gtd_analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": false
   },
   "source": [
    "## Global Terrorism Database Analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Model which predicts responsible groups for terrorist attacks using historical data.\n",
    "\n",
    "The open-source database(GTD) has been maintained at : http://www.start.umd.edu/gtd/ by National Consortium for the\n",
    "Study of Terrorism And Responses of Terrorism (START). \n",
    "\n",
    "The database includes information about all the terrorist attacks from 1970 to 2014."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Description"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The database details information about all the terrorist attacks all over the world from 1970 to 2014 including information about description of the attack, terrorist groups invloved, weapons used, attack type etc.for each terrorist attack. \n",
    "\n",
    "However certain incidents have not been attributed to any particular terrorist group. This model attempts to predict the terorist group responsible for such attacks.\n",
    "\n",
    "Information about the various features of the database has been detailed at: https://www.start.umd.edu/gtd/downloads/Codebook.pdf"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The dataset has been split into two groups - one for training and evaluating the model and the other(test dataset) - to run the model on.\n",
    "\n",
    "The training dataset has information about the terrorist group involved for the attack - the model trained on this attack will be used to generate predictions for the test dataset."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Setup Imports and Variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import pandas as pd\n",
    "import matplotlib\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "sns.set(style=\"whitegrid\", color_codes=True)\n",
    "np.random.seed(sum(map(ord, \"categorical\")))\n",
    "matplotlib.style.use('ggplot')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Explore the Data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Read the data: <b>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The database has been downloaded as a csv file and is loaded into a Pandas DataFrame."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Keshav\\Anaconda\\lib\\site-packages\\IPython\\core\\interactiveshell.py:2902: DtypeWarning: Columns (4,6,13,31,33,53,61,62,63,75,78,91,93,95,113,114,120) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "  interactivity=interactivity, compiler=compiler, result=result)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>eventid</th>\n",
       "      <th>iyear</th>\n",
       "      <th>imonth</th>\n",
       "      <th>iday</th>\n",
       "      <th>approxdate</th>\n",
       "      <th>extended</th>\n",
       "      <th>resolution</th>\n",
       "      <th>country</th>\n",
       "      <th>country_txt</th>\n",
       "      <th>region</th>\n",
       "      <th>...</th>\n",
       "      <th>addnotes</th>\n",
       "      <th>scite1</th>\n",
       "      <th>scite2</th>\n",
       "      <th>scite3</th>\n",
       "      <th>dbsource</th>\n",
       "      <th>INT_LOG</th>\n",
       "      <th>INT_IDEO</th>\n",
       "      <th>INT_MISC</th>\n",
       "      <th>INT_ANY</th>\n",
       "      <th>related</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>197000000001</td>\n",
       "      <td>1970</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>58</td>\n",
       "      <td>Dominican Republic</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>PGIS</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>197000000002</td>\n",
       "      <td>1970</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>130</td>\n",
       "      <td>Mexico</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>PGIS</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>197001000001</td>\n",
       "      <td>1970</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>160</td>\n",
       "      <td>Philippines</td>\n",
       "      <td>5</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>PGIS</td>\n",
       "      <td>-9</td>\n",
       "      <td>-9</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>197001000002</td>\n",
       "      <td>1970</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>78</td>\n",
       "      <td>Greece</td>\n",
       "      <td>8</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>PGIS</td>\n",
       "      <td>-9</td>\n",
       "      <td>-9</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>197001000003</td>\n",
       "      <td>1970</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>101</td>\n",
       "      <td>Japan</td>\n",
       "      <td>4</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>PGIS</td>\n",
       "      <td>-9</td>\n",
       "      <td>-9</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 134 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        eventid  iyear  imonth  iday approxdate  extended resolution  country  \\\n",
       "0  197000000001   1970       0     0        NaN         0        NaN       58   \n",
       "1  197000000002   1970       0     0        NaN         0        NaN      130   \n",
       "2  197001000001   1970       1     0        NaN         0        NaN      160   \n",
       "3  197001000002   1970       1     0        NaN         0        NaN       78   \n",
       "4  197001000003   1970       1     0        NaN         0        NaN      101   \n",
       "\n",
       "          country_txt  region   ...    addnotes scite1 scite2 scite3  \\\n",
       "0  Dominican Republic       2   ...         NaN    NaN    NaN    NaN   \n",
       "1              Mexico       1   ...         NaN    NaN    NaN    NaN   \n",
       "2         Philippines       5   ...         NaN    NaN    NaN    NaN   \n",
       "3              Greece       8   ...         NaN    NaN    NaN    NaN   \n",
       "4               Japan       4   ...         NaN    NaN    NaN    NaN   \n",
       "\n",
       "   dbsource  INT_LOG  INT_IDEO INT_MISC INT_ANY  related  \n",
       "0      PGIS        0         0        0       0      NaN  \n",
       "1      PGIS        0         1        1       1      NaN  \n",
       "2      PGIS       -9        -9        1       1      NaN  \n",
       "3      PGIS       -9        -9        1       1      NaN  \n",
       "4      PGIS       -9        -9        1       1      NaN  \n",
       "\n",
       "[5 rows x 134 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_gtd = pd.read_csv('gtd_0615dist/globalterrorismdb_0615dist.csv')\n",
    "df_gtd.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Information about the DataFrame: <b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "141966"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df_gtd)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are over 141k terrorist attacks that has been recorded in this database."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "134"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df_gtd.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A total of 134 attributes for each attack has been recorded."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 141966 entries, 0 to 141965\n",
      "Columns: 134 entries, eventid to related\n",
      "dtypes: float64(52), int64(23), object(59)\n",
      "memory usage: 146.2+ MB\n"
     ]
    }
   ],
   "source": [
    "df_gtd.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Many of the column datatypes are of type object (52)  which have be to changed to number representations as models usually don't work very well on string objects."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Descriptive Statistics about the dataframe:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>eventid</th>\n",
       "      <th>iyear</th>\n",
       "      <th>imonth</th>\n",
       "      <th>iday</th>\n",
       "      <th>extended</th>\n",
       "      <th>country</th>\n",
       "      <th>region</th>\n",
       "      <th>longitude</th>\n",
       "      <th>specificity</th>\n",
       "      <th>vicinity</th>\n",
       "      <th>...</th>\n",
       "      <th>ransomamt</th>\n",
       "      <th>ransomamtus</th>\n",
       "      <th>ransompaid</th>\n",
       "      <th>ransompaidus</th>\n",
       "      <th>hostkidoutcome</th>\n",
       "      <th>nreleased</th>\n",
       "      <th>INT_LOG</th>\n",
       "      <th>INT_IDEO</th>\n",
       "      <th>INT_MISC</th>\n",
       "      <th>INT_ANY</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1.419660e+05</td>\n",
       "      <td>141966.000000</td>\n",
       "      <td>141966.000000</td>\n",
       "      <td>141966.000000</td>\n",
       "      <td>141966.000000</td>\n",
       "      <td>141966.000000</td>\n",
       "      <td>141966.000000</td>\n",
       "      <td>115631.000000</td>\n",
       "      <td>118632.000000</td>\n",
       "      <td>141966.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>1.119000e+03</td>\n",
       "      <td>3.380000e+02</td>\n",
       "      <td>548.000000</td>\n",
       "      <td>329.000000</td>\n",
       "      <td>7347.000000</td>\n",
       "      <td>6757.000000</td>\n",
       "      <td>141966.000000</td>\n",
       "      <td>141966.000000</td>\n",
       "      <td>141966.000000</td>\n",
       "      <td>141966.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>1.999023e+11</td>\n",
       "      <td>1998.955982</td>\n",
       "      <td>6.503684</td>\n",
       "      <td>15.425581</td>\n",
       "      <td>0.038467</td>\n",
       "      <td>134.094973</td>\n",
       "      <td>6.829713</td>\n",
       "      <td>14.858269</td>\n",
       "      <td>1.438583</td>\n",
       "      <td>0.068685</td>\n",
       "      <td>...</td>\n",
       "      <td>3.475641e+06</td>\n",
       "      <td>6.632483e+05</td>\n",
       "      <td>488555.027482</td>\n",
       "      <td>372.914894</td>\n",
       "      <td>4.551382</td>\n",
       "      <td>-21.153914</td>\n",
       "      <td>-4.881134</td>\n",
       "      <td>-4.837848</td>\n",
       "      <td>0.097678</td>\n",
       "      <td>-4.231471</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.272072e+09</td>\n",
       "      <td>12.719985</td>\n",
       "      <td>3.383358</td>\n",
       "      <td>8.816347</td>\n",
       "      <td>0.192321</td>\n",
       "      <td>115.506200</td>\n",
       "      <td>2.996252</td>\n",
       "      <td>63.854924</td>\n",
       "      <td>1.021746</td>\n",
       "      <td>0.326257</td>\n",
       "      <td>...</td>\n",
       "      <td>3.291150e+07</td>\n",
       "      <td>7.347255e+06</td>\n",
       "      <td>2756693.467399</td>\n",
       "      <td>3765.982637</td>\n",
       "      <td>2.055365</td>\n",
       "      <td>90.021073</td>\n",
       "      <td>4.525994</td>\n",
       "      <td>4.585612</td>\n",
       "      <td>0.629478</td>\n",
       "      <td>4.693714</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.970000e+11</td>\n",
       "      <td>1970.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-176.176447</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-9.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>-9.900000e+01</td>\n",
       "      <td>-9.900000e+01</td>\n",
       "      <td>-99.000000</td>\n",
       "      <td>-99.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-99.000000</td>\n",
       "      <td>-9.000000</td>\n",
       "      <td>-9.000000</td>\n",
       "      <td>-9.000000</td>\n",
       "      <td>-9.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.988090e+11</td>\n",
       "      <td>1988.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>69.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>-61.516667</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>-99.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>-99.000000</td>\n",
       "      <td>-9.000000</td>\n",
       "      <td>-9.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-9.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>2.000092e+11</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>101.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>36.420664</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>1.500000e+04</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-9.000000</td>\n",
       "      <td>-9.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>2.012032e+11</td>\n",
       "      <td>2012.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>23.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>160.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>65.665857</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>5.000000e+05</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>12000.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>2.015022e+11</td>\n",
       "      <td>2014.000000</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>31.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1004.000000</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>179.366667</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>1.000000e+09</td>\n",
       "      <td>1.320000e+08</td>\n",
       "      <td>41000000.000000</td>\n",
       "      <td>48000.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>1201.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 75 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            eventid          iyear         imonth           iday  \\\n",
       "count  1.419660e+05  141966.000000  141966.000000  141966.000000   \n",
       "mean   1.999023e+11    1998.955982       6.503684      15.425581   \n",
       "std    1.272072e+09      12.719985       3.383358       8.816347   \n",
       "min    1.970000e+11    1970.000000       0.000000       0.000000   \n",
       "25%    1.988090e+11    1988.000000       4.000000       8.000000   \n",
       "50%    2.000092e+11    2000.000000       7.000000      15.000000   \n",
       "75%    2.012032e+11    2012.000000       9.000000      23.000000   \n",
       "max    2.015022e+11    2014.000000      12.000000      31.000000   \n",
       "\n",
       "            extended        country         region      longitude  \\\n",
       "count  141966.000000  141966.000000  141966.000000  115631.000000   \n",
       "mean        0.038467     134.094973       6.829713      14.858269   \n",
       "std         0.192321     115.506200       2.996252      63.854924   \n",
       "min         0.000000       4.000000       1.000000    -176.176447   \n",
       "25%         0.000000      69.000000       5.000000     -61.516667   \n",
       "50%         0.000000     101.000000       6.000000      36.420664   \n",
       "75%         0.000000     160.000000      10.000000      65.665857   \n",
       "max         1.000000    1004.000000      12.000000     179.366667   \n",
       "\n",
       "         specificity       vicinity      ...           ransomamt  \\\n",
       "count  118632.000000  141966.000000      ...        1.119000e+03   \n",
       "mean        1.438583       0.068685      ...        3.475641e+06   \n",
       "std         1.021746       0.326257      ...        3.291150e+07   \n",
       "min         1.000000      -9.000000      ...       -9.900000e+01   \n",
       "25%         1.000000       0.000000      ...        0.000000e+00   \n",
       "50%         1.000000       0.000000      ...        1.500000e+04   \n",
       "75%         1.000000       0.000000      ...        5.000000e+05   \n",
       "max         5.000000       1.000000      ...        1.000000e+09   \n",
       "\n",
       "        ransomamtus       ransompaid  ransompaidus  hostkidoutcome  \\\n",
       "count  3.380000e+02       548.000000    329.000000     7347.000000   \n",
       "mean   6.632483e+05    488555.027482    372.914894        4.551382   \n",
       "std    7.347255e+06   2756693.467399   3765.982637        2.055365   \n",
       "min   -9.900000e+01       -99.000000    -99.000000        1.000000   \n",
       "25%    0.000000e+00       -99.000000      0.000000        2.000000   \n",
       "50%    0.000000e+00         0.000000      0.000000        4.000000   \n",
       "75%    0.000000e+00     12000.000000      0.000000        7.000000   \n",
       "max    1.320000e+08  41000000.000000  48000.000000        7.000000   \n",
       "\n",
       "         nreleased        INT_LOG       INT_IDEO       INT_MISC        INT_ANY  \n",
       "count  6757.000000  141966.000000  141966.000000  141966.000000  141966.000000  \n",
       "mean    -21.153914      -4.881134      -4.837848       0.097678      -4.231471  \n",
       "std      90.021073       4.525994       4.585612       0.629478       4.693714  \n",
       "min     -99.000000      -9.000000      -9.000000      -9.000000      -9.000000  \n",
       "25%     -99.000000      -9.000000      -9.000000       0.000000      -9.000000  \n",
       "50%       0.000000      -9.000000      -9.000000       0.000000       0.000000  \n",
       "75%       2.000000       0.000000       0.000000       0.000000       0.000000  \n",
       "max    1201.000000       1.000000       1.000000       1.000000       1.000000  \n",
       "\n",
       "[8 rows x 75 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_gtd.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Preprocessing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "Real-life data can never be clean and optimal for machine learning. It must be preprocessed, cleaned and presented in a form suitable to train the model on. \n",
    "\n",
    "There are features which may be redundant/ noisy. Such features must be removed as they can potentially lead to <b> overfitting. </b> \n",
    "\n",
    "As per <b> Occam's Razor </b> , a model with lower complexity must always be preferred. Such models can prevent overfitting and have <b> low bias and low variance. </b>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Handling Missing Values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Many features have too many missing information about them - More than 80% of the instances for these features are missing. \n",
    "\n",
    "These columns are dropped as the model won't require them. Imputing these features will not be effective due to the high number of missing values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def remove_columns_missing_values(df, min_threshold):\n",
    "    \"\"\"\n",
    "    removes the columns with missing values below a certain threshold\n",
    "    \"\"\"\n",
    "    for col in df.columns:\n",
    "        rate = sum(df[col].notnull())/float(len(df)) * 100\n",
    "        if rate <= min_threshold:\n",
    "            df = df.drop(col,1)\n",
    "    return df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Removing features having proportion of missing values more than 80%."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_gtd = remove_columns_missing_values(df_gtd, 20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "65"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df_gtd.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "69 features have been removed as they contain a high proportion of missing values."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Removing Features of type Object - String Form"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Some features have text representations. Such features are redundant as they exist in categorical form too."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "columns_to_drop = df_gtd.columns[df_gtd.columns.map(lambda x: 'txt' in x)]\n",
    "df_gtd = df_gtd.drop(columns_to_drop,1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "provstate      object\n",
       "city           object\n",
       "latitude       object\n",
       "location       object\n",
       "summary        object\n",
       "corp1          object\n",
       "target1        object\n",
       "gname          object\n",
       "motive         object\n",
       "weapdetail     object\n",
       "propcomment    object\n",
       "scite1         object\n",
       "scite2         object\n",
       "scite3         object\n",
       "dbsource       object\n",
       "dtype: object"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_gtd.dtypes[df_gtd.dtypes.map(lambda x: x == 'object')]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Dropping other noisy/redundant 'Object' type Features."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "columns_to_drop = ['summary', 'scite1' , 'scite2' , 'scite3' , 'dbsource' , 'provstate', 'location', 'latitude', 'city', \n",
    "                   'propcomment', 'weapdetail', 'corp1', 'motive', 'target1']\n",
    "df_gtd = df_gtd.drop(columns_to_drop,1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "These features are either redundant being already represented in categorical form or simply noisy. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "42"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df_gtd.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Removing Noisy/Redundant categorical features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Features : INT_LOG, INT_IDEO, INT_MISC, INT_ANY are removed as they don't add much value to the data when required to predict the group involved."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "columns_to_drop = ['INT_LOG' , 'INT_MISC', 'INT_ANY', 'INT_IDEO']\n",
    "df_gtd = df_gtd.drop(columns_to_drop,1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Features such as Longitude and Specificity are removed as they won't be used in the model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "columns_to_drop = ['longitude','specificity']\n",
    "df_gtd = df_gtd.drop(columns_to_drop,1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Feature: 'eventid'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "'eventid' is unique to each instance of the data and hence will not be useful while training the model on. The information in this feature has already been represented in other features."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_gtd = df_gtd.drop('eventid',1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Feature: 'extended'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This binary categorical feature tells us if the attack extended over 24 hours or not."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.038466956877\n"
     ]
    }
   ],
   "source": [
    "print sum(df_gtd.extended) / float(len(df_gtd))\n",
    "df_gtd = df_gtd.drop('extended', 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since less than 4% of the data extends over 24 hours - it has been removed."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Other features with very less obvious 'feature information'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Features 'nkillus' and 'nwoundus' are also removed as they are not present for many records and they add much value. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_gtd = df_gtd.drop(['nwoundus','nkillus'], 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Features 'nperpcap', 'guncertain1', 'nwoundte', 'propextent', 'nkillter' are removed as they don't add much value."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_gtd = df_gtd.drop(['nwoundte','propextent','nkillter', 'guncertain1', 'nperpcap'], 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature Creation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Feaures: Inclusion Criteria ( crit1, crit2, crit3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>crit1</th>\n",
       "      <th>crit2</th>\n",
       "      <th>crit3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   crit1  crit2  crit3\n",
       "0      1      1      1\n",
       "1      1      1      1\n",
       "2      1      1      1\n",
       "3      1      1      1\n",
       "4      1      1      1"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_gtd[['crit1','crit2','crit3']].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "These features can be better represented if they are grouped into the following:\n",
    "    1. crit1 and crit2\n",
    "    2. crit2 and crit3\n",
    "    3. crit3 and crit1\n",
    "    4. crit1, crit2 and crit3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "group_crit12 = lambda df: 1 if df.crit1 == 1 and df.crit2 == 1 else 0\n",
    "df_gtd['crit1andcrit2'] = df_gtd.apply(group_crit12,axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "group_crit23 = lambda df: 1 if df.crit2 == 1 and df.crit3 == 1 else 0\n",
    "df_gtd['crit2andcrit3'] = df_gtd.apply(group_crit23,axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "group_crit13 = lambda df: 1 if df.crit1 == 1 and df.crit3 == 1 else 0\n",
    "df_gtd['crit1andcrit3'] = df_gtd.apply(group_crit13,axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "group_crit123 = lambda df: 1 if df.crit1 == 1 and df.crit2 == 1 and df.crit3 == 1 else 0\n",
    "df_gtd['crit1andcrit2andcrit3'] = df_gtd.apply(group_crit123,axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Dropping the original columns - 'crit1' , 'crit2' and 'crit3'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_gtd = df_gtd.drop(['crit1','crit2','crit3'],1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Imputing Missing Values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A Class DataFrameImputer is used to impute the columns of the dataframe - \n",
    "\n",
    "Columns of datatype object are replaced by the most frequent value while the missing values in the column of other datatypes are replaced by their median.\n",
    "\n",
    "Source: http://stackoverflow.com/questions/25239958/impute-categorical-missing-values-in-scikit-learn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.base import TransformerMixin\n",
    "\n",
    "class DataFrameImputer(TransformerMixin):\n",
    "    def __init__(self):\n",
    "        \"\"\"Impute missing values.\n",
    "        Columns of dtype object are imputed with the most frequent value \n",
    "        in column.\n",
    "        Columns of other types are imputed with median of column.\n",
    "        \"\"\"\n",
    "    def fit(self, X, y=None):\n",
    "\n",
    "        self.fill = pd.Series([X[c].value_counts().index[0]\n",
    "            if X[c].dtype == np.dtype('O') else X[c].median() for c in X],\n",
    "            index=X.columns)\n",
    "        return self\n",
    "    \n",
    "    def transform(self, X, y=None):\n",
    "        return X.fillna(self.fill)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_gtd_imp = DataFrameImputer().fit_transform(df_gtd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Saving these to pickle\n",
    "df_gtd.to_pickle('df_gtd.pkl')\n",
    "df_gtd_imp.to_pickle('df_gtd_imp.pkl')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The missing values in the columns are now imputed. Imputation is necessary as the scikit-learn models don't work with NaN values."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Final Data Preparation for Learning"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The class label in the dataset is 'gname'.  The model has to be trained and predict for those instances which have their gname as 'Unknown'."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Test Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_test = df_gtd_imp.query(\"gname == 'Unknown'\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_test = df_test.drop('gname',1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Training Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_train = df_gtd_imp.query(\"gname != 'Unknown'\")\n",
    "df_train_x = pd.DataFrame(df_train)\n",
    "df_train_x = df_train_x.drop('gname',1)\n",
    "df_train_y = df_train.gname"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Dimensions of the train and test data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We notice that the train data is further split into two - one with only the features and the other with the class labels. \n",
    "\n",
    "The number of columns in test data is the same as that of train data which has features - This is expected. The number of instances in both vary. It must also be noticed that these lengths are comparable. \n",
    "\n",
    "Thus we have a training and test set of similar proportions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(76304, 27)"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train_x.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(76304L,)"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train_y.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(65662, 27)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature Selection"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The data we have is cleaned and does not have any noisy/redundant features. Ranking the features gives us an idea of how relevant each feature is."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are many tests which can find out the relative importance of each feature. One such method is to train the data with a \n",
    "decision tree of Scikit-learn and find the feature importances."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Fitting the data with Decision Tree Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "clf = DecisionTreeClassifier()\n",
    "clf = clf.fit(df_train_x, df_train_y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Feature Importance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The relative feature importances (scaled and normalized to 1) are printed below"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('iyear', 0.18348871571000372)\n",
      "('imonth', 0.067977529192091588)\n",
      "('iday', 0.074042012094153759)\n",
      "('country', 0.30807879229332291)\n",
      "('region', 0.063312744239232374)\n",
      "('vicinity', 0.0042723613643706184)\n",
      "('doubtterr', 0.01891376390033524)\n",
      "('multiple', 0.011872068608749268)\n",
      "('success', 0.004768784334976634)\n",
      "('suicide', 0.00078752378440137051)\n",
      "('attacktype1', 0.018765862413359171)\n",
      "('targtype1', 0.024023426684241487)\n",
      "('targsubtype1', 0.055460267319541297)\n",
      "('natlty1', 0.020858792704342165)\n",
      "('nperps', 0.018400452529178987)\n",
      "('claimed', 0.0087261432864760916)\n",
      "('weaptype1', 0.011967126740969797)\n",
      "('weapsubtype1', 0.03087376258186772)\n",
      "('nkill', 0.025925697211306283)\n",
      "('nwound', 0.020654623692805442)\n",
      "('property', 0.012008366985784188)\n",
      "('ishostkid', 0.003791181462787039)\n",
      "('ransom', 0.002609976372511084)\n",
      "('crit1andcrit2', 0.0020971855539101024)\n",
      "('crit2andcrit3', 0.0020214752821982067)\n",
      "('crit1andcrit3', 0.0023442306644099677)\n",
      "('crit1andcrit2andcrit3', 0.0019571329926733683)\n"
     ]
    }
   ],
   "source": [
    "for index_feature, feature_rank in enumerate(list(clf.feature_importances_)):\n",
    "        print (df_train_x.columns[index_feature], feature_rank)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Features which form upper 70% percentile are selected. The rest are discarded as their relevance is too low."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('iyear', 0.18348871571000372)\n",
      "('imonth', 0.067977529192091588)\n",
      "('iday', 0.074042012094153759)\n",
      "('country', 0.30807879229332291)\n",
      "('region', 0.063312744239232374)\n",
      "('targsubtype1', 0.055460267319541297)\n",
      "('weapsubtype1', 0.03087376258186772)\n",
      "('nkill', 0.025925697211306283)\n"
     ]
    }
   ],
   "source": [
    "perc_val = np.percentile(clf.feature_importances_,70)\n",
    "index_relevant_features = list()\n",
    "for index_feature, feature_rank in enumerate(list(clf.feature_importances_)):\n",
    "        if feature_rank >= perc_val:\n",
    "            print (df_train_x.columns[index_feature], feature_rank)\n",
    "            index_relevant_features.append(index_feature)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The relevant features alone are present in the training and test data sets. \n",
    "\n",
    "We see that the number of features have been reduced to only 8. \n",
    "Thus, the features which are used to train the model are:\n",
    "\n",
    "1. Year\n",
    "2. Month                    } Date Feature - Easier to train when separated\n",
    "3. Day\n",
    "4. Country\n",
    "5. Region\n",
    "6. Specific Primary Target\n",
    "7. Specific Primary Weapon\n",
    "8. Number of people killed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_train_x = pd.DataFrame(df_train_x[df_train_x.columns[index_relevant_features]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(76304, 8)"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train_x.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(76304L,)"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train_y.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_test = df_test[df_test.columns[index_relevant_features]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(65662, 8)"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature Ranking of final Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.20078094,  0.09575658,  0.11413252,  0.32778058,  0.05070187,\n",
       "        0.10693487,  0.05945306,  0.04445958])"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "clf = DecisionTreeClassifier()\n",
    "clf = clf.fit(df_train_x, df_train_y)\n",
    "clf.feature_importances_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "data_feature_rank = pd.DataFrame(clf.feature_importances_)\n",
    "data_feature_rank.columns = ['rank']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x22936550>"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfYAAAFVCAYAAAAdY838AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3Xt8FPW9//H3bnY3IdwCwYQssHqQJNyEgBgBi2LSUym2\ntMW2h8bjrYJNlIs/a6GIcUGCwKk+uByEg8KDE6Be4KDS9ogCQo6PeCNgMUUaAkaTQAgBEsIl2c1l\n9/cHZUsEwkYzLIyv51/Mzn5nPl9md9/znZnMWPx+v18AAMAUrKEuAAAAtB6CHQAAEyHYAQAwEYId\nAAATIdgBADARgh0AABOxGblwv9+vFStWqLi4WHa7Xenp6YqNjQ3M//jjj7Vx40ZZrVbddtttGj16\ntCRp2rRpioyMlCTFxMQoIyPDyDIBADANQ4M9Ly9P9fX1ysrK0v79+5Wdna2pU6dKknw+n1599VXN\nnz9fDodDTzzxhG6//XaFh4dLktxut5GlAQBgSoYGe0FBgZKSkiRJ8fHxKioqCsyzWq1asGCBrFar\nqqur5ff7ZbPZVFxcLK/Xqzlz5sjn82ncuHGKj483skwAAEzD0GCvqakJHFKXpLCwMPl8PlmtZ0/t\nW61W7dixQytXrtTgwYMVHh4uh8OhMWPGKCUlRYcPH9Zzzz2nRYsWBdoAAIBLMzQtIyMj5fF4AtPn\nh/o5ycnJWr58uerr6/V///d/cjqd+t73vidJiouLU/v27XXixAkjywQAwDQMHbEnJiZq165dGjp0\nqAoLC+VyuQLzamtrNX/+fD399NOy2WyKiIiQ1WrVtm3bVFJSovHjx6uyslK1tbWKioq67LrKysqM\n7AoAAFcNp9N5yXkWIx8Cc+6q+JKSEklSRkaGioqK5PV6lZqaqvfee0/btm2TzWaTy+XSr3/9azU2\nNmrp0qU6duyYLBaL7r33XiUkJFx2XQQ7AOC7ImTBfiUR7ACA74rmgp0r0gAAMBGCHQAAEyHYAQAw\nEYIdAAATIdgBADARgh0AABMh2AEAMBGCHQAAEyHYAQAwEYIdAAATIdgBADARgh0AABMh2AEAMBGC\nHQAAEyHYAQAwEVuoCwC+i6oaTquy7lSoywhaZ0d7dbK1C3UZAIJAsAMhUFl3SnP3vBLqMoI2vX8a\nwQ5cIzgUDwCAiRDsAACYCMEOAICJEOwAAJgIwQ4AgIkQ7AAAmAjBDgCAiRDsAACYCMEOAICJEOwA\nAJgIwQ4AgIkQ7AAAmAjBDgCAiRDsAACYCMEOAICJEOwAAJgIwQ4AgIkQ7AAAmAjBDgCAidiMXLjf\n79eKFStUXFwsu92u9PR0xcbGBuZ//PHH2rhxo6xWq2677TaNHj36sm0AAMClGTpiz8vLU319vbKy\nspSWlqbs7OzAPJ/Pp1dffVVut1uzZ8/W5s2bdfr06WbbAACA5hk6Yi8oKFBSUpIkKT4+XkVFRYF5\nVqtVCxYskNVqVXV1tfx+v2w2W7NtAABA8wwN9pqaGkVGRgamw8LC5PP5ZLWePVBgtVq1Y8cOrVy5\nUoMHD5bD4bhsGwAAcGmGBntkZKQ8Hk9g+mIBnZycrOTkZC1ZskTvv/9+UG0uxul0tl7hgMFKSytD\nXUKLOMLD+Y4B1whDgz0xMVG7du3S0KFDVVhYKJfLFZhXW1ur+fPn6+mnn5bNZlNERISsVqt69+6t\nnTt3XrRNc8rKyozqBtDq6rzeUJfQInVeL98x4CrS3I62ocGenJys/Px8ZWZmSpIyMjKUm5srr9er\n1NRUjRgxQm63WzabTS6XSyNGjJAkffbZZ03aAACA4Fj8fr8/1EW0BkYTuJZ8UXNYc/e8Euoygja9\nf5pujIwLdRkA/qG5ETtXpAEAYCIEOwAAJkKwAwBgIgQ7AAAmQrADAGAiBDsAACZCsAMAYCIEOwAA\nJkKwAwBgIgQ7AAAmQrADAGAiBDsAACZCsAMAYCIEOwAAJkKwAwBgIgQ7AAAmQrADAGAiBDsAACZC\nsAMAYCIEOwAAJkKwAwBgIgQ7AAAmQrADAGAiBDsAACZCsAMAYCIEOwAAJkKwAwBgIgQ7AAAmQrAD\nAGAiBDsAACZCsAMAYCIEOwAAJkKwAwBgIgQ7AAAmQrADAGAiBDsAACZiM3Lhfr9fK1asUHFxsex2\nu9LT0xUbGxuYn5ubq02bNiksLEwul0vjx4+XJE2bNk2RkZGSpJiYGGVkZBhZJgAApmFosOfl5am+\nvl5ZWVnav3+/srOzNXXqVElSXV2d1q1bpxdeeEF2u12LFi3Srl27NGDAAEmS2+02sjQAAEzJ0GAv\nKChQUlKSJCk+Pl5FRUWBeXa7XVlZWbLb7ZKkxsZG2e12FRcXy+v1as6cOfL5fBo3bpzi4+ONLBMA\nANMw9Bx7TU1N4JC6JIWFhcnn80mSLBaLOnToIEnatGmTvF6vBgwYIIfDoTFjxmjGjBkaP368Fi9e\nHGgDAACaZ+iIPTIyUh6PJzDt8/lktf5zX8Lv92vt2rU6fPiwnnzySUmS0+lU165dJUlxcXFq3769\nTpw4oc6dOze7LqfTaUAPAGOUllaGuoQWcYSH8x0DrhGGBntiYqJ27dqloUOHqrCwUC6Xq8n85cuX\ny+FwBM67S9K2bdtUUlKi8ePHq7KyUrW1tYqKirrsusrKylq9fsAodV5vqEtokTqvl+8YcBVpbkfb\n4vf7/Uat+NxV8SUlJZKkjIwMFRUVyev1qmfPnpo+fbr69OkTeP/o0aM1ePBgvfjiizp27JgsFovu\nvfdeJSQkXHZd/OjgWvJFzWHN3fNKqMsI2vT+aboxMi7UZQD4h5AF+5VEsONaQrAD+DaaC3ZuUAMA\ngIkQ7AAAmAjBDgCAiRDsAACYCMEOAICJEOwAAJgIwQ4AgIkQ7AAAmAjBDgCAiRDsAACYCMEOAICJ\nEOwAAJgIwQ4AgIkQ7AAAmAjBDgCAiRDsAACYCMEOAICJEOwAAJgIwQ4AgIkQ7AAAmAjBDgCAiRDs\nAACYCMEOAICJEOwAAJgIwQ4AgIkQ7AAAmAjBDgCAiRDsAACYCMEOAICJEOwAAJgIwQ4AgIkQ7AAA\nmAjBDgCAiRDsAACYiC2YN+3cuVM7duxQWVmZrFarnE6nhg0bpoEDBxpdHwAAaIFmg72srEwvvvii\n2rVrp8GDB2v48OHy+XyqqKjQpk2btH79eqWnp6t79+5Xql4AANCMZoN9w4YNmjJlimJiYi6YN2rU\nKJWXl2vdunWaPHnyRdv7/X6tWLFCxcXFstvtSk9PV2xsbGB+bm6uNm3apLCwMLlcLo0fP/6ybQAA\nwKU1e4590qRJFw11SWpsbFTXrl0vGeqSlJeXp/r6emVlZSktLU3Z2dmBeXV1dVq3bp1mzpypZ599\nVmfOnNGuXbuabQMAAJoX1MVz7777bpPpgwcP6qmnnrpsu4KCAiUlJUmS4uPjVVRUFJhnt9uVlZUl\nu90u6eyOgt1ub7YNAABoXlAXz23evFnt2rXTbbfdpr/85S/asGGD7rnnnsu2q6mpUWRkZGA6LCxM\nPp9PVqtVFotFHTp0kCRt2rRJXq9XAwYM0IcffnjJNs1xOp1Npk8fOarG02eC6d5VI6xdW7WLvS7U\nZeAKKC2tDHUJLeIID7/gOwbg6hRUsGdmZmr27Nl6++23ZbFYNGfOnKC+5JGRkfJ4PIHprwe03+/X\n2rVrdfjwYT355JNBtbmUsrKyJtP2E6ek/8u7bLuryh236GRjfairwBVQ5/WGuoQWqfN6L/iOAQid\n5jK42cQ8ffq0Tp8+LZvNpkmTJun48eMaO3asOnTooNOnT192xYmJifr0008lSYWFhXK5XE3mL1++\nXPX19Zo6dWrgkPzl2gAAgEtrdsT+8MMPX/Da/PnzA/9+/fXXm114cnKy8vPzlZmZKUnKyMhQbm6u\nvF6vevbsqZycHPXp00ezZs2SJI0ePfqibQAAQHAsfr/fH+oiWoNZDsXXR7UPdRW4Ar6oOay5e14J\ndRlBm94/TTdGxoW6DAD/0Nyh+KDOsXu9Xn300UcXHH7/0Y9+9O0qAwAArSqoYF+wYIGqqqrkcrlk\nsViMrgkAAHxDQQV7WVmZFixYoLCwMKPrAQAA30JQN6iJjo42ug4AANAKghqxu1wuzZo1S0lJSXI4\nHIHXOccOAMDVJahgr62tVdeuXVVeXm50PQAA4FsIKtgfffRRo+sAAACtIKhgLyws1FtvvSWPxyO/\n3x94JvuyZcuMrg8AALRAUBfP/dd//ZcSEhJUW1urESNGqE2bNrr11luNrg0AALRQUMFusVj005/+\nVH379pXT6dQTTzyhv//970bXBgAAWiioYI+IiJAkxcbGqrS0VA6HQz6fz9DCAABAywUV7L169dKC\nBQvUv39//fnPf9bq1auDepQqAAC4soJK5/vuu0933323nE6nHnzwQfl8Pk2ZMsXo2gAAQAsFFexP\nPfWUEhISJEmDBw/Wgw8+2OyTZQAAQGgEFezh4eE6fvy40bUAAIBvKai/Y/d4PJo4caKio6MDF9JJ\n0vPPP29YYQAAoOWCCvaHHnrI6DoAAEAraDbY8/PzNWDAAPXt2/eS7/nss880cODAVi8MAAC0XLPB\nvmfPHr3zzjv6/ve/r4EDBwaex97Q0BCY53Q6CXYAAK4SzQZ7Wlqa9u/fr/Xr12vhwoXq1KmTfD6f\nqqur1adPH91zzz2Bq+UBAEDoXfYce3x8vJ566inV1NSovLxcFotFsbGxioyMvBL1AQCAFgj69nFn\nzpyRxWKR3+9XeXm5ioqKjKwLAAB8A0FdFb927Vq988476tixY+A1i8WiJUuWGFYYAABouaCC/aOP\nPtLixYvVuXNno+sBAADfQlCH4rt06UKoAwBwDQhqxN6/f3+tXbtWQ4YMkcPhCLzes2dPwwoDAAAt\nF1Sw5+TkSDp7SP4czrEDAHD1CSrYX3zxRaPrAAAArSCoYD958qTef/99eTweSZLP51N5ebkmT55s\naHEAAKBlggr2BQsWyOFw6ODBg7rpppv0t7/9Tb179za6NgAA0EJBXRV/7NgxTZ8+XYMGDdKoUaM0\ne/ZslZeXG10bAABooaCCPSoqSpLUtWtXlZaWqnPnzvL5fIYWBgAAWi6oQ/EdOnTQn/70JyUkJGjd\nunVq06aNampqjK4NAAC0UFAj9kceeUQ2m029e/dWz549tW7dOt17771G1wYAAFooqBF7x44d9f3v\nf18lJSVKS0vTL3/5yyY3qgEAAFeHoEbshYWFmjRpkubOnauqqiqlp6dr3759RtcGAABaKOinu2Vm\nZmrx4sWKjo7WxIkT9d///d+aO3dus+38fr9WrFih4uJi2e12paenKzY2tsl7vF6vsrKylJGRIafT\nKUmaNm1a4HnvMTExysjI+CZ9AwDgOyeoYPd6verevXtgevDgwXrttdcu2y4vL0/19fXKysrS/v37\nlZ2dralTpwbmFxUV6eWXX1ZlZWXgtfr6ekmS2+0OuhMAAOCsoA7F22w2nT59WhaLRZJUVlYW1MIL\nCgqUlJQkSYqPj1dRUVGT+Q0NDfrd734XGKlLUnFxsbxer+bMmaPZs2dr//79Qa0LAAAEOWL/2c9+\nppkzZ6qqqkoLFy5Ufn6+Hnnkkcu2q6mpCRxSl6SwsDD5fD5ZrWf3JxISEi5o43A4NGbMGKWkpOjw\n4cN67rnntGjRokCbSzl/50CSqmu/Ul0wnbuKOMLDdd3X+gFzKi2tvPybriKO8PALvmMArk5BBfuQ\nIUMUGRmp4uJiWSwWjR079rJBK0mRkZGB+8tLahLql+J0OtW1a1dJUlxcnNq3b68TJ05c9nnwXz+K\nYPd6L1vf1abO6w36aAiubXXX2OeTzyZwdWluRzuoYM/Ozta7776rNm3aBF6zWCxasWJFs+0SExO1\na9cuDR06VIWFhXK5XJdd17Zt21RSUqLx48ersrJStbW1gTvfAQCA5gUV7Dt27NDy5cvVvn37Fi08\nOTlZ+fn5yszMlCRlZGQoNzdXXq9XqampF22TkpKipUuX6plnnpHFYlFGRkZQRwcAAECQwR4XF6e2\nbdu2eOEWi0UTJkxo8trFDh+cfwW8zWbjcbAAAHxDQQX7D3/4Q7ndbvXr10822z+b/PznPzesMAAA\n0HJBBfu6desUFRXFg18AALjKBRXsdXV1mj59utG1AACAbymoq9K6d++u4uJio2sBAADfUlAj9qqq\nKv3+979XTEyM7HZ74PXnn3/esMIAAEDLBRXsaWlpRtcBAABaQVDB3rdvX6PrAAAArYA7vwAAYCIE\nOwAAJkKwAwBgIgQ7AAAmQrADAGAiBDsAACZCsAMAYCIEOwAAJkKwAwBgIgQ7AAAmQrADAGAiBDsA\nACZCsAMAYCIEOwAAJkKwAwBgIgQ7AAAmQrADAGAiBDsAACZCsAMAYCIEOwAAJmILdQHAxXjrq+Tx\nHA91GS0SERGtcHunUJcB4DuOYMdVyeM5rvd3zAl1GS1ye/IMgh1AyHEoHgAAEyHYAQAwEYIdAAAT\nIdgBADARLp4DgBY40SBV1flCXUaLdHJYFcWv/XcGmxoAWqCqzqcFn1WEuowW+X8DYxRl4wDtd4Wh\nwe73+7VixQoVFxfLbrcrPT1dsbGxTd7j9XqVlZWljIwMOZ3OoNoAAICLM3QXLi8vT/X19crKylJa\nWpqys7ObzC8qKtLMmTNVUVERdBsAAHBphgZ7QUGBkpKSJEnx8fEqKipqMr+hoUG/+93v5HQ6g24D\nAAAuzdBD8TU1NYqMjAxMh4WFyefzyWo9uz+RkJDQ4jYAAODSDA32yMhIeTyewHQwAf1N2khqMuqX\npOrar1TXwnpDzREeruu+1o/vqtqvSkNdQos5wsMv+BxeSmlppcHVtK6W9M3sykqurQvnJCnc4ZDT\nGRPqMnCFGBrsiYmJ2rVrl4YOHarCwkK5XC5D2khSWVlZk2m71/uNag6lOq/3gn58V9WZfPtda/3j\ns/lP3mvsT90kyVtXx/YzmeZ2tA0N9uTkZOXn5yszM1OSlJGRodzcXHm9XqWmpgbdBgAABMfQYLdY\nLJowYUKT1y62l+F2u5ttAwAAgsMNaq5V3io1eqpCXUWLhEV0ksJ5rCkAGIlgv0Y1eqpU/v78UJfR\nIl1vn6Ywgh0ADMXfkAEAYCIEOwAAJkKwAwBgIgQ7AAAmQrADAGAiBDsAACZCsAMAYCIEOwAAJkKw\nAwBgIgQ7AAAmQrADAGAiBDsAACZCsAMAYCIEOwAAJkKwAwBgIgQ7AAAmQrADAGAiBDsAACZCsAMA\nYCIEOwAAJkKwAwBgIgQ7AAAmQrADAGAiBDsAACZCsAMAYCIEOwAAJkKwAwBgIgQ7AAAmQrADAGAi\nBDsAACZiC3UBAICrR2O9TfV119aYz+7wKczeEOoyrhoEOwAgoL7Oqn2fhrqKlkkcbFWYPdRVXD2u\nrd0yAADQLIIdAAATMfRQvN/v14oVK1RcXCy73a709HTFxsYG5u/cuVMbNmyQzWbTyJEjlZqaKkma\nNm2aIiMjJUkxMTHKyMgwskwAAEzD0GDPy8tTfX29srKytH//fmVnZ2vq1KmSpMbGRq1evVrz5s2T\nw+FQZmambrnlFrVp00aS5Ha7jSwNAABTMjTYCwoKlJSUJEmKj49XUVFRYN6hQ4cUFxcXGJn37t1b\ne/fuVZcuXeT1ejVnzhz5fD6NGzdO8fHxRpYJAIBpGBrsNTU1geCWpLCwMPl8PlmtVtXU1ARG55LU\npk0b1dTUKDw8XGPGjFFKSooOHz6s5557TosWLZLVyuUAwLWiqqFelV5vqMsIWufwcHWycVk1zMHQ\nYI+MjJTH4wlMnwv1c/Nqa2sD82pra9W2bVvFxcUFzsPHxcWpffv2OnHihDp37tzsupxOZ5Pp6tqv\nVNdaHblCHOHhuu5r/biUitqDBlfT+sIdDsUE2b/ar0oNrqb1OcLDL/gcXkppaaXB1bSulvRNkg6W\nfKX5n+UZWFHreib5NvULsn9lJRUGV9P6wh0OOZ0xQb33UHG1pHpjC2pl4Q67nM4uoS7jqmFosCcm\nJmrXrl0aOnSoCgsL5XK5AvO6deum8vJynTlzRuHh4SooKNCYMWO0bds2lZSUaPz48aqsrFRtba2i\noqIuu66ysrIm0/ZraLRwTp3Xe0E/LqWx7lrbbZG8dXVB96/O5NvvWutfS/omSV7vtfX59HqD/2x6\n63wGV9P6WvLd89Y5DK6m9Xnr6lv0+TSD5na0DQ325ORk5efnKzMzU5KUkZGh3Nxceb1epaam6oEH\nHlBWVpYk6c4771SnTp2UkpKipUuX6plnnpHFYlFGRgaH4QEACJKhwW6xWDRhwoQmr52/lzF48GAN\nHjy4aUE2myZPnmxkWQAAmBZDYQAATIR7xQMAvjNsHqssNdfWdRL+SKsaIoKvmWAHAHxnWGp8anz3\nZKjLaJGwuzpIEcG/n0PxAACYCMEOAICJEOwAAJgIwQ4AgIkQ7AAAmAjBDgCAiRDsAACYCMEOAICJ\nEOwAAJgIwQ4AgIkQ7AAAmAjBDgCAiRDsAACYCMEOAICJEOwAAJgIwQ4AgIkQ7AAAmAjBDgCAiRDs\nAACYCMEOAICJEOwAAJgIwQ4AgIkQ7AAAmAjBDgCAiRDsAACYCMEOAICJEOwAAJgIwQ4AgIkQ7AAA\nmAjBDgCAiRDsAACYCMEOAICJEOwAAJiIzciF+/1+rVixQsXFxbLb7UpPT1dsbGxg/s6dO7VhwwbZ\nbDaNHDlSqampl20DAAAuzdARe15enurr65WVlaW0tDRlZ2cH5jU2Nmr16tXKzMyU2+3W1q1bdfLk\nyWbbAACA5hk6Yi8oKFBSUpIkKT4+XkVFRYF5hw4dUlxcnCIjIyVJffr00d69e1VYWHjJNgAAoHmG\njthramoCwS1JYWFh8vl8gXlt2rQJzIuIiFBNTY1qa2sv2QYAADTP0BF7ZGSkPB5PYNrn88lqtQbm\n1dbWBubV1taqbdu2atOmzSXbNMfpdH7tBUl9E79dB65mTqd69PleqKswjFNO9Um8PdRlGMYpp0b0\nujnUZRjGKWlEfEKoyzCEU9JtvbqHugxD9YwPdQUGckrq77zs265lho7YExMT9emnn0qSCgsL5XK5\nAvO6deum8vJynTlzRg0NDSooKFBCQkKzbQAAQPMsfr/fb9TCz13hXlJSIknKyMhQUVGRvF6vUlNT\n9emnn2r9+vWSpDvvvFM/+MEPLtrmgtE4AAC4KEODHQAAXFncoAYAABMh2AEAMBGCHQAAEyHYAQAw\nEYL9H3bv3q333nsv1GW0GiP6c/r0aeXm5kqSli5dqs8++6xVl98aLtbvGTNm6NixYyGqKPTO325m\nsGjRIjU2Nhq+nvr6em3bts3w9UjS0aNHNWPGjAteP3bsmHbt2tWq68rOztbWrVtbdZktMWvWLJWV\nlV3w+jvvvNOq69m/f79mzZrVqsv8uov1JTs7W8ePH9f69eu1detW7d27VwsXLjS0jq8j2P8hKSlJ\nqampoS6j1RjRn+LiYu3cubNVl9nazLYdW8O1sN1aYsqUKQoLCzN8PVVVVVd0Z99isVzw2p49e7Rv\n375WWf7Jkyc1d+7cVt9RaC1vvPFGqy3rT3/6k5YvX676+vpWW2awHnjgAUVHRzd57WLb1kiG3nnu\nWpKTk6PVq1crJSVF//7v/y6fz6epU6dq3rx52rJliz744ANZLBbddtttGjVqlEpLS7V69Wr5fD6d\nOnVK48ePV0JCgh599FF1795d3bt31/333x/S/uzevVtHjx5VdHS0jh49quHDh6u0tFRffvmlBg8e\nrF/96lf68ssvtWrVKoWFhclut+s3v/mNfD6fFi1apC5duqi8vFy9evXS+PHj9eabb6q4uDjwY7dl\nyxZt3LhRtbW1Gj9+vG688caQ9fecnJwclZWVyWq1avfu3YqOjtbp06clSZWVlXr55ZfV0NCgqqoq\njRs3Tt26ddN//ud/6rnnnpMkLVy4UD/+8Y9D1pe6ujotXbpUx44dU2Njox544AFt2bJFFRUV8vl8\n+tGPfqRhw4Zp1qxZmjBhgpxOp7Zs2aLq6mrdcccdl91u+/bt06lTp3T69Gldf/316tGjh+666y6d\nOXNGs2fP1rx5865YX3NycrR9+3ZJ0l133aW3335bYWFhSkxMVFpamk6dOqXFixeroaFBcXFx+vzz\nz7Vo0SI99thjWrRokaqqqrRs2TL5fD5ZLBY99NBDcrlcmjJlihITE1VWVqaoqCj99re//UY/rG++\n+aYOHTqkDRs26MCBA00+N0OGDNFvf/tbxcXFyW6369e//nWTWvfs2aPFixfr1Vdf1d69e+Xz+XTr\nrbdqzJgxl9x21dXV+o//+A9VV1fr5ptv1k9/+lNt3LhRdXV1SkhIUHZ2thYvXiyLxaI//vGP6tmz\npzZv3iyn0xkYNT7++OPq2LGjXnnlFe3bt08+n0933323hg4dqlmzZun+++/Xvn37tGrVKvXq1Us3\n3HCDpk2bpjvuuEMffvhhUL9xEydOVEJCgsrLy+VyuZSenq6CggKtWbNGNptN4eHheuKJJ/Txxx+r\nrKxMaWlpqq+v1+OPP64XX3xRkvT666/r1KlTstvteuyxx7R161adOXNGK1eu1JkzZzRixAgNGjRI\nhw4d0po1azR06FDl5eXJ4/Ho1KlTuueee3Trrbdq7969eu211xQWFqbY2Fg98sgjslqt6tq1q558\n8kktWbKkVT+vf/3rX+X1elVRUaExY8YE5u3cuVNvv/22nnzySf3hD3/QhAkTWm293wTBfp6RI0cq\nLy9P9957r3bv3q3+/furvLxcH330kWbPni2/36+srCwNHDhQpaWluv/++9WjRw/l5uYqJydHCQkJ\nOn78uP7whz+obdu2oe6OJKmiokKZmZnyeDyaOHGiXnrppcCX6Ve/+pVeeuklZWRkyOVyaefOncrO\nztZ9992nw4cPKzMzU3a7XZMmTVJ1dbXGjh2rLVu2KDU1Vfv27VPPnj01duxY5eTkKCcn56oIdkn6\n4osv1NDQoHnz5qm2tlZTpkyRdPbBQz/+8Y/Vt29fFRYWav369ZoxY4bCw8N16NAhdezYUUePHg1p\nP7Zs2aJPxqJfAAAJjklEQVSYmBg9/vjjKi8v14cffqiOHTtq0qRJ8ng8mjZtmvr373/J9sFst5tu\nukmjR49WRUWFFi1apLvuuku5ubkaMWLEFezpWe3atdOjjz6qzMxMzZs3Tw6HQ0uWLFF+fr7++te/\n6pZbbtEPfvAD5efnKz8/X9I/Rz9r1qzR3XffrZtvvllfffWVli1bprlz5+rIkSNyu93q3LmzMjMz\n9cUXX6hXr14trm3s2LEqLS1VYmKi+vTp0+RzM2TIEHk8Hv3iF7/Q9ddfr+zs7IvW+sEHH2jmzJmK\niopSTk5Os+vzer2aPHmyHA6H3G63hgwZop/85CcqKyvTkCFDtGPHDn322WcaMGCAdu/erXHjxmnz\n5s3q3bu3JkyYoM2bN+uNN97QoEGDVFFRoVmzZqm+vl4zZszQgAEDNHz4cJ08eVJHjx5Vhw4dlJ+f\nL5vNpq5du+qTTz4J+jeusrJS48aNU0xMjBYuXKgdO3aosLBQw4cP1+jRo7Vr1y6dOXPmgv6dv3M1\ndOhQDRs2TJs3b9Zbb72l+++/X++++64efvhhff7559q8ebMGDRqk7du3KyUlRTU1Naqrq1NmZqaq\nq6v11FNPaciQIVq+fLlmz56tDh066PXXX1dOTo5SUlKUnJyso0ePtnibX05tba2eeuoplZeXa/78\n+YqKitInn3yivXv36ve//70cDkerr/ObINjPY7PZ1LdvX+3evVs5OTn6+c9/rtLSUh09elTPPvus\nJOnMmTM6fPiwOnfurP/5n/9ReHh4k4fddOjQ4aoJdUmKiYlRRESEwsLCFBUVFajz3JessrIycNve\nPn366JVXXpEkde3aVeHh4ZKkTp06XfSQVs+ePSVJUVFR8nq9hvclWMeOHdPNN5+9D3ubNm3Uo0cP\nSWf7sWHDhsB504aGBklSSkqKtm/fri5duoQk3M5XVlamQYMGSTq7DaqqqjRgwABJZx+U1K1bNx05\ncqRJm/PvMRXMdjt3J8eYmBi1adNGBw8eVG5urqZNm2ZIn5rjdDp1+PDhwGFiSfJ4PKqoqFBZWZlG\njhwp6exn83x+v1+HDh0KvH7DDTfo+PHjkqT27durc+fOkqTo6GjV1dV9qxqjoqIu+rmRpLi4OEln\ndxovVuukSZO0du1aVVdXB7br1/txzvXXX6+IiAhJ0o033njBuduUlBRt2rRJPp9PN910U+B0RL9+\n/SRJCQkJysvLU3R0tL788svA+eXGxkYdPXpUycnJeuONN3T8+HHdcsst+tvf/ia/369bb71Va9as\nCfo3rkuXLoqJiQmss6ysTGPHjtWGDRv07LPPKjo6+oIdqa/fB+3c/1FCQkLgFuLn3tOvXz+tWrVK\nJ0+eVH5+vtLS0vT+++8H2nTs2FHt2rVTVVWVTpw4oQULFkg6e7Tr3HfFKDfccIOkpp+rPXv2qLa2\n9oqcHgoW59i/JjU1Ve+9955Onjwpl8slp9Mpl8slt9stt9utkSNHyuVyadWqVfq3f/s3Pfroo03u\nZ3+lz6Vczvn1XOwmg507dw7cvvfzzz8P/FCd71w7i8XSZBlXW1/PiY2N1YEDB+T3++XxeHTw4EFJ\n0muvvaY77rhDEydObDLqHTp0qPLz85WXlxfyYO/WrZsOHDggSTpy5Ig++OAD/f3vf5d0drRQWlqq\nmJgY2e12nThxQpL05ZdfXnRZwWy31NRUbdiwQdHR0WrXrp0hfWqOxWJRTEyMunTposzMTLndbt11\n112Kj49Xjx49AueXCwsLA238fr8sFou6deumvXv3SpK++uorRUVFBZbZWrX5fL5Lfm4kBR5Q5XK5\nLqi1oaFBH330kR5//HG53W5t375dx44dk8PhuOi2O3jwoLxerxobG3XgwAH16NFDVqs18HTL3r17\n68iRI9q2bZtSUlIC7c492nrfvn3q0aOHunXrpn79+sntduuZZ57RsGHDFBsbqx49eqiioiKwM+/x\neLRz50517969Rb9xlZWVqq6ulnT20dw9evTQ+++/rzvvvFNut1vdu3fX1q1b5XA4VFVV1aTGc859\nxgsKCgLLPv8zevvtt2vVqlUaMGBA4P/43DJOnDihmpoaRUdHKzo6WlOnTpXb7dbPfvazC7bPlbix\n6sMPP6yBAwfq9ddfN3xdwWLE/jW9evXSkSNHNGrUKEln96L79eunzMxMNTQ0qFevXoqOjtaIESP0\nwgsvqF27doqOjtapU6ckXV1h9/VaLlbbb37zG61cuVLS2UfkpqenX/Dec/+OjY1VSUmJ3n77baNK\nbhU33HCDIiIiNH36dHXq1Cnwgz9s2DCtWbNGb731ljp37qyTJ09Kkux2u/r06aNTp06F/GjLv/7r\nv2rZsmWaOXOm/H6/ZsyYoXfeeUfPPPOM6uvr9Ytf/EIdOnTQD3/4Q7388su67rrrAqNTqeXb7ZZb\nbtHKlSs1efJk4zt3CR06dNDdd98tt9stn8+n6667TsOHD9dPfvITLVmyRB9//LGioqICI6Jz/brv\nvvu0fPly/fnPf5bP51NGRkaT+V//d0t17NhRjY2NgfO85z43F/uuX6xWm82mdu3aacaMGXI4HEpK\nSlKXLl00atSoi2679u3ba8GCBTp16pSGDx+ubt26qb6+Xm+++aZ69uyp4cOH63vf+54++eQTde/+\nz6fL5eTk6C9/+YsiIiI0ceJEtWvXTp9//rncbrc8Ho+Sk5MDRwL69u0bGCH36dNHZWVlcrlcLfqN\ns9vtWrlypY4fP674+HjdfPPNOnDggJYtW6aIiAhZrVY98sgjatu2rTZv3iy3261/+Zd/afI47h07\nduh///d/1aZNG02cOFGS1KNHDy1ZskQTJ07UHXfcoddee00vvPBCoM2JEyc0e/Zs1dTUaMKECbJY\nLHrwwQc1d+5c+f1+RUZG6rHHHmuyDY38PT5/2ffcc49mzJgROFJo9Lovy48mGhsb/U8//bS/trY2\n1KXgClqxYoV/z549oS7jivN4PP7p06eHuoyL+vTTT/1ffPGF3+/3+/Pz8/3PPvtsiCu6tCtV68aN\nG/3bt28PTM+cOdN/6NAhQ9Z1KRMmTDB8HcePH2/yf7h9+3b/H//4R8PXaxaM2M9TUVGh559/Xikp\nKYE9XJjfnDlz1L59+8C5yu+KwsJCvfTSS/rlL38Z6lIuKiYmRsuWLVNYWJh8Pp8eeuihUJd0SVei\n1qVLl6qqqiok10Kcz+iR6I4dO7R+/fqQX1l+LePpbgAAmAgXzwEAYCIEOwAAJkKwAwBgIgQ7AAAm\nQrADAGAi/x9X/pjuTP9HJwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x224ba8d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(x = df_train_x.columns, y = 'rank' , data = data_feature_rank, order = df_train_x.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As you can see, the feature 'country' is ranked the highest. The rest of the features are plotted accordingly."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model Evaluation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The dataset has large number of instances and the features are categorical. Using a <b> Random Forest </b> will be a good choice for the model. Other model choices include Support Vector Machines, Naive Bayes Classifiers or k-NN classifiers."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A Random Forest is an ensemble of decision trees, where each tree is constructed by subsets of the training data. \n",
    "\n",
    "There are many advantages of using a Random Forest, the primary of them being:\n",
    "1. It can handle large data sets with high dimensionality. It trains well even on high number of input features. \n",
    "2. It can handle categorical and continous variables. Other estimators in Scikit-learn need encoding of categorical\n",
    "features which is not the case in Random Forest models.\n",
    "3. It is more robust and accurate compared to Decision Trees\n",
    "4. They perform well in multi-class classification problems and can work in regression cases too."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Other classifiers can give similar performance but a Random Forest in Scikit-Learn is faster and more accurate than the other\n",
    "classifiers mentioned above. \n",
    "\n",
    "Boosted ensemble classifiers can also be tried - though will probably be more accurate, they can be slower and occupy more memory.\n",
    "\n",
    "Also rest of the classifiers need the categorical features to be encoded to avoid them being taken in order - this is not required in the case of Random Forests."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Training and testing the model on the train data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.99677605368001676"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "clf_dtree = DecisionTreeClassifier()\n",
    "clf_dtree = clf_dtree.fit(df_train_x, df_train_y)\n",
    "clf_dtree.score(df_train_x,df_train_y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We notice that the accuracy of the classifier(taken as the performance metric in this case) is <b> 99.67 %. </b> \n",
    "\n",
    "This is not surprising as the model is trained and tested on the same data. Though this is just for illustratory purposes - This is a classic example of Overfitting where the trained model is tested on the same dataset. The high accuracy is a red-flag. \n",
    "\n",
    "The model will not <b> generalize </b> well on <b> unseen data </b> . The following run of model shows the model run on new data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Evaluate Model Accuracy"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The data is split into two - training set for training the model and validation/test set to evaluate the model accuracy in the ratio: 70% (train) and 30% (validation)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(53412, 8) (53412L,)\n",
      "(22892, 8) (22892L,)\n"
     ]
    }
   ],
   "source": [
    "from sklearn.cross_validation import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(df_train_x, df_train_y, test_size=0.3, random_state=0)\n",
    "print X_train.shape, y_train.shape\n",
    "print X_test.shape, y_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "Fitting the Random Forest Classifier onto the training dataset."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The number of estimators of the Random Forest used to train the model is 10. Normally the accuracy of the classifier increases\n",
    "with increase in number of estimators, but at the cost of memory and speed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "            max_depth=None, max_features='auto', max_leaf_nodes=None,\n",
       "            min_samples_leaf=1, min_samples_split=2,\n",
       "            min_weight_fraction_leaf=0.0, n_estimators=10, n_jobs=1,\n",
       "            oob_score=False, random_state=None, verbose=0,\n",
       "            warm_start=False)"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "clf_rf = RandomForestClassifier(n_estimators=10)\n",
    "clf_rf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "Evaluating the model accuracy on the validation set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.69757994059059936"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score\n",
    "y_pred = clf_rf.predict(X_test)\n",
    "accuracy_score(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "The accuracy of the classifier is 69.75%"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Further model evaluation methods such as k-Fold Cross Validation can be used but the split method is much faster when compared to the other methods."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Model Predicting"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The model is now run on the test dataset and the class labels are predicted."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "test_y = clf_rf.predict(df_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "65662"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(test_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df_test['gname'] = test_y"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "The predicted labels are merged with the original dataset as shown above."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Displaying the top 5 rows of the test set dataframe."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>iyear</th>\n",
       "      <th>imonth</th>\n",
       "      <th>iday</th>\n",
       "      <th>country</th>\n",
       "      <th>region</th>\n",
       "      <th>targsubtype1</th>\n",
       "      <th>weapsubtype1</th>\n",
       "      <th>nkill</th>\n",
       "      <th>gname</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1970</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>160</td>\n",
       "      <td>5</td>\n",
       "      <td>54</td>\n",
       "      <td>12</td>\n",
       "      <td>1</td>\n",
       "      <td>New People's Army (NPA)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1970</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>78</td>\n",
       "      <td>8</td>\n",
       "      <td>46</td>\n",
       "      <td>16</td>\n",
       "      <td>0</td>\n",
       "      <td>Armenian Secret Army for the Liberation of Arm...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1970</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>101</td>\n",
       "      <td>4</td>\n",
       "      <td>46</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>Armed Proletarian Power</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1970</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>217</td>\n",
       "      <td>1</td>\n",
       "      <td>107</td>\n",
       "      <td>16</td>\n",
       "      <td>0</td>\n",
       "      <td>New World Liberation Front (NWLF)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>1970</td>\n",
       "      <td>1</td>\n",
       "      <td>8</td>\n",
       "      <td>98</td>\n",
       "      <td>8</td>\n",
       "      <td>42</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>Red Brigades</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    iyear  imonth  iday  country  region  targsubtype1  weapsubtype1  nkill  \\\n",
       "2    1970       1     0      160       5            54            12      1   \n",
       "3    1970       1     0       78       8            46            16      0   \n",
       "4    1970       1     0      101       4            46            12      0   \n",
       "7    1970       1     2      217       1           107            16      0   \n",
       "12   1970       1     8       98       8            42             4      0   \n",
       "\n",
       "                                                gname  \n",
       "2                             New People's Army (NPA)  \n",
       "3   Armenian Secret Army for the Liberation of Arm...  \n",
       "4                             Armed Proletarian Power  \n",
       "7                   New World Liberation Front (NWLF)  \n",
       "12                                       Red Brigades  "
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_test.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Thus a predictive model has been built to predict the terror groups for those incidents unclassified."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Appendix"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Various features of the data are analyzed to understand more about the terror incidents. \n",
    "More detailed analysis invloving effect of each feature on the class label - Terror group can also be done in future."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Analysis of the predicions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A plot of the top 10 predicted class labels show that the AlQaida in Iraq, ISIL, TTP and Taliban are the groups having the highest count amongst the predicted class labels.\n",
    "\n",
    "This shows that the terrorist incidents classified as 'Unknown' have a good probability of being classified under these 4 groups.\n",
    "       \n",
    "We can also say that a large portions of the attacks by these groups have gone unattributed.\n",
    "\n",
    "<b>(Subject to accuracy of the predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x44e700b8>"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAr4AAAFgCAYAAABdW3q3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nREQsQcFXRERERCzh/wD2UN99tsKkogAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x44b50b70>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(y= df_test.gname.value_counts().index[:10], x = df_test.gname.value_counts()[:10] ,data=df_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Analysis of the Terrorism incidents Country-wise"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The following histogram of the countrywise terror incidents show that most incidents have taken place in Iraq, Pakistand and India. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x4701d470>"
      ]
     },
     "execution_count": 145,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAikAAAFgCAYAAAB+LOknAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x4701dbe0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(y= df_gtd.country_txt.value_counts().index[:10], x = df_gtd.country.value_counts()[:10])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Analysis of the Terrorism incidents Region-wise"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The following histogram of the regionwise terror incidents show that major proportion of incidents have taken place in \n",
    "Middle East & North Africa, South Asia and South America. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x4ab61f98>"
      ]
     },
     "execution_count": 146,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmMAAAFgCAYAAAD3tH5OAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x4a216320>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(y= df_gtd.region_txt.value_counts().index[:10], x = df_gtd.region.value_counts()[:10])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Analysis of the Terrorism incidents Target-wise"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The following histogram show that major proportion of targets have been at the Military or the Police or at Unnamed civilians."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x53adc278>"
      ]
     },
     "execution_count": 166,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsQAAAFRCAYAAABzFrvuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nocce2gAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x5330f7f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(y= df_gtd.targsubtype1_txt.value_counts().index[:10], x = df_train.targsubtype1.value_counts().values[:10])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Analysis of the Terrorism incidents Weapon type-wise"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The following histogram show that major proportion of weapons used have been either Explosives, Automatic Weapons or Guns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x50f8b438>"
      ]
     },
     "execution_count": 167,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAApIAAAFRCAYAAADHHE3hAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x507eb438>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(y= df_gtd.weapsubtype1_txt.value_counts().index[:10], x = df_train.weapsubtype1.value_counts().values[:10])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}