analisis deudores BCRA - parte 2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Análisis deudores BCRA (parte 2)\n",
    "## junio de 2020\n",
    "\n",
    "Autor: Carlos A. Crespo (lvccrespo@gmail.com)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A pedido del público ¿? retomamos el análisis de las empresas deudoras en Argentina y profundizamos en algunos aspectos.\n",
    "\n",
    "[Aquí](https://github.com/cacrespo/analisis_BCRA/blob/master/analisis%20deudores%20BCRA.ipynb) se encuentra el desarrollo de la primera parte.\n",
    "\n",
    "Contábamos con un dataset con la info de las empresas a febrero de 2020. Cada CUIT con su deuda en cada institución financiera de la República Argentina."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Distintos enfoques para ver ¿lo mismo?\n",
    "En el trabajo anterior **obtuvimos resultados para cada institución**: \"deuda total\", \"deuda promedio\", etc. Se podría decir que obtuvimos resultados \"mirando\" la información desde los ojos de los bancos. \n",
    "\n",
    "Ahora bien, podríamos **revisar los mismos datos desde una perspectiva distinta**, la perspectiva de las empresas: **¿a cuántos bancos le deben? ¿cuál es la deuda promedio por empresa?**\n",
    "\n",
    "Para hacer todo esto **es necesario tener claro cuál es el nivel de detalle**, de \"granularidad\" de los datos, y **cuál es el nivel que necesitamos**. Y finalmente *cómo dialogan cada uno de estos niveles*.\n",
    "\n",
    "Teníamos millones de filas y agregamos todo para cada una de las entidades financieras. Para calcular el total de deuda, tomamos cada fila con el valor de deuda con la entidad en cuestión y para todo ese grupo de filas obtuvimos un sólo valor final con la suma. Ahora vamos a dar un paso más y trabajar vinculando estos niveles que mencionaba anteriormente.\n",
    "\n",
    "- Si quisierás hacer esto en SPSS deberías explorar el comando `AGGREGATE` (que tiene todo el sentido del mundo porque es una \"agregación\").\n",
    "- Si lo trabajás en Tableau, hay que bucear en lo que ellos denominan \"expresiones de nivel de detalle\" (*LOD* por las siglas en inglés). Acá el comando no es tan intuitivo: `FIXED`.\n",
    "- Si lo hacés a mano con un ábaco: te deseo suerte.\n",
    "\n",
    "En pandas combinamos `groupby` y `agg` (casi casi SQL). Vamos a generar tablas __agrupando por__ y __agregando__ con las operaciones aritméticas que necesitemos:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "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>cuitcuil</th>\n",
       "      <th>Deuda máxima en MM</th>\n",
       "      <th>Prestamos Total en $ MM</th>\n",
       "      <th>Número de Acreedores</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>30086400096</td>\n",
       "      <td>0.021</td>\n",
       "      <td>0.021</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>30259329996</td>\n",
       "      <td>0.007</td>\n",
       "      <td>0.007</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>30322242812</td>\n",
       "      <td>0.012</td>\n",
       "      <td>0.012</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>30500000127</td>\n",
       "      <td>4.165</td>\n",
       "      <td>6.213</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>30500001115</td>\n",
       "      <td>1.486</td>\n",
       "      <td>2.209</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      cuitcuil  Deuda máxima en MM  Prestamos Total en $ MM  \\\n",
       "0  30086400096               0.021                    0.021   \n",
       "1  30259329996               0.007                    0.007   \n",
       "2  30322242812               0.012                    0.012   \n",
       "3  30500000127               4.165                    6.213   \n",
       "4  30500001115               1.486                    2.209   \n",
       "\n",
       "   Número de Acreedores  \n",
       "0                     1  \n",
       "1                     1  \n",
       "2                     1  \n",
       "3                     5  \n",
       "4                     2  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "# Levamtamos el dataset que armamos previamente (¿no vas a correr todo de nuevo?)\n",
    "df = pd.read_csv(\"deudores_juridicas_202002.csv\")\n",
    "\n",
    "# Calculamos para cada CUIT con cuántos bancos tiene deuda y el volumen.\n",
    "df_empresas = df.groupby('cuitcuil').agg({'totalprestamos':['max', 'sum']  , 'entidad': 'nunique'}) # Incluimos 'max' para un cálculo posterior\n",
    "\n",
    "# Ajustamos \"etiquetas\" de columnas para visualizar mejor\n",
    "df_empresas = df_empresas.reset_index()\n",
    "df_empresas.columns = [\"cuitcuil\",  \"Deuda máxima en MM\", \"Prestamos Total en $ MM\",\"Número de Acreedores\"]\n",
    "\n",
    "# Revisamos los resultados\n",
    "df_empresas.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ya tenemos una tabla con todos los cuits, el volumen total de deuda, la deuda máxima entre todas las entidades y cuántos acreedores tienen. Por ejemplo para la última fila que visualizamos debe un total de \\\\$ 2.209 MM entre dos bancos y al que más le debe lo hace por \\\\$ 1.486 MM.\n",
    "\n",
    "#### Comencemos a explorar los resultados: ¿a cuántas entidades le debe cada empresa en promedio?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "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>Prestamos Total en $ MM</th>\n",
       "      <th>Número de Acreedores</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>201338.000000</td>\n",
       "      <td>201338.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>7.718074</td>\n",
       "      <td>1.613372</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>228.121328</td>\n",
       "      <td>1.366594</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.003000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.045000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.334000</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>53826.226000</td>\n",
       "      <td>31.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Prestamos Total en $ MM  Número de Acreedores\n",
       "count            201338.000000         201338.000000\n",
       "mean                  7.718074              1.613372\n",
       "std                 228.121328              1.366594\n",
       "min                   0.000000              1.000000\n",
       "25%                   0.003000              1.000000\n",
       "50%                   0.045000              1.000000\n",
       "75%                   0.334000              2.000000\n",
       "max               53826.226000             31.000000"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Nos ubicamos a nivel detalle empresas y obtenemos algunos valores descriptivos mediante describe()\n",
    "\n",
    "df_empresas[df_empresas.columns[2:]].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Perfecto. Ya tenemos las respuestas! El promedio de deuda por empresa es $ 7.7 MM y se endeudan con 1.6 bancos.\n",
    "\n",
    "Momeeeeeento! Si observamos con detenimiento vemos que **para ambos indicadores hay valores \"[muy extremos](https://es.wikipedia.org/wiki/Valor_at%C3%ADpico)\"**. \n",
    "Por ej. el 75% de los prestamos tiene valores menores a \\\\$ 334.000 y ¿el promedio es \\\\$ 7.7 MM?. El máximo en toda la muestra es casi \\\\$ 54.000 millones (WTF!).\n",
    "\n",
    "Veamos cómo se ve en un [boxplot](https://es.wikipedia.org/wiki/Diagrama_de_caja):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig, ax =plt.subplots(2,1) # dos gráficos en una sola columna\n",
    "fig.suptitle('Diagrama de Caja Prestamos Total por empresa')\n",
    "\n",
    "sns.boxplot(data=df_empresas,x=\"Prestamos Total en $ MM\", orient=\"h\",showfliers=True , palette=\"Set2\",ax=ax[0]); \n",
    "sns.boxplot(data=df_empresas,x=\"Prestamos Total en $ MM\", orient=\"h\",showfliers=False , palette=\"Set2\",ax=ax[1]); "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`\"showfliers=False\"` excluye de la visualización los outliers. En este caso el gráfico de abajo sería como un zoom del gráfico de arriba."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "En número de acreedores pasa algo parecido: casi toda la muestra tiene menos de 3 acreedores y el máximo 31. Pero vamos a enfocarnos en el volumen de deuda.\n",
    "\n",
    "Intuitivamente -entre otras cosas- iría a ver esos casos puntuales: qué CUITS son, qué valores tienen en el resto de los campos. ¿Será un problema de carga? ¿habrá algún problema en la base? \n",
    "\n",
    "Hay numerosas maneras de abordar el tratamiento de los \"outliers\". Podríamos por ejemplo quedarnos con la [media recortada](https://es.wikipedia.org/wiki/Media_truncada), quitar casos según algún criterio externo a los datos, etc.\n",
    "\n",
    "A titulo ilustrativo veamos cómo se transforma la muestra luego de excluir todos los casos que se desvíen mucho de la media. Básicamente normalizamos los resultados y después filtramos con la condición hasta +-1.96 desvíos (¿te suena de algún lado ese valor?). "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "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>Prestamos Total en $ MM</th>\n",
       "      <th>Número de Acreedores</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>200887.000000</td>\n",
       "      <td>200887.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>2.585876</td>\n",
       "      <td>1.596390</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>17.677062</td>\n",
       "      <td>1.299253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.003000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.045000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.329000</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>453.890000</td>\n",
       "      <td>31.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Prestamos Total en $ MM  Número de Acreedores\n",
       "count            200887.000000         200887.000000\n",
       "mean                  2.585876              1.596390\n",
       "std                  17.677062              1.299253\n",
       "min                   0.000000              1.000000\n",
       "25%                   0.003000              1.000000\n",
       "50%                   0.045000              1.000000\n",
       "75%                   0.329000              2.000000\n",
       "max                 453.890000             31.000000"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from scipy import stats\n",
    "import numpy as np\n",
    "\n",
    "c = np.abs(stats.zscore(df_empresas[\"Prestamos Total en $ MM\"])) < 1.96 # creamos una serie con True False para filtrar el dataset\n",
    "\n",
    "# con la nueva selección de casos vemos de nuevo el resumen de resultados. \n",
    "# el [2:] es porque tomamos desde la 3er columna en adelante (el orden arranca en 0)\n",
    "\n",
    "df_empresas[df_empresas.columns[2:]][c].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Parece mucho mejor. *Ahora el promedio de deuda es \\\\$ 2.5 MM y el máximo en \\\\$ 454 MM*. El número de acreedores (1.6) no parece alterarse demasiado.\n",
    "\n",
    "Sin embargo, dado que el 75% de los casos poseen valores hasta \\\\$ 329.000 y ***tenemos valores tan extremos: ¿tiene sentido enfocarnos en la media? ¿es adecuado considerar todos los casos \"dentro de la misma bolsa\"?***\n",
    "\n",
    "En situaciones como esta recomiendo el enfoque de [Lerner](https://www.youtube.com/watch?v=plLaNCLG-JE).\n",
    "\n",
    "Para terminar vale la pena hacer un cruce entre los dos indicadores. Parece que los valores extremos no son exclusividad de las empresas endeudadas con muchos bancos:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax =plt.subplots(2,1,figsize=(20, 10)) # dos gráficos en una sola columna y agrandamos un poco.\n",
    "fig.suptitle('Diagrama de Caja Prestamos Total por empresa según Número de Acreedores')\n",
    "\n",
    "sns.boxplot(data=df_empresas,y='Número de Acreedores',x=\"Prestamos Total en $ MM\", orient=\"h\",\n",
    "            showfliers=True , palette=\"Set2\",ax=ax[0]);\n",
    "sns.boxplot(data=df_empresas,y='Número de Acreedores',x=\"Prestamos Total en $ MM\", orient=\"h\",\n",
    "            showfliers=False , palette=\"Set2\",ax=ax[1]);\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### ¿Cuál es el banco que se lleva el mayor volumen de deuda de cada empresa?\n",
    "Ahora bien, resulta que dificilmente una empresa tome deuda equitativamente con cada banco. Incluimos en los cálculos de agregación el valor máximo de deuda, y con ello hacemos el ejercicio de identificar cuál es el banco que se lleva el mayor volumen de cada CUIT (podría ser más de uno. Si fuera 50% y 50% entrarían las dos instituciones).\n",
    "\n",
    "Para abordar este interrogante **tenemos que hacer dialogar niveles de detalle distintos: necesitamos identificar el máximo deudor por empresa y luego consolidar para cada banco todos sus deudores**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "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>cuitcuil</th>\n",
       "      <th>entidad</th>\n",
       "      <th>totalprestamos</th>\n",
       "      <th>Deuda máxima en $ MM</th>\n",
       "      <th>Número de Acreedores</th>\n",
       "      <th>máximo acreedor</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>30500041443</td>\n",
       "      <td>7</td>\n",
       "      <td>0.002</td>\n",
       "      <td>0.074</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>30500064230</td>\n",
       "      <td>7</td>\n",
       "      <td>0.002</td>\n",
       "      <td>38.426</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>30500095764</td>\n",
       "      <td>7</td>\n",
       "      <td>0.002</td>\n",
       "      <td>0.658</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>30500145966</td>\n",
       "      <td>7</td>\n",
       "      <td>0.002</td>\n",
       "      <td>0.002</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>30500689443</td>\n",
       "      <td>7</td>\n",
       "      <td>0.020</td>\n",
       "      <td>5.574</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      cuitcuil  entidad  totalprestamos  Deuda máxima en $ MM  \\\n",
       "0  30500041443        7           0.002                 0.074   \n",
       "1  30500064230        7           0.002                38.426   \n",
       "2  30500095764        7           0.002                 0.658   \n",
       "3  30500145966        7           0.002                 0.002   \n",
       "4  30500689443        7           0.020                 5.574   \n",
       "\n",
       "   Número de Acreedores  máximo acreedor  \n",
       "0                     3                0  \n",
       "1                     5                0  \n",
       "2                     3                0  \n",
       "3                     1                1  \n",
       "4                     3                0  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# el dataset con las empresas y el detalle deuda máxima y número de acreedores que contruimos [\"NIVEL 1\"]\n",
    "df1 = df_empresas[['cuitcuil', \"Deuda máxima en $ MM\",\"Número de Acreedores\"]]\n",
    "\n",
    "# el dataset original con los campos entidad y monto de cada una de las deudas. [\"NIVEL 0\"]\n",
    "df2 = df[['cuitcuil', \"entidad\", \"totalprestamos\"]]\n",
    "\n",
    "# los juntamos en df3\n",
    "df3 = pd.merge(df2,df1,on='cuitcuil',how='left')\n",
    "\n",
    "# agregamos una columna donde identificamos si la entidad es máximo acreedor\n",
    "df3[\"máximo acreedor\"] = (df3[\"Deuda máxima en $ MM\"] == df3[\"totalprestamos\"]) * 1\n",
    "\n",
    "df3.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "En esta linea:\n",
    "`(df3[\"Deuda máxima en $ MM\"] == df3[\"totalprestamos\"]) * 1`\n",
    "\n",
    "multiplicamos por 1 un resultado *booleano* (true o false): convertimos de una manera simple y directa en 0 y 1. (Ya van a ver para qué)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Con esta tabla ahora necesitamos consolidar la información por Banco y obtener los resultados que necesitamos:\n",
    "- % de deudores donde el banco es máximo acreedor\n",
    "- promedio de acreedores por empresa (por banco)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Calculo de agregados por banco [\"NIVEL 2\"]\n",
    "\n",
    "df4 = df3.groupby('entidad').agg({'totalprestamos': 'sum', 'máximo acreedor': 'mean', 'Número de Acreedores': 'mean', \n",
    "                                  'cuitcuil': 'nunique'})\n",
    "df4 = df4.reset_index()\n",
    "\n",
    "df4.columns = ['entidad', 'vol deuda', '% máximo acreedor', 'prom acreedores por emp','cant de deudores']\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Como habíamos convertido en 1 y 0 la condición \"máximo deudor\", al pedir la media\n",
    "`'máximo acreedor': 'mean'`\n",
    "obtenemos por resultado el porcentaje de clientes máximo deudores de cada banco.\n",
    "\n",
    "Del mismo modo que hicimos en la primera parte, nos quedamos con las 15 instituciones con mayor deuda consolidada y añadimos etiquetas correspondientes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "df4 = df4.sort_values(by = df4.columns[1], ascending = False)[0:14]\n",
    "\n",
    "# pasar nombres de entidades\n",
    "df4[\"entidad\"].replace({7: \"Galicia\",\n",
    "                   11: \"Nación\",\n",
    "                   14: \"Bapro\",\n",
    "                   15: \"ICBC\",\n",
    "                   16: \"Citi\",\n",
    "                   17: \"BBVA\",\n",
    "                   27: \"Supervielle\",\n",
    "                   29: \"Ciudad\",\n",
    "                   34: \"Patagonia\",\n",
    "                   72: \"Santander\",\n",
    "                   150: \"HSBC\",\n",
    "                   191: \"Credicoop\",\n",
    "                   259: \"Itaú\",\n",
    "                   285: \"Macro\",\n",
    "                   300: \"BICE\"}, inplace = True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Ya estamos en condiciones de repasar resultados"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "%matplotlib inline\n",
    "sns.set_style(\"darkgrid\")\n",
    "\n",
    "df4 = df4.sort_values(by = df4.columns[3], ascending = False) # ordenamos por acreedores promedio\n",
    "\n",
    "g = sns.barplot(x='entidad', y='prom acreedores por emp', data=df4)\n",
    "g.set_xticklabels(g.get_xticklabels(), rotation=90); # el \";\" evita texto en el output"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Habíamos visto que las empresas **toman crédito con 1.6 instituciones en promedio**.\n",
    "\n",
    "Ahora bien, si lo vemos en cada cartera deudora encontramos que **Citi y BICE** son los bancos de mayor exposición. Por el contrario, **Galicia** parece compartir menos sus deudores."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df4 = df4.sort_values(by = df4.columns[2], ascending = False)\n",
    "\n",
    "g = sns.barplot(x='entidad', y='% máximo acreedor', data=df4)\n",
    "g.set_xticklabels(g.get_xticklabels(), rotation=90); # el \";\" evita texto en el output"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "¿Cuál es el banco con mayor porcentaje de empresas \"máximo\" deudor\"? **Galicia**. En el otro extremo encontramos a **Ciudad**."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Más interrogantes...\n",
    "A todas las inquietudes de la publicación anterior, le sumamos nuevas cuestiones:\n",
    "\n",
    "¿Es correcto analizar conjuntamente las empresas \"mono-deudoras\" con las que que toman deuda en varias instituciones? ¿será que Galicia detenta una porción grande de deudores exclusivos?\n",
    "\n",
    "¿Tiene el mismo \"valor\" ser máximo deudor en una empresa que debe a dos instituciones que en otra que debe a cinco?\n",
    "\n",
    "¿Tiene el mismo \"valor\" ser máximo deudor prestando \\\\$ 1 MM que prestando \\\\$ 10.000? \n",
    "\n",
    "Si algún día llega la parte tres, vamos a trabajar en la mejor manera de visualizar toda esta información. Áspecto fundamental de todo esto. Muchas gracias!"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}