CIA_y_Matching.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "CIA y Matching.ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyOo4nogOVJj0GQtkogGglY0",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/rpasquini/herramientas-econometricas/blob/main/CIA_y_Matching.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3CShQ-M6gGnx"
      },
      "source": [
        "import pandas as pd\n",
        "import statsmodels.api as sm  \n",
        "from scipy import stats  \n",
        "import numpy as np\n",
        "from sklearn.neighbors import KNeighborsRegressor"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LlNTXFb1yOgS"
      },
      "source": [
        "df=pd.read_csv('https://raw.githubusercontent.com/rpasquini/econometrics_and_causality/master/happiness_on_tap_extract.csv')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AxASTzaeZEup"
      },
      "source": [
        "df['dsexojefe']=np.where(df['sexojefe']==1,1,0)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZRO_QZ1IaBkt"
      },
      "source": [
        "Extracto de datos de:\n",
        "\n",
        "[Devoto, F., Duflo, E., Dupas, P., Parienté, W., & Pons, V. (2012). Happiness on tap: piped water adoption in urban Morocco. American Economic Journal: Economic Policy, 4(4), 68-99.](https://web.stanford.edu/~pdupas/MoroccoWaterConnections.pdf)\n",
        "\n",
        "Las variables fueron modificadas para el proposito de este ejercicio.\n",
        "\n",
        "Descripción de variables\n",
        "\n",
        "motivacion: Dummy = 1 si el hogar recibió motivación para conexión \\\n",
        "conectado: Dummy = 1 si el hogar se conectó a la red de agua \\\n",
        "sexojefe:  1 Masculino, 2 Femenino \\\n",
        "edadjefe: Edad Jefe de Hogar \\\n",
        "jefesabeleer: Dummy=1 Jefe sabe leer \\\n",
        "jefesabeescribir: Dummy=1 Jefe sabe escribir \\\n",
        "pdiasfueaescuela: Porcentaje dias hijos fueron a la escuela."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o5XP7fAagc0Z"
      },
      "source": [
        "df=df.dropna()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "35w6FE8Cgnlt"
      },
      "source": [
        "Diferencia entre conectados y no conectados sugiere "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6NrVr9riaTBK",
        "outputId": "c9348401-3bee-4e5b-f2ef-c0a7f6da69bf",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        }
      },
      "source": [
        "df.loc[df.conectado==1].pdiasfueaescuela.mean()-df.loc[df.conectado==0].pdiasfueaescuela.mean()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "-3.523604421369349"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 169
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YUnI65PMBl23"
      },
      "source": [
        "Antes de continuar verificamos si hay diferencias en las caracteristicas observables\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZtnMo51uzGqz",
        "outputId": "41722825-5b45-4152-b218-d4c59290ce26",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 281
        }
      },
      "source": [
        "for variable in ['sexojefe','edadjefe','jefesabeleer','jefesabeescribir','pdiasfueaescuela']:\n",
        "  var_tratados=df[variable].loc[(df.conectado==1) ]\n",
        "  var_control=df[variable].loc[(df.conectado==0) ]\n",
        "  print(variable)\n",
        "  print('dif: ',var_tratados.mean()-var_control.mean())\n",
        "  resultadottest=stats.ttest_ind(var_tratados, var_control, equal_var=True)\n",
        "  print('statistic', resultadottest.statistic, 'pvalue', resultadottest.pvalue)\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "sexojefe\n",
            "dif:  -0.014426908974972008\n",
            "statistic -0.3985894717651943 pvalue 0.6904657650625463\n",
            "edadjefe\n",
            "dif:  -0.7358919177427055\n",
            "statistic -0.6786088936491492 pvalue 0.4978839468444204\n",
            "jefesabeleer\n",
            "dif:  -0.06260959668420218\n",
            "statistic -1.2386725523291628 pvalue 0.21638902860325182\n",
            "jefesabeescribir\n",
            "dif:  -0.05519687549816665\n",
            "statistic -1.0737778188107645 pvalue 0.28374413297112716\n",
            "pdiasfueaescuela\n",
            "dif:  -3.523604421369349\n",
            "statistic -1.224783170253449 pvalue 0.22157203698942263\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JyKygcbi-ac9"
      },
      "source": [
        "# Recuperando el ATT usando la CIA.\n",
        "A los fines de un primer ejemplo, vamos a suponer que es suficiente con el control por el sexo del jefe de hogar para eliminar el posible sesgo de selección. \n",
        "El estimador que propone la teoría es:\n",
        "\n",
        "$$ATT=E[Y_{1i}∣T_{i}=1]-E[Y_{0i}∣T_{i}=1]=\\sum_x\\delta_xP(X_i=x|T_i=1)$$"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jh6W3T0D1y7V",
        "outputId": "93ef2ac3-1b24-498e-eb6a-6225050435bb",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        }
      },
      "source": [
        "# Diferencia para jefes hombres\n",
        "dif_hombres=df.loc[(df.conectado==1)& (df.sexojefe==1)].pdiasfueaescuela.mean() \\\n",
        "-df.loc[(df.conectado==0)&(df.sexojefe==1)].pdiasfueaescuela.mean()\n",
        "print('dif_hombres ',dif_hombres)\n",
        "# Diferencia para jefes mujeres\n",
        "\n",
        "dif_mujeres=df.loc[(df.conectado==1)& (df.sexojefe==2)].pdiasfueaescuela.mean() \\\n",
        "-df.loc[(df.conectado==0)&(df.sexojefe==2)].pdiasfueaescuela.mean()\n",
        "print('dif_mujeres ',dif_mujeres)\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "dif_hombres  -3.090223847986074\n",
            "dif_mujeres  -6.808688535087732\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "64dThVBC8UOI",
        "outputId": "9c4989f8-9603-4650-856a-3acb0451f343",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        }
      },
      "source": [
        "total_hombres=df.loc[(df.sexojefe==1) ].shape[0]\n",
        "total_mujeres= df.loc[(df.sexojefe==2)].shape[0]\n",
        "print('total hombres  ', total_hombres, ', total mujeres ', total_mujeres)\n",
        "\n",
        "\n",
        "total_hombres_tratados=df.loc[(df.sexojefe==1) & (df.conectado==1)].shape[0]\n",
        "total_mujeres_tratadas= df.loc[(df.sexojefe==2) & (df.conectado==1)].shape[0]\n",
        "print('total hombres tratados ', total_hombres_tratados, ', total mujeres tratadas ', total_mujeres_tratadas)\n",
        "\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "total hombres   280 , total mujeres  37\n",
            "total hombres tratados  146 , total mujeres tratadas  18\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YUB_m7mN8eDv",
        "outputId": "55a325c6-026e-44de-f1a6-8920d62fcb94",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        }
      },
      "source": [
        "# ATT\n",
        "(dif_hombres*total_hombres_tratados+dif_mujeres*total_mujeres_tratadas)/(146+18)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "-3.4983480209606466"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 173
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hyng7-e0BA7e",
        "outputId": "b84755b9-727b-4623-edf1-d975918f91de",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        }
      },
      "source": [
        "# ATE\n",
        "(dif_hombres*total_hombres+dif_mujeres*total_mujeres)/(280+37)\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "-3.5242402310231764"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 174
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RjwcxE5LTs3f",
        "outputId": "1ff6e7f7-709c-496f-8aed-f2ca8f57550a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        }
      },
      "source": [
        "df.pdiasfueaescuela.mean()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "59.80171416088331"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 175
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_MCgAgfsMC9V"
      },
      "source": [
        "# Algoritmo de Matching\n",
        "\n",
        "Para el ATT, la idea es asignarle a cada tratado una pareja en el control, basandome en todas las caracteristicas observables.\n",
        "Vamos a usar una función de la libreria sklearn, que busca a partir de un conjunto de caracteristicas, la observación más similar, y devuelve el resultado de esa observación.\n",
        "El metodo de Kneighbors se utiliza para hacer predicciones en general y esta es una adaptacion a nuestro caso de interes.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_hsQTzQ-NGDA"
      },
      "source": [
        "# Llamemos X a la lista de variables que vamos a considerar para el matching\n",
        "X=['dsexojefe', 'edadjefe', 'jefesabeleer', 'jefesabeescribir']\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YmkR2iwfIrlp",
        "outputId": "13c277a0-9e7b-4f44-d02a-3206d93ad727",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 417
        }
      },
      "source": [
        "# Separo las observaciones de conectados y no conectados\n",
        "conectados = df.loc[(df.conectado==1)]\n",
        "noconectados = df.loc[(df.conectado==0)]\n",
        "\n",
        "# KNeighborsRegressor necesita que 'entrene' un modelo, que va a aprender a buscar el resultado de un 'no conectado' cuando le pida que 'prediga' uno.\n",
        "# El 'entrenamiento' es equivalente al 'ajuste' de nuestro modelo de regresion.\n",
        "\n",
        "modelo_no_conectados = KNeighborsRegressor(n_neighbors=1).fit(noconectados[X], noconectados.pdiasfueaescuela)\n",
        "\n",
        "# le pido al modelo que obtenga la pareja de no conectado (predecir el no conectado correspondiente)\n",
        "\n",
        "df_matcheado=conectados.assign(matchnoconectado=modelo_no_conectados.predict(conectados[X]))\n",
        "df_matcheado"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>random_id</th>\n",
              "      <th>motivacion</th>\n",
              "      <th>conectado</th>\n",
              "      <th>sexojefe</th>\n",
              "      <th>edadjefe</th>\n",
              "      <th>jefesabeleer</th>\n",
              "      <th>jefesabeescribir</th>\n",
              "      <th>pdiasfueaescuela</th>\n",
              "      <th>dsexojefe</th>\n",
              "      <th>matchnoconectado</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>45.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>57.142860</td>\n",
              "      <td>1</td>\n",
              "      <td>14.285715</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>7</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>54.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>85.714287</td>\n",
              "      <td>1</td>\n",
              "      <td>85.714287</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>11</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>52.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>71.428574</td>\n",
              "      <td>1</td>\n",
              "      <td>71.428574</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>21</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>51.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>71.428574</td>\n",
              "      <td>1</td>\n",
              "      <td>71.428574</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>18</th>\n",
              "      <td>26</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>45.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>57.142860</td>\n",
              "      <td>1</td>\n",
              "      <td>28.571430</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>766</th>\n",
              "      <td>1012</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>41.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>42.857143</td>\n",
              "      <td>0</td>\n",
              "      <td>71.428574</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>768</th>\n",
              "      <td>1014</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>36.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>14.285715</td>\n",
              "      <td>1</td>\n",
              "      <td>71.428574</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>772</th>\n",
              "      <td>1018</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>51.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>85.714287</td>\n",
              "      <td>1</td>\n",
              "      <td>71.428574</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>778</th>\n",
              "      <td>1025</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>39.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>14.285715</td>\n",
              "      <td>1</td>\n",
              "      <td>71.428574</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>784</th>\n",
              "      <td>1031</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>37.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>14.285715</td>\n",
              "      <td>1</td>\n",
              "      <td>71.428574</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>164 rows × 10 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "     random_id  motivacion  ...  dsexojefe  matchnoconectado\n",
              "1            3           1  ...          1         14.285715\n",
              "3            7           1  ...          1         85.714287\n",
              "7           11           0  ...          1         71.428574\n",
              "13          21           1  ...          1         71.428574\n",
              "18          26           1  ...          1         28.571430\n",
              "..         ...         ...  ...        ...               ...\n",
              "766       1012           1  ...          0         71.428574\n",
              "768       1014           0  ...          1         71.428574\n",
              "772       1018           1  ...          1         71.428574\n",
              "778       1025           0  ...          1         71.428574\n",
              "784       1031           1  ...          1         71.428574\n",
              "\n",
              "[164 rows x 10 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 177
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rLTS76y0Qn_B",
        "outputId": "62fbf230-a2bf-482d-877d-ccc4c3a9cc68",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        }
      },
      "source": [
        "np.mean(df_matcheado.pdiasfueaescuela-df_matcheado.matchnoconectado)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "-5.313588926829269"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 178
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hPlfWxUBYRtf"
      },
      "source": [
        "# Propensity Score Matching \n",
        "\n",
        "Mostramos la idea del propensity score utilizando el mismo procedimiento que para el matching, pero en este caso vamos a matchear sobre la probabilidad"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lkIAgI1hYP4F"
      },
      "source": [
        "def logit(lista, data):\n",
        "  \"\"\" \n",
        "  Funcion que realiza la regresion Logit\n",
        "  La primera variable de la lista es la dependiente, el resto son las explicativas\n",
        "  data es la base de datos\n",
        "  \"\"\"\n",
        "  datos=data[lista]\n",
        "  # el metodo dropna() me permite eliminar las filas que tienen algun valor missing\n",
        "  datos=datos.dropna()\n",
        "  Y=datos[lista[0]]\n",
        "  X=datos[lista[1:]]\n",
        "  X=sm.add_constant(X)\n",
        "  modelo=sm.Logit(Y,X)\n",
        "  modeloajustado=modelo.fit()\n",
        "  print(modeloajustado.summary())\n",
        "  return modeloajustado"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "x3hS8B6WYWLq",
        "outputId": "25fa65cd-7d1d-4d8f-b894-57aaae4d4084",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 387
        }
      },
      "source": [
        "modelo=logit(['conectado','dsexojefe', 'edadjefe', 'jefesabeleer', 'jefesabeescribir'], data=df)\n",
        "df['probabilidades']=modelo.predict()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Optimization terminated successfully.\n",
            "         Current function value: 0.687622\n",
            "         Iterations 4\n",
            "                           Logit Regression Results                           \n",
            "==============================================================================\n",
            "Dep. Variable:              conectado   No. Observations:                  317\n",
            "Model:                          Logit   Df Residuals:                      312\n",
            "Method:                           MLE   Df Model:                            4\n",
            "Date:                Fri, 07 Aug 2020   Pseudo R-squ.:                0.007109\n",
            "Time:                        00:38:24   Log-Likelihood:                -217.98\n",
            "converged:                       True   LL-Null:                       -219.54\n",
            "Covariance Type:            nonrobust   LLR p-value:                    0.5377\n",
            "====================================================================================\n",
            "                       coef    std err          z      P>|z|      [0.025      0.975]\n",
            "------------------------------------------------------------------------------------\n",
            "const                0.5142      0.678      0.758      0.448      -0.815       1.843\n",
            "dsexojefe            0.3258      0.376      0.867      0.386      -0.410       1.062\n",
            "edadjefe            -0.0092      0.012     -0.774      0.439      -0.032       0.014\n",
            "jefesabeleer        -0.7785      0.946     -0.823      0.410      -2.632       1.075\n",
            "jefesabeescribir     0.3716      0.925      0.402      0.688      -1.441       2.185\n",
            "====================================================================================\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FxgmH18TZZX-",
        "outputId": "e3abedcb-6330-4a83-d366-3cea9dfe89c3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 203
        }
      },
      "source": [
        "df.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>random_id</th>\n",
              "      <th>motivacion</th>\n",
              "      <th>conectado</th>\n",
              "      <th>sexojefe</th>\n",
              "      <th>edadjefe</th>\n",
              "      <th>jefesabeleer</th>\n",
              "      <th>jefesabeescribir</th>\n",
              "      <th>pdiasfueaescuela</th>\n",
              "      <th>dsexojefe</th>\n",
              "      <th>probabilidades</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>45.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>57.142860</td>\n",
              "      <td>1</td>\n",
              "      <td>0.605246</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>5</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>50.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>85.714287</td>\n",
              "      <td>1</td>\n",
              "      <td>0.493662</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>7</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>54.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>85.714287</td>\n",
              "      <td>1</td>\n",
              "      <td>0.484496</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>11</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>52.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>71.428574</td>\n",
              "      <td>1</td>\n",
              "      <td>0.489078</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>14</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>42.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>42.857143</td>\n",
              "      <td>1</td>\n",
              "      <td>0.511999</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   random_id  motivacion  ...  dsexojefe  probabilidades\n",
              "1          3           1  ...          1        0.605246\n",
              "2          5           0  ...          1        0.493662\n",
              "3          7           1  ...          1        0.484496\n",
              "7         11           0  ...          1        0.489078\n",
              "9         14           0  ...          1        0.511999\n",
              "\n",
              "[5 rows x 10 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 189
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lcjWpqoRZ7N-",
        "outputId": "babb9e4c-62bc-40a3-a6e7-fc8ea51f5fa3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        }
      },
      "source": [
        "# Vemos que haya un soporte comun en las distribuciones\n",
        "df.loc[df.conectado==1]['probabilidades'].hist( alpha=0.5)\n",
        "df.loc[df.conectado==0]['probabilidades'].hist(alpha=0.5)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7fa738c87c88>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 181
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0T2UPwa8aAp9",
        "outputId": "735007cf-7a09-4b20-ad2c-e3aa0c2b44cd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 417
        }
      },
      "source": [
        "# KNeighborsRegressor para buscar parejas\n",
        "# Separo las observaciones de conectados y no conectados\n",
        "conectados = df.loc[(df.conectado==1)]\n",
        "noconectados = df.loc[(df.conectado==0)]\n",
        "\n",
        "\n",
        "modelo_no_conectados = KNeighborsRegressor(n_neighbors=1).fit(noconectados['probabilidades'].values.reshape(-1, 1), noconectados.pdiasfueaescuela)\n",
        "\n",
        "# le pido al modelo que obtenga la pareja de no conectado (predecir el no conectado correspondiente)\n",
        "\n",
        "df_matcheado=conectados.assign(matchnoconectado=modelo_no_conectados.predict(conectados['probabilidades'].values.reshape(-1, 1)))\n",
        "\n",
        "df_matcheado"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>random_id</th>\n",
              "      <th>motivacion</th>\n",
              "      <th>conectado</th>\n",
              "      <th>sexojefe</th>\n",
              "      <th>edadjefe</th>\n",
              "      <th>jefesabeleer</th>\n",
              "      <th>jefesabeescribir</th>\n",
              "      <th>pdiasfueaescuela</th>\n",
              "      <th>dsexojefe</th>\n",
              "      <th>probabilidades</th>\n",
              "      <th>matchnoconectado</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>45.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>57.142860</td>\n",
              "      <td>1</td>\n",
              "      <td>0.605246</td>\n",
              "      <td>14.285715</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>7</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>54.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>85.714287</td>\n",
              "      <td>1</td>\n",
              "      <td>0.484496</td>\n",
              "      <td>85.714287</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>11</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>52.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>71.428574</td>\n",
              "      <td>1</td>\n",
              "      <td>0.489078</td>\n",
              "      <td>85.714287</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>21</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>51.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>71.428574</td>\n",
              "      <td>1</td>\n",
              "      <td>0.491370</td>\n",
              "      <td>85.714287</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>18</th>\n",
              "      <td>26</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>45.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>57.142860</td>\n",
              "      <td>1</td>\n",
              "      <td>0.505124</td>\n",
              "      <td>42.857143</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>766</th>\n",
              "      <td>1012</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>41.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>42.857143</td>\n",
              "      <td>0</td>\n",
              "      <td>0.534512</td>\n",
              "      <td>42.857143</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>768</th>\n",
              "      <td>1014</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>36.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>14.285715</td>\n",
              "      <td>1</td>\n",
              "      <td>0.525734</td>\n",
              "      <td>71.428574</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>772</th>\n",
              "      <td>1018</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>51.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>85.714287</td>\n",
              "      <td>1</td>\n",
              "      <td>0.491370</td>\n",
              "      <td>85.714287</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>778</th>\n",
              "      <td>1025</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>39.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>14.285715</td>\n",
              "      <td>1</td>\n",
              "      <td>0.618313</td>\n",
              "      <td>71.428574</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>784</th>\n",
              "      <td>1031</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>37.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>14.285715</td>\n",
              "      <td>1</td>\n",
              "      <td>0.523447</td>\n",
              "      <td>71.428574</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>164 rows × 11 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "     random_id  motivacion  ...  probabilidades  matchnoconectado\n",
              "1            3           1  ...        0.605246         14.285715\n",
              "3            7           1  ...        0.484496         85.714287\n",
              "7           11           0  ...        0.489078         85.714287\n",
              "13          21           1  ...        0.491370         85.714287\n",
              "18          26           1  ...        0.505124         42.857143\n",
              "..         ...         ...  ...             ...               ...\n",
              "766       1012           1  ...        0.534512         42.857143\n",
              "768       1014           0  ...        0.525734         71.428574\n",
              "772       1018           1  ...        0.491370         85.714287\n",
              "778       1025           0  ...        0.618313         71.428574\n",
              "784       1031           1  ...        0.523447         71.428574\n",
              "\n",
              "[164 rows x 11 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 182
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "txOKd9qWaaw1",
        "outputId": "418a668a-2eb0-4e5b-f270-154e735ad687",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        }
      },
      "source": [
        "np.mean(df_matcheado.pdiasfueaescuela-df_matcheado.matchnoconectado)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "-2.613240280487805"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 183
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zaC3ljDBdThX"
      },
      "source": [
        "# Utilizando la libreria Causal Inference\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lOdUEtLdSQp_",
        "outputId": "300f950e-0562-4f7c-8f24-9da065da3511",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        }
      },
      "source": [
        "!pip install causalinference"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: causalinference in /usr/local/lib/python3.6/dist-packages (0.1.3)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Y04F5l9uSKv_"
      },
      "source": [
        "from causalinference import CausalModel"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FqfiPgEfZl5a"
      },
      "source": [
        "X=['dsexojefe', 'edadjefe', 'jefesabeleer', 'jefesabeescribir']"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "T4A0WJT6SOPI",
        "outputId": "03df6979-a115-4d07-fc38-081d9e689b70",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 248
        }
      },
      "source": [
        "cm = CausalModel( Y=df[\"pdiasfueaescuela\"].values,  D=df[\"conectado\"].values,  X=df[X].values )\n",
        "\n",
        "cm.est_via_matching(matches=1, bias_adj=True)\n",
        "\n",
        "print(cm.estimates)\n",
        "cm.reset()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Treatment Effect Estimates: Matching\n",
            "\n",
            "                     Est.       S.e.          z      P>|z|      [95% Conf. int.]\n",
            "--------------------------------------------------------------------------------\n",
            "           ATE     -0.709      3.913     -0.181      0.856     -8.378      6.959\n",
            "           ATC      0.221      4.235      0.052      0.958     -8.080      8.522\n",
            "           ATT     -1.577      4.224     -0.373      0.709     -9.856      6.701\n",
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/causalinference/estimators/matching.py:100: FutureWarning: `rcond` parameter will change to the default of machine precision times ``max(M, N)`` where M and N are the input matrix dimensions.\n",
            "To use the future default and silence this warning we advise to pass `rcond=None`, to keep using the old, explicitly pass `rcond=-1`.\n",
            "  return np.linalg.lstsq(X, Y)[0][1:]  # don't need intercept coef\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WumzvmDXV6K4"
      },
      "source": [
        "# Propensity Score using Causal Model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aq45We3sTqTE",
        "outputId": "8a791c52-50cc-4da4-dd6f-22b9f153ed01",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 299
        }
      },
      "source": [
        "causal = CausalModel(\n",
        "    Y=df[\"pdiasfueaescuela\"].values, \n",
        "    D=df[\"conectado\"].values, \n",
        "    X=df[X].values\n",
        ")\n",
        "causal.est_propensity_s()\n",
        "print(causal.propensity)\n",
        "causal.stratify_s()\n",
        "print(causal.strata)\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Estimated Parameters of Propensity Score\n",
            "\n",
            "                    Coef.       S.e.          z      P>|z|      [95% Conf. int.]\n",
            "--------------------------------------------------------------------------------\n",
            "     Intercept      0.294      0.214      1.373      0.170     -0.126      0.714\n",
            "            X2     -0.312      0.252     -1.238      0.216     -0.806      0.182\n",
            "\n",
            "\n",
            "Stratification Summary\n",
            "\n",
            "              Propensity Score         Sample Size     Ave. Propensity   Outcome\n",
            "   Stratum      Min.      Max.  Controls   Treated  Controls   Treated  Raw-diff\n",
            "--------------------------------------------------------------------------------\n",
            "         1     0.496     0.573       153       164     0.515     0.520    -3.524\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}