXg58KEk9wW+DfwhXSJ4SpIXA98FDp1mGZIkSZK0YE0raauqbwDj9bs8eDrLlSRJkiR1pvs7bZIkSZKkWWTSJkmSJEk9ZtImSZIkST1m0iZJkiRJPWbSJkmSJEk9Nt1H/kuSJGkG+aPkksaypU2SJEmSesykTZIkSZJ6zKRNkiRJknrMpE2SJEmSesykTZIkSZJ6zKRNkiRJknrMpE2SJEmSesykTZIkSZJ6zKRNkiRJknrMpE2SJEmSesykTZIkSZJ6zKRNkiRJknrMpE2SJEmSesykTZIkSZJ6zKRNkiRJknps2klbkq2SfD3J6e39bknOTnJl+7/r9KspSZIkSQvTTLS0vRK4fOD9kcA5VbUvcE57L0mSJEkawrSStiR7A08D3jcw+BDgxPb6ROBZ0ylDkiRJkhay6ba0vRN4LXDXwLDFVXUDQPu/5zTLkCRJkqQFK1U13IzJ04Hfqar/m2Q58OqqenqSH1XVLgPT3VJVm9zXlmQVsApg8eLFB65evXqoesymtdfdOmdlL94Obrx9bspetmTnac1v3IZj3KbOmA3HuA3HuE2dMRuOcRuOcZtf1q9fz6JFi+a6Gr2yYsWKi6pqZLxx00na/hZ4IbAB2BbYCfgY8ChgeVXdkGQvYE1VPXyyZY2MjNSFF144VD1m09Ijz5izso9YtoFj1249J2WvO+Zp05rfuA3HuE2dMRuOcRuOcZs6YzYc4zYc4za/rFmzhuXLl891NXolyYRJ29DdI6vqqKrau6qWAocBn62qFwCnAYe3yQ4HPjlsGZIkSZK00M3G77QdAzwxyZXAE9t7SZIkSdIQZqQdt6rWAGva6x8CB8/EciVJkiRpoZubzreSJEmS5tRc3we4co7Kn4/3Ac5G90hJkiRJ0gwxaZMkSZKkHjNpkyRJkqQeM2mTJEmSpB4zaZMkSZKkHjNpkyRJkqQeM2mTJEmSpB4zaZMkSZKkHjNpkyRJkqQeM2mTJEmSpB4zaZMkSZKkHjNpkyRJkqQeM2mTJEmSpB4zaZMkSZKkHjNpkyRJkqQeM2mTJEmSpB4zaZMkSZKkHjNpkyRJkqQeM2mTJEmSpB4zaZMkSZKkHjNpkyRJkqQeGzppS7JPks8luTzJpUle2YbvluTsJFe2/7vOXHUlSZIkaWGZTkvbBuCIqvpl4NHAy5LsBxwJnFNV+wLntPeSJEmSpCEMnbRV1Q1V9bX2+sfA5cAS4BDgxDbZicCzpllHSZIkSVqwZuSetiRLgV8FLgAWV9UN0CV2wJ4zUYYkSZIkLUSpquktIFkEnAscXVUfS/KjqtplYPwtVbXJfW1JVgGrABYvXnzg6tWrp1WP2bD2ulvnrOzF28GNt89N2cuW7Dyt+Y3bcIzb1Bmz4Ri34Ri3qTNmwzFuwzFuU2fM+mXFihUXVdXIeOOmlbQl2QY4Hfh0Vb2jDbsCWF5VNyTZC1hTVQ+fbDkjIyN14YUXDl2P2bL0yDPmrOwjlm3g2LVbz0nZ64552rTmN27DMW5TZ8yGY9yGY9ymzpgNx7gNx7hNnTHrlyQTJm3TeXpkgPcDl48mbM1pwOHt9eHAJ4ctQ5IkSZIWuumkt78JvBBYm+QbbdhfAscApyR5MfBd4NBp1VCSJEmSFrChk7aq+jyQCUYfPOxyJUmSJEkbzcjTIyVJkiRJs8OkTZIkSZJ6zKRNkiRJknrMpE2SJEmSesykTZIkSZJ6zKRNkiRJknrMpE2SJEmSesykTZIkSZJ6zKRNkiRJknrMpE2SJEmSesykTZIkSZJ6zKRNkiRJknrMpE2SJEmSesykTZIkSZJ6zKRNkiRJknrMpE2SJEmSesykTZIkSZJ6zKRNkiRJknrMpE2SJEmSesykTZIkSZJ6zKRNkiRJknrMpE2SJEmSemzWkrYkT0lyRZKrkhw5W+VIkiRJ0r3ZrCRtSbYC3g08FdgPeF6S/WajLEmSJEm6N5utlraDgKuq6ttV9TNgNXDILJUlSZIkSfdas5W0LQGuGXh/bRsmSZIkSZqCVNXMLzQ5FHhyVb2kvX8hcFBVvXxgmlXAqvb24cAVM16R+W134AdzXYl5yLgNx7hNnTEbjnEbjnGbOmM2HOM2HOM2dcZsUw+qqj3GG7H1LBV4LbDPwPu9gesHJ6iq44DjZqn8eS/JhVU1Mtf1mG+M23CM29QZs+EYt+EYt6kzZsMxbsMxblNnzKZmtrpHfhXYN8mDk9wXOAw4bZbKkiRJkqR7rVlpaauqDUn+FPg0sBVwfFVdOhtlSZIkSdK92Wx1j6SqzgTOnK3lLwB2HR2OcRuOcZs6YzYc4zYc4zZ1xmw4xm04xm3qjNkUzMqDSCRJkiRJM2O27mmTJEmSJM0Ak7YtkOSLc1z+m5O8egrTL01yyWzWSfPPlqwXSdYl2f2eqtN8lmRlkgcOvH9Vku3nsk5zJckJSZ4z1/XoqyTr57oO81WSZyY5cq7r0Tdbur8Z3acn2SXJ/51mmWuSLLgn/bn9qi9M2rZAVf3GXNfhnpJk1u5znIp0XD/nib6sN/ewlcADB96/CphS0pZkqxmsj3SvU1WnVdUxc12PHnoVU9vf7AJMK2mbjgV6jLhX8zu953lSvAWSrE+yKMk5Sb6WZG2SQ9q4pUm+leR9SS5J8qEkT0jyhSRXJjmoTbdDkuOTfDXJ1wfm3z/JV5J8I8nFSfadoBq/kuSzbZkvbfOOW6cxdX9IK+9RSR6a5FNJLkpyfpJfatOckOQdST4H/N1sxHBLtFhenuQ9wNeAN7R4XZzkrwam+0T7DJe2H2kfHf7iJP/Vrga+N8m72vBnJLmgxeEzSRa34W9u38maJN9O8op7+jNPV5K/G7x62j7TEUne1tbHtUmeO858WyV5ext/cZKXjxm/XVtXXjrJursyyUeS/Adw1qx/2BmS5C1JXjnw/ugkr0jymrHrW8a0TiZ5dYvxc4AR4ENt230lXQL3ubYdkeRJSb7Uts+PJFnUhq9L8sYknwcOvQc/+hYb2Bbf27azs9o68dIWo28mOTXjXOlv8T0hyX2S/L8kF7ZlDG7DxyS5rMX67W3Yg9r+7OL2/xfa8BOS/FOSL7bt9DkDy9nkO+uzifbZSV47uv9J8g9JPtteH5zkpLa9njCwTf9ZG//IJF9un//jSXZtw9e0fcNX2j7xsXP1mSeTLTh+tv3Mu5LsmOTqJNu0eXdq29I2E62Xbdsc/bs9yePbMr/Y9mVfTPLwNu2ZSR7RXn89yRvb67ckeclE3909FKcdkpzRPt8lSd7Epvub57V6XZJkvOP4McBDWyze1uaZaJ+3ybY/sJwXtLhdks2f39ztGJFk+ySntPI+nO64PG9a7pJ8cPB7b+vrM9vn/Fi6Y+aVSf5+YJpxz0vmgyRvaNvn2UlOTnf8W5Pkb5KcC7wyyYFJzk13TvbpJHu1eSc739xkf55kryTntfXzktF9Vrrz72PbdndOkj3a8Im2+QmPF/cKVeXfZv6A9XRP2typvd8duAoIsBTYACyjS4IvAo5v4w4BPtHm+RvgBe31LsB/ATsA/ww8vw2/L7DdOOW/GfgmsF0r+xq6HfZkdboEeDjwdeCRbZpzgH3b618HPttenwCcDmw1x3FeCtwFPBp4Et1ThdLiejrwuDbdbu3/du1z3r/FYx2wG7ANcD7wrjbdrmx86M5LgGMH4vpF4H4tfj8Etpnr9W2KMftV4NyB95cBhwNn0/3cxmLgu8Beo+tFm+5PgFOBrcfEdF2b7jPAizaz7q4Erh2dd778tc/3tfb6PsB/A88db30bjFmb/tXAm9vrNcDIwLh1wO7t9e7AecAO7f1fAG8cmO61cx2HLYjRBjbuO04BXgDcf2CatwIvb69PAJ4D/D3wrwPb2+h6tVWL1yPaNnrFwDS7tP//ARzeXv8RG/edJwAfad/LfsBVbfiE+4i+/QHr2/+J9tmPBj7Shp8PfIVuP/Ym4P8ABwJnDyxvNGYXA49vr/8aeOfAujm6n/sd4DNzHYPNrGcTHj/p9jOj+/IPAM9qr1cNfMZx18uBYc9ocd0G2ImN+70nAKe210cCL2vjvwp8ug3/HN2xdNzv7h6K07OB9w6835m7728eSLef36PV87MDcVrX6ruUu+/Lxt1+mGDbH1iv3tteP46Nx5MtOkbQ7T//tb0+oJUzMpOxmqX4j26/j2fjfmln4OoW75XAt9uwbYHvAPswyXlJ3//oLkp+g+48a0fgyvb9rQHe06bZhu4cao/2/rl0P/EFk59vjrc/PwJ4XXu9FbBje11sPEd+Ixv3BZMdizZZ/r3lz6bNLRfgb5I8ji6xWEJ3QgxwdVWtBUhyKXBOVVWStXQ7QOh2kM/MxnvTtgV+AfgS8LokewMfq6orJyj/k1V1O3B7u7J2EHDGJHXaA/gk8OyqujTdVf7fAD6SZHSZ9xtY/keq6s4pR2Xmfaeqvpzu6vuT6JJOgEXAvnQnwq9I8rtt+D5t+APokpebAZJ8BHhYm2Zv4MPtCtB96Xa0o86oqjuAO5LcRBe/a2ft082wqvp6kj3T3Vu1B3AL8Ejg5PZ93tiuiD2K7gRv1BOAf6mqDW05Nw+M+yTw91X1ofZ+onUXuhPJwXl7r6rWJflhkl+l+76/Thef8da37w5ZzKPpDhhfaNvbfem29VEfHnK596Srq+ob7fVFdPuyA5K8le7EbBHdb3GOegNwQVWtGhj2++law7emu3CwH92FhZ8C70tyBt3JIsBjgN9rrz9IlwCO+kRV3QVcltZSTvd9TbSP6KuJjiMXAQcm2RG4g66nwQjwWOAVwA3AQ5L8M91+/6wkO9Mlb+e2ZZ9Id7Iy6mPt/+h311dbcvwc9T7gtXTJ3B8CL23DJ1wv0/VeeRvw21X18yQPAE5sw4vuxBO6E+pX0B0fzgCe2K7eL62qK9K18I333X1vpgIxibXA21sL2ulVdf7AcRy6/deaqvo+dC1AdEnVJyZZ5kTbz3cZf9sfdTJAVZ2XrrVzF7b8GPFbwD+2+S9JMnhM6r2qOjfJu5PsSbevOrW63yWGbr29FSDJZcCD6JLlic5L+u632HjeSWstHTV6/Ho4XfJ9dovBVsANW3C+Od7+/KvA8W07+8TA+nfXQHknsXG/NtmxaLzl3yuYtG2559OdFB/Ydvzr6HZM0B1kR9018P4uNsY4dAnUFWOWe3mSC4CnAZ9O8hLgl9l4MPqd9r/GzFebqdOtdC1yvwlcSnfV4UdV9cgJPt9tEwy/p43WI8DfVtW/Do5Mspwu4XhMVf0kyRq6z3y3I9gY/wy8o6pOa/O/eWDc4Hd3J/Nzm/goXSvHA4DVwEO3YJ6w6To16gvAU5P8e3WXrsZdd5P8Ov1Zb6bqfXRXRx9Ad2X/YMZf3/bm7t3It2XLhO5k5XkTjJ8PcRu7bWxHdxXzWVX1zSQrgeUD03yVLvHYrapuTvJguiuzj6qqW5KcAGzbTnIOoov5YcCfAr89TvmD6+dgXTLwf5PvrOfG3WcPvP5DuivXFwMr6Lbly1sS8yvAk+lag34f+LPNlDUas77v17bk+AlAVX0hXfe9x9P1DBntunwC46yXSXagayl6aVVd36Z9C/C5qvrdJEvpWg6gW39H6FpMzqY74X4pXdICkx9vZ1VV/VeSA+nOB/42ydju6JMd/yYy0TF2KeNv+/9bnbHVY8uPEcPUs28+SLcuHEbXI2DUeOcS8/nzTlb3wfO0S6vqMXebMdmJyc83N9mft4sAj6M7F/5gkrdV1b+NM+/o+ncCEx+Lxjte3Ct4T9uW2xm4qe2sV9BdRZmKTwMvT7vs0K7yk+QhwLer6p+A04BHVNW7q+qR7W/0QHNIkm2T3J9u5fzqZur0M+BZwIuS/EFV/Q9wdZJDW7lpJwF99Wngj7LxPqAl7erWzsAtLWH7JboWDei6Ez0+ya7pbo599sCydgaua68Pv2eqf49aTXcAeQ5dAnce8Nx098HsQXfF9Stj5jkL+OMWK5LsNjDujXRdRd/T3o+77s5zHweeQneF+tNMvL7dCOyZ5P5J7gc8fWAZP6brNjLe+y8Dv5nkF9vytk8yX66wTmZHuiup29CduAz6FN19M2e0FqOd6A7ut7arnU+F7r4uYOeqOpPuYQqPbPN/kW49pi3785upy0TfWZ9Nts8+jy7JPY+u1eePgW+0hG134D5VdSpdi+avtav6t2Tj/WovBM7l3u/f6Fp7PjAwbKL18gPAB6rq/IFhg8eDlaMDq+pndBc6f59u+z2f7vs4f2C+6ZwDDC1dT4qfVNVJwNuBX+Pu+5sL6I5/u6d7uNHz2HRdGLu/Gnb7eW6b/reAW9t6uKXHiM/TxZck+9F1i51vTqDbb1FVl25m2snOS/ru88Az2nnnIrpkaqwrgD2SPAYg3f2l+w9zvpnkQXTb13uB99Ot49DlKaP3pf0BG48Lkx2L7rX6fPWtTwr4EPAfSS6k6+f7rSku4y3AO4GL245tHd0J4HPpbuz9OV03i7+eYP6v0HXZ+AXgLVV1fbouEBPWqapuS/J0uqbr2+hW7P+X5PV0XUJW090r1ztVdVaSXwa+1I4D6+nuqfkUXbJxMd0O48tt+uuS/A3dwet6ui5Yt7bFvZmumf66Nv2D78GPMuta99cdgeuq6oYkH6fravZNunX3tVX1vXYFddT76LppXNzWvfcCgzdIv4quq8Lf091X8042XXfnrar6Wbpuxj9q3UjHXd+q6qYkf023Xl3N3bexE4B/SXI7XbyPA/4zyQ1VtaJd/Tu5JXsAr6e712M+ewNdLL5D12Vr8CSQqvpIWxdPo2sV+DpdS/+36VpwafN8MsloC/loi9Er6Na51wDfp2t1mtAk+4ibpvkZZ9Nk++zzgdcBX2r77p+yMWFYAnwgG5+oe1T7fzjdOrg9XYwnjdm9xIfo7mE5eWDYJutlOwl8DvCwJKMtIi+h63Z7YpI/p7v3a9D5wMHtouD5dF3rR7+D6Z4DTMcy4G1J7gJ+TndP8mO4+/7mKLr77wKcWVWfHFxAVf0w3QNeLgH+s6peM8H2s7nbJG5J9zNIO7GxpWmi85ux3kMX+4vp9g0Xs/E4PS9U1Y1JLmfyrqej0052XtJrVfXVJKfRnUd8B7iQMXVvx9HnAP+Urrv21nTrwaVM/XxzOfCadj6yHnhRG34bsH+Si1r5ow9Wm/RYdG81eiO4JpCuZetrVXWPXVXTcJIsqqr17YrWx+luiP34XNdL/dNOfr8GHFoT30cqqWfaSeIhVfXCua6Lpqa1Am5TVT9N8lC6h1U8rLVyzgvtAslaNrZ2b276eXteMlD37el6AKyqqq/dw3VYX1WL7sky+8yWtkm0Lglr6LojqP/enOQJdPcZnMUWXAnTwtO65ZwOfNyETZo/0j2I5alsvNdb88v2dD9TsA1di+CfzLOE7Ql090C/Y0sStmY+n5cc146X2wIn3tMJmzZlS5skSZIk9ZgPIpEkSZKkHjNpkyRJkqQeM2mTJEmSpB4zaZMkSZKkHjNpkyRJkqQeM2mTJEmSpB77/y1pUCFJL+YNAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "most_freq_people = {}\n",
    "for p in people:\n",
    "    most_freq_people[p] = dataframe.PEOPLE.str.contains(p).sum()\n",
    "\n",
    "del most_freq_people['']\n",
    "people_map_top_10 = dict(sorted(most_freq_people.items(), key=lambda x: -x[1])[0:10]) #sortiramo od najčešćeg topica prema manje čestima\n",
    "\n",
    "plt.figure(figsize=(15, 5))\n",
    "plt.bar(people_map_top_10.keys(), people_map_top_10.values())\n",
    "plt.title(\"10 ukupno najčešćih people-a\")\n",
    "plt.grid()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "most_freq_places = {}\n",
    "for p in places:\n",
    "    most_freq_places[p] = dataframe.PEOPLE.str.contains(p).sum()\n",
    "\n",
    "del most_freq_places['']\n",
    "places_map_top_10 = dict(sorted(most_freq_places.items(), key=lambda x: -x[1])[0:10]) #sortiramo od najčešćeg topica prema manje čestima\n",
    "\n",
    "plt.figure(figsize=(15, 5))\n",
    "plt.bar(places_map_top_10.keys(), places_map_top_10.values())\n",
    "plt.title(\"10 ukupno najčešćih places-a\")\n",
    "plt.grid()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA20AAAE/CAYAAADVKysfAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAAlCElEQVR4nO3de5hddX3v8fdHoIAGuQjEGKihltqCVHpI0VaxSbHeK1il4kFLrC31qfXS4gV6rNVWTtEeerHiabFa6UGNaRGlQhVEA16Qm3IXCgpKJAUviMYLNfg9f6xf7HayJ5m9M5O9Jnm/nmee2Xtdf+s7e9aez/791ppUFZIkSZKkfnrApBsgSZIkSZqeoU2SJEmSeszQJkmSJEk9ZmiTJEmSpB4ztEmSJElSjxnaJEmSJKnHDG2SpBlJsiRJJdlx0m2ZC0mOS3LBwPMHJvloksVJzk7yqBlu57VJViQ5OsnfDJn/gCTnJvnd9vyIJDcPzL89yRNn4ZAkSdsIQ5skzXNJ/iDJlUnuS/KuIfOPTHJTku8m+XiSh0+gmb1XVe+uqicNPP8ucCJwBXBxVV0/w039b+B/Aq8CXjNk/inARVX19rafT1TVI7eo8ZKkbVr859qSNL8l+Q3gh8CTgV2rasXAvL2BLwC/A/wb8OfAEVX12DH2swS4DdipqtZvecs1TJLbgd+pqo/O8nZ39OcmSfOTPW2SNM9V1fur6gPA14fM/g3ghqr6l6r6PvB64NFJfnbYtqYOzUvy+iRnTbPss9vyj0ryriRvHJi3LMmaKds9OcmNSe5J8k9JdhlcNsmJSe5OsjbJCwfW3T3JPyf5apIvteGHQ9+/WntXteW/neSGJEsH5p+U5Att3o1JnjUwb0WSTw48PzjJhUm+keSuJH/cpj9gYDtfb/vbq83bJclZbfo3k1yRZGGbt1c77jtbDT4wrFbNoUmuTXJvkvdtqNWQ431Aq8eXWu3+Ocnubd6G4awvSvJl4GNJdkhyWpKvJbmt9dL+aMhrq8EXW31uS3LcsP22Zf82yR1JvpXkqiRHTLesJGnLGNokadt2MHDNhidV9R26nreDt2SjLVS9CXjiCMMGj6PrDXwE8DPAawfmPRTYHVgMvAg4Pcmebd7ftXk/BfwK8FvAC5neM4GVwB7AucBbB+Z9ATiibe8NwFlJFg05vt2AjwIfBh4G/DRwUZv9MuDo1paHAfcAp7d5x7dt7w88BHgx8L027/8BD6Sr/b7AX2/iGH4TeApwAPDzwIppllvRvpbT1WfBlOOltfPn6Gr/u8BTgUOB/9GOY8MxPwh4C/DUqtoN+GXg6k208Yq2nb2A9wD/Ml24lCRtGUObJG3bFgD3Tpl2L7DbFmzzFXTXay2rqltHWO+tVXVHVX2D7rqu5w3M+wHwZ1X1g6o6H1gHPDLJDsBzgZOr6ttVdTtwGvCCTeznk1V1flXdTxeUHr1hRutxvLOqflhV7wNuAQ4fso1nAP9ZVadV1ffbvi9r834P+F9Vtaaq7qPrvXxO6636AV1Y++mqur+qrqqqb7Vg+FTgxVV1TzvOizdxDG9p7fwG3bDWQ6dZ7jjgr6rqi1W1DjgZODY/frOY11fVd6rqe3Rh8G9b2+8BTp2yvR8Cj0qya1WtraobpmtgVZ1VVV+vqvVVdRqwM+C1eZI0BwxtkrRtWwc8eMq0BwPf3oJtvgo4vaqmDunbnDsGHn+Jrpdqg69Pud7qu3SBc2/gJ9ryg+su3sR+/nPKdnYZGP73W0mubkMXvwk8qu1jqv3peuWGeThwzsA2Pg/cDyykC4kfAVa2YZBvTrJT2943WlCaianHsGCa5R7GxrXZsbVlgzumLH/HsHmtF/a5dL2Da5Oct2EYbRtmuq59HdGmnZjk820I5zfpehiH1VKStIUMbZK0bbuBgZ6mNgTuEW36MN+hG8K3wUOHLPMk4LVJnj3ievsPPP5J4M7pm/0jX6PrvRq84+VPAl+Zwbo/Jt1dM98O/AHwkKraA7geyJDF76Cr0zB30A0h3GPga5eq+krrQXtDVR1EN7zwGXTDOe8A9kqyx6jt3ow72bg264G7BqYN3nFsLbDfwPPBnwlV9ZGq+jVgEXATXb2oqoOrakH7+kQLbq+h67nbs9XyXobXUpK0hQxtkjTPJdmxXUu0A7BDuxnGhuFx59ANd3t2W+Z1wLVVddM0m7uabnjdTu0GHs8ZsswNdNdbnZ7kmQPrPa3dbOOhdEMop3pJkv3aTTv+GHjf5o6tDXFcBZySZLcWvP4IGHpzlM14EF2A+Sr86Lq86f732oeAhyZ5RZKd274f0+b9fWvPw9t29klyVHu8PMkhbVjnt+gC5/1VtRb4d+BtSfZs9X3CGMcw1XuBP0xyQJIFdP9u4H2buEvkKuDl6f733B4M/EuCJAuTPLMF+/voemnvn2Y7u9GFw68COyZ5HRv36EqSZomhTZLmv9fS3eziJOD57fFrAarqq8Cz6a4huwd4DHDsJrb1J3Q9TPfQ3ajjPcMWqqpr6HqR3p7kqXTDAq8BbgcuYHgge0+b98X29cYhywzzUrqevC8Cn2zbeecM1x1s841018NdStcTdQjwqWmW/Tbwa8Cv0w1VvIXuZh8Af0t3g5MLknwb+AxdXaHrYfxXusD2eeBi/jtgvoAuxN0E3M3wYDuqd9LV/hK6f8fwfbp6TeftdD+Da4HPAefTha/76f4mOJGu9+4bdDcw+f1ptvMRuhD6H3RDMr/Pjw+7lCTNIv9PmyRpzmWO/vfYbEry28Dzq+pXJ92WraUF7r+vKv/huiT1mD1tkiR1DqbrrdpmJdk1ydPakNrFwJ/SDaGVJPXYjptfRJKkbVv7R9cHAsdMuClzLXTDXt9HN4z2PLrrHCVJPebwSEmSJEnqMYdHSpIkSVKPGdokSZIkqcd6cU3b3nvvXUuWLJl0M3rlO9/5Dg960IMm3Yx5x7qNx7qNzpqNx7qNx7qNzpqNx7qNx7qNzppt7KqrrvpaVe0zbF4vQtuSJUu48sorJ92MXlm9ejXLli2bdDPmHes2Hus2Oms2Hus2Hus2Oms2Hus2Hus2Omu2sSRfmm6ewyMlSZIkqcdmFNqS3J7kuiRXJ7myTdsryYVJbmnf9xxY/uQktya5OcmT56rxkiRJkrStG6WnbXlVHVpVS9vzk4CLqupA4KL2nCQHAcfS/ZPSpwBvS7LDLLZZkiRJkrYbWzI88ijgzPb4TODogekrq+q+qroNuBU4fAv2I0mSJEnbrZmGtgIuSHJVkhPatIVVtRagfd+3TV8M3DGw7po2TZIkSZI0olTV5hdKHlZVdybZF7gQeClwblXtMbDMPVW1Z5LTgUur6qw2/R3A+VV19pRtngCcALBw4cLDVq5cOVvHtE1Yt24dCxYsmHQz5h3rNh7rNjprNh7rNh7rNjprNh7rNh7rNjprtrHly5dfNXAp2o+Z0S3/q+rO9v3uJOfQDXe8K8miqlqbZBFwd1t8DbD/wOr7AXcO2eYZwBkAS5cuLW/5+eO8Dep4rNt4rNvorNl4rNt4rNvorNl4rNt4rNvorNloNjs8MsmDkuy24THwJOB64Fzg+LbY8cAH2+NzgWOT7JzkAOBA4PLZbrgkSZIkbQ9m0tO2EDgnyYbl31NVH05yBbAqyYuALwPHAFTVDUlWATcC64GXVNX9c9J6SZIkSdrGbTa0VdUXgUcPmf514Mhp1jkFOGWLWydJkiRJ27ktueW/JEmSJGmOzehGJNurJSedN7F9n3jIelZMaP+3n/r0iexXkiRJ0sbsaZMkSZKkHjO0SZIkSVKPGdokSZIkqccMbZIkSZLUY4Y2SZIkSeoxQ5skSZIk9ZihTZIkSZJ6zNAmSZIkST1maJMkSZKkHjO0SZIkSVKPGdokSZIkqccMbZIkSZLUY4Y2SZIkSeoxQ5skSZIk9ZihTZIkSZJ6zNAmSZIkST1maJMkSZKkHjO0SZIkSVKPGdokSZIkqccMbZIkSZLUY4Y2SZIkSeoxQ5skSZIk9ZihTZIkSZJ6zNAmSZIkST1maJMkSZKkHjO0SZIkSVKPGdokSZIkqccMbZIkSZLUY4Y2SZIkSeoxQ5skSZIk9ZihTZIkSZJ6zNAmSZIkST1maJMkSZKkHjO0SZIkSVKPGdokSZIkqccMbZIkSZLUY4Y2SZIkSeoxQ5skSZIk9ZihTZIkSZJ6zNAmSZIkST1maJMkSZKkHjO0SZIkSVKPzTi0JdkhyeeSfKg93yvJhUluad/3HFj25CS3Jrk5yZPnouGSJEmStD0Ypaft5cDnB56fBFxUVQcCF7XnJDkIOBY4GHgK8LYkO8xOcyVJkiRp+zKj0JZkP+DpwD8OTD4KOLM9PhM4emD6yqq6r6puA24FDp+V1kqSJEnSdmamPW1/A7wa+OHAtIVVtRagfd+3TV8M3DGw3Jo2TZIkSZI0olTVphdIngE8rap+P8ky4JVV9Ywk36yqPQaWu6eq9kxyOnBpVZ3Vpr8DOL+qzp6y3ROAEwAWLlx42MqVK2fxsGbHdV+5d2L7Xrgr3PW9yez7kMW7T2bHs2DdunUsWLBg0s2Yd6zb6KzZeKzbeKzb6KzZeKzbeKzb6KzZxpYvX35VVS0dNm/HGaz/OOCZSZ4G7AI8OMlZwF1JFlXV2iSLgLvb8muA/QfW3w+4c+pGq+oM4AyApUuX1rJly2Z6PFvNipPOm9i+TzxkPaddN5Mfz+y7/bhlE9nvbFi9ejV9fC31nXUbnTUbj3Ubj3UbnTUbj3Ubj3UbnTUbzWaHR1bVyVW1X1UtobvByMeq6vnAucDxbbHjgQ+2x+cCxybZOckBwIHA5bPeckmSJEnaDmxJV86pwKokLwK+DBwDUFU3JFkF3AisB15SVfdvcUslSZIkaTs0UmirqtXA6vb468CR0yx3CnDKFrZNkiRJkrZ7o/yfNkmSJEnSVmZokyRJkqQeM7RJkiRJUo8Z2iRJkiSpxwxtkiRJktRjhjZJkiRJ6jFDmyRJkiT1mKFNkiRJknrM0CZJkiRJPWZokyRJkqQeM7RJkiRJUo8Z2iRJkiSpxwxtkiRJktRjhjZJkiRJ6jFDmyRJkiT1mKFNkiRJknrM0CZJkiRJPWZokyRJkqQeM7RJkiRJUo8Z2iRJkiSpxwxtkiRJktRjhjZJkiRJ6jFDmyRJkiT1mKFNkiRJknrM0CZJkiRJPWZokyRJkqQeM7RJkiRJUo8Z2iRJkiSpxwxtkiRJktRjhjZJkiRJ6jFDmyRJkiT1mKFNkiRJknrM0CZJkiRJPWZokyRJkqQeM7RJkiRJUo8Z2iRJkiSpxwxtkiRJktRjhjZJkiRJ6jFDmyRJkiT1mKFNkiRJknrM0CZJkiRJPWZokyRJkqQeM7RJkiRJUo8Z2iRJkiSpxzYb2pLskuTyJNckuSHJG9r0vZJcmOSW9n3PgXVOTnJrkpuTPHkuD0CSJEmStmUz6Wm7D/jVqno0cCjwlCSPBU4CLqqqA4GL2nOSHAQcCxwMPAV4W5Id5qDtkiRJkrTN22xoq8669nSn9lXAUcCZbfqZwNHt8VHAyqq6r6puA24FDp/NRkuSJEnS9mJG17Ql2SHJ1cDdwIVVdRmwsKrWArTv+7bFFwN3DKy+pk2TJEmSJI0oVTXzhZM9gHOAlwKfrKo9BubdU1V7JjkduLSqzmrT3wGcX1VnT9nWCcAJAAsXLjxs5cqVW3gos++6r9w7sX0v3BXu+t5k9n3I4t0ns+NZsG7dOhYsWDDpZsw71m101mw81m081m101mw81m081m101mxjy5cvv6qqlg6bt+MoG6qqbyZZTXet2l1JFlXV2iSL6HrhoOtZ239gtf2AO4ds6wzgDIClS5fWsmXLRmnKVrHipPMmtu8TD1nPadeN9OOZNbcft2wi+50Nq1evpo+vpb6zbqOzZuOxbuOxbqOzZuOxbuOxbqOzZqOZyd0j92k9bCTZFXgicBNwLnB8W+x44IPt8bnAsUl2TnIAcCBw+Sy3W5IkSZK2CzPpylkEnNnuAPkAYFVVfSjJpcCqJC8CvgwcA1BVNyRZBdwIrAdeUlX3z03zJUmSJGnbttnQVlXXAr8wZPrXgSOnWecU4JQtbp0kSZIkbedmdPdISZIkSdJkGNokSZIkqccMbZIkSZLUY4Y2SZIkSeoxQ5skSZIk9ZihTZIkSZJ6zNAmSZIkST1maJMkSZKkHjO0SZIkSVKPGdokSZIkqccMbZIkSZLUY4Y2SZIkSeoxQ5skSZIk9ZihTZIkSZJ6zNAmSZIkST1maJMkSZKkHjO0SZIkSVKPGdokSZIkqccMbZIkSZLUY4Y2SZIkSeoxQ5skSZIk9ZihTZIkSZJ6zNAmSZIkST1maJMkSZKkHjO0SZIkSVKPGdokSZIkqccMbZIkSZLUY4Y2SZIkSeoxQ5skSZIk9ZihTZIkSZJ6zNAmSZIkST1maJMkSZKkHjO0SZIkSVKPGdokSZIkqccMbZIkSZLUY4Y2SZIkSeoxQ5skSZIk9ZihTZIkSZJ6zNAmSZIkST1maJMkSZKkHjO0SZIkSVKPGdokSZIkqccMbZIkSZLUY4Y2SZIkSeqxzYa2JPsn+XiSzye5IcnL2/S9klyY5Jb2fc+BdU5OcmuSm5M8eS4PQJIkSZK2ZTPpaVsPnFhVPwc8FnhJkoOAk4CLqupA4KL2nDbvWOBg4CnA25LsMBeNlyRJkqRt3WZDW1WtrarPtsffBj4PLAaOAs5si50JHN0eHwWsrKr7quo24Fbg8FlutyRJkiRtF0a6pi3JEuAXgMuAhVW1FrpgB+zbFlsM3DGw2po2TZIkSZI0olTVzBZMFgAXA6dU1fuTfLOq9hiYf09V7ZnkdODSqjqrTX8HcH5VnT1leycAJwAsXLjwsJUrV87KAc2m675y78T2vXBXuOt7k9n3IYt3n8yOZ8G6detYsGDBpJsx71i30Vmz8Vi38Vi30Vmz8Vi38Vi30VmzjS1fvvyqqlo6bN6OM9lAkp2As4F3V9X72+S7kiyqqrVJFgF3t+lrgP0HVt8PuHPqNqvqDOAMgKVLl9ayZctm0pStasVJ501s3ycesp7TrpvRj2fW3X7csonsdzasXr2aPr6W+s66jc6ajce6jce6jc6ajce6jce6jc6ajWYmd48M8A7g81X1VwOzzgWOb4+PBz44MP3YJDsnOQA4ELh89posSZIkSduPmXTlPA54AXBdkqvbtD8GTgVWJXkR8GXgGICquiHJKuBGujtPvqSq7p/thkuSJEnS9mCzoa2qPglkmtlHTrPOKcApW9AuSZIkSRIj3j1SkiRJkrR1GdokSZIkqccMbZIkSZLUY4Y2SZIkSeoxQ5skSZIk9ZihTZIkSZJ6zNAmSZIkST1maJMkSZKkHjO0SZIkSVKPGdokSZIkqccMbZIkSZLUY4Y2SZIkSeoxQ5skSZIk9ZihTZIkSZJ6zNAmSZIkST1maJMkSZKkHjO0SZIkSVKPGdokSZIkqccMbZIkSZLUY4Y2SZIkSeoxQ5skSZIk9ZihTZIkSZJ6zNAmSZIkST1maJMkSZKkHjO0SZIkSVKPGdokSZIkqccMbZIkSZLUY4Y2SZIkSeoxQ5skSZIk9ZihTZIkSZJ6zNAmSZIkST1maJMkSZKkHjO0SZIkSVKPGdokSZIkqccMbZIkSZLUY4Y2SZIkSeoxQ5skSZIk9ZihTZIkSZJ6zNAmSZIkST1maJMkSZKkHjO0SZIkSVKPGdokSZIkqccMbZIkSZLUY4Y2SZIkSeqxzYa2JO9McneS6wem7ZXkwiS3tO97Dsw7OcmtSW5O8uS5argkSZIkbQ92nMEy7wLeCvzzwLSTgIuq6tQkJ7Xnr0lyEHAscDDwMOCjSX6mqu6f3Warz5acdN7E9n3iIetZMaH9337q0yeyX0mSJG3bNtvTVlWXAN+YMvko4Mz2+Ezg6IHpK6vqvqq6DbgVOHx2mipJkiRJ259xr2lbWFVrAdr3fdv0xcAdA8utadMkSZIkSWNIVW1+oWQJ8KGqelR7/s2q2mNg/j1VtWeS04FLq+qsNv0dwPlVdfaQbZ4AnACwcOHCw1auXDkLhzO7rvvKvRPb98Jd4a7vTWbfhyzefYvWt27zz7p161iwYMGkmzGvWLPxWLfxWLfRWbPxWLfxWLfRWbONLV++/KqqWjps3kyuaRvmriSLqmptkkXA3W36GmD/geX2A+4ctoGqOgM4A2Dp0qW1bNmyMZsydyZ1bRR012addt24P54tc/txy7Zofes2/6xevZo+/g72mTUbj3Ubj3UbnTUbj3Ubj3UbnTUbzbjDI88Fjm+Pjwc+ODD92CQ7JzkAOBC4fMuaKEmSJEnbr812SSR5L7AM2DvJGuBPgVOBVUleBHwZOAagqm5Isgq4EVgPvMQ7R0qSJEnS+DYb2qrqedPMOnKa5U8BTtmSRknbI/9VgiRJkoYZd3ikJEmSJGkrMLRJkiRJUo9N5jZ7kjQLHFIqSZK2B/a0SZIkSVKPGdokSZIkqcccHilJ2xmHlUqSNL/Y0yZJkiRJPWZokyRJkqQeM7RJkiRJUo8Z2iRJkiSpxwxtkiRJktRjhjZJkiRJ6jFDmyRJkiT1mKFNkiRJknrM0CZJkiRJPWZokyRJkqQeM7RJkiRJUo8Z2iRJkiSpx3acdAMkSZoPlpx03sT2feIh61kxof3ffurTJ7JfSdJ/s6dNkiRJknrM0CZJkiRJPebwSEmSNCccUipJs8OeNkmSJEnqMUObJEmSJPWYoU2SJEmSesxr2iRJknrEawElTWVPmyRJkiT1mKFNkiRJknrM0CZJkiRJPWZokyRJkqQeM7RJkiRJUo8Z2iRJkiSpxwxtkiRJktRjhjZJkiRJ6jFDmyRJkiT1mKFNkiRJknrM0CZJkiRJPbbjpBsgSZIkbaklJ503sX2feMh6Vkxo/7ef+vSJ7Fdbl6FNkiRJ2g4ZdOcPh0dKkiRJUo8Z2iRJkiSpxwxtkiRJktRjhjZJkiRJ6jFDmyRJkiT12JyFtiRPSXJzkluTnDRX+5EkSZKkbdmchLYkOwCnA08FDgKel+SgudiXJEmSJG3L5qqn7XDg1qr6YlX9F7ASOGqO9iVJkiRJ26y5Cm2LgTsGnq9p0yRJkiRJI0hVzf5Gk2OAJ1fV77TnLwAOr6qXDixzAnBCe/pI4OZZb8j8tjfwtUk3Yh6ybuOxbqOzZuOxbuOxbqOzZuOxbuOxbqOzZht7eFXtM2zGjnO0wzXA/gPP9wPuHFygqs4Azpij/c97Sa6sqqWTbsd8Y93GY91GZ83GY93GY91GZ83GY93GY91GZ81GM1fDI68ADkxyQJKfAI4Fzp2jfUmSJEnSNmtOetqqan2SPwA+AuwAvLOqbpiLfUmSJEnStmyuhkdSVecD58/V9rcDDh0dj3Ubj3UbnTUbj3Ubj3UbnTUbj3Ubj3UbnTUbwZzciESSJEmSNDvm6po2SZIkSdIsMLRp3kny6RGX3yfJZUk+l+SIuWqXtj1JVifZ6M5WSVYkeWt7/K4kz5ml/d2eZO/Z2FbfJTk6yUEDz1ckedgk27QlRj0vzdI+d07y0SRXJ3lukiOS3NCe77q12zNfbWe/d1v9dTpfzdV5fmD7S5O8ZTa3qW2boU3zTlX98oirHAncVFW/UFWfmIs2aduTZIdJt2EbdzRw0MDzFcC8DW1jnJdmwy8AO1XVoVX1PuA44P+059+bQHvUcxN6nWqIqrqyql426XZo/jC09USS5ye5vH1C+g9JdkjylCSfTXJNkosm3ca+SLKufV+W5OIkq5L8R5JTkxzX6nhdkkckORR4M/C07e3T5yR/lOT69vWKJEuS3JTkzCTXJvnXJA9syx7WanlVko8kWdSm/3T7JP+a9lp8xGSPamaSvDrJy9rjv07ysfb4yCRnJXlee41cn+RNA+utS/JnSS4DfmnKNl/YXmcXA4+bsssnJvlEm/+MtvySNu2z7euX2/RlrQfvX9vP491JMmVfuyb5cJLfnfXizKEkf9KO6cIk703yyiS/m+SK9ho6O8kDWy2eCfxl+718DbAUePd8/T3dcF5qj1/dXl/XJDm1TTs0yWfa7945SfacZjsbnfeT7JXkA23dzyT5+ST7AmcBh7aa/R7wm8Drkry7rfeqVvtrk7xhYB8bvd/MXWVm39RzW5s29JimqedDklyQbvTFPwCZfm/blhm8Tjc656fzl63e1yV57uSOYPa036mr0vVOn9CmjXSeH7LNjc53bfoxrX7XJLmkTVuW5EPt8eFJPt1ek59O8si5O/LJSPeeeP3A81cmeX2698M3td/f/4gjoqZXVX5N+Av4OeDf6D4xBXgbcDxwB3BAm7bXpNvZly9gXfu+DPgmsAjYGfgK8IY27+XA37THK4C3TrrdW7lGhwHXAQ8CFgA30H0qX8Dj2jLvBF4J7AR8GtinTX8u3b/pALgMeFZ7vAvwwEkf2wyP/7HAv7THnwAub8f5p+3ry8A+dHfQ/RhwdFu2gN8c2M5qujCxaGCdnwA+teE1BbwL+DDdh2AHAms21ArYpS1zIHDlwOv2XmC/ts6lwOPbvNuBJcBHgd+adB1HrPlS4GpgV2A34Jb2+nrIwDJvBF46ULfnTK31pI9jC45/w3npqe336YHt+V7t+7XAr7THf7bh/DRlG/sw5LwP/B3wp+3xrwJXD7yWPjSw/o9qCjyJ7s5saa+zDwFPYPj7zbx5rTH83PZLw45pE/V8C/C69vjpdL/3e0/62HryOt3onA88G7iQ7l84LaQ7Fy6a9LHMQi02HPOuwPXAYkY8zw/Z5nTnu+uAxe3xHu37j35/gQcDO7bHTwTOnnR95qDeS4DrB56/Eng93bn/tDbtacBHJ93Wvn7N2S3/NZIj6d6Irkj3gfuuwGOAS6rqNoCq+sbkmtdrV1TVWoAkXwAuaNOvA5ZPrFWT93jgnKr6DkCS9wNHAHdU1afaMmcBL6N7I3oUcGF7/e0ArE2yG92bzDkAVfX9rXsIW+Qq4LB2DPcBn6ULFUfQ/XG3uqq+CtB6JZ4AfAC4Hzh7yPYeM2Wd9wE/MzB/VVX9ELglyReBnwVuA96arrf3/inLX15Va9q2rqZ7M/tkm/dB4M1V9e7xD38iHg98sNqwvCT/1qY/KskbgT3o/sj+yGSat9U8EfinqvoudOfuJLvT/aF2cVvmTOBfhqz7WIaf9x9P94czVfWx1lO0+2ba8aT29bn2fAHdH5s/z8bvN3ePfJSTM+zc9osMP6bp6vkE4DfatPOS3LNVj6Afhr1Oh57zkzweeG9V3Q/c1XqhfhE4dzJNnzUvS/Ks9nh/4AWMfp6/eso2pzvffQp4V5JVwPuHtGV34MwkB9J9iLDTFh7bfLOhJlfRvR9qCENbPwQ4s6pO/tGE5Jl0Q120afcNPP7hwPMfsn2/vqcb7jP1f3xUW/aGqpo6HPDBc9GwraGqfpDkduCFdJ8mX0sX4h9B90nqYdOs+v32h8nQzW5ql0Oe/yFwF/Bouk9nB0Pv4Ov2fn78tfop4KlJ3lPto8d5YrrX3LvoejKvSbKC7tPlbVnY9GvlvxfshvBd1Z6eC1w5zbrDaru5fQT4i6r6hyn7fClT3m/mmWG12Og9FH70PjpdnebT79ZcGPY6ne53eJsbPppkGV1w/aWq+m6S1cBNdD3R09noPJ/kn+hGsdxZVU9jmvNdVb04yWPoenavbh/mDfpz4ONV9awkS+h6n7Y16/nxy7J2GXi84T1x6vuhBnhNWz9cBDwn3TUKJNkLuAb4lSQHDEyTZuoS4Oh01w89CHgW3TDBn0yyIZw9j65352Zgnw3Tk+yU5OCq+hawJsnRbfrOG8bnzxOX0A2/uITu2F9M96noZ+h+t/ZufzQ/D7h4uo00lwHLWg/HTsAxU+Yfk+QB6a75+ym6mu4OrG2fzL6ArgdzJl4HfJ1uiNd88kng15PskmQB3R8n0A2VXNvqdtzA8t9u86Z7Pl9dAPz2wLUse1XVvcA9A9dqvAC4uKrur+6mIYdW1evohsoOO+9fQqtd+2Pza+33c1M+0tqxoK23uL3HbPR+k+Ths3LkW8ewc9uVDD+mmdTzqcDQ6wu3ccNep9Od8y8BnpvuWvt96HoqL59Qu2fL7sA9LbD9LF2v7K6MeJ6vqhe239+ntWWGnu+SPKKqLmu/51+j69mb2p6vtMcrZvE4++QuYN9W352BodcFanqGth6oqhuB1wIXJLmWbuz4IuAE4P1JrgHeN8Emap6pqs/SfeJ3OV3g+EfgHuDzwPHtdbYX8H+r6r+A5wBvaq+1q4ENdxh7Ad0QkmvpeqweuhUPY0t9gu736NKquouup+sTbTjtycDH6T4c+WxVfXBTG2rrvJ7uj8CP0g23HHQzXfD7d+DFbVjR2+hq/Rm6ITbfGaHtrwB2SfLmEdaZqKq6gq636Bq6oS5X0l279yd0r8EL6T7J3mAl8Kp24f0j6F6vf595eiOSDarqw7Reszb09ZVt1vF0N165FjiU7rq2qet+leHn/dcDS9u6p7Ztba4dFwDvAS5Nch3wr8Bum3i/mReGndvakO+NjmkT9XwD8IQkn6UbQvrlrXsUk7eJ1+mwc/45dKMVrqG7BvjVVfWfW73Rs+vDwI7tOP+c7sO8cc7zU013vvvLtJtf0YXga6as92bgL5J8ipl/wDevVNUP6M57l9FdY3vTptfQVJlfo28kjasNufhQVT1q0m3RtinJgqpaN/Dp/Antj2xJkrQFHDcqSZotZ6T7h9m70F1jZGCTJGkW2NMmSZIkST3mNW2SJEmS1GOGNkmSJEnqMUObJEmSJPWYoU2SJEmSeszQJkmSJEk9ZmiTJEmSpB77/+NMvrhfVMz3AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "most_freq_orgs = {}\n",
    "for o in orgs:\n",
    "    most_freq_orgs[o] = dataframe.ORGS.str.contains(o).sum()\n",
    "\n",
    "del most_freq_orgs['']\n",
    "orgs_map_top_10 = dict(sorted(most_freq_orgs.items(), key=lambda x: -x[1])[0:10]) #sortiramo od najčešćeg topica prema manje čestima\n",
    "\n",
    "plt.figure(figsize=(15, 5))\n",
    "plt.bar(orgs_map_top_10.keys(), orgs_map_top_10.values())\n",
    "plt.title(\"10 ukupno najčešćih orgs-a\")\n",
    "plt.grid()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Dodavanje stupca"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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>TOPICS</th>\n",
       "      <th>PLACES</th>\n",
       "      <th>PEOPLE</th>\n",
       "      <th>ORGS</th>\n",
       "      <th>EXCHANGES</th>\n",
       "      <th>TITLE</th>\n",
       "      <th>DATELINE</th>\n",
       "      <th>TOPICS_ENUM</th>\n",
       "      <th>FILENAME</th>\n",
       "      <th>DATE</th>\n",
       "      <th>BODY</th>\n",
       "      <th>TEXT TYPE</th>\n",
       "      <th>LEWISSPLIT</th>\n",
       "      <th>CGISPLIT</th>\n",
       "      <th>WORD_COUNT</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>['cocoa']</td>\n",
       "      <td>['el-salvador', 'usa', 'uruguay']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>BAHIA COCOA REVIEW</td>\n",
       "      <td>SALVADOR, Feb 26 -</td>\n",
       "      <td>2</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:01:01.79</td>\n",
       "      <td>Showers continued throughout the week in\n",
       "the B...</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2864</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>STANDARD OIL &amp;lt;SRD&gt; TO FORM FINANCIAL UNIT</td>\n",
       "      <td>CLEVELAND, Feb 26 -</td>\n",
       "      <td>1</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:02:20.00</td>\n",
       "      <td>Standard Oil Co and BP North America\n",
       "Inc said ...</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>445</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>TEXAS COMMERCE BANCSHARES &amp;lt;TCB&gt; FILES PLAN</td>\n",
       "      <td>HOUSTON, Feb 26 -</td>\n",
       "      <td>1</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:03:27.51</td>\n",
       "      <td>Texas Commerce Bancshares Inc's Texas\n",
       "Commerce...</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>334</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>['usa', 'brazil']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>TALKING POINT/BANKAMERICA &amp;lt;BAC&gt; EQUITY OFFER</td>\n",
       "      <td>LOS ANGELES, Feb 26 -</td>\n",
       "      <td>1</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:07:13.72</td>\n",
       "      <td>BankAmerica Corp is not under\n",
       "pressure to act ...</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2853</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>['grain', 'wheat', 'corn', 'barley', 'oat', 's...</td>\n",
       "      <td>['usa']</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>&lt;NA&gt;</td>\n",
       "      <td>NATIONAL AVERAGE PRICES FOR FARMER-OWNED RESERVE</td>\n",
       "      <td>WASHINGTON, Feb 26 -</td>\n",
       "      <td>2</td>\n",
       "      <td>reut2-000.sgm</td>\n",
       "      <td>26-FEB-1987 15:10:44.60</td>\n",
       "      <td>The U.S. Agriculture Department\n",
       "reported the f...</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1145</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                              TOPICS  \\\n",
       "0                                          ['cocoa']   \n",
       "1                                               <NA>   \n",
       "2                                               <NA>   \n",
       "3                                               <NA>   \n",
       "4  ['grain', 'wheat', 'corn', 'barley', 'oat', 's...   \n",
       "\n",
       "                              PLACES PEOPLE  ORGS EXCHANGES  \\\n",
       "0  ['el-salvador', 'usa', 'uruguay']   <NA>  <NA>      <NA>   \n",
       "1                            ['usa']   <NA>  <NA>      <NA>   \n",
       "2                            ['usa']   <NA>  <NA>      <NA>   \n",
       "3                  ['usa', 'brazil']   <NA>  <NA>      <NA>   \n",
       "4                            ['usa']   <NA>  <NA>      <NA>   \n",
       "\n",
       "                                              TITLE  \\\n",
       "0                                BAHIA COCOA REVIEW   \n",
       "1      STANDARD OIL &lt;SRD> TO FORM FINANCIAL UNIT   \n",
       "2     TEXAS COMMERCE BANCSHARES &lt;TCB> FILES PLAN   \n",
       "3   TALKING POINT/BANKAMERICA &lt;BAC> EQUITY OFFER   \n",
       "4  NATIONAL AVERAGE PRICES FOR FARMER-OWNED RESERVE   \n",
       "\n",
       "                     DATELINE  TOPICS_ENUM       FILENAME  \\\n",
       "0         SALVADOR, Feb 26 -             2  reut2-000.sgm   \n",
       "1        CLEVELAND, Feb 26 -             1  reut2-000.sgm   \n",
       "2          HOUSTON, Feb 26 -             1  reut2-000.sgm   \n",
       "3      LOS ANGELES, Feb 26 -             1  reut2-000.sgm   \n",
       "4       WASHINGTON, Feb 26 -             2  reut2-000.sgm   \n",
       "\n",
       "                      DATE                                               BODY  \\\n",
       "0  26-FEB-1987 15:01:01.79  Showers continued throughout the week in\n",
       "the B...   \n",
       "1  26-FEB-1987 15:02:20.00  Standard Oil Co and BP North America\n",
       "Inc said ...   \n",
       "2  26-FEB-1987 15:03:27.51  Texas Commerce Bancshares Inc's Texas\n",
       "Commerce...   \n",
       "3  26-FEB-1987 15:07:13.72  BankAmerica Corp is not under\n",
       "pressure to act ...   \n",
       "4  26-FEB-1987 15:10:44.60  The U.S. Agriculture Department\n",
       "reported the f...   \n",
       "\n",
       "   TEXT TYPE  LEWISSPLIT  CGISPLIT  WORD_COUNT  \n",
       "0          0           2         1        2864  \n",
       "1          0           2         1         445  \n",
       "2          0           2         1         334  \n",
       "3          0           2         1        2853  \n",
       "4          0           2         1        1145  "
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "podaci_added = podaci.copy()\n",
    "podaci_added['WORD_COUNT'] = podaci_added.apply(lambda row: len(row.BODY), axis=1)\n",
    "podaci_added.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Histogram"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "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": [
    "interval = (podaci_added.WORD_COUNT.max() - podaci_added.WORD_COUNT.min())/10\n",
    "bins = [podaci_added.WORD_COUNT.min()]\n",
    "xticks = [podaci_added.WORD_COUNT.min()]\n",
    "names = []\n",
    "for i in range(1, 10):\n",
    "    bins.append(int(podaci_added.WORD_COUNT.min()+interval*i))\n",
    "    xticks.append(int(podaci_added.WORD_COUNT.min()+interval*i))\n",
    "    names.append(\" \" + str(int(podaci_added.WORD_COUNT.min() + (i-1)*interval)) + \" - \" + str(int(podaci_added.WORD_COUNT.min() + i*interval)))\n",
    "\n",
    "podaci_added['WORD_COUNT_RANGE'] = pd.cut(podaci_added['WORD_COUNT'], bins, labels=names)\n",
    "\n",
    "plt.hist(podaci_added.WORD_COUNT, color='blue', alpha=0.5)\n",
    "plt.xlabel(\"Broj riječi\")\n",
    "plt.ylabel(\"Broj članaka\")\n",
    "plt.xticks(xticks)\n",
    "plt.xlim(xmax = 7000)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Kutijasti graf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "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": [
    "plt.boxplot(podaci_added.WORD_COUNT) # kružići su outliersi, whiskers su min i max vrijednosti, a narančasta crta je srednja vrijednost\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Podaci o riječima u tijelima članaka"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[nltk_data] Downloading package stopwords to\n",
      "[nltk_data]     C:\\Users\\MajaJuric\\AppData\\Roaming\\nltk_data...\n",
      "[nltk_data]   Unzipping corpora\\stopwords.zip.\n",
      "[nltk_data] Downloading package punkt to\n",
      "[nltk_data]     C:\\Users\\MajaJuric\\AppData\\Roaming\\nltk_data...\n",
      "[nltk_data]   Unzipping tokenizers\\punkt.zip.\n"
     ]
    }
   ],
   "source": [
    "#!pip install nltk\n",
    "import nltk as nltk\n",
    "nltk.download('stopwords')\n",
    "nltk.download('punkt')\n",
    "from nltk.corpus import stopwords\n",
    "from nltk.tokenize import word_tokenize"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataframe = podaci.loc[podaci.BODY.notnull(), :]\n",
    "\n",
    "dataframe['BODY'] = dataframe['BODY'].replace('[\\n\\t0-9]',' ', regex=True) # uklanjamo sve prijelaze u novi red, tabulatore i brojeve iz tijela članaka\n",
    "\n",
    "dataframe['BODY'] = dataframe['BODY'].replace('[\\.\\-\"\\#$%&!/,;]',' ', regex=True) #\n",
    "\n",
    "dataframe['BODY'] = dataframe['BODY'].str.lower()\n",
    "\n",
    "dataframe['BODY'] = dataframe['BODY'].replace('[\\s]', ' ', regex=True)\n",
    "\n",
    "dataframe['BODY'] = dataframe['BODY'].replace('[^a-zA-Z0-9\\n \\.]', '', regex=True)\n",
    "\n",
    "#dataframe['BODY'] = dataframe['BODY'].replace('[^\\x00-\\x7f]', '', regex=True)\n",
    "\n",
    "#dataframe['BODY'] = dataframe['BODY'].encode()\n",
    "\n",
    "from collections import Counter\n",
    "count = Counter()\n",
    "\n",
    "exclude = set(stopwords.words('english'))\n",
    "#print(exclude)\n",
    "bodies=[]\n",
    "topics=[]\n",
    "\n",
    "dataframe = dataframe.loc[dataframe.TOPICS.notnull(), :]\n",
    "\n",
    "for index, row in dataframe.iterrows():\n",
    "    body = \"\"\n",
    "    words_per_row = word_tokenize(row['BODY'])\n",
    "    for w in words_per_row:\n",
    "        if w not in exclude:\n",
    "            body += w\n",
    "            body += \" \"\n",
    "    for w in body.split(\" \"):\n",
    "        count[w] += 1\n",
    "    bodies.append(body)\n",
    "    topics.append(row.TOPICS)\n",
    "\n",
    "d = {'TOPICS': pd.Series(topics), 'BODY': pd.Series(bodies)}\n",
    "dataframe = pd.DataFrame(d)\n",
    "\n",
    "dataframe.to_csv(\"my_data/data/clanci_stripped.csv\")\n",
    "\n",
    "plt.figure(figsize=(15, 5))\n",
    "plt.title(\"10 najčešćih riječi\")\n",
    "for item in count.most_common(10):\n",
    "    plt.bar(item[0], item[1])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}