diff_with_cc_multiple_baselines_separate_impacts.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import os\n",
    "from lib_for_growth_model import *\n",
    "from decimal import Decimal\n",
    "import numpy as np\n",
    "from scipy.stats import gaussian_kde\n",
    "from pandas_helper import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from pandas.io import wb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "codes = read_csv('wbccodes2014.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "ini_data = pd.read_csv(\"ini_data_info_may14.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['countrycode', 'wbreg', 'shareemp', 'shareag', 'sharemanu', 'poorag',\n",
       "       'poorindus', 'poorserv', 'pooremp', 'poorold', 'poorunemp',\n",
       "       'poorchildren', 'below125', 'pooradult', 'skillpag', 'skillpserv',\n",
       "       'skillpmanu', 'gdp', 'gdppc', 'quintile2', 'quintilepc2', 'below4',\n",
       "       'below6', 'below8', 'below10', 'gini', 'tot_pop', 'gap', 'incbott20',\n",
       "       'inc2040', 'incbott40', 'avincome', 'shbottom20ag', 'cat7workers',\n",
       "       'old', 'nonaginc', 'aginc'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ini_data.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def correct_countrycode(countrycode):\n",
    "    '''\n",
    "    Corrects countrycodes in the database that don't correspond to official 3 letters codes.\n",
    "    '''\n",
    "    if countrycode=='TMP':\n",
    "        countrycode='TLS'\n",
    "    if countrycode=='ZAR':\n",
    "        countrycode='COD'\n",
    "    if countrycode=='ROM':\n",
    "        countrycode='ROU'\n",
    "    return countrycode"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "codes['country'] = codes.country.apply(correct_countrycode)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "nameofthisround = 'sept2016_separate_impacts'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "model = os.getcwd()\n",
    "bau_folder   = \"{}/baselines_{}/\".format(model,nameofthisround)\n",
    "cc_folder    = \"{}/with_cc_{}/\".format(model,nameofthisround)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "outputs_spark_cc_0_100.csv\n",
      "outputs_spark_cc_100_200.csv\n",
      "outputs_spark_cc_200_300.csv\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Miniconda3\\lib\\site-packages\\pandas\\io\\parsers.py:1170: DtypeWarning: Columns (6,7,8,9,10) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "  data = self._reader.read(nrows)\n"
     ]
    }
   ],
   "source": [
    "list_bau = os.listdir(bau_folder)\n",
    "list_cc  = os.listdir(cc_folder)\n",
    "\n",
    "bau = pd.DataFrame()\n",
    "cc  = pd.DataFrame()\n",
    "\n",
    "for file in list_bau:\n",
    "    bau = bau.append(pd.read_csv(bau_folder+file).drop('Unnamed: 0',axis=1))\n",
    "    \n",
    "for file in list_cc:\n",
    "    temp = pd.read_csv(cc_folder+file).drop('Unnamed: 0',axis=1)\n",
    "    if \"disasters\" not in file:\n",
    "        print(file)\n",
    "        temp = temp.ix[temp.switch_disasters==0,:]\n",
    "    cc = cc.append(temp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "bau = bau.drop_duplicates(['country', 'scenar', 'ssp'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "cc = cc.dropna()\n",
    "bau = bau.dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "for col_switch in ['switch_ag_rev','switch_temp', 'switch_ag_prices', 'switch_disasters', 'switch_health']:\n",
    "    cc.ix[cc[col_switch],\"switch\"] = col_switch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "cc['issp5'] = cc.ssp=='ssp5'\n",
    "bau['issp5'] = bau.ssp=='ssp5'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "bau['countryname'] = bau.country.replace(codes.set_index('country').country_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "cc['countryname'] = cc.country.replace(codes.set_index('country').country_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "ini_data['countryname'] = ini_data.countrycode.replace(codes.set_index('country').country_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "gni = wb.download(indicator='NY.GNP.PCAP.PP.CD',start=2015,end=2015,country='all').reset_index().rename(columns={'NY.GNP.PCAP.PP.CD':'gni'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "cc['inigdppc'] = cc.countryname.replace(gni.set_index('country').gni)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "cc['inigdp_survey'] = cc.countryname.replace(ini_data.set_index('countryname').gdp)\n",
    "cc['inigdppc_survey'] = cc.countryname.replace(ini_data.set_index('countryname').gdppc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "cc['ini_poor'] = cc.countryname.replace(ini_data.set_index('countryname').below125)\n",
    "cc['ini_pop']  = cc.countryname.replace(ini_data.set_index('countryname').tot_pop)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "hop = cc.set_index(['countryname', 'scenar', 'ssp','ccint','switch'])\n",
    "hip = broadcast_simple(bau.set_index(['countryname', 'scenar', 'ssp']),\\\n",
    "                       hop.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "diff_poor_ag = (hop['below125']-hip['below125']).sum(level=['ccint','ssp','scenar','switch'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "diff_poor = (hop['below125']-hip['below125'])\n",
    "diff_poor_pc = (hop['below125']-hip['below125'])/hop['tot_pop']\n",
    "diff_poor_pb = (hop['below125']-hip['below125'])/hip['below125']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "hop['inc20diffpc'] = (hop.incbott20-hip.incbott20)/hip.incbott20\n",
    "hop['ginidiff'] = (hop.gini-hip.gini)\n",
    "hop['diff_poor'] = diff_poor\n",
    "hop['diff_poor_pc'] = diff_poor_pc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count      552000.000000\n",
       "mean       140301.251413\n",
       "std        941178.224553\n",
       "min      -9293217.537954\n",
       "25%             0.000000\n",
       "50%           765.043596\n",
       "75%         26129.532364\n",
       "max      49693549.555117\n",
       "Name: diff_poor, dtype: float64"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hop.diff_poor.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "hop['incbott40diff'] = (hop.incbott40-hip.incbott40)/hip.incbott40\n",
    "hop['avincomediff'] = (hop.avincome-hip.avincome)/hip.avincome"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "hop['GDPpcdiff'] = (hop.GDP/hop.tot_pop-hip.GDP/hip.tot_pop)/(hip.GDP/hip.tot_pop)\n",
    "hop['GDPdiff'] = (hop.GDP-hip.GDP)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "hop['GDPbau'] = hip.GDP\n",
    "hop['poorbau'] = hip.below125\n",
    "hop['poorbaupc'] = hip.below125/hip.tot_pop\n",
    "hop['GDPpcbau'] = hip.GDP/hip.tot_pop"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "hop['growthpc_bau'] = (hop.GDPpcbau/hop.inigdppc_survey)**(1/(2030-2007))-1\n",
    "hop['growthpc_cc'] = (hop.GDP/hop.tot_pop/hop.inigdppc_survey)**(1/(2030-2007))-1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "hop['pov_reduc_bau'] = hop.ini_poor/hop.ini_pop-hop.below125/hop.tot_pop"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "hop['diff_below10'] = (hop['below10']-hip['below10'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['country', 'year', 'switch_ag_rev', 'switch_temp', 'switch_ag_prices',\n",
       "       'switch_disasters', 'switch_health', 'GDP', 'avincome', 'incbott10',\n",
       "       'incbott20', 'inc2040', 'incbott40', 'quintilecum1', 'quintilecum2',\n",
       "       'quintilepc1', 'quintilepc2', 'below125', 'below2', 'below4', 'below6',\n",
       "       'below8', 'below10', 'gini', 'tot_pop', 'gap125', 'gap2', 'gap4',\n",
       "       'gap6', 'gap8', 'gap10', 'childrenag', 'childrenonag', 'peopleag',\n",
       "       'peoplenonag', 'avincomeag', 'avincomenonag', 'shareag', 'sharemanu',\n",
       "       'shareemp', 'grserv', 'grag', 'grmanu', 'skillpserv', 'skillpag',\n",
       "       'skillpmanu', 'p', 'b', 'voice', 'fprice_increase', 'farmer_rev_change',\n",
       "       'transmission', 'food_cursor', 'temp_impact', 'paramshstunt',\n",
       "       'malaria_cursor', 'diarrhea_cursor', 'disasters_cursor',\n",
       "       'malaria_share', 'diarrhea_share', 'flood_share_poor',\n",
       "       'flood_share_nonpoor', 'drought_share', 'wind_share', 'surge_share',\n",
       "       'prod_gr_ag', 'prod_gr_serv', 'prod_gr_manu', 'issp5', 'inigdppc',\n",
       "       'inigdp_survey', 'inigdppc_survey', 'ini_poor', 'ini_pop',\n",
       "       'inc20diffpc', 'ginidiff', 'diff_poor', 'diff_poor_pc', 'incbott40diff',\n",
       "       'avincomediff', 'GDPpcdiff', 'GDPdiff', 'GDPbau', 'poorbau',\n",
       "       'poorbaupc', 'GDPpcbau', 'growthpc_bau', 'growthpc_cc', 'pov_reduc_bau',\n",
       "       'diff_below10'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hop.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "hop[['country', 'year', 'switch_ag_rev', 'switch_temp', 'switch_ag_prices',\\\n",
    "       'switch_disasters', 'switch_health', 'shareag', 'sharemanu',\\\n",
    "       'shareemp', 'grserv', 'grag', 'grmanu', 'skillpserv', 'skillpag',\\\n",
    "       'skillpmanu', 'p', 'b', 'voice', 'fprice_increase', 'farmer_rev_change',\\\n",
    "       'transmission', 'food_cursor', 'temp_impact', 'paramshstunt',\\\n",
    "       'malaria_cursor', 'diarrhea_cursor', 'disasters_cursor',\\\n",
    "       'malaria_share', 'diarrhea_share', 'flood_share_poor',\\\n",
    "       'flood_share_nonpoor', 'drought_share', 'wind_share', 'surge_share',\\\n",
    "       'prod_gr_ag', 'prod_gr_serv', 'prod_gr_manu', 'issp5', 'inc20diffpc', 'ginidiff', 'diff_poor', 'diff_poor_pc', 'incbott40diff',\n",
    "       'diff_below10']].to_csv(\"all_results_separate_impacts.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from scipy.interpolate import interp1d"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def sample_output(var,countryname,ccint,switch,rand):\n",
    "    \n",
    "    Z = hop[var].unstack(\"ccint\").unstack(\"switch\").unstack(\"countryname\")[ccint][switch][countryname].values\n",
    "\n",
    "    H,X1 = np.histogram( Z, bins = 500, normed = True )\n",
    "    dx = X1[1] - X1[0]\n",
    "    F1 = np.cumsum(H)*dx\n",
    "    \n",
    "    out = interp1d(F1,X1[1:], fill_value=F1[-1], bounds_error=False)(rand)\n",
    "    return out"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 228,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "list_countries = hop.index.get_level_values('countryname').unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 229,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "rand=np.random.rand(1*10**(4),len(list_countries))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 230,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def build_global_df(var,ccint,switch):\n",
    "    output_0 = pd.DataFrame(columns=list_countries)\n",
    "    for (i,countryname) in enumerate(list_countries):\n",
    "        output_0[countryname] = sample_output(var,countryname,ccint,switch,rand[:,i])\n",
    "    output = pd.DataFrame(output_0.sum(axis=1),columns=[var])\n",
    "    output.index.name=\"scenar\"\n",
    "    output['ccint'] = ccint\n",
    "    output['switch'] = switch\n",
    "    return output"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 231,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "a = pd.DataFrame()\n",
    "for ccint in [0,1]:\n",
    "    for switch in ['switch_ag_rev','switch_temp', 'switch_ag_prices', 'switch_disasters', 'switch_health']:\n",
    "        a = pd.concat([a,build_global_df('GDPdiff',ccint,switch)],axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 232,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "b = pd.DataFrame()\n",
    "for ccint in [0,1]:\n",
    "    for switch in ['switch_ag_rev','switch_temp', 'switch_ag_prices', 'switch_disasters', 'switch_health']:\n",
    "        b = pd.concat([b,build_global_df('diff_poor',ccint,switch)],axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 233,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "world = pd.concat([a.reset_index().set_index([\"scenar\",\"switch\",\"ccint\"]),b.reset_index().set_index([\"scenar\",\"switch\",\"ccint\"])],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "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></th>\n",
       "      <th></th>\n",
       "      <th>GDPdiff</th>\n",
       "      <th>diff_poor</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>scenar</th>\n",
       "      <th>switch</th>\n",
       "      <th>ccint</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>9990</th>\n",
       "      <th>switch_health</th>\n",
       "      <th>1</th>\n",
       "      <td>-2.000712e+10</td>\n",
       "      <td>33996953.668139</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9991</th>\n",
       "      <th>switch_health</th>\n",
       "      <th>1</th>\n",
       "      <td>-2.192225e+10</td>\n",
       "      <td>31425767.799913</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9992</th>\n",
       "      <th>switch_health</th>\n",
       "      <th>1</th>\n",
       "      <td>-2.796952e+10</td>\n",
       "      <td>34684351.663752</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9993</th>\n",
       "      <th>switch_health</th>\n",
       "      <th>1</th>\n",
       "      <td>-2.576566e+10</td>\n",
       "      <td>33004513.893345</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9994</th>\n",
       "      <th>switch_health</th>\n",
       "      <th>1</th>\n",
       "      <td>-2.680103e+10</td>\n",
       "      <td>27859473.654297</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9995</th>\n",
       "      <th>switch_health</th>\n",
       "      <th>1</th>\n",
       "      <td>-3.542346e+10</td>\n",
       "      <td>24058820.722997</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9996</th>\n",
       "      <th>switch_health</th>\n",
       "      <th>1</th>\n",
       "      <td>-3.110348e+10</td>\n",
       "      <td>32684049.958394</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9997</th>\n",
       "      <th>switch_health</th>\n",
       "      <th>1</th>\n",
       "      <td>-1.961155e+10</td>\n",
       "      <td>31577851.996707</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9998</th>\n",
       "      <th>switch_health</th>\n",
       "      <th>1</th>\n",
       "      <td>-2.517932e+10</td>\n",
       "      <td>28174223.534819</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9999</th>\n",
       "      <th>switch_health</th>\n",
       "      <th>1</th>\n",
       "      <td>-1.924988e+10</td>\n",
       "      <td>39867606.020999</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                 GDPdiff        diff_poor\n",
       "scenar switch        ccint                               \n",
       "9990   switch_health 1     -2.000712e+10  33996953.668139\n",
       "9991   switch_health 1     -2.192225e+10  31425767.799913\n",
       "9992   switch_health 1     -2.796952e+10  34684351.663752\n",
       "9993   switch_health 1     -2.576566e+10  33004513.893345\n",
       "9994   switch_health 1     -2.680103e+10  27859473.654297\n",
       "9995   switch_health 1     -3.542346e+10  24058820.722997\n",
       "9996   switch_health 1     -3.110348e+10  32684049.958394\n",
       "9997   switch_health 1     -1.961155e+10  31577851.996707\n",
       "9998   switch_health 1     -2.517932e+10  28174223.534819\n",
       "9999   switch_health 1     -1.924988e+10  39867606.020999"
      ]
     },
     "execution_count": 234,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "world.tail(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 235,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "world.to_csv(\"world_distributions.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "world = pd.read_csv(\"world_distributions.csv\").set_index([\"scenar\",\"switch\",\"ccint\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      ":0: FutureWarning: IPython widgets are experimental and may change in the future.\n"
     ]
    }
   ],
   "source": [
    "from lib_for_plot import *\n",
    "import seaborn as sns\n",
    "\n",
    "sns.set_context(\"poster\",rc={\"font.size\": 18})\n",
    "sns.set_style(\"whitegrid\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import matplotlib.gridspec as gridspec"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAvsAAAFjCAYAAACjXMT9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\njX1JkiRJkiRJyqf+Dx7pvsK4eAvWAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xa7f8438>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "font = {'family' : 'sans serif',\n",
    "    'size'   : 16}\n",
    "plt.rc('font', **font)\n",
    "\n",
    "titles = [\"Agriculture\\n revenues\",\"Labor\\nproductivity\",\"Food prices\",\"Disasters\",\"Health\"]\n",
    "\n",
    "fig = plt.figure(figsize=(12,5))\n",
    "gs1 = gridspec.GridSpec(1, 5)\n",
    "gs1.update(wspace=0.025, hspace=0.05)\n",
    "i=0\n",
    "\n",
    "for switch in ['switch_ag_rev','switch_temp', 'switch_ag_prices', 'switch_disasters', 'switch_health']:\n",
    "    ax = fig.add_subplot(gs1[i])\n",
    "    ax.get_xaxis().set_ticks([])\n",
    "    ax.set_ylim([-10,120])\n",
    "    spine_and_ticks(ax)\n",
    "    #ax.set_aspect('equal')\n",
    "\n",
    "    to_plot = pd.DataFrame(world.diff_poor.unstack(\"switch\")[switch].reset_index().replace(dict({0:'Low',1:'High'})))\n",
    "    to_plot[switch] = 10**(-6)*to_plot[switch]\n",
    "\n",
    "    to_plot['prout']=\"\"\n",
    "\n",
    "    sns.violinplot(x='prout',y=switch,data=to_plot,hue='ccint',split=True,ax=ax,palette=\"Set2\",legend=False)\n",
    "    plt.xlabel(titles[i])\n",
    "    if i==0:\n",
    "        ax.legend(title=\"Climate change\\nimpacts\",prop=font,bbox_to_anchor=(1.12, 1.02))\n",
    "    else:\n",
    "        ax.legend_.remove()\n",
    "    sns.despine(bottom=True)\n",
    "    if i==0:\n",
    "        plt.ylabel(\"People living in extreme poverty\\ndue to climate change (million, 2030)\")\n",
    "    else:\n",
    "        plt.ylabel(\"\")\n",
    "        ax.set_yticklabels([])\n",
    "        #ax.get_yaxis().set_visible(False)\n",
    "    \n",
    "    #plt.xlabel(\" \".join(switch.split('_')[1::]))\n",
    "    i+=1\n",
    "plt.savefig(\"graphs_separate_impacts/global_poor_all_switches_new.png\",bbox_inches=\"tight\",dpi=300)\n",
    "plt.savefig(\"graphs_separate_impacts/global_poor_all_switches_new.pdf\",bbox_inches=\"tight\",dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "hop.ix[hop.avincomediff<hop.incbott40diff,:].reset_index().countryname.value_counts();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [default]",
   "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": "3.5.2"
  },
  "toc": {
   "colors": {
    "hover_highlight": "#DAA520",
    "navigate_num": "#000000",
    "navigate_text": "#333333",
    "running_highlight": "#FF0000",
    "selected_highlight": "#FFD700",
    "sidebar_border": "#EEEEEE",
    "wrapper_background": "#FFFFFF"
   },
   "moveMenuLeft": true,
   "nav_menu": {
    "height": "12px",
    "width": "252px"
   },
   "navigate_menu": true,
   "number_sections": true,
   "sideBar": true,
   "threshold": 4,
   "toc_cell": false,
   "toc_section_display": "block",
   "toc_window_display": false,
   "widenNotebook": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}