Script to get Global Peace Index (GPI) data from WIkipedia as pandas DF

get_gpi_from_wiki.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import requests\n",
    "from bs4 import BeautifulSoup\n",
    "\n",
    "response = requests.get(url='https://en.wikipedia.org/wiki/Global_Peace_Index')\n",
    "\n",
    "soup = BeautifulSoup(response.text, 'html.parser')\n",
    "\n",
    "base_year = 2016\n",
    "years = 9\n",
    "\n",
    "def get_countries_by_gpi():\n",
    "    for table in soup.find_all('table'):\n",
    "        if table.find_all('th')[0].get_text() == 'Country':\n",
    "            for tr in table.find_all('tr'):\n",
    "                country_name = tr.find_all('a')[0].get_text()\n",
    "                if not country_name.startswith('['):\n",
    "                    row = {'country': country_name}\n",
    "                    for year, index in zip(range(base_year-years+1,base_year+1),range(2*(years),0,-2)):\n",
    "                        score = tr.find_all('td')[index].get_text()\n",
    "                        if score != '':\n",
    "                            row['score_%s' % year] = float(score)\n",
    "                    yield row\n",
    "\n",
    "import pandas as pd\n",
    "gpi = pd.DataFrame.from_dict(list(get_countries_by_gpi()))\n",
    "gpi.to_csv('gpi_%s-%s.csv' % (base_year-years+1,base_year),index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "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>country</th>\n",
       "      <th>score_2008</th>\n",
       "      <th>score_2009</th>\n",
       "      <th>score_2010</th>\n",
       "      <th>score_2011</th>\n",
       "      <th>score_2012</th>\n",
       "      <th>score_2013</th>\n",
       "      <th>score_2014</th>\n",
       "      <th>score_2015</th>\n",
       "      <th>score_2016</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Iceland</td>\n",
       "      <td>1.107</td>\n",
       "      <td>1.203</td>\n",
       "      <td>1.212</td>\n",
       "      <td>1.148</td>\n",
       "      <td>1.113</td>\n",
       "      <td>1.162</td>\n",
       "      <td>1.189</td>\n",
       "      <td>1.148</td>\n",
       "      <td>1.192</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Denmark</td>\n",
       "      <td>1.238</td>\n",
       "      <td>1.263</td>\n",
       "      <td>1.341</td>\n",
       "      <td>1.289</td>\n",
       "      <td>1.239</td>\n",
       "      <td>1.207</td>\n",
       "      <td>1.193</td>\n",
       "      <td>1.150</td>\n",
       "      <td>1.246</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Austria</td>\n",
       "      <td>1.291</td>\n",
       "      <td>1.240</td>\n",
       "      <td>1.290</td>\n",
       "      <td>1.337</td>\n",
       "      <td>1.328</td>\n",
       "      <td>1.250</td>\n",
       "      <td>1.200</td>\n",
       "      <td>1.198</td>\n",
       "      <td>1.278</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>New Zealand</td>\n",
       "      <td>1.190</td>\n",
       "      <td>1.227</td>\n",
       "      <td>1.188</td>\n",
       "      <td>1.279</td>\n",
       "      <td>1.239</td>\n",
       "      <td>1.237</td>\n",
       "      <td>1.236</td>\n",
       "      <td>1.221</td>\n",
       "      <td>1.287</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Portugal</td>\n",
       "      <td>1.385</td>\n",
       "      <td>1.426</td>\n",
       "      <td>1.366</td>\n",
       "      <td>1.453</td>\n",
       "      <td>1.470</td>\n",
       "      <td>1.467</td>\n",
       "      <td>1.425</td>\n",
       "      <td>1.344</td>\n",
       "      <td>1.356</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       country  score_2008  score_2009  score_2010  score_2011  score_2012  \\\n",
       "0      Iceland       1.107       1.203       1.212       1.148       1.113   \n",
       "1      Denmark       1.238       1.263       1.341       1.289       1.239   \n",
       "2      Austria       1.291       1.240       1.290       1.337       1.328   \n",
       "3  New Zealand       1.190       1.227       1.188       1.279       1.239   \n",
       "4     Portugal       1.385       1.426       1.366       1.453       1.470   \n",
       "\n",
       "   score_2013  score_2014  score_2015  score_2016  \n",
       "0       1.162       1.189       1.148       1.192  \n",
       "1       1.207       1.193       1.150       1.246  \n",
       "2       1.250       1.200       1.198       1.278  \n",
       "3       1.237       1.236       1.221       1.287  \n",
       "4       1.467       1.425       1.344       1.356  "
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gpi[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}