Microeconometrics 2, part a

gistfile1.txt

{
 "metadata": {
  "name": "Ejercicio 2- Parte a - con familia-pscore"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import pandas as pd\n",
      "from scipy import stats as st\n",
      "from statsmodels.stats.weightstats import CompareMeans, DescrStatsW\n",
      "from statsmodels.iolib import genfromdta\n",
      "import statsmodels.api as sm\n",
      "import matplotlib.cm as cm\n",
      "import matplotlib as mpl\n",
      "import brewer2mpl\n",
      "import matplotlib.pyplot as plt\n",
      "import numpy as np\n",
      "from IPython.display import display, HTML\n",
      "import pscore_match as ps\n",
      "pd.set_option('display.max_columns', 10)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 3
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "aguas = pd.DataFrame(genfromdta('baseaguas2.dta'))\n",
      "bid = pd.read_csv(\"base_bid.csv\")\n",
      "bidaguas = pd.read_csv(\"matching_bidaguas.csv\")\n",
      "bidaguas.sexo = (bidaguas.sexo == 'masculino').astype(int)\n",
      "bidaguas.bedad = bidaguas.bedad.astype(int)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 4
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Mean Equality Test"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def mean_equality_test():\n",
      "    rows = []\n",
      "    covariates = bid.loc[bid.jefe==1, ['idhogar', 'treatment2', 'jefe', 'gender', 's', 'casado', 'extranjero', \\\n",
      "            'empleoHH', 'desempleoHH', 'bedad', 'electricidad', 'techo_malo', 'Ypcf']]\n",
      "    treated, control = covariates.treatment2 ==1, covariates.treatment2 ==0\n",
      "    for column in ['gender', 's', 'casado', 'extranjero', 'empleoHH', 'desempleoHH', 'bedad', 'electricidad', 'techo_malo', 'Ypcf' ]:\n",
      "        x = covariates[column]\n",
      "        t, c = x[treated].dropna(), x[control].dropna()\n",
      "        rows.append([column, t.mean(), c.mean(), t.mean() - c.mean(), st.ttest_ind(t, c)[1]])\n",
      "    return pd.DataFrame(rows, columns= ['name', 'treated', 'control', 'diff', 'p-value'])"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 1
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "mean_equality_test()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
        "<table border=\"1\" class=\"dataframe\">\n",
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th></th>\n",
        "      <th>name</th>\n",
        "      <th>treated</th>\n",
        "      <th>control</th>\n",
        "      <th>diff</th>\n",
        "      <th>p-value</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>0</th>\n",
        "      <td>       gender</td>\n",
        "      <td>   0.789054</td>\n",
        "      <td>   0.823691</td>\n",
        "      <td> -0.034638</td>\n",
        "      <td> 0.166190</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>1</th>\n",
        "      <td>            s</td>\n",
        "      <td>   7.297330</td>\n",
        "      <td>   7.338150</td>\n",
        "      <td> -0.040820</td>\n",
        "      <td> 0.805589</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>2</th>\n",
        "      <td>       casado</td>\n",
        "      <td>   0.729761</td>\n",
        "      <td>   0.779614</td>\n",
        "      <td> -0.049854</td>\n",
        "      <td> 0.067164</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>3</th>\n",
        "      <td>   extranjero</td>\n",
        "      <td>   0.135690</td>\n",
        "      <td>   0.082645</td>\n",
        "      <td>  0.053045</td>\n",
        "      <td> 0.008924</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>4</th>\n",
        "      <td>     empleoHH</td>\n",
        "      <td>   0.895833</td>\n",
        "      <td>   0.891892</td>\n",
        "      <td>  0.003941</td>\n",
        "      <td> 0.852647</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>5</th>\n",
        "      <td>  desempleoHH</td>\n",
        "      <td>   0.104167</td>\n",
        "      <td>   0.108108</td>\n",
        "      <td> -0.003941</td>\n",
        "      <td> 0.852647</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>6</th>\n",
        "      <td>        bedad</td>\n",
        "      <td>  41.964652</td>\n",
        "      <td>  45.757576</td>\n",
        "      <td> -3.792924</td>\n",
        "      <td> 0.000007</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>7</th>\n",
        "      <td> electricidad</td>\n",
        "      <td>   0.972727</td>\n",
        "      <td>   0.984536</td>\n",
        "      <td> -0.011809</td>\n",
        "      <td> 0.368092</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>8</th>\n",
        "      <td>   techo_malo</td>\n",
        "      <td>   0.056911</td>\n",
        "      <td>   0.065934</td>\n",
        "      <td> -0.009023</td>\n",
        "      <td> 0.675279</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>9</th>\n",
        "      <td>         Ypcf</td>\n",
        "      <td> 123.601713</td>\n",
        "      <td> 165.438165</td>\n",
        "      <td>-41.836453</td>\n",
        "      <td> 0.000011</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
        "</div>"
       ],
       "output_type": "pyout",
       "prompt_number": 5,
       "text": [
        "           name     treated     control       diff   p-value\n",
        "0        gender    0.789054    0.823691  -0.034638  0.166190\n",
        "1             s    7.297330    7.338150  -0.040820  0.805589\n",
        "2        casado    0.729761    0.779614  -0.049854  0.067164\n",
        "3    extranjero    0.135690    0.082645   0.053045  0.008924\n",
        "4      empleoHH    0.895833    0.891892   0.003941  0.852647\n",
        "5   desempleoHH    0.104167    0.108108  -0.003941  0.852647\n",
        "6         bedad   41.964652   45.757576  -3.792924  0.000007\n",
        "7  electricidad    0.972727    0.984536  -0.011809  0.368092\n",
        "8    techo_malo    0.056911    0.065934  -0.009023  0.675279\n",
        "9          Ypcf  123.601713  165.438165 -41.836453  0.000011"
       ]
      }
     ],
     "prompt_number": 5
    },
    {
     "cell_type": "heading",
     "level": 1,
     "metadata": {},
     "source": [
      "Selecci\u00f3n de ni\u00f1o de hasta 6 a\u00f1os"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "ninos = bid[bid.bedad <= 6].copy()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 5
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def extract_result(endog, result, ingreso):\n",
      "    return endog.name, endog.mean(), result.params['ty'], result.pvalues['ty'], result.nobs, result.rsquared, ingreso"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 12
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def dummyfy(dataframe, column):\n",
      "    answer = dataframe.drop(column, axis=1)\n",
      "    for value in dataframe[column].unique()[1:]:\n",
      "        answer[str(value)] = dataframe[column] == value\n",
      "    return answer"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 49
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def regress(dataset, column, ingreso, rivera = True):\n",
      "    if rivera:\n",
      "        treatment = 'treatment2'\n",
      "    else:\n",
      "        treatment = 'treatment1'\n",
      "    columns = [column, treatment, 'year', 'idhogar']\n",
      "    if ingreso:\n",
      "        columns.append('Ypcf')\n",
      "    filtered = dataset.loc[:, columns].dropna()\n",
      "    endog = filtered[column]\n",
      "    exog = sm.add_constant(filtered, prepend=True)\n",
      "    exog['year2'] = exog.year==2005\n",
      "    exog['ty'] = exog.year2*exog[treatment]\n",
      "    exog = dummyfy(exog, 'idhogar').drop([column, 'year'], axis=1)\n",
      "    ols = sm.OLS(endog, exog)\n",
      "    return extract_result(endog, ols.fit(), ingreso)\n",
      "\n",
      "def regress_multiple_columns(dataset, columns, rivera):\n",
      "    results = []\n",
      "    for column in columns:\n",
      "        results.extend([regress(dataset, column, ingreso, rivera) for ingreso in [True, False]])\n",
      "    return results\n",
      "    "
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 67
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def replicate(rivera, dataset=bid):\n",
      "    ninos = dataset[dataset.bedad <= 6]\n",
      "    hogares = dataset[dataset.jefe==1]\n",
      "    ninos_columns = ['sangre', 'hijodi', 'durdiarrea1']\n",
      "    hogares_columns = ['distancia', 'gastobidon1', 'gbidon1']\n",
      "    results = regress_multiple_columns(ninos, ninos_columns, rivera)\n",
      "    results += regress_multiple_columns(hogares, hogares_columns, rivera)\n",
      "    columns = ['Dependent variable', 'Mean dependent variable', '$I_{it}$', 'p-value', 'Observations', 'R-squared', 'Per capita income']\n",
      "    return pd.DataFrame(results, columns = columns)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 70
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Replicaci\u00f3n del an\u00e1nlisis sobre los efectos del programa MPG incluyendo el barrio La Rivera"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "replications = replicate(rivera = True)\n",
      "HTML(replications.to_html())"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<table border=\"1\" class=\"dataframe\">\n",
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th></th>\n",
        "      <th>Dependent variable</th>\n",
        "      <th>Mean dependent variable</th>\n",
        "      <th>$I_{it}$</th>\n",
        "      <th>p-value</th>\n",
        "      <th>Observations</th>\n",
        "      <th>R-squared</th>\n",
        "      <th>Per capita income</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>0 </th>\n",
        "      <td>      sangre</td>\n",
        "      <td> 0.050926</td>\n",
        "      <td> -0.081423</td>\n",
        "      <td> 0.079276</td>\n",
        "      <td> 648</td>\n",
        "      <td> 0.466693</td>\n",
        "      <td>  True</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>1 </th>\n",
        "      <td>      sangre</td>\n",
        "      <td> 0.045232</td>\n",
        "      <td> -0.064633</td>\n",
        "      <td> 0.048597</td>\n",
        "      <td> 818</td>\n",
        "      <td> 0.457392</td>\n",
        "      <td> False</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>2 </th>\n",
        "      <td>      hijodi</td>\n",
        "      <td> 0.154321</td>\n",
        "      <td> -0.127408</td>\n",
        "      <td> 0.067109</td>\n",
        "      <td> 648</td>\n",
        "      <td> 0.555897</td>\n",
        "      <td>  True</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>3 </th>\n",
        "      <td>      hijodi</td>\n",
        "      <td> 0.144254</td>\n",
        "      <td> -0.105212</td>\n",
        "      <td> 0.041804</td>\n",
        "      <td> 818</td>\n",
        "      <td> 0.527932</td>\n",
        "      <td> False</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>4 </th>\n",
        "      <td> durdiarrea1</td>\n",
        "      <td> 1.010802</td>\n",
        "      <td> -1.124065</td>\n",
        "      <td> 0.192419</td>\n",
        "      <td> 648</td>\n",
        "      <td> 0.520855</td>\n",
        "      <td>  True</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>5 </th>\n",
        "      <td> durdiarrea1</td>\n",
        "      <td> 0.925428</td>\n",
        "      <td> -0.941136</td>\n",
        "      <td> 0.116084</td>\n",
        "      <td> 818</td>\n",
        "      <td> 0.504519</td>\n",
        "      <td> False</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>6 </th>\n",
        "      <td>   distancia</td>\n",
        "      <td> 0.790055</td>\n",
        "      <td> -0.880561</td>\n",
        "      <td> 0.001376</td>\n",
        "      <td> 724</td>\n",
        "      <td> 0.979351</td>\n",
        "      <td>  True</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>7 </th>\n",
        "      <td>   distancia</td>\n",
        "      <td> 0.783726</td>\n",
        "      <td> -0.837642</td>\n",
        "      <td> 0.000227</td>\n",
        "      <td> 934</td>\n",
        "      <td> 0.971832</td>\n",
        "      <td> False</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>8 </th>\n",
        "      <td> gastobidon1</td>\n",
        "      <td> 9.457995</td>\n",
        "      <td>-19.723242</td>\n",
        "      <td> 0.000207</td>\n",
        "      <td> 369</td>\n",
        "      <td> 0.879240</td>\n",
        "      <td>  True</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>9 </th>\n",
        "      <td> gastobidon1</td>\n",
        "      <td> 8.873706</td>\n",
        "      <td>-21.867824</td>\n",
        "      <td> 0.000001</td>\n",
        "      <td> 483</td>\n",
        "      <td> 0.847855</td>\n",
        "      <td> False</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>10</th>\n",
        "      <td>     gbidon1</td>\n",
        "      <td> 9.120407</td>\n",
        "      <td>-17.107765</td>\n",
        "      <td> 0.000988</td>\n",
        "      <td> 369</td>\n",
        "      <td> 0.878281</td>\n",
        "      <td>  True</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>11</th>\n",
        "      <td>     gbidon1</td>\n",
        "      <td> 8.574741</td>\n",
        "      <td>-19.246921</td>\n",
        "      <td> 0.000012</td>\n",
        "      <td> 483</td>\n",
        "      <td> 0.846867</td>\n",
        "      <td> False</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>"
       ],
       "output_type": "pyout",
       "prompt_number": 71,
       "text": [
        "<IPython.core.display.HTML at 0x108fefd10>"
       ]
      }
     ],
     "prompt_number": 71
    },
    {
     "cell_type": "heading",
     "level": 1,
     "metadata": {},
     "source": [
      "Replicaci\u00f3n del an\u00e1nlisis sobre los efectos del programa MPG sin incluir el barrio La Rivera"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "replications = replicate(rivera = False)\n",
      "HTML(replications.to_html())"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<table border=\"1\" class=\"dataframe\">\n",
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th></th>\n",
        "      <th>Dependent variable</th>\n",
        "      <th>Mean dependent variable</th>\n",
        "      <th>$I_{it}$</th>\n",
        "      <th>p-value</th>\n",
        "      <th>Observations</th>\n",
        "      <th>R-squared</th>\n",
        "      <th>Per capita income</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>0 </th>\n",
        "      <td>      sangre</td>\n",
        "      <td> 0.044883</td>\n",
        "      <td> -0.044141</td>\n",
        "      <td> 0.334344</td>\n",
        "      <td> 557</td>\n",
        "      <td> 0.491076</td>\n",
        "      <td>  True</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>1 </th>\n",
        "      <td>      sangre</td>\n",
        "      <td> 0.037762</td>\n",
        "      <td> -0.034826</td>\n",
        "      <td> 0.267953</td>\n",
        "      <td> 715</td>\n",
        "      <td> 0.460464</td>\n",
        "      <td> False</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>2 </th>\n",
        "      <td>      hijodi</td>\n",
        "      <td> 0.138241</td>\n",
        "      <td> -0.089605</td>\n",
        "      <td> 0.200103</td>\n",
        "      <td> 557</td>\n",
        "      <td> 0.572033</td>\n",
        "      <td>  True</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>3 </th>\n",
        "      <td>      hijodi</td>\n",
        "      <td> 0.124476</td>\n",
        "      <td> -0.085952</td>\n",
        "      <td> 0.088159</td>\n",
        "      <td> 715</td>\n",
        "      <td> 0.538402</td>\n",
        "      <td> False</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>4 </th>\n",
        "      <td> durdiarrea1</td>\n",
        "      <td> 0.903052</td>\n",
        "      <td> -0.937809</td>\n",
        "      <td> 0.300049</td>\n",
        "      <td> 557</td>\n",
        "      <td> 0.541725</td>\n",
        "      <td>  True</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>5 </th>\n",
        "      <td> durdiarrea1</td>\n",
        "      <td> 0.820979</td>\n",
        "      <td> -0.832825</td>\n",
        "      <td> 0.174150</td>\n",
        "      <td> 715</td>\n",
        "      <td> 0.524452</td>\n",
        "      <td> False</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>6 </th>\n",
        "      <td>   distancia</td>\n",
        "      <td> 0.788162</td>\n",
        "      <td> -0.696224</td>\n",
        "      <td> 0.004643</td>\n",
        "      <td> 642</td>\n",
        "      <td> 0.986442</td>\n",
        "      <td>  True</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>7 </th>\n",
        "      <td>   distancia</td>\n",
        "      <td> 0.799520</td>\n",
        "      <td> -0.736462</td>\n",
        "      <td> 0.000946</td>\n",
        "      <td> 833</td>\n",
        "      <td> 0.977270</td>\n",
        "      <td> False</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>8 </th>\n",
        "      <td> gastobidon1</td>\n",
        "      <td> 9.528701</td>\n",
        "      <td>-15.681490</td>\n",
        "      <td> 0.000076</td>\n",
        "      <td> 331</td>\n",
        "      <td> 0.919842</td>\n",
        "      <td>  True</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>9 </th>\n",
        "      <td> gastobidon1</td>\n",
        "      <td> 9.143519</td>\n",
        "      <td>-17.639268</td>\n",
        "      <td> 0.000000</td>\n",
        "      <td> 432</td>\n",
        "      <td> 0.903737</td>\n",
        "      <td> False</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>10</th>\n",
        "      <td>     gbidon1</td>\n",
        "      <td> 9.235438</td>\n",
        "      <td>-13.313769</td>\n",
        "      <td> 0.000644</td>\n",
        "      <td> 331</td>\n",
        "      <td> 0.918678</td>\n",
        "      <td>  True</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>11</th>\n",
        "      <td>     gbidon1</td>\n",
        "      <td> 8.872917</td>\n",
        "      <td>-15.304367</td>\n",
        "      <td> 0.000004</td>\n",
        "      <td> 432</td>\n",
        "      <td> 0.902217</td>\n",
        "      <td> False</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>"
       ],
       "output_type": "pyout",
       "prompt_number": 72,
       "text": [
        "<IPython.core.display.HTML at 0x108fdd050>"
       ]
      }
     ],
     "prompt_number": 72
    },
    {
     "cell_type": "heading",
     "level": 1,
     "metadata": {},
     "source": [
      "Distribuci\u00f3n del ingreso per capita en ni\u00f1os"
     ]
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Indicios de la necesidad de utilizar una transformaci\u00f3n $\\log$ sobre el ingreso per capita"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "plt.hist(bid[bid.bedad<=6].loc[:, ['sangre', 'treatment2', 'year', 'Ypcf', 'idhogar']].dropna().Ypcf, label=u'Ingreso per capita en ni\u00f1os')\n",
      "plt.legend()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "pyout",
       "prompt_number": 74,
       "text": [
        "<matplotlib.legend.Legend at 0x108fe3050>"
       ]
      },
      {
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAXsAAAD+CAYAAADbNKXUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJsCyJyIyAZY9EZEJsOyJiEzg/wC05i6jr2FuoAAAAABJRU5ErkJggg==\n"
      }
     ],
     "prompt_number": 74
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "plt.hist(bid[bid.bedad <= 6].loc[:, ['sangre', 'treatment2', 'year', 'Ypcf', 'idhogar']].dropna().Ypcf.apply(lambda x: np.log10(x+1)), \\\n",
      "    label='$\\log$' + u' del ingreso percapita en ni\u00f1os')\n",
      "plt.legend()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "pyout",
       "prompt_number": 75,
       "text": [
        "<matplotlib.legend.Legend at 0x109004a90>"
       ]
      },
      {
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAXsAAAD+CAYAAADbNKXUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n+H9LvkVCiS63VQAAAABJRU5ErkJggg==\n"
      }
     ],
     "prompt_number": 75
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Gragico de diferencias en la distrbuci\u00f3n de ingreso per capita entre el grupo de tratados y el grupo de control"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "filtered = bid[bid.bedad <=6].loc[:, ['sangre', 'treatment2', 'year', 'Ypcf', 'idhogar']].dropna()\n",
      "ypcf_t, ypcf_c = [filtered.Ypcf[x].apply(lambda x: np.log10(x+1)) for x in [filtered.treatment2 == 1, filtered.treatment2 == 0]]\n",
      "plt.hist(ypcf_t, label=\"treated\", alpha=0.5, normed=True)\n",
      "plt.hist(ypcf_c, label=\"control\", alpha=0.5, normed=True)\n",
      "plt.legend()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "pyout",
       "prompt_number": 76,
       "text": [
        "<matplotlib.legend.Legend at 0x1096bdf90>"
       ]
      },
      {
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAXsAAAD+CAYAAADbNKXUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nnohIAxjsiYg04P8Bj2PoHX6z1ckAAAAASUVORK5CYII=\n"
      }
     ],
     "prompt_number": 76
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "filtered = bid[bid.bedad<=6].loc[:, ['sangre', 'treatment2', 'year', 'Ypcf', 'idhogar']].dropna()\n",
      "ypcf_t, ypcf_c = [filtered.Ypcf[x].apply(lambda x: np.log10(x+1)) for x in [filtered.treatment2 == 1, filtered.treatment2 == 0]]\n",
      "plt.hist(ypcf_t, label=\"treated\", alpha=0.5, normed=True, cumulative=True)\n",
      "plt.hist(ypcf_c, label=\"control\", alpha=0.5, normed=True, cumulative=True)\n",
      "plt.legend()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "pyout",
       "prompt_number": 77,
       "text": [
        "<matplotlib.legend.Legend at 0x109acfc90>"
       ]
      },
      {
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAXsAAAD+CAYAAADbNKXUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nzZ6ISAPY7ImINIDNnohIA/4Pz1NYxE+REJwAAAAASUVORK5CYII=\n"
      }
     ],
     "prompt_number": 77
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Replicaci\u00f3n utilizando $\\log(Ingreso)$"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "bid_log = bid.copy()\n",
      "bid_log.Ypcf = bid_log.Ypcf.apply(lambda x: np.log10(x+1))\n",
      "replications = replicate(rivera = True, dataset=bid_log)\n",
      "HTML(replications.to_html())"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<table border=\"1\" class=\"dataframe\">\n",
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th></th>\n",
        "      <th>Dependent variable</th>\n",
        "      <th>Mean dependent variable</th>\n",
        "      <th>$I_{it}$</th>\n",
        "      <th>p-value</th>\n",
        "      <th>Observations</th>\n",
        "      <th>R-squared</th>\n",
        "      <th>Per capita income</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>0 </th>\n",
        "      <td>      sangre</td>\n",
        "      <td> 0.050926</td>\n",
        "      <td> -0.080773</td>\n",
        "      <td> 0.081674</td>\n",
        "      <td> 648</td>\n",
        "      <td> 0.467056</td>\n",
        "      <td>  True</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>1 </th>\n",
        "      <td>      sangre</td>\n",
        "      <td> 0.045232</td>\n",
        "      <td> -0.064633</td>\n",
        "      <td> 0.048597</td>\n",
        "      <td> 818</td>\n",
        "      <td> 0.457392</td>\n",
        "      <td> False</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>2 </th>\n",
        "      <td>      hijodi</td>\n",
        "      <td> 0.154321</td>\n",
        "      <td> -0.124856</td>\n",
        "      <td> 0.072520</td>\n",
        "      <td> 648</td>\n",
        "      <td> 0.556843</td>\n",
        "      <td>  True</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>3 </th>\n",
        "      <td>      hijodi</td>\n",
        "      <td> 0.144254</td>\n",
        "      <td> -0.105212</td>\n",
        "      <td> 0.041804</td>\n",
        "      <td> 818</td>\n",
        "      <td> 0.527932</td>\n",
        "      <td> False</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>4 </th>\n",
        "      <td> durdiarrea1</td>\n",
        "      <td> 1.010802</td>\n",
        "      <td> -1.121961</td>\n",
        "      <td> 0.193375</td>\n",
        "      <td> 648</td>\n",
        "      <td> 0.520852</td>\n",
        "      <td>  True</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>5 </th>\n",
        "      <td> durdiarrea1</td>\n",
        "      <td> 0.925428</td>\n",
        "      <td> -0.941136</td>\n",
        "      <td> 0.116084</td>\n",
        "      <td> 818</td>\n",
        "      <td> 0.504519</td>\n",
        "      <td> False</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>6 </th>\n",
        "      <td>   distancia</td>\n",
        "      <td> 0.790055</td>\n",
        "      <td> -0.888537</td>\n",
        "      <td> 0.001202</td>\n",
        "      <td> 724</td>\n",
        "      <td> 0.979362</td>\n",
        "      <td>  True</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>7 </th>\n",
        "      <td>   distancia</td>\n",
        "      <td> 0.783726</td>\n",
        "      <td> -0.837642</td>\n",
        "      <td> 0.000227</td>\n",
        "      <td> 934</td>\n",
        "      <td> 0.971832</td>\n",
        "      <td> False</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>8 </th>\n",
        "      <td> gastobidon1</td>\n",
        "      <td> 9.457995</td>\n",
        "      <td>-19.710362</td>\n",
        "      <td> 0.000205</td>\n",
        "      <td> 369</td>\n",
        "      <td> 0.879310</td>\n",
        "      <td>  True</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>9 </th>\n",
        "      <td> gastobidon1</td>\n",
        "      <td> 8.873706</td>\n",
        "      <td>-21.867824</td>\n",
        "      <td> 0.000001</td>\n",
        "      <td> 483</td>\n",
        "      <td> 0.847855</td>\n",
        "      <td> False</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>10</th>\n",
        "      <td>     gbidon1</td>\n",
        "      <td> 9.120407</td>\n",
        "      <td>-17.094885</td>\n",
        "      <td> 0.000980</td>\n",
        "      <td> 369</td>\n",
        "      <td> 0.878366</td>\n",
        "      <td>  True</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>11</th>\n",
        "      <td>     gbidon1</td>\n",
        "      <td> 8.574741</td>\n",
        "      <td>-19.246921</td>\n",
        "      <td> 0.000012</td>\n",
        "      <td> 483</td>\n",
        "      <td> 0.846867</td>\n",
        "      <td> False</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>"
       ],
       "output_type": "pyout",
       "prompt_number": 79,
       "text": [
        "<IPython.core.display.HTML at 0x108efc1d0>"
       ]
      }
     ],
     "prompt_number": 79
    },
    {
     "cell_type": "heading",
     "level": 1,
     "metadata": {},
     "source": [
      "Propensity Score"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def treatment_effect(column, title):\n",
      "    interval = np.logspace(-2,-3)\n",
      "    means = []\n",
      "    ci_min,ci_max = [], []\n",
      "\n",
      "    ninos = bidaguas.loc[bidaguas.bedad <= 6, ['idhogar', column, 'ypcf_pre1', 'treatment2', 'sexo']].dropna() #, #'sexo', 'bedad']].copy()\n",
      "    for radius in interval:\n",
      "        psmatch = ps.PropensityScoreMatch(ninos.treatment2, ninos.loc[:, ['ypcf_pre1', 'sexo']], ninos[column], algo='radius', radius=radius)\n",
      "        psmatch.fit()\n",
      "        means.append(psmatch.treatment_effect())\n",
      "        ci = psmatch.confidence_interval()\n",
      "        ci_min.append(ci[0])\n",
      "        ci_max.append(ci[1])\n",
      "    plt.plot(interval, means)\n",
      "    plt.ylabel('treatment effect')\n",
      "    plt.xlabel('radius')\n",
      "    plt.title(title + ' (obs: %d)' % ninos[column].count())\n",
      "    fig = plt.fill_between(interval, ci_min,ci_max, alpha=0.2)\n",
      "    #plt.savefig('pscore.png', format='png')\n",
      "    display(fig)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 6
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "treatment_effect('dhijodi', 'presencia de diarrea')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "display_data",
       "text": [
        "<matplotlib.collections.PolyCollection at 0x108aaa350>"
       ]
      },
      {
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAZsAAAEWCAYAAACwtjr+AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEZHqWGyIiEh1/wAlcm9fKho5AgAAAABJRU5ErkJggg==\n"
      }
     ],
     "prompt_number": 7
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "treatment_effect('dsangre', 'severidad de los episodios')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "display_data",
       "text": [
        "<matplotlib.collections.PolyCollection at 0x10967b0d0>"
       ]
      },
      {
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAZsAAAEWCAYAAACwtjr+AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEQnHYkNERMKx2BARkXD/AGsXuNKBgCGjAAAAAElFTkSuQmCC\n"
      }
     ],
     "prompt_number": 5
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "treatment_effect('ddurdiarrea1', u'duraci\u00f3n de los episodios')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "display_data",
       "text": [
        "<matplotlib.collections.PolyCollection at 0x109a1c450>"
       ]
      },
      {
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAZUAAAEWCAYAAACufwpNAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzs3Xl8FPX9P/DXHDt7ZnOQE3JySyAgqIgiIlahohRs0VKL\neFC1Wr/VWrSeqBTq+fWoiAeKaC1+G6pi9Wc9sEgREAWSgNxyBZJNCLk22ew18/n9scmSTbLZSbLH\nJHk/H488wszOzr6yCfPe+RwzHGOMgRBCCAkDPtYBCCGE9B1UVAghhIQNFRVCCCFhQ0WFEEJI2FBR\nIYQQEjZUVAghhIQNFZU+qKGhAUOHDsWf//znWEchpEt2794d6wia5/V6cfDgwVjHCIqKSiterxc8\nz+P48eOxjtKpxx9/HCtWrAj6+D333IP7778fDz30UFTyrFy5EpdccknY9/uzn/0MWVlZ0Ol0WL9+\nfdj3HytvvvkmHnzwwai9XmZmJjZu3NhuvcvlwrRp03Ds2LGoZenMU089haVLl3b5eTzP4/DhwxFI\n1DVff/01Ro0ahaysLGRnZ+MPf/gDPB5PwDZvvfUWzGYzsrKy/NudPn26w/2tWbMGPM/j4YcfDlhf\nW1uLBQsW4JtvvonYz9ITYqwDkK575JFHOn381VdfjVKSyFq3bh0AIC8vDxzHxThN+Nx0001Rfb1g\n751er8dXX30V1SzBrF27Fh9//DE2bNgQ6yjdNnbsWGzevBkJCQmw2Wy49NJLkZCQEPD/leM4XHvt\ntXjzzTc73Vd1dTUefPBBXHLJJe1+f8nJySgsLMSll16K//73v0hJSYnIz9Nd/fpMxel04pZbbkFa\nWhrGjBmDN954o902ubm5AZ+S33rrLVx00UUB2zz66KOYP38+Hn/8ceTl5WHgwIH49NNP/Y/X1NTg\nrrvuQn5+PjIyMlBQUIBNmzYF7MPj8eDJJ59Efn4+srKykJubizVr1gRsM2vWLGRlZcFisbT79AIA\n9fX1uO2225CXl4ecnBzMmzcPlZWV/sePHj0Knufxn//8BxdeeCHS0tIwffp0NDY2dul9KykpwcSJ\nE5Geno5p06bhwIED7bbZt28ffvKTnyAzMxNnn302/vOf/3TpNdT629/+hjFjxiAzMxPnnnsuvvzy\ny4DH7XY7Fi5ciJycHGRnZ2PYsGHdajr4+OOPMXbsWGRlZWH69OntzmZzc3PxwgsvYOrUqUhLS8NF\nF12EH3/8MWCbO+64A1lZWYiPj8f8+fPbvYbH48GiRYswePBgZGdnY/Dgwe3OMA4ePIif/vSnyMrK\nwrBhw/DQQw/B6/X6H2eM4YEHHkBGRgZGjBiBZcuWtXudH3/80f9JOdin/O+++w5TpkxBVlYWRo4c\nieeff77dNk888QRGjBiB7Oxs5Obm4r333uv8TQzC5XLh9ttvx8svvwyeDzwkud1uPPDAAxg2bBiy\ns7Mxc+bMDvN+/fXXmDhxIlJTU3HFFVegoqIi4PE333wTY8aMQXZ2NrKzs/Hss892mOXqq6/GPffc\n062fIyEhAQkJCQCA9PR0XHnlldi2bVvANowxqLmIyR//+EfceeedyM7O7nD7QYMG4bbbbotaa0SX\nsH7s/vvvZ+eddx6rqalhXq+XPfTQQ4zjOHbs2DH/Nrm5uWz9+vX+5VWrVrHJkycH7Gfx4sUsKSmJ\nLV68mCmKwpqampjH4/E/7nQ62WeffcZkWWaMMfbII4+wUaNGBezjuuuuY9OnT2fl5eWMMcZcLher\nq6vrMPcNN9zAHn744Xbr58yZw66//nrmcrmY1+tlixYtYueff77/8SNHjjCO49jChQtZXV0dczgc\nbMSIEWzlypVq3zLmcrlYTk4OW7p0KWOMsZMnT7IJEyawSy65xL+N3W5ngwYNYi+++CJjjLGioiKW\nkpLCSktLVb9Oa21/By3WrVvHsrKy2P79+xljjG3atIklJiaykpIS/zYPPPAAmz59OnM4HIwxxhwO\nB1MUpUuv//3337O4uDi2ZcsWxhhjK1euZAUFBQH7yc3NZTNmzGDV1dVMURT2+9//np177rkd7u/R\nRx9lv/71r9utf+2119jYsWNZTU0NY4wxt9vN3G63//GGhgaWlZXFVqxYwRhjrLa2lk2bNo3de++9\n/m1effVVlpeX53+vX3nlFcbzPPv66687zMJxHPvxxx8D1p08eZIlJiayjz76iDHG2IkTJ9iYMWPY\nyy+/7N/m888/ZxkZGezEiROMMcZkWfa/x1313nvvsbPOOqvDx+6++2522WWXsbq6OqYoCvvrX//K\n8vLyAl6L4zh2ww03MIfDwdxuN5s7dy6bM2eO//EDBw4wg8HAfvjhB/+6hoaGDl9v0aJF7IUXXujW\nz9HWVVddxe6///6AdX//+99ZfHw8y8nJYRdddBHbsGFDu+d99dVX7JxzzmGyLLMFCxawhx56qMP9\n22w2FhcXx5xOZ1jyhku/Lip5eXnsk08+8S97vd5uF5WZM2eqft2SkhImCIJ/ubS0lPE8zyoqKlQ9\n/4Ybbmj3h2az2RjP86y2tta/zuPxsOTkZLZt2zbG2Jmi0lLcGPMVs0ceeUR19g0bNrC0tLSAA+rK\nlSvZ1KlT/ctr1qxhI0eODHjerbfeypYtW6b6dVoLVlRmzJjR7gDwP//zP+z222/3L69YsYKNGDGC\n/fvf/+72Qe/WW29lt912W8C6ESNGsM2bNwfNaLfbmSAIAX9LLRYvXtxhUfn0009Zeno6++c//8nq\n6+vbPf7ee++xs88+O2Ddjh07mNls9i9PnTqVLV++PGCbzMzMLhWVJ554IuCgzBhj77//fsAHoZKS\nEhYfH8/eeustVlVV1eG+1brvvvvY7Nmz261XFIWZzWZWXFwcsL6goID94x//CPoz7N27l4mi6P87\nr6ioYPHx8ey5557zF8FIKy4uZpmZmR2+N5WVlYwxxtavX88SEhICsjc1NbGzzjqLbd++nTEW/ANk\nC6vVGlAstaBfN3/ZbDbk5eX5l1kPrq2ZmJgY9DHGGJ577jlMmTIFkydPxh133AFFUaAoCgDg2LFj\nSEhIQGpqardf/9ixY0hKSkJ8fLx/nSiKyMnJ6XTggU6n8+dQw2azITs7O6Cdt+37VlpaitLSUuTl\n5fm/1q1bB5vN1oWfKLTjx49j6NChAeuGDh0a0PF82223YdmyZXjzzTeRl5eHefPmobq6ukuvU1pa\nisLCwoCfp7q6GidOnAj6HIvFgsTExHbNMJ2ZMWMG3n33XXz00UcYOXIkZs6ciaNHj/ofP3bsGIYM\nGRLwnKFDh8LhcPg7eysqKgL+prvj+PHjHb5O6/d1zJgxWL9+PbZu3YoJEyZgypQpKCoq6tbr2e12\nGI3GdutPnToFh8PR7nc8bNiwTgcXZGZmQpZl/+85NTUV27Ztw/HjxzFt2jSMHz8+ogM/amtrMX/+\nfKxevRoDBgxo93hLH8i0adMwZcoUfPzxx/7HHn/8cVx++eUYP348gNDNZWazGfX19WH+CXqmXxeV\nzMzMgPZZWZbbbcPzfMiDbqhO5L/+9a9Yu3YtCgsLsWnTJqxevTrg8ezsbNTU1PRoFE52djaqq6sD\nRpK4XC4cOXIEOTk53d5vW5mZmTh+/HjAH3rb923IkCE4++yzceTIEf9XeXk5XnjhhbDlAHw/8759\n+wLW7du3D7m5uQHrrr76avzf//0fjh07Bp7nu9xmPmTIEPz2t78N+HkqKysxd+7cgO1aj/SpqalB\ndXU1srOz2+2vs7+XadOm4a233kJpaSny8/Nx4403+h/Lyclp13+1b98+mEwm/8Gr7d800PHfdWey\ns7Oxf//+dq/T9n2dMGECVqxYgaNHj+Kaa67BnDlzuvQ6LXJzcwP6/lokJyfDaDSq+h23fu8PHjwI\no9GI5ORk/7rhw4fjf//3f7F//348+uijuOqqq+ByubqVtzMNDQ2YNWsWHnroIUybNi3k9m63O6Cg\nrl69Gm+99RYSExORmJiINWvW4JlnnsHVV1/d7rmKoqCqqiqs/7/DoV8XlXnz5uEvf/kLGhoa0NjY\niFtuuaXdNpmZmf6x8xUVFXjttdfabRPqDOfEiRNIT09HamoqqqqqsGjRIgBn/iNkZWXh2muvxa9+\n9St/YfF4PEE/UXf0eunp6Zg1axZ+97vfwel0wuv14t5778WwYcNwzjnnBM3W1bOzSZMmwWw2+wtE\ncXExnnrqqYAD5cyZM1FVVYWnn37a/zM2NTXB7XZ36bVC5bztttvw7LPPYs+ePQB8nbV/+9vfAn6P\nVVVVsNvtAHwHV4/HA5PJ1KXXvv322/Hqq6/is88+86+rqalpl+++++5DRUUFZFnGvffei0svvRRp\naWmqfhbA94m95UNBR1mvuuoqVFdX44UXXgBjDNXV1fjDH/6A22+/3b/NvHnz8OKLL6KyshJutxt/\n+tOfunS2BADz58/H119/jffffx+A78xl8eLFAa/j8Xhw8uRJAL6Dm9vt7vL72uJnP/sZvvvuu3YH\neZ7nccstt+Cee+5BbW0tFEXBs88+i8bGRlx55ZUB2955552or69HU1MT7r//ftxwww0Bj7cutG63\nG3q9HoIgtMuyaNGibn/4aWxsxKxZs3D77be3+8DR4vjx4/4i/8UXX6CkpASzZ8/2P37y5EnU1tai\npqYGNTU1+NWvfoVFixb5fxetffPNNygoKEBGRka38kZKvy4qf/rTn5CVlYXMzExMnDgRM2bMaPcp\ncsmSJVixYgUuu+wyLFq0CJdffnm7bTiO6/TT5z333IOqqioMHDgQM2fOxC233AJRFP3/KQHfJ5SZ\nM2dixowZyMrKwogRI/xDatsK9nqrV69GXFwcRo4cicGDB6OysrLdPrqavS2e57FmzRq8/vrrSE1N\nxZ/+9Cdcd911Advo9XqsX78excXFGD58OHJzczFx4kT/wb875s2bh6ysrICRWz/72c+wbNkyzJ07\nF4MGDcIf//hHFBYWoqCgwL/N119/7R+1NWrUKCQmJuLJJ5/s0muPHDkS//rXv/Dkk08iOzsbeXl5\nmDVrVsBBkOM4XHfddZg9ezZSU1Nx4MABrFy5ssP9BXvPS0pKcMEFFyArKwtDhgyBzWYLGJFoMpnw\n5Zdf4tNPP0V2djbOPfdcXHTRRQFzOxYsWIDLL78cI0aM8I+KGzRoUNCfraMcAwcOxOeff47nnnsO\nmZmZuOyyy/Cb3/wGv/3tb/3bHDt2DNOnT0dWVhZycnLw1VdfdXjgU2P48OG4/PLLOxwK/+STT2Li\nxImYMGGC/3W++OILGAyGgO2uvvpqXHzxxRg4cCCMRiOeeOIJ/2P19fW45pprkJmZiaysLCxfvhyf\nfvopRLH9jIrDhw+jtLS0Wz/Hn//8Z2zevBmLFi3yj66bOHFiwDbPPvus/z176qmn8Pnnn3d7SPBz\nzz2He++9t1vPjSSO9aQjgRACwDeX5o033lDV5EHaq6mpwZQpU7By5cp2B2LS3ksvvYRt27bh7bff\njnWUdqI6+bGkpARr164FAFxzzTUYPXp00G2XL1+OsrIySJKEiy++GFOnTo1SSkJItCUmJuKLL77A\n/fffT0UlBJvNhj179uCtt96KdZQORa2oKIqCwsJC/6S9pUuXIj8/P2jTC8dxuPvuuwM62wghfVd6\nejpWrVoV6xial56ejpdffjnWMYKKWlGx2WzIyMiAJEkAgLS0NP+6YKhljvQWR44ciXUEQjQhakWl\noaEBZrPZP5zWZDLBbrcHLSpGoxEvvvgiLBYLFixYgPT09A6360sXGiSEkGi69NJLw77PqBUVi8WC\nxsZGLFy4EIwxrFy5ElarNej2LePzjx49infeecc/DLcjLROFCCGEqLNjx46I7DdqQ4rT09NRXl7u\nX7bZbEHPPlrT6XQdDv3TsrYXi9QCyqSeFnNRJnUoU+xF7WjN8zx+8YtfYMmSJQAQMDloy5Yt0Ov1\nAWcczz//PGpqamA0GnHzzTdHKyYhhJAe6PXzVNavX0/NX4QQ0kU7duyISJ9Kv55RTwghJLyoqESA\nFttQKZN6WsxFmdShTLFHRYUQQkjYUJ8KIYT0Q9SnQgghRPOoqESAFttQKZN6WsxFmdShTLFHRYUQ\nQkjYUJ8KIYT0Q9SnQgghRPOoqESAFttQKZN6WsxFmdShTLFHRYUQQkjYUJ8KIUQTFMbQ6JKhMAYG\nQGEAAwNj8H01/7tlvcJ8N/JjAMySgDhJhCTS52S1ItWn0ruuKU8I6ZPqnV5UN3nglrv3Gdfp9eK0\nwwuLxMOqF2GWhKC3KieRRWU9ArTYhkqZ1NNirr6ayeGWUVrbBFuDu9sFJSDTth0os7txrNaJGocH\nHlnp8T57nEmDv7tIojMVQkjUubwKqh1u2N2ROei7ZYZTDg9OOzyw6AVY9SJMkhCR1yKBqE+FEBI1\nXoWhxuFBndOLaJ9DGAQOcQYRVr0IgaemMepTIYT0WgpjqHN6Ud3ogRyjDE6ZwdnowelGD6wGAXF6\nETqB93f2gwHM9w1oXtfykZu1XW6zzvc81rxt4MCCM9ucGXTAcYAk8JAEDjqBhyTyEPtIoaOiEgGb\nNm3C5MmTYx0jAGVST4u5enMmu9OL0z3ohO+KoqIijBs3rtNtFAC1Thm1ThkcmotIDDI1es6cq3EA\nRJ6DJHDQizx0vK/Y6EW+151VUVEhhESEwy2j2uGBwxv7zvJgtNL2zwB4FAaPwtoVG53gKza+Mxu+\neZkHx/ke19ooN+pTIYSEXaPLi3K7O+r9Jv0F12aBR0uB8a3wF5zmx33/5vz/5jmg/NAe6lMhhGif\nwy1TQYkw1mZBDngg2HlCdM4faJ5KBGhxXDplUk+LuXpLpiaPjPJ6V8wKSlFRUYxeOTgtZookKiqE\nkLBwehWU17tiNrqLaAP1qRBCesztVXCi3gWv0qsPJ/1Kw/F9dD8VQoj2uGUFZXYqKMSHikoE9Jb2\n71jTYiZAm7m0mskj+5q8ojEHRQ0t9l9oMVMkUVEhhHQLxwsot7vg0khBIdpAfSqEkC6TFYayeiea\nvL368NGvUZ8KIUQTFMZQbndRQSEdoqISAVpt/9YaLWYCtJlLK5kUxmCzu+DwKJrsK6BMsUcz6uGb\nAWx3eQH4bktq0dPbQkhbjDFU2N1oiNA9UEjf0Gf7VLwKU3UpaYUxnKh1wtnc2Rgn8ciwGsKek5De\nTFYYqhrdqHPR1Ma+IlJ9Kn32I3md04sBJl3o7Zq8/oICAA6PAsaY5q78SUg0McbQ5FHg9MpwuBU0\neRXNXNGXaFuf7VNpcHnhDHHJbbesoNrhCVgnM6DJ07PTe620f7dGmdTraS63V4Ec5omA0XivXF4F\ntU0elNc7cbi6CSfqXahyeOEIUlC02FdAmWKvT56pKIzBLTM0ur0wiFLQ7WocHd+FzumV6X7WfYBH\nVuCRme+7wqDjOVgNYsTOQmWF4bTD7b/5k0HkYNQJ0As8DDoeOkFbn+G8CkOTW0aTV4bDo2hmAiPp\n3fpkn4rTq+B4rRN6gUNOorHD53kVhqPVTR1eTdUk8shMoH6VaHO4ZXhkBXrRd3tVXsXBv+UDhMfr\nKxweWYFb9h0gOzpGGgQOiUYd4gzh/TxV7/SiyuHp9FIlksDBKPIwiDz0OgEGMTpFRlYYXF7FX1zd\nsgKXl8GrMGrS6seoT6UTTR4ZRt2ZMwtPc7OXS2btHmthd3mDXp7b2dx80dtu49lbNbi8qG3yBtwh\nkAOgFzhIIu+/l7fAc/DKDG5FgdvrKyAuuWsHRqfMUN7gRq3TgwEmqcdnpC6vgqpGd8Dd+oJxywxu\nWW7u7Pb4b5bEc7679/FA8w2XfDdZ4pvvsOR7HOC5lpsvNa+Db13L8znOd//zluLh8vqKKxUPEk1R\nLSolJSVYu3YtAOCaa67B6NGjw7LtqQY3MhMM/k+2nlafFk87PEgyot3Bw+70Bt2fAl+h6u7Q4t58\nP/Gu6MmABsYYNm79DnlnjYGzg0l0DL4C4JRlIAIXU2/yMpyod8Gk45Fk1AX8fah5rxTGUNPkQY0j\n+IeTUBh8fXhyuxsrtX8/1Nx7PdookzpazBRJUSsqiqKgsLAQDz/8MABg6dKlyM/P7/Cg1JVtAd/B\np9rhQbLZ13/ikc/8N3d4FDg8LphEHhlWPQSeg8MtB4z46kiTR4FF360ftc9zeRU0uL2od8rgOcCq\nF2HWC5BU9BnICvOdmTi9qGj0IiPGs7Jb/j7iJB7xBh1EgQMvdH720ujyNXXRNa8IaS9qRcVmsyEj\nIwOS5Dvwp6Wl+df1ZFsAKC4qwjnjx2GASYdvvvkGdYqEYaN8ZzYtIy/GjRuHsnonju8pgoMJyBnR\n/vHWyxMnnA3gzKibyZMngzGG/279DgKTceEFk9o93nq5RbDHtbZ8zvkXoK7Jg+93FkMSgPPGnw1J\n5LHzu21gTMH5ky5Ag8uLzd/vRINbxtixYwPer7PHjYNJx+PQnhLomIwLLpgExhi+2boNXsZhVMFY\nOD0KvttZDFlRMG7cOIwbNy7o+x+LZbvbheLiYnAccLDKAZEHfti9CwLHYdzYsRB4YEdRMexOL8Zq\nIG+sl7X2+2vR+swg1nnajvzSSp5InjlFraP+wIED2LJli3+ZMYYLLrgAw4cP79G269evhyV7JABg\nUJwEs17Ej6cdHXbSAr5OeJdXUdWgMsB0puZ6ZIZGtwyZ+fYxMF6vqiNZ6xxuGbVOT9BZ0jwASeTg\n8TLVjVACB+hb3mf6ME+IJvX6C0paLBY0NjZi3rx5+OUvf4nGxkZYrdYeb9tag1uGW+78QOZQWVAA\n4LTD6/+qd8n+/Tq8CirsLgSrxx3NKWjdJNcVjDE4vQrqmjw47XCjqtGNCrsL5fVOnKxzosLuQrXD\njbomDxxuOej8iNaZGGOwO70orfXNRejsshsKAGcXCgrg6yNweEIXFK2O39diLsqkDmWKvag1f6Wn\np6O8vNy/bLPZkJ6e3uNtW2twyzBHaX6J3a1AaHQj2Sz5Rtk0D2nVCzw4wfe2Ks0H73qXF04vg1Hk\nYTWIsOiFDs9yvArz7ad59I7To8DpVbrUEcwDsBoExOnFdqPevIovT53LS3MSCCEREdV5KsXFxf4R\nXXPnzkVBQQEAYMuWLdDr9QHzTYJt21br5i8AsOgENHiid30iHmh30G8ZDutVGDrqh9bxHEw6HgoD\nZMagKAwyY/AqHY376T6jyCHeoIMkcKh3+TrW6VKAhBAgcs1ffWLyY+uiQgghJLRe36fSn2ixDZUy\nqafFXJRJHcoUe1RUCCGEhA01fxFCSD9EzV+EEEI0j4pKBGixDZUyqafFXJRJHcoUe1RUCCGEhE3I\nPpXvv/8e55xzjn9ZURS89dZbuOmmmyIeTg3qUyGEkK6LWZ/KRx99FPgEnkdpaWnYgxBCCOn9ghaV\nEydOYOvWrbDb7fj222+xdetWfPvtt/j8889RVVUVzYy9jhbbUCmTelrMRZnUoUyxF/TaX+Xl5di+\nfTsaGhqwfft2/3qdToc77rgjKuEIIYT0LiH7VF555RXcdttt0crTZdSnQgghXRezPhUtFxRCCCHa\nQkOKI0CLbaiUST0t5qJM6lCm2AtZVI4cOdJu3d69eyMShhBCSO8Wsqi88cYb7datWbMmImH6ikje\n/7m7KJN6WsxFmdShTLEXsqjwfPtNevk1KAkhhERIyKIiCELAvJTy8vIOCw05Q4ttqJRJPS3mokzq\nUKbYC3mP+rlz5+Lxxx/HpEmTIMsyNm/eTPNUCCGEdEjV/VQqKyuxc+dOcByHcePGITU1NRrZVKF5\nKoQQ0nWRmqcS8kwFAFJTUzF9+vSwvzghhJC+RVXnSMuZSgun0xmxQH2BFttQKZN6WsxFmdShTLEX\nsqhs3LgRL7zwAv7+978D8I38WrZsWcSDEUII6X1CFpXPPvsMjz76KCwWCwCA47iIh+rttDgunTKp\np8VclEkdyhR7qoYU63Q6/7LT6YTb7Y5oKEIIIb1TyKIybNgwvPvuu3A4HPj++++xbNkyTJ48ORrZ\nei0ttqFSJvW0mIsyqUOZYi9kUbnuuuuQkpKClJQU/Pe//8Xll1+OK6+8MhrZCCGE9DKq5qloGc1T\nIYSQrov6/VQURQn7ixFCCOnbghaVp59+GgDw3HPPRS1MX6HFNlTKpJ4Wc1EmdShT7AUtKnV1dQCA\nmpqaqIUhhBDSuwXtU3n++edx8OBB1NfXt7vWF8dxeOaZZ6ISMBTqUyGEkK6L+rW/7rrrLtTV1eGJ\nJ57AH/7wB7qHCiGEkJA6HVIcHx+P8ePHIyUlBampqQFfJDgttqFSJvW0mIsyqUOZYi/kPJW5c+dG\nIwchhJA+gOapEEJIPxT1eSpr1qwBAHzwwQdhf1FCCCF9U9Cism/fPgAIuI9KT5SUlOCRRx7BI488\ngt27d4fcfvny5XjwwQfx2GOPYcOGDWHJEC1abEOlTOppMRdlUocyxV7Q0V9utxvLly9HRUUFVq1a\nFTD6i+M43HjjjapfRFEUFBYW4uGHHwYALF26FPn5+Z1eRp/jONx9991ITk5W/TqEEEJiK2hRuf/+\n+7F7924cOHAAeXl5PXoRm82GjIwMSJIEAEhLS/Ov60xv7e7R4v0TKJN6WsxFmdShTLEXtKhYrVZc\ncMEF+OabbzB16lTVOywpKcG6desC1v385z+H2WzG6tWrAQAmkwl2u73TomI0GvHiiy/CYrFgwYIF\nSE9PD7ptcVERxjb/4lpONcfRMi3TMi3TcqfLkRCV0V9lZWX48MMPsXDhQjDGsHLlSvz85z/vtFC0\nOHr0KAoLC7Fo0aIOH9fi6K+ioiLNfTqhTOppMRdlUocyqRf10V/hlJ6ejvLycv+yzWZTVVAAQKfT\nQRSDnlARQgjREFVnKhs3boTNZsM111wDxhj279+PkSO7dnZQXFyMtWvXAvBNqCwoKPA/tmXLFuj1\neowfP96/7vnnn0dNTQ2MRiNuvvlmpKSkdLhfLZ6pEEKI1kXqTCVkUVm9ejVkWcahQ4ewbNkyAMDD\nDz+MJUuWhD1Md1BRISQyGGOoaHBjX6UD+081Ismkw+z8FOiEqDRwBOXyKqh3elHX/GV3ychPNyPF\nLMU0V2+IsSyOAAAgAElEQVQT9QtKtjh06BCWLFmCxx57LOwv3ldpsQ2VMqmnxVzRymR3ef1FZN8p\nBzgAI1PMmJBpRXFZA57/bykWnjcQiSZd2DMxxlDl8KCu6UzB8BUP2b9c5/RCVhisBhHxBhHxBgEG\nUcCn+09jwqA4ZHnLMWlC//zdaYWqzgpZlv3/ttlsdFdIQjpR2eBGnF6AUSfEOkpIblnBj6ebsK/S\nV0SqHR4MSzZhRIoJlw0bgFSLzj+fbMKgOHx5qAbPbDyGG84ZGNYclQ1uvFdUgYoGN5LNOlj1IuKN\nIuL1AgZa9a2KiAijjm83x22Wy4vPD1Rj7UkDTpuqcOnQxF7x/vdFIZu/Nm7ciK+++gpVVVU499xz\nsXXrVtx6662aqbzU/EW04lSDGx/8cApHqpvg8iqIN4jITjAgK8HQ/F0f0wOdy6vAZneh3O6Grd6F\n0joXjtY0YZDVgJGpJoxMMSMn0QCBDz4pGQD2VTZi9fZyTB8+ABcPTuh0EnMoHlnBlwer8fXhWswY\nMQBTBieA78H+Tjs8+GRvFfZWNuLy4QMwOTc+5s11WhWzPhUAKC0txa5duyCKIsaNG6epS99TUSGx\n1uSR8dn+09hyvB4/GZqIqUMSwXMcKhrcKK114nitE6W1TpyocyHeIPqLTHqcBL3IQ+Q56AQOOp73\nfRc4iDwHSeDBc+jyQbvJI6PC7vYVj+YiUl7vQoNbRppFQnqchIw4PQbF6zFkgLFbha6q0Y3Xt5Vh\noFWPeWPTIIldP3AfrHLgvaIKpMVJmDsmFYkmXZf3EczJOhc+2nMKNrsbM0cOwDlZ1h4Vq74opkVF\ny7RYVLTYhtpfMjHGUOeUUdHgQmWDBxV2Fyoa3KhocGOASYcrz0rGkAGmsORSGMOWY3X4ZG8V8tMs\nuGpUMqyG4C3KCmOosLv9RaaiwQ2PzOBRmO+7rMDb8m9FgUdmYAzQCRyYokAnCuA5DhwHCM3feY4D\n3+q706ug0S0jPU5Cepze/z3DKmGASRfWA+v3O4rwg5KGcrsLv5k4CANUFoVGt4wPdldi/ykHflGQ\nirEZcWHL1PZ3d6jKgXV7TsHlZZg1Khn5aeYenVmFI5NWxKyjnhAt88gKDlY1YZetAcdqnKhscEMn\ncEi1SP5P5WelmpFqkXC4ugmrt5cjPU6PWaOSkRlv6PbrHqxy4J+7KqEXedw2KRPZCaH3xXMcMqx6\nZFj1mJgdr+p1ZIXBqzAUl5Qgf/QYKIxBYQj4zpq/ywqgFzkkhbl4BCPywPVnp2PD4Ro88/UxLJiQ\ngZGp5qDbM8bwXWk9PvjhFCZkWvHgtDwYdJFtmhqabMIfLsrGLlsDPvzhFL44WI3Z+SnISzJG9HX7\nMzpTIZqgMIbdtkbUOb1Is0hItUiINwgdfqpscHnxQ0UjSsobsP+UAwOteoxJN2PIABPS4iSYpeDN\nOR5ZweZjdfhs/2kMTTbhyrOSkWpRPxS1qtGND384heO1TszOT8HZA+Oi/slXiw5WObDquzJMG5qE\nS4cmtntPWjrimzwyfjkuHTmJ3S/o3SUrDNtK6/HJvipkJxhw1VnJyLDq223HGIPMfNvLCoPMfIXd\ntwzIjPmLvf9LbrOsMHgVBV4F8Dafgbb+EnkO6XF6DLL6ziJj0ddGzV9BUFGJjdomD74rrUeKRcLY\nDEu3D6wKY9h50o7PDpyGyHMYFG9AZYMbFXY3vApDqkWH1OYiI/IcfqhoxMk6F0akmDAmw4L8NDPi\n9F0/4XZ5FWz4sQb/+bEGBRkW/HTEgIA2fVlhONXoazarsLths/v+farBjWlDkzBtaCIk6gAOUO3w\nYOW2k0g2S7ju7HToRd7fEb/hcC1mjEjClLzEkAMBIs0tK9h4uBZfHqyGyHO+gtGqUCgM4DlA4H19\nWwLHQeA5CLyv2VHkW5Z9/2791dIf1vrLv53A+9d5ZAXl9W6U1btQbnfB0jzKbZBVj4HNX6kWKaLv\nFRWVILRYVLTYhqomU7XDg9MODwZa9R1+2vcqDLttDdh8rA5Hq5swbmAcDlc3Idmsw7UFaV3qaJUV\nhn9u3o19TjPMkoAZIwZgVGpge7fDLaOiwY3K5i+XV8FZaWYMTzaFbUSPwy3ji4PV+OZoLcZkWOD0\nKDhWVY8GWUCCUUR6nK8ZLS1OQroldp8qe8vflFtW8I/iChyvdWH68CT8v32nkWrRYW5BGpLC2BHf\nlUzBuLwKHB65XaFo6aOKViaFMVQ1elBW7/J/naxzobbJi1SLhIHxLYVGwkCrHgkGMSxnx9SnQiLm\ntMODz/afRlGZHakWCTa7GwYdj0FW3wihjDg9jtc68V1pPTKsEiZlx2PhuQMhtfok+sSGY/ipiiGh\nHlnBttJ6fH6gGpIi4NoJaRiebOrwP4lJEpCXZIxo+7dJEvCz/BRMHZKIHSfrEW/QYQhXhYsmjKGh\nqN0gCTyuOzsd/z1Si88PVGPWqGQU9OBMNpL0Ig99N0athRvPcf6z8XEDzwxacHsVlNvd/kKzr7IR\nZfUuyApDepweerG5EHKtv/v211IgheYzrjPbwP/YhAgd/elMpR9rXUwuykvAtKFJMEsCFMZQ7fDg\nZL0LZXW+P+hUi4Tzs+OREqT/wWZ3YU1RBWSF4Vdnp2Ngc1t1g8uLI9VOHK5uwuHqJpTWOjF0gAnT\nRwzAkAHUWUpIV9U7vahocMMtK5CVlkEaLf0/8PcDtV7uaJsrEmuo+asjVFS6ziMr+OeuSuw4GVhM\nekphDJuP1uHjvVUYMsAIm92NOqcXuYkGDB5gxOAkI3ISDTTTmRAN0NSl710uV7hz9ClavCd1SyaP\nrOD1b8vQ4Jax+LLBuGpUSlgKCuA77Z6cl4A/XZKL0ekW3HhuBp6aORS/uzALV4xMxshUc0BB0eL7\nBGgzF2VShzLFXsiiUlhYGLCsKAqeffbZiAUikdNSUAw6HjeeMzBsxaStBKOISTnxyIw30CxmQvqZ\nkEVl165dgU/geTQ1NUUsUF+gtVE6AJA/pgCvfXsSBh2PBRMyYj6sE9Dm+wRoMxdlUocyxV7Q/v+d\nO3di586dqKiowKpVq9DS9VJXV0fNXxrQ4PLix9NNiNOLGGDWIU4vBD0rcMsKXv/2JIw6QTMFhRDS\nNwUtKomJiRg8eDCKi4uRl5fnXy9JEsaMGROVcL1VpOYUeGQFP1Q0YltpPQ5WOZCXaESjR0a1wwOn\nR0GSSYdEo4g4g4g4SfB91wv4vrQebkc9brt0lKYKihbnXgDazEWZ1KFMsRe0qOTm5iI3NxdOpxNT\np06NYiTSWsu1rYrL7Sgqa8BAq4TzsuIxf3x6QKe3y6ug2uFBTZMXdpfvbnh2lxdl9S5kxuuRbfJo\nqqAQQvqmfj+kmDGGl7ecwL5KBwAg2azDg5fmQYzhAbi2yYsfTztQUt6APZWNyIiTUJARh7MHxam+\nEiwhhHSGZtRHyC5bI2qbvHh+1nBwHPDCplLsqWhAQRgvxw0AZfUumCUBVn3gRRLdXgUn6104VuPE\nkeomHKlpgsvLMDjJgNHpFvx8TGqnl1MnhBAtCXm02r9/Pz7//HM4HI6A9ffdd1/EQkWLwhg+2nMK\ns/NT/E1DE7Ot2Hq8vkdFpW0b6qYjtfjX3ir/8kCrhDi9iLJ6F047PMiIk5CVYMBZaWbMPCsZKWZd\n2C9rocV2XS1mArSZizKpQ5liL2RRWb58Oa6++mqkpKT412nxOj7d8e3xepglAflpZ+4BcfbAOHyw\n+xQaXF5YunH127Z2nLTj0/2n8ccp2Ug261DvknGyzgW7y4vLhychPU4f06Y2QggJp5BHzbS0tD7Z\nUe+RFfy/fVW48ZyBAUXSqBMwOs2C70/YMXVIYrvnnahzQmi+2VJbpxrdSDLq/J9K9lU2orCkAndc\nkOm/Zla8QUR8DJqztPhJSYuZAG3mokzqUKbYC3l0mzhxIrZt24bzzjsvGnmiZk9FI1LMEgZ3cFHD\nidlWrPvhVEBRURjD+kM1WH+wGgIP3HVRNlLMZy6u+Nn+0/jiYDUA+K9x9d8jtfjNxIE9usMgIYT0\nJiGLyurVq+H1eqHTnRl1xHEcVq9eHdFgPaEwhq8P12Dq4PZ3oGvxQ0UjxmR0fOvT4Skm2N0yyupd\nSI+TUGF34/3dlXB5Ge6dmoMfKhrx8uYTuGdKNix6EV8erMbW43V4+Ce+UWOff/cDmrzpuOGcjJD3\nQ48WLbbrajEToM1clEkdyhR7IYvKO++8E40cYVXt8OCfu04hN7Hje3EwxvBDRSN+Miypw+fzHIfz\nsqx4ZesJONwKzJKAidlWTB8+AALP4aK8BFQ7PHj125MoyIjDN0dr8fvJWf5mrTyzgnFjUiP6MxJC\niBb1yXkqeysbsXzzCZyfbcWvx2e0e05prRNvfleGxZcNDrpfh1vG8VonMuP1HXbYK4zh7e3lOFzd\nhLsmZ0flrnaEEBIuMb30/caNG/GPf/wDgO9T/r59+8IeJJxONbgxNsOC4vIGNHnkdo//UNGI/HRL\np/swSQJGppqDjgDjOQ7XT8jAg9PyqKAQQkizkEVl9erVOHTokP+eABzH4d133414sJ441ehBXpIR\nI1PM+K60vt3jP1Q0BAwj7i6e4zq8HakW759AmdTTYi7KpA5lir2QReXQoUO46aaboNe3H0KrVVWN\nbqSYdbgwNx4bj9Tih4oGlNe7oDCGBpcX5fVuDKVb2RJCSNipmjAhy2eakGw2GxRFiVigcDjV6EGy\nWUKGVUJ+mhkbfqzBqUYPOADZCQYMSzZBJ3TrppeqaHGkB2VST4u5KJM6lCn2QhaVyy67DEuWLEFV\nVRVWr16NrVu34tZbb41Gtm5RGMPpRg+SzTrwHIc5o32jsBhjOFLtxDfHanFuljXGKQkhpG8K+XF9\nypQpuPnmm3HFFVcgIyMDjz76qOYq767yBv+/65q8MElCu74OjuMweIAR88dnYHSITvqe0mIbKmVS\nT4u5KJM6lCn2VDV/ZWVlISsrK9JZuu3dnTbM59KRn27BqUYPUsw0GosQQmIh5DyVjRs3Yvv27XC7\n3QHrtXKV4vXr16PKkoNXtp7EPRdn48ApB45WO3Hd+PRYRyOEEM2K2f1U1q1bh1/+8pcwm3s+BDdS\ncpOMmJQTj42HayHwHFIsdKZCCCGxELJPZe7cuTh8+DDsdjvq6+tRX18Pu90ejWxdclFeAr49XoeT\ndU4kt7rQYyxosQ2VMqmnxVyUSR3KFHshz1T+/ve/IysrC6dPnw5YP3HixC690N69e/H2229j1KhR\nmD9/fsjtS0pKsHbtWgDANddcg9GjR3e6fZJJh+EpJhSVNWDWqJROtyWEEBIZIYvKpEmTcMkllyA9\nvWd9FB6PB3PmzMH+/ftDbqsoCgoLC/Hwww8DAJYuXYr8/PyQNwebOjgRRWUNSI5xR73WRscBlKkr\ntJiLMqlDmWIvZFH58ssv8dFHH/X40vcFBQXYs2ePqm1tNhsyMjIgSb5mrLS0NP+6zgwZYMRvzx8E\no07oUjZCCCHhEbKovPHGG13aYUlJCdatWxew7vrrr0dOTo7qfTQ0NMBsNvsLl8lkgt1uD1pUiouK\nMHbcOHAcB4/tEIpsZz4dtLRnRnP50MFD+MXcX8Ts9TtablmnlTyts2glT8sy/f567+9vbeFaDB02\nVDN5tPr3FMmzp25d+t7tdvvPIrpiz5492L59e8g+lbKyMnz44YdYuHAhGGNYuXIlfv7zn3fYBNfR\npe9jTYs35aFM6mkxF2VShzKpF7NL3xcWFgYsK4qCZ555plsvprZ+paeno7y83L9ss9l63KcTTVr8\nA6JM6mkxF2VShzLFXsjmr127dmHu3Ln+ZZ7n0dTU1OUX+vDDD1FUVITa2lo0NTXhlltu8T+2ZcsW\n6PV6jB8/3v8av/jFL7BkyRIACHh9Qggh2hW0qOzcuRM7d+5ERUUFVq1a5T/LqKurg8vl6vILzZ49\nG7Nnz+7wsUmTJrVbN3bsWIwdO7bLr6MFWjzdpUzqaTEXZVKHMsVe0KKSmJiIwYMHo7i4GHl5ef71\nkiRhzJgxUQlHCCGkdwnZUf/vf/8bM2bMiFaeLtNiRz0hhGhdzDrqtVxQCCGEaEvkbn/Yj2nxWj+U\nST0t5qJM6lCm2As5+qusrAwff/wxampqAPiGBdfV1eEvf/lLxMMRQgjpXUL2qdx3332YMmUKysrK\nMHjwYBw+fBiDBg3CFVdcEa2MnaI+FUII6bqY9alIkoSZM2di+PDhSExMxM0334zvv/8+7EEIIYT0\nfiGLitFoBADk5ORg69at8Hq97S6DTwJpsQ2VMqmnxVyUSR3KFHshi8oll1wCu92O3NxcAMCtt96K\nyy67LNK5CCGE9ELduqCkllCfCiGEdF3M+lQIIYQQtVQVlY0bN+If//gHAN+Q4n379kU0VG+nxTZU\nyqSeFnPFMpPAAXESD7HNjVfpfVJHi5kiKWRRWb16NQ4dOuR/YziOw7vvvhvxYISQ2OABmEQeySYR\nmVY9BicZkWE1wGoIOa2NkNBF5dChQ7jpppug1+ujkadP0OIVSSmTelrMFelMeoFDokHEwDgJeUlG\nZCYYkGSSYJIEcJzvFMUiBd6muz++T92hxUyRpOqjhyzL/n/bbDYoihKxQIT0djqeg8wYFI0OgeEB\nSAIHvcjDIPIwSgIkIXRLuEEnwCBycHo1+oMRTQj5l3TZZZdhyZIlOHXqFFavXo3HHnuMbpoVghbb\nUCmTet3NJQkc0i0SchMNGJJkRG6CAYPiJKSYdEgwCDCJPAQu9H7ClYkHYBA5WPUCkk06ZFgkZCcY\nMHiAEdmJRqTF6RFv1KkqKC2s+jOfQ7X4+6NMsRfyTGXKlCnIy8vDrl27IIoiHnvsMaSmpkYjGyG9\ngsgDiUYd4g0ieO5M1ZBEDpLIw9xme7eswCszeGQFCgMYAAaGlsH9CvP9m8E3MIYBsBpExEk8AA4c\nB3AAWl6Ka14HADzHQRI46AQeksD5m67CxaIXUdXoAbVVkGBongoh3SQASDSJiDfoIPDhPXhrWXm9\nC3a3HHpDomkxm6dy6tSpsL8oIdHU0gzUdkhsT/aXaBCRk2REkknqVwUFAKx6IfRGpN8KWVSeeuqp\naOToU7TYhtofMul4wKTjkWAQkGLSYVCchNwEA4YMMCI7wYicRCMGmESEOiQGy8UDiNcLyE4wIMUi\nQYxiMdm0aVPUXisUkyRAx3P94m8qHLSYKZJC9qlIkhSNHIR0Cw/AohcQbxBh1HVeLgSewwCThHiD\nDrVNHtQ2eVX1DRgEDnEGEXF6MaqFRKs4jkMcna2QIEL2qaxfvx4nT57E1VdfHbDeYrFENJha1KcS\nOzwQsw5bgQPiDSKsBrFLo5dac8uKr7g42/cPCADiDALi9KGLVX/k8io4VuuMdQzSA5HqUwl5pvL+\n++8DAL799lv/Oo7j8NJLL4U9DOkdjCKHOL3vk3uD24vKBg/CMdrDrOMhCTy8igKPzOBWzoyCaqEX\nOFgNIqx6scd9GZLAI9WiR7xBQbXDgwa3DKPII04vwBKG/fdlepGHSeTh8NI4MBIoZFFZvnx5NHL0\nKUVFRZqbRRsqEw8gwShCL/DwKL7hri5ZgdvLoMA3bNaqF2GRBBhafXKPN+ggchxsdje6Oh6oJZNZ\nxyPRqINJan9G4G3O4pUZOA4wt5rhHS56kUeGVQ+vwiDyHDZt2oTJkyeH9TV6SouZDu0pwcDho2Md\nI0Bv/L/X19DFfAji9QISTR1PgmOMwS0z6AQuYA5Ga2a9iIE8h/J6F7oy2TpO77u2VEfFpIXIcxB5\nAdCp3293UX9J1+ggx7QJlGhTt+apuFwuzVwLjPpUus8i8Ugy6gLOPHrC7VVQbnfBJXf+J9XZmQnp\nXSrsLtS5aM5KbxSzeSqFhYUBy4qi4Nlnnw17EBJdaRYdBloNYSsoACCJPAbFG2BsNSGEg2+OSKJB\nREbzJUwGxRuooPQRcXpq7CCBQhaVXbt2BT6B59HU1BSxQH2BFselt85kkXjEGyLTniTyHAZaDf55\nIoOTfHNEUiwS4tqM1NLS3IvWtJhLq5mMOt/lYLRC6//3+oOgHzN27tyJnTt3oqKiAqtWrUJLK1ld\nXR1cLlfUApLwEgAkmyM790jgOSSaotAJQmKuZc7KaYc31lGIRgTtUzl69CiOHj2KDz74AHPmzPGv\nlyQJY8aMQVxcXNRCdob6VLom1aJDQoTOUkj/5PYqOEpzVnqdqM9Tyc3NRW5uLpxOJ6ZOnRr2FybR\nZxJ5Kigk7CSRh0nHw+GhcWBERZ/KjBkzopGjT9FiG+qukmKkWLR1yR0t9hMA2syl9UxxGhl4ocX/\ne1rMFEndu74F6XUSDSL0Iv26SWRY9KEv1Bkp9FetLXQ/lX7AIHLIijeEfSY6Ia1VNrg6vI5auAgc\noBd4SGLzTch4DjqRh47nUOfy4nSDp8tXdejPYnbtL9J7SQIHo8gj3qijgkIiziKJPS4qHHyXBJIE\n33XgdALn/67r5MKhCQYdTKKAykY39e3EGJ05RkA02lDjJAG5CQZkx+sxME5CilmHJKMYMMkwt/k+\n5AaR13ybvJZoMVdvyGSSBOh7MGclwSAgN9GAvCQTBsX77lmT0Hzlhc4KSgtJ5HFs93akWnQxa4rr\nSH/rU6EzlV5G5DmkmHSIM9CvjmiPVS/ilMPTpeeIHIdUiw6WMMzOZ4z5zlp0Ak41uNFIZy1RF7U+\nlb179+Ltt9/GqFGjMH/+/JDbL1++HGVlZZAkCRdffHHQYc39qU/FqheQbI7uHQcJ6Qq3rOBojfo5\nK3ESjxSLPmJ/03VOD6qor6VDvb5PxePxYM6cOdi/f7+q7TmOw913343k5OQIJ9O2ljsbWvUiXS+L\naJ4k8DDr+JBnCAKAZIsuYpcLahFv0MFIZy1RFbU+lYKCgi7fLVLtSVRxqzbLoqKigDbMWCyvLVzb\n4/3peA4DTCIqDpbgUPF3/oKyadOmgLZstcst67r7/Egst80W6zwtyytWrNBUnt72+zu0p8S/3NHf\n9497dyMrwYB4gy7s+VasWNHu8W1bNmNQvAFpFh12l5T0yuNBpJYjIezNXyUlJVi3bl3Auuuvvx45\nOTnYs2cPtm/frqr5a9WqVTh8+DAsFgsWLFiA9PT0DrfTYvNXd27K03I1X5MkwCgKMOr4sI7Y0uJN\nnrSYCdBmrt6USWEMR6qb0PYOCDyAAWYdEgxixEYjhnqf3LKCqkY3GtzRO2vR6k26ItX8FdV5Kl0p\nKi2OHj2KwsJCLFq0qMPHtVhU1OIBmCQeZp2geoQLIb1B2zkrBpFDqlkK660WeqKrfS0dlUCO8/0f\n5jgOHOfb5sx3rtVy28c5MDC4vApcMoMSo5mCvb5PBVDfnNWaTqeDKPb+kU5GkYNe5MFzvjso6njf\nWQndB530Ra3nrCQZRSSZdEHvHBoL8c0jxNxeJeCsiW8+8nNoLiQdFgmE7Uyr5c6qbq8Ct6LA7VXg\n9DJ4FYbeOis9ah+NP/zwQxQWFmL79u147bXXAh7bsmULduzYEbDu+eefx+LFi/HOO+/g17/+dbRi\nhkXbNksBQHqcHqkWPZLNEpKahwRHs6D0hnkOWqHFXL0tk0kSYJEEZFp9f/PRKihdeZ90Ag9z8wCY\nli+DToBB5KEXeUhiywRMHiLPQeB9Hwi7WlA6y8Rxvg+bcQYRA0wSMqwG5CUZkZdkRKZVjxSzDgkG\nAQaR6zWTCqN2CjB79mzMnj27w8cmTZrUbt1dd90V6UhRk2TSUdMW6XcGWrVxy/HeSOQ5iJIAU6tp\nnEqrsxqPokBRgJbzGcYA1vy9ZW3LOjDfmjP/RkTPgujaXxEmCRyyEwyaOvUnhJAdO3bE5h71pGcG\nGLXVlkwIIZFERSUCWvpUTM1tpVrQ29rkY0mLuSiTOpQp9rRxxOsj0i06yAywGkSInG9MPiGE9CfU\npxImcRKPDKvBv8wYo8vNE0I0K1J9KnSmAt9VUk2SryXQLfvGiXe2LQB429TiJFPgrXqpoBBC+iPq\nUwGQbNYhPU7v/wo251cAkGGVkGGVArZJMAgBt+rVYhsqZVJPi7kokzqUKfb6fVHRCxzi9GdKhCTw\nSArSF5IWJ8GoE2DUCUi1+M5MBACJRuo7IYQQgPpUkGbWIb5NUWCMobTWCWerK+KlmnVIaLNdbZMH\nMmMY0KbpixBCtI76VCJAErgOh/xyHIdks4RyuwtxegFxehHGDi6El2DUQendNZkQQsKqXzd/JRjE\noBMTTZKA3EQjUi36DgtKi46er8U2VMqknhZzUSZ1KFPs9euiEupOinQFYUII6Zp+26ciCRxyE40R\nSEQIIdpH1/4KM4PYb390QgiJmH57ZI1kUdFiGyplUk+LuSiTOpQp9qioEEIICZt+2acicMDgJCNd\nSoUQ0m9Rn0oXSZ2M3DLqeCoohBASAX22qCSZxKDX8DKInQ8l7ikttqFSJvW0mIsyqUOZYq/PFhVJ\nFGANcoMs6k8hhJDI6JN9Khx8fSaywnCs1onWP6DAAbmJRprYSAjp1+jaX10g8r7Z8ALPId4gQFYY\n9DoBBoGHXuSpoBBCSIT0yXYgsVXRSLXokWE1IMmog0kSolJQtNiGSpnU02IuyqQOZYq9PllUdEKf\n/LEIIUTz+mSfSpJRRLKZ7nFCCCHB0DyVLtBRnwkhhMREnywqYoyLihbbUCmTelrMRZnUoUyx1zeL\nCvWpEEJITPSJPpXE3LPgUXw/RsscFRo2TAghwVGfSieyEwxIMPguvSI2z08hhBASfX2iqAg8hxSz\nBIPAaaKTXottqJRJPS3mokzqUKbY6xNFBQA4jkOSSQedEPuiQggh/VWf6FMZP368f7nB7YVF6pNX\nnyGEkLChPhWVqKAQQkjs9LmiogVabEOlTOppMRdlUocyxR4VFUIIIWETtT6V119/HWVlZVAUBbff\nfjvS0tI63b6kpARr164FAFxzzTUYPXp0h9u17VMhhBASWq+/n8pvfvMbAMDu3bvx0Ucf+Zc7oigK\nClNE2mwAAAgjSURBVAsL8fDDDwMAli5divz8fLqvPCGEaFzUm78MBgNEsfNaZrPZkJGRAUmSIEkS\n0tLSYLPZopSw57TYhkqZ1NNiLsqkDmWKvbA3f5WUlGDdunUB666//nrk5OQA8DWDXXHFFRg0aFDQ\nfRw4cABbtmzxLzPGcMEFF2D48OHttl2/fn2YkhNCSP/SK5q/CgoKUFBQ0OFj33//PQYOHNhpQQEA\ni8WCxsZGLFy4EIwxrFy5ElartcNtI/GmEEII6Z6oNX8dPnwYe/fuxcyZM0Num56ejvLycv+yzWZD\nenp6JOMRQggJg6iN/vrd736HAQMGgOd5ZGdn48Ybb/Q/tmXLFuj1+oBRXMXFxf7RX3Pnzg169kMI\nIUQ7ev1lWgghhGgHTX4khBASNlRUCCGEhI2mrr6odhZ9Z9sGW7937168/fbbGDVqFObPn6+ZXF29\n0kA0Mr333nvYv38/eJ7HLbfcoolMAODxePD73/8es2bNwowZM2Keafny5SgrK4MkSbj44osxderU\nmGc6ffo0XnrpJciyjCFDhmDBggUxzeRwOPD000/7n3v48GGsXr1aVaZI5gKAr7/+Gp999hkEQcC1\n117b6b6jlemLL77Ahg0bYDAYsHDhQmRkZEQtU7BjZFf2DQBgGiHLMnvooYeYy+ViLpeLPfLII0xR\nFNXbdraeMcaKi4vZt99+y95++21N5Gq7j127drHXXntNU5n27t3LXn31Vc1k+uSTT9jTTz/N/v3v\nf8c0U4vly5ezU6dOqcoSrUzPPfcc27dvnyYytd3H0aNH2SuvvBLzXC3uueceJssya2xsZA888EDM\nMzmdTn+Ouro69uyzz0YtE2MdHyO7su8Wmmn+6sos+o62LS8vD7oe8M2fsVgsmsnVdh9qrjQQ7UwH\nDx4MOacoWplcLhdKSkpwzjnngKkcWxLpvykAqrNEI5OiKKioqMCIESM0kantPj799FPVZ5iRzNXy\n+8vMzMSePXuwY8eODidWRzsTYwxerxcejwdmsxm1tbXwer1RyQR0fIzsztVNNNP81dDQALPZ7D81\nNplMsNvtHZ7+BdsWgOp9aC3Xf/7zH1xxxRWaybR48WLU19fj8ccf10SmlgNSbW2tqjzRyGQ0GvHi\niy/CYrFgwYIFquZSRTKT0WiE2+3G008/DYfDgZ/+9Kc477zzYv4+AYDdbsfp06f9V9ZQI9K5CgoK\n8Mknn8Dr9WL69OmayDRnzhwsW7YMRqMRjY2NcDgcQSd+hzNTsGNkV7cHNNRR3zKLft68efjlL3+J\nxsbGoG9msG27sg8t5VJ7pYFoZnrsscdwxx134KWXXop5JofDgX379mHcuHGqskTrfbrxxhuxZMkS\nXHvttXjnnXdinsliscBkMuGee+7Bgw8+iA8++AButzvm7xMAfPnll12++kUkc1VUVGDHjh247777\n8OCDD+Jf//qXJt6r888/H4sXL8a9994LURRVHb/CkSkc+26hmTOVrsyiD7atoiid7qOrTRXRyNVy\npYGuDB6IxnsFAAkJCVAUJeaZduzYAY/HgxdeeAGVlZWQZRmjR49GZmZmzDK1ptPpVDddRjpTcnIy\namtrkZSUpJlMsixjx44deOyxx1TliUausrIyyLIMwHdcUFNQIp2ptR07diA3NzdqmVq0PUZ25+om\nmpr8GGwWfVdm3Adb/+GHH6KoqAi1tbUYNWoUbrnlFk3k6uxKA7HK9Nxzz8Fut0MURdx4442qmw8j\nmanFhg0b4HK5VDdXRDLT888/j5qaGhiNRtx8881ISUmJeaaqqiq8/vrrcDgcmDRpkuom1Uhm2rp1\nK2w2G2bPnq0qS7Ryvf/++9i/fz8URcGFF16oevReJDOtWLECZWVlMBgMuPPOO1W3tIQjU7BjZFev\nbqKpokIIIaR300yfCiGEkN6PigohhJCwoaJCCCEkbKioEEIICRsqKoSE2UcffYTCwsJ26wsLC1FW\nVhaDRIREj2bmqRDSV3Ac1+H6uXPnRjkJIdFHRYWQViorK/Hkk0/ivPPOQ3FxMfR6PRYvXoympias\nWrUK1dXVOHXqFM4//3zMmzfP/7xVq1Zhz549SEpKQnx8fMCclc8++wzffPMNjh8/jkceeQSDBw/2\nP/boo4/i+uuv96+bP3++f3a+2+3Gm2++idLSUiiKgoKCgoDXJESLqKgQ0obNZkN2djauvfZa/zqj\n0Yjrr78eFosFbrcbd955J2bMmIHExERs3boVx48fx5NPPgkAeOqpp5Camup/7vTp0zF9+vQOZ5S3\nPatpvVxcXIz6+nosXbo03D8iIRFDfSqEtJGeno5Jkya1W8/zPLZv346vvvoKOp3Of3HLffv2YcqU\nKeB5HjzPIz8/v1uXBGprxIgRsNvt+Otf/4rNmzfD4/H0eJ+ERBoVFUJUOHbsGBYvXozTp08jNzcX\nVqvVXzh4ng8oIuG6SIXVasWSJUswZ84cHDt2DA8++GBY9ktIJFFRIUSFXbt2Yfz48bj88sthMplQ\nWVnpfyw/Px9btmwBYwxOpxNFRUWq92s2m1FXVwcA2L9/f8BjjDEwxpCZmYk5c+agpqYGTqczPD8Q\nIRFCfSqEtNHR6K0LL7wQTz/9NHbv3o1BgwbhrLPO8jd/TZgwAbt27cJ9992H+Ph4JCf//3bu2Iah\nEAagoCUaBvgNGzEEezEB8zAFU9D8OpHSOd1d74LqSZbF8/MC7FvvPdZasfeO1trH3Dkn5pxRSol7\nb4wxotaa80j4Ex9KApDG+guANKICQBpRASCNqACQRlQASCMqAKR5AZoeMJWCknrRAAAAAElFTkSu\nQmCC\n"
      }
     ],
     "prompt_number": 10
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def family_treatment_effect(column, title):\n",
      "    interval = np.logspace(-2,-3)\n",
      "    means = []\n",
      "    ci_min,ci_max = [], []\n",
      "\n",
      "    column = 'dgastobidon2'\n",
      "    families = bidaguas.loc[:, ['idhogar', 'ypcf_pre1', 'treatment2', column ]].dropna().drop_duplicates()\n",
      "    for radius in interval:\n",
      "        psmatch = ps.PropensityScoreMatch(families.treatment2, families.loc[:, ['ypcf_pre1']], families[column], algo='radius', radius=radius)\n",
      "        psmatch.fit()\n",
      "        means.append(psmatch.treatment_effect())\n",
      "        ci = psmatch.confidence_interval()\n",
      "        ci_min.append(ci[0])\n",
      "        ci_max.append(ci[1])\n",
      "    plt.plot(interval, means)\n",
      "    plt.ylabel('treatment effect')\n",
      "    plt.xlabel('radius')\n",
      "    plt.title(title + ' (obs: %d)' % families[column].count())\n",
      "    fig = plt.fill_between(interval, ci_min,ci_max, alpha=0.2)\n",
      "    #plt.savefig('pscore.png', format='png')\n",
      "    display(fig)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 15
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "family_treatment_effect('dgastobidon2', 'gastos relacionados con agua')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "display_data",
       "text": [
        "<matplotlib.collections.PolyCollection at 0x1099cf6d0>"
       ]
      },
      {
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAY8AAAEWCAYAAACe8xtsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nFg8iIsoYiwcREWXsf/HmpvZxkkdWAAAAAElFTkSuQmCC\n"
      }
     ],
     "prompt_number": 16
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Distibuci\u00f3n del effecto de tratamiento local en funci\u00f3n del pscore de los tratados y matcheados\n"
     ]
    },
    {
     "cell_type": "heading",
     "level": 3,
     "metadata": {},
     "source": [
      "duraci\u00f3n de los episodios"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def plot_scores(psmatch):\n",
      "    scores=[]\n",
      "    effects=[]\n",
      "    algo = psmatch.matching_algo\n",
      "    for key, value in algo.matched.items():\n",
      "        scores.append(psmatch.scores[key])\n",
      "        effects.append(algo.treat_effect_per_treat(key, value))\n",
      "    plt.scatter(scores, effects)\n",
      "    plt.ylabel(\"effect size\")\n",
      "    plt.xlabel(\"pscore\")\n",
      "    return plt.title(\"radius = \" + str(algo.radius))\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 24
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "def effect_size_distribution():\n",
      "    ninos = bidaguas.loc[:, ['idhogar', 'ddurdiarrea1', 'ypcf_pre1', 'treatment2', 'sexo']].dropna() #, #'sexo', 'bedad']].copy()\n",
      "    psmatch = ps.PropensityScoreMatch(ninos.treatment2, ninos.loc[:, ['ypcf_pre1', 'sexo']], ninos.ddurdiarrea1, algo='radius', radius=0.001)\n",
      "    psmatch.fit()\n",
      "    plot_scores(psmatch)\n",
      "effect_size_distribution()\n",
      "#plt.savefig('radius001', format='png')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAYkAAAEVCAYAAAAVeRmFAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEgAAAABJRU5ErkJggg==\n"
      }
     ],
     "prompt_number": 25
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": []
    }
   ],
   "metadata": {}
  }
 ]
}