exploracion.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Análisis exploratorio de datos \n",
    "### Contenido\n",
    "1. [Introducción](#introducción)\n",
    "2. [Librearías e importación de datos](#librerías-e-importación-de-datos)\n",
    "3. [Tabla clientes](#tabla-clientes)\n",
    "4. [Tabla transacciones](#tabla-transacciones)\n",
    "5. [Variable objetivo](#variable-objetivo)\n",
    "5. [Resumen](#Resumen)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Introducción"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "En este notebook realizamos un análisis exploratorio de los datos brindados para construir el modelo pedictivo de impagos de tarjetas de crédito. \n",
    "\n",
    "Para cada campo de las tablas de *clientes* y *transacciones* se analizan características como: \n",
    "- El tipo de variable, si es nominal, ordinal, contínua, etc. \n",
    "- La consistencia e inconsistencia de los datos.\n",
    "- Valores faltantes y valores atípicos.\n",
    "- Para los campos numericos se hallan medidas estadísticas básicas como distribución, rango, media, etc. \n",
    "- Para los campos cualitativos se realizan conteos básicos.\n",
    "\n",
    "Al final, se presenta un resúmen con los hallazgos más significativos."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Librerías e importación de datos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importar librerías y funciones\n",
    "import pandas as pd \n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "# Funciones provenientes de otro módulo\n",
    "from utils import dar_estilo_visual, grafico_distribucion"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Configurar parámetros globales de matplotlib y seaborn \n",
    "dar_estilo_visual()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Cargar datos\n",
    "transacciones = pd.read_csv(\"data/historial_transacciones.csv\", delimiter=\"|\")\n",
    "clientes = pd.read_csv(\"data/informacion_clientes.csv\", delimiter=\"|\")\n",
    "etiquetas = pd.read_csv(\"data/etiquetas.csv\", delimiter=\"|\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Tabla clientes"
   ]
  },
  {
   "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>CLIENT_ID</th>\n",
       "      <th>Cupo_TC</th>\n",
       "      <th>SEX</th>\n",
       "      <th>EDUCATION</th>\n",
       "      <th>MARRIAGE</th>\n",
       "      <th>AGE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>29c55f5f-0da7-43ec-ba1b-b60592689825</td>\n",
       "      <td>270000</td>\n",
       "      <td>Femenino</td>\n",
       "      <td>universidad</td>\n",
       "      <td>Casado</td>\n",
       "      <td>28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>f818136a-138f-4578-a0bb-41605b67e0e5</td>\n",
       "      <td>80000</td>\n",
       "      <td>Femenino</td>\n",
       "      <td>colegio</td>\n",
       "      <td>soltero_a</td>\n",
       "      <td>52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>bef609a9-f9c5-4870-be7f-cb20e70f7e92</td>\n",
       "      <td>210000</td>\n",
       "      <td>Femenino</td>\n",
       "      <td>universidad</td>\n",
       "      <td>Casado</td>\n",
       "      <td>36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2ba3a218-4792-4672-82c7-4b0953101d41</td>\n",
       "      <td>230000</td>\n",
       "      <td>Femenino</td>\n",
       "      <td>colegio</td>\n",
       "      <td>Casado</td>\n",
       "      <td>44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>723e61b4-d151-4368-b129-468417fd930c</td>\n",
       "      <td>130000</td>\n",
       "      <td>Femenino</td>\n",
       "      <td>Posgrado</td>\n",
       "      <td>Casado</td>\n",
       "      <td>36</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                              CLIENT_ID  Cupo_TC       SEX    EDUCATION  \\\n",
       "0  29c55f5f-0da7-43ec-ba1b-b60592689825   270000  Femenino  universidad   \n",
       "1  f818136a-138f-4578-a0bb-41605b67e0e5    80000  Femenino      colegio   \n",
       "2  bef609a9-f9c5-4870-be7f-cb20e70f7e92   210000  Femenino  universidad   \n",
       "3  2ba3a218-4792-4672-82c7-4b0953101d41   230000  Femenino      colegio   \n",
       "4  723e61b4-d151-4368-b129-468417fd930c   130000  Femenino     Posgrado   \n",
       "\n",
       "    MARRIAGE  AGE  \n",
       "0     Casado   28  \n",
       "1  soltero_a   52  \n",
       "2     Casado   36  \n",
       "3     Casado   44  \n",
       "4     Casado   36  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clientes.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "La tabla contiene 24000 registros y 6 columnas\n"
     ]
    }
   ],
   "source": [
    "print(f\"La tabla contiene {clientes.shape[0]} registros y {clientes.shape[1]} columnas\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 24000 entries, 0 to 23999\n",
      "Data columns (total 6 columns):\n",
      " #   Column     Non-Null Count  Dtype \n",
      "---  ------     --------------  ----- \n",
      " 0   CLIENT_ID  24000 non-null  object\n",
      " 1   Cupo_TC    24000 non-null  int64 \n",
      " 2   SEX        24000 non-null  object\n",
      " 3   EDUCATION  24000 non-null  object\n",
      " 4   MARRIAGE   24000 non-null  object\n",
      " 5   AGE        24000 non-null  int64 \n",
      "dtypes: int64(2), object(4)\n",
      "memory usage: 1.1+ MB\n"
     ]
    }
   ],
   "source": [
    "clientes.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "24000"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Verificar si la columna CLIENT_ID no presenta valores repetidos\n",
    "clientes[\"CLIENT_ID\"].nunique()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "La tabla Clientes presenta información socio demográfica de los clientes en cuestión, posee 24 mil registros y los siguientes campos:\n",
    "- **CLIENT_ID:** Identificador único para cada cliente.\n",
    "- **Cupo_TC:** Cupo fijo original de la tarjeta de crédito. \n",
    "- **SEX:** Sexo biológico. \n",
    "- **EDUCATION:** Nivel educativo \n",
    "- **MARRIAGE:** Estado civil\n",
    "- **AGE:** Edad\n",
    "\n",
    "Adicionalmente, no se presentan missing values en ningún campo."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Cupo tarjeta de crédito"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count     24000.000000\n",
       "mean     167828.736667\n",
       "std      129957.242751\n",
       "min       10000.000000\n",
       "25%       50000.000000\n",
       "50%      140000.000000\n",
       "75%      240000.000000\n",
       "max      800000.000000\n",
       "Name: Cupo_TC, dtype: float64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Medidas estadísticas básicas\n",
    "clientes[\"Cupo_TC\"].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Graficar un histograma y boxplot para conocer la distribución de la variable\n",
    "fig, ax = grafico_distribucion(clientes[\"Cupo_TC\"])\n",
    "plt.xlabel(\"\\nCupo tarjeta de crédito\")\n",
    "fig.suptitle(\"Distribución de Cupo tarjeta de crédito\")\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Observamos la variable Cupo_TC es numérica con valores enteros, tiene una distribución de probabilidad sesgada a la izquierda, presenta un rango de $790.000 y no contiene datos inconsistentes."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Sexo biológico"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_a28b4\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_a28b4_level0_col0\" class=\"col_heading level0 col0\" >Cantidad</th>\n",
       "      <th id=\"T_a28b4_level0_col1\" class=\"col_heading level0 col1\" >Proporción</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >SEX</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_a28b4_level0_row0\" class=\"row_heading level0 row0\" >Femenino</th>\n",
       "      <td id=\"T_a28b4_row0_col0\" class=\"data row0 col0\" >14507</td>\n",
       "      <td id=\"T_a28b4_row0_col1\" class=\"data row0 col1\" >60.45%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a28b4_level0_row1\" class=\"row_heading level0 row1\" >Masculino</th>\n",
       "      <td id=\"T_a28b4_row1_col0\" class=\"data row1 col0\" >9493</td>\n",
       "      <td id=\"T_a28b4_row1_col1\" class=\"data row1 col1\" >39.55%</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x1e111b86990>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Calcular el porcentaje de cada categoría \n",
    "sex_summary = clientes.groupby(\"SEX\").count()[\"CLIENT_ID\"]\n",
    "share = sex_summary / sex_summary.sum()\n",
    "sex_summary = pd.concat([sex_summary, share], axis=1)\n",
    "sex_summary.columns = [\"Cantidad\", \"Proporción\"]\n",
    "sex_summary.style.format({\"Proporción\": '{:,.2%}'.format})"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "El campo SEX es una variable nominal binaria cuyos valores son *Masculino* y *Femenino*. No presenta ningún dato inconsistente. Aproximadamente el 60% de la muestra son mujeres y el 40% son hombres."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Nivel de escolaridad"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "EDUCATION\n",
       "universidad    11151\n",
       "Posgrado        8493\n",
       "colegio         3967\n",
       "5                230\n",
       "otro             107\n",
       "6                 41\n",
       "0                 11\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Categorías y número de clientes por categoría\n",
    "clientes.EDUCATION.value_counts()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Consideramos que hay datos inconsistentes, ya que los valores $0$, $5$ y $6$ no tienen coherencia con el campo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_dbbbe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_dbbbe_level0_col0\" class=\"col_heading level0 col0\" >Cantidad</th>\n",
       "      <th id=\"T_dbbbe_level0_col1\" class=\"col_heading level0 col1\" >Proporción</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >EDUCATION</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_dbbbe_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_dbbbe_row0_col0\" class=\"data row0 col0\" >11</td>\n",
       "      <td id=\"T_dbbbe_row0_col1\" class=\"data row0 col1\" >0.05%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_dbbbe_level0_row1\" class=\"row_heading level0 row1\" >5</th>\n",
       "      <td id=\"T_dbbbe_row1_col0\" class=\"data row1 col0\" >230</td>\n",
       "      <td id=\"T_dbbbe_row1_col1\" class=\"data row1 col1\" >0.96%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_dbbbe_level0_row2\" class=\"row_heading level0 row2\" >6</th>\n",
       "      <td id=\"T_dbbbe_row2_col0\" class=\"data row2 col0\" >41</td>\n",
       "      <td id=\"T_dbbbe_row2_col1\" class=\"data row2 col1\" >0.17%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_dbbbe_level0_row3\" class=\"row_heading level0 row3\" >Posgrado</th>\n",
       "      <td id=\"T_dbbbe_row3_col0\" class=\"data row3 col0\" >8493</td>\n",
       "      <td id=\"T_dbbbe_row3_col1\" class=\"data row3 col1\" >35.39%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_dbbbe_level0_row4\" class=\"row_heading level0 row4\" >colegio</th>\n",
       "      <td id=\"T_dbbbe_row4_col0\" class=\"data row4 col0\" >3967</td>\n",
       "      <td id=\"T_dbbbe_row4_col1\" class=\"data row4 col1\" >16.53%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_dbbbe_level0_row5\" class=\"row_heading level0 row5\" >otro</th>\n",
       "      <td id=\"T_dbbbe_row5_col0\" class=\"data row5 col0\" >107</td>\n",
       "      <td id=\"T_dbbbe_row5_col1\" class=\"data row5 col1\" >0.45%</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_dbbbe_level0_row6\" class=\"row_heading level0 row6\" >universidad</th>\n",
       "      <td id=\"T_dbbbe_row6_col0\" class=\"data row6 col0\" >11151</td>\n",
       "      <td id=\"T_dbbbe_row6_col1\" class=\"data row6 col1\" >46.46%</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x1e1231c2550>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Calcular el porcentaje de cada categoría \n",
    "ed_summary = clientes.groupby(\"EDUCATION\").count()[\"CLIENT_ID\"]\n",
    "share = ed_summary / ed_summary.sum()\n",
    "ed_summary = pd.concat([ed_summary, share], axis=1)\n",
    "ed_summary.columns = [\"Cantidad\", \"Proporción\"]\n",
    "ed_summary.style.format({\"Proporción\": '{:,.2%}'.format})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Hay 282 datos inconsistentes en la columna EDUCATION lo que representa el 1.18%\n"
     ]
    }
   ],
   "source": [
    "# Calcular porcentaje de datos inonsistentes\n",
    "num_inconsistentes = len(clientes.query('EDUCATION in [\"0\",\"5\",\"6\"]'))\n",
    "prop_inconsistentes = num_inconsistentes / len(clientes)\n",
    "print('Hay {} datos inconsistentes en la columna EDUCATION lo que representa el {:.2%}'.format(num_inconsistentes, prop_inconsistentes))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Puesto que la cantidad de datos inconsistentes es solo del 1.18% , para el modelo predictivo eliminaremos estos registros.\n",
    "\n",
    "La variable EDUCATION es una columna categórica nominal cuyas categorías son *colegio*, *universidad*, *posgrado* y *otro*\n",
    "\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Estado civil"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "MARRIAGE\n",
       "soltero_a    12752\n",
       "Casado       10954\n",
       "otro           245\n",
       "0               49\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Categorías y número de clientes por categoría\n",
    "clientes.MARRIAGE.value_counts()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Observamos que hay 49 datos inconsistes en esta columna cuyo valor es $0$. Puesto que se trata de un porcentaje muy bajo de la muestra, prescindiremos de estos registros para la construcción del modelo predictivo.\n",
    "\n",
    "La columna MARRIAGE representa una variable nominal cuyas categorías son *solotero_a*, *Casado* y *otro*"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Edad"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    24000.000000\n",
       "mean        35.477542\n",
       "std          9.210014\n",
       "min         21.000000\n",
       "25%         28.000000\n",
       "50%         34.000000\n",
       "75%         41.000000\n",
       "max         79.000000\n",
       "Name: AGE, dtype: float64"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Medidas estadísticas básicas\n",
    "clientes.AGE.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Graficar un histograma y boxplot para conocer la distribución de la variable\n",
    "fig, ax = grafico_distribucion(clientes[\"AGE\"])\n",
    "plt.xlabel(\"\\nEdad\")\n",
    "fig.suptitle(\"Distribución de la edad de los clientes\")\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Observamos que la columna AGE representa una variable numérica con valores enteros, cuya distribución es sesgada a la izquierda. Los clientes presentan un rango de edad entre 21 y 79 años, aunque la mayoría de clientes se encuentran entre los 25 y 45 años de edad. La columna no presenta datos inconsistentes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Tabla transacciones"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>MONTH</th>\n",
       "      <th>CLIENT_ID</th>\n",
       "      <th>RETRASO_PAGO_ESTADO</th>\n",
       "      <th>DEUDA_MES</th>\n",
       "      <th>PAGO_MES</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>9</td>\n",
       "      <td>29c55f5f-0da7-43ec-ba1b-b60592689825</td>\n",
       "      <td>0</td>\n",
       "      <td>322249</td>\n",
       "      <td>10000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>9</td>\n",
       "      <td>f818136a-138f-4578-a0bb-41605b67e0e5</td>\n",
       "      <td>0</td>\n",
       "      <td>42635</td>\n",
       "      <td>1623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9</td>\n",
       "      <td>bef609a9-f9c5-4870-be7f-cb20e70f7e92</td>\n",
       "      <td>0</td>\n",
       "      <td>143225</td>\n",
       "      <td>6483</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>9</td>\n",
       "      <td>2ba3a218-4792-4672-82c7-4b0953101d41</td>\n",
       "      <td>-1</td>\n",
       "      <td>54798</td>\n",
       "      <td>11632</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>9</td>\n",
       "      <td>723e61b4-d151-4368-b129-468417fd930c</td>\n",
       "      <td>0</td>\n",
       "      <td>129689</td>\n",
       "      <td>5000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   MONTH                             CLIENT_ID  RETRASO_PAGO_ESTADO  \\\n",
       "0      9  29c55f5f-0da7-43ec-ba1b-b60592689825                    0   \n",
       "1      9  f818136a-138f-4578-a0bb-41605b67e0e5                    0   \n",
       "2      9  bef609a9-f9c5-4870-be7f-cb20e70f7e92                    0   \n",
       "3      9  2ba3a218-4792-4672-82c7-4b0953101d41                   -1   \n",
       "4      9  723e61b4-d151-4368-b129-468417fd930c                    0   \n",
       "\n",
       "   DEUDA_MES  PAGO_MES  \n",
       "0     322249     10000  \n",
       "1      42635      1623  \n",
       "2     143225      6483  \n",
       "3      54798     11632  \n",
       "4     129689      5000  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "transacciones.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "La tabla contiene 144000 registros y 5 columnas\n"
     ]
    }
   ],
   "source": [
    "print(f\"La tabla contiene {transacciones.shape[0]} registros y {transacciones.shape[1]} columnas\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "MONTH\n",
       "9    24000\n",
       "8    24000\n",
       "7    24000\n",
       "6    24000\n",
       "5    24000\n",
       "4    24000\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Recuento columna MONTH\n",
    "transacciones.MONTH.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Número de IDs únicos: 24000\n"
     ]
    }
   ],
   "source": [
    "print(f\"Número de IDs únicos: {transacciones.CLIENT_ID.nunique()}\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Observamos que la tabla transacciones tiene 144000 registros con observaciones temporales de los campos *RETRASO_PAGO_ESTADO*, *DEUDA_MES* y *PAGO_MES*. \n",
    "\n",
    "Tiene información de 24 mil clientes a lo largo de 6 meses, cada mes corresponde a una observación temporal, tomaremos la siguiente convención:\n",
    "\n",
    "$$\n",
    "\\begin{array}{l | c | c | c | c | c | c}\n",
    "\\text{Mes} & 9 & 8 & 7 & 6 & 5 & 4 \\\\\n",
    "\\hline\n",
    "\\text{Marca temporal} & t & t-1 & t-2 & t-3 & t-4 & t-5\n",
    "\\end{array}\n",
    "$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Aplicar el mapeo de acuerdo a la convención\n",
    "convencion = {9-i: f't-{i}' for i in range(6)}\n",
    "transacciones[\"MONTH\"] = transacciones[\"MONTH\"].map(convencion)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Para analizar adecuadamente la información de las tres variables como series de tiempo asociada a cada cliente, utilizaremos tablas de pivote para cada una variables."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Estado de retraso de pago"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Recuento de todos los valores únicos \n",
    "ax = sns.countplot(data = transacciones, y=\"RETRASO_PAGO_ESTADO\")\n",
    "plt.title(\"Recuento de estados de retraso de pago\\na lo largo del semestre\\n\")\n",
    "plt.ylabel(\"Etiqueta\\nde estado\", rotation=0,)\n",
    "plt.xlabel(\"Cantidad\")\n",
    "ax.yaxis.set_label_coords(-0.2,0.5)\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Observamos que el estado más recurrente es $0$, que indica que no hubo retrasos en los pagos. Adicionalmente, los números del $1$ al $8$ sugieren que hay o hubo un retraso de esa cantidad de meses en el pago. Conjeturamos que las etiquetas $-1$ y $-2$ no son datos inconsistentes ya que existe una alta proporción de estas etiquetas, probablemente indiquen un pago adelantado."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "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>MONTH</th>\n",
       "      <th>t-0</th>\n",
       "      <th>t-1</th>\n",
       "      <th>t-2</th>\n",
       "      <th>t-3</th>\n",
       "      <th>t-4</th>\n",
       "      <th>t-5</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CLIENT_ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>00004d47-2e99-461f-b6bf-59802fc72a8c</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>000af878-0c8d-4d31-901a-d4b5c584e08f</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>000c3e2f-013e-4894-81aa-aa96357a1cf2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>000e7e42-2d80-441c-9ed5-df0e9dbb4c02</th>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0011d46a-bf48-482d-aab0-526fe8b6e254</th>\n",
       "      <td>1</td>\n",
       "      <td>-2</td>\n",
       "      <td>-2</td>\n",
       "      <td>-2</td>\n",
       "      <td>-2</td>\n",
       "      <td>-2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "MONTH                                 t-0  t-1  t-2  t-3  t-4  t-5\n",
       "CLIENT_ID                                                         \n",
       "00004d47-2e99-461f-b6bf-59802fc72a8c    0    0    0    0    0    0\n",
       "000af878-0c8d-4d31-901a-d4b5c584e08f    0    0    0    0    0    0\n",
       "000c3e2f-013e-4894-81aa-aa96357a1cf2    0    0   -1   -1   -1   -1\n",
       "000e7e42-2d80-441c-9ed5-df0e9dbb4c02   -1   -1   -1   -1   -1   -1\n",
       "0011d46a-bf48-482d-aab0-526fe8b6e254    1   -2   -2   -2   -2   -2"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Transponer meses para la analizar variable RETRASO_PAGO_ESTADO como serie de tiempo para cada cliene\n",
    "retraso = transacciones.pivot(index=\"CLIENT_ID\", columns=\"MONTH\", values=\"RETRASO_PAGO_ESTADO\")\n",
    "retraso.head()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "El campo RETRASO_PAGO_ESTADO es una variable numérica ordinal con valores enteros que induce una serie temporal para cada cliente. No presenta datos inconsistentes."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Deudas mensuales"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "La variable DEUDA_MES corresponde al monto total de la deuda en la tarjeta de crédito del cliente en un determinado mes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    1.440000e+05\n",
       "mean     4.466619e+04\n",
       "std      6.630685e+04\n",
       "min     -3.396030e+05\n",
       "25%      2.301750e+03\n",
       "50%      1.919100e+04\n",
       "75%      5.704425e+04\n",
       "max      1.664089e+06\n",
       "Name: DEUDA_MES, dtype: float64"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Medidas estadísticas básicas\n",
    "transacciones[\"DEUDA_MES\"].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Graficar un histograma y boxplot para conocer la distribución de la variable\n",
    "fig, ax = grafico_distribucion(transacciones[\"DEUDA_MES\"])\n",
    "plt.xlabel(\"\\nMonto de deuda ($)\")\n",
    "fig.suptitle(\"Distribución de la variable monto de deuda en todo el semestre\")\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Observamos que la variable DEUDA_MES presenta valores enteros, sin embargo,\n",
    "- Existen valores **negativos**, puede que sean datos inconsistentes o que representen algún otro hecho relacionado con la deuda. \n",
    "- Observamos la presencia de **outliers** estadísticos de acuerdo a la **regla intercuartil**.\n",
    "- Observamos también la presencia de datos **inconsistentes**, pues el monto de deuda no puede ser mayor al cupo de la tarjeta, reordemos que el cupo mayor era de $\\$800.000$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Existen 3163 transacciones con deudas negativas\n"
     ]
    }
   ],
   "source": [
    "num_deudas_neg = len(transacciones.query('DEUDA_MES<0'))\n",
    "print(f\"Existen {num_deudas_neg} transacciones con deudas negativas\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ahora observemos las deudas a nivel de cliente. Para ello transponemos los meses en columnas."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "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>MONTH</th>\n",
       "      <th>t-0</th>\n",
       "      <th>t-1</th>\n",
       "      <th>t-2</th>\n",
       "      <th>t-3</th>\n",
       "      <th>t-4</th>\n",
       "      <th>t-5</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CLIENT_ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>00004d47-2e99-461f-b6bf-59802fc72a8c</th>\n",
       "      <td>81787</td>\n",
       "      <td>77845</td>\n",
       "      <td>60159</td>\n",
       "      <td>54151</td>\n",
       "      <td>45108</td>\n",
       "      <td>44490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>000af878-0c8d-4d31-901a-d4b5c584e08f</th>\n",
       "      <td>21023</td>\n",
       "      <td>27270</td>\n",
       "      <td>11342</td>\n",
       "      <td>8305</td>\n",
       "      <td>7288</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>000c3e2f-013e-4894-81aa-aa96357a1cf2</th>\n",
       "      <td>19805</td>\n",
       "      <td>19900</td>\n",
       "      <td>390</td>\n",
       "      <td>390</td>\n",
       "      <td>0</td>\n",
       "      <td>780</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>000e7e42-2d80-441c-9ed5-df0e9dbb4c02</th>\n",
       "      <td>25487</td>\n",
       "      <td>6655</td>\n",
       "      <td>6355</td>\n",
       "      <td>18543</td>\n",
       "      <td>6022</td>\n",
       "      <td>1334</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0011d46a-bf48-482d-aab0-526fe8b6e254</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ffeeee1b-ad11-4bca-b461-2955c872a158</th>\n",
       "      <td>47937</td>\n",
       "      <td>33816</td>\n",
       "      <td>60320</td>\n",
       "      <td>36249</td>\n",
       "      <td>780</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fff43052-7f1f-473f-8a56-99fff7ed0018</th>\n",
       "      <td>22778</td>\n",
       "      <td>6405</td>\n",
       "      <td>1683</td>\n",
       "      <td>-885</td>\n",
       "      <td>3019</td>\n",
       "      <td>-177</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fff71109-4aae-49d5-9233-3efd3c7f45b1</th>\n",
       "      <td>268537</td>\n",
       "      <td>207118</td>\n",
       "      <td>202412</td>\n",
       "      <td>207496</td>\n",
       "      <td>250337</td>\n",
       "      <td>255277</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fff95170-fb6d-4561-91b2-94d81f4eebb2</th>\n",
       "      <td>1100</td>\n",
       "      <td>1183</td>\n",
       "      <td>2736</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fffa4fb3-4495-4622-8cce-2652ec71e2e6</th>\n",
       "      <td>24526</td>\n",
       "      <td>25532</td>\n",
       "      <td>25971</td>\n",
       "      <td>14859</td>\n",
       "      <td>14930</td>\n",
       "      <td>15091</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>24000 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "MONTH                                    t-0     t-1     t-2     t-3     t-4  \\\n",
       "CLIENT_ID                                                                      \n",
       "00004d47-2e99-461f-b6bf-59802fc72a8c   81787   77845   60159   54151   45108   \n",
       "000af878-0c8d-4d31-901a-d4b5c584e08f   21023   27270   11342    8305    7288   \n",
       "000c3e2f-013e-4894-81aa-aa96357a1cf2   19805   19900     390     390       0   \n",
       "000e7e42-2d80-441c-9ed5-df0e9dbb4c02   25487    6655    6355   18543    6022   \n",
       "0011d46a-bf48-482d-aab0-526fe8b6e254       0       0       0       0       0   \n",
       "...                                      ...     ...     ...     ...     ...   \n",
       "ffeeee1b-ad11-4bca-b461-2955c872a158   47937   33816   60320   36249     780   \n",
       "fff43052-7f1f-473f-8a56-99fff7ed0018   22778    6405    1683    -885    3019   \n",
       "fff71109-4aae-49d5-9233-3efd3c7f45b1  268537  207118  202412  207496  250337   \n",
       "fff95170-fb6d-4561-91b2-94d81f4eebb2    1100    1183    2736       0       0   \n",
       "fffa4fb3-4495-4622-8cce-2652ec71e2e6   24526   25532   25971   14859   14930   \n",
       "\n",
       "MONTH                                    t-5  \n",
       "CLIENT_ID                                     \n",
       "00004d47-2e99-461f-b6bf-59802fc72a8c   44490  \n",
       "000af878-0c8d-4d31-901a-d4b5c584e08f       0  \n",
       "000c3e2f-013e-4894-81aa-aa96357a1cf2     780  \n",
       "000e7e42-2d80-441c-9ed5-df0e9dbb4c02    1334  \n",
       "0011d46a-bf48-482d-aab0-526fe8b6e254       0  \n",
       "...                                      ...  \n",
       "ffeeee1b-ad11-4bca-b461-2955c872a158       0  \n",
       "fff43052-7f1f-473f-8a56-99fff7ed0018    -177  \n",
       "fff71109-4aae-49d5-9233-3efd3c7f45b1  255277  \n",
       "fff95170-fb6d-4561-91b2-94d81f4eebb2       0  \n",
       "fffa4fb3-4495-4622-8cce-2652ec71e2e6   15091  \n",
       "\n",
       "[24000 rows x 6 columns]"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "deudas = pd.pivot(transacciones, index=\"CLIENT_ID\", columns=\"MONTH\", values=\"DEUDA_MES\")\n",
    "deudas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "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>MONTH</th>\n",
       "      <th>t-0</th>\n",
       "      <th>t-1</th>\n",
       "      <th>t-2</th>\n",
       "      <th>t-3</th>\n",
       "      <th>t-4</th>\n",
       "      <th>t-5</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CLIENT_ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0025e4aa-2ad1-49a2-8ee5-fb6e65791502</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-101.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>004580bc-a753-4d62-8134-6dfa72b133e9</th>\n",
       "      <td>NaN</td>\n",
       "      <td>-2000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>004e5527-dd4a-48ef-add0-8fc797c4eef6</th>\n",
       "      <td>-89.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>009aae1c-e7fc-4ac8-8cc5-215a35d7ebdf</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-27.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>00c22be3-34dc-47cc-8fef-180da277f0f7</th>\n",
       "      <td>NaN</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>NaN</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ff4b718f-cd71-4e0b-ae52-faf5b36c4338</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-721.0</td>\n",
       "      <td>-710.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ff9871c0-0587-46fe-a95e-567d9db6c86d</th>\n",
       "      <td>NaN</td>\n",
       "      <td>-627.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ffb0196a-d064-4632-a8c0-0798c17d6933</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-29.0</td>\n",
       "      <td>-29.0</td>\n",
       "      <td>-29.0</td>\n",
       "      <td>-29.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ffd23e07-5d75-449b-ab32-ab0cf852668c</th>\n",
       "      <td>NaN</td>\n",
       "      <td>-4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fff43052-7f1f-473f-8a56-99fff7ed0018</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-885.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-177.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1524 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "MONTH                                  t-0     t-1    t-2    t-3   t-4    t-5\n",
       "CLIENT_ID                                                                    \n",
       "0025e4aa-2ad1-49a2-8ee5-fb6e65791502   NaN     NaN    NaN -101.0   NaN    NaN\n",
       "004580bc-a753-4d62-8134-6dfa72b133e9   NaN -2000.0    NaN    NaN   NaN    NaN\n",
       "004e5527-dd4a-48ef-add0-8fc797c4eef6 -89.0     NaN    NaN    NaN   NaN    NaN\n",
       "009aae1c-e7fc-4ac8-8cc5-215a35d7ebdf   NaN     NaN  -27.0    NaN   NaN    NaN\n",
       "00c22be3-34dc-47cc-8fef-180da277f0f7   NaN    -1.0   -1.0   -1.0  -1.0    NaN\n",
       "...                                    ...     ...    ...    ...   ...    ...\n",
       "ff4b718f-cd71-4e0b-ae52-faf5b36c4338   NaN     NaN -721.0 -710.0   NaN    NaN\n",
       "ff9871c0-0587-46fe-a95e-567d9db6c86d   NaN  -627.0    NaN    NaN   NaN    NaN\n",
       "ffb0196a-d064-4632-a8c0-0798c17d6933   NaN     NaN  -29.0  -29.0 -29.0  -29.0\n",
       "ffd23e07-5d75-449b-ab32-ab0cf852668c   NaN    -4.0    NaN    NaN   NaN    NaN\n",
       "fff43052-7f1f-473f-8a56-99fff7ed0018   NaN     NaN    NaN -885.0   NaN -177.0\n",
       "\n",
       "[1524 rows x 6 columns]"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Clientes con deudas negativas\n",
    "deudas[deudas<0].dropna(how='all')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Hay 1524 clientes con deudas negativas, lo que representa el 6.35%\n"
     ]
    }
   ],
   "source": [
    "num_clientes_deudas_neg = len(deudas[deudas<0].dropna(how='all'))\n",
    "prop_deudas_neg = num_clientes_deudas_neg/len(clientes)\n",
    "print(\"Hay {} clientes con deudas negativas, lo que representa el {:.2%}\".format(num_clientes_deudas_neg, prop_deudas_neg))\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Observamos que existen 1524 clientes que tienen por lo menos un monto de deuda negativo en algún mes, lo que representa un 6.35% de la muestra de todos los clientes. Para corregir este problema se podrían eliminar estos registros o reemplazarlos usando alguna técnica de imputación de datos."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**¿Imputación de datos?**\n",
    "\n",
    "Una aproximación a la imputación de datos para esta variable consiste en reemplazar los valores negativos con la media de deuda a lo largo del semestre o reemplazar el dato negativo realizando una interpolación con base en los datos positivos. Sin embargo, para realizarlo, tendríamos que asegurarnos de que la mayoría de datos de deudas del cliente fuesen consistentes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "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>MONTH</th>\n",
       "      <th>t-0</th>\n",
       "      <th>t-1</th>\n",
       "      <th>t-2</th>\n",
       "      <th>t-3</th>\n",
       "      <th>t-4</th>\n",
       "      <th>t-5</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CLIENT_ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>01b4f798-0901-44cf-9c63-63e087da56d1</th>\n",
       "      <td>-1590.0</td>\n",
       "      <td>-6256.0</td>\n",
       "      <td>-10922.0</td>\n",
       "      <td>-15588.0</td>\n",
       "      <td>-20254.0</td>\n",
       "      <td>-24920.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>07d20a27-570c-4ee9-94df-5ce595bc8259</th>\n",
       "      <td>-5.0</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>-5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0a53f40f-ed7a-4297-9b30-d07d3fad3d93</th>\n",
       "      <td>-212.0</td>\n",
       "      <td>-212.0</td>\n",
       "      <td>-212.0</td>\n",
       "      <td>-212.0</td>\n",
       "      <td>-212.0</td>\n",
       "      <td>-212.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11746ec8-1a75-4250-8d9f-07eccb5998eb</th>\n",
       "      <td>-3.0</td>\n",
       "      <td>-3.0</td>\n",
       "      <td>-3.0</td>\n",
       "      <td>-3.0</td>\n",
       "      <td>-3.0</td>\n",
       "      <td>-3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11cc56fe-19da-4042-9c6a-6bf4baaf7162</th>\n",
       "      <td>-12.0</td>\n",
       "      <td>-12.0</td>\n",
       "      <td>-12.0</td>\n",
       "      <td>-12.0</td>\n",
       "      <td>-12.0</td>\n",
       "      <td>-12.0</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ec0a1e13-85c0-4f7c-b8e4-ee0b559a4115</th>\n",
       "      <td>-10.0</td>\n",
       "      <td>-10.0</td>\n",
       "      <td>-10.0</td>\n",
       "      <td>-10.0</td>\n",
       "      <td>-10.0</td>\n",
       "      <td>-10.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f15aa97e-4d5e-4cb2-8537-3b641b7ce6b6</th>\n",
       "      <td>-702.0</td>\n",
       "      <td>-808.0</td>\n",
       "      <td>-1141.0</td>\n",
       "      <td>-1247.0</td>\n",
       "      <td>-1353.0</td>\n",
       "      <td>-1459.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f46c196b-a311-4f1e-a565-caf6d6065cd5</th>\n",
       "      <td>-10.0</td>\n",
       "      <td>-10.0</td>\n",
       "      <td>-10.0</td>\n",
       "      <td>-10.0</td>\n",
       "      <td>-10.0</td>\n",
       "      <td>-10.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fc1688fc-f318-4deb-8914-e9f14ea06f1b</th>\n",
       "      <td>-18.0</td>\n",
       "      <td>-18.0</td>\n",
       "      <td>-18.0</td>\n",
       "      <td>-18.0</td>\n",
       "      <td>-18.0</td>\n",
       "      <td>-18.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fcd2c327-2aac-4606-90eb-785be7328325</th>\n",
       "      <td>-897.0</td>\n",
       "      <td>-897.0</td>\n",
       "      <td>-897.0</td>\n",
       "      <td>-897.0</td>\n",
       "      <td>-897.0</td>\n",
       "      <td>-897.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>74 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "MONTH                                    t-0     t-1      t-2      t-3  \\\n",
       "CLIENT_ID                                                                \n",
       "01b4f798-0901-44cf-9c63-63e087da56d1 -1590.0 -6256.0 -10922.0 -15588.0   \n",
       "07d20a27-570c-4ee9-94df-5ce595bc8259    -5.0    -5.0     -5.0     -5.0   \n",
       "0a53f40f-ed7a-4297-9b30-d07d3fad3d93  -212.0  -212.0   -212.0   -212.0   \n",
       "11746ec8-1a75-4250-8d9f-07eccb5998eb    -3.0    -3.0     -3.0     -3.0   \n",
       "11cc56fe-19da-4042-9c6a-6bf4baaf7162   -12.0   -12.0    -12.0    -12.0   \n",
       "...                                      ...     ...      ...      ...   \n",
       "ec0a1e13-85c0-4f7c-b8e4-ee0b559a4115   -10.0   -10.0    -10.0    -10.0   \n",
       "f15aa97e-4d5e-4cb2-8537-3b641b7ce6b6  -702.0  -808.0  -1141.0  -1247.0   \n",
       "f46c196b-a311-4f1e-a565-caf6d6065cd5   -10.0   -10.0    -10.0    -10.0   \n",
       "fc1688fc-f318-4deb-8914-e9f14ea06f1b   -18.0   -18.0    -18.0    -18.0   \n",
       "fcd2c327-2aac-4606-90eb-785be7328325  -897.0  -897.0   -897.0   -897.0   \n",
       "\n",
       "MONTH                                     t-4      t-5  \n",
       "CLIENT_ID                                               \n",
       "01b4f798-0901-44cf-9c63-63e087da56d1 -20254.0 -24920.0  \n",
       "07d20a27-570c-4ee9-94df-5ce595bc8259     -5.0     -5.0  \n",
       "0a53f40f-ed7a-4297-9b30-d07d3fad3d93   -212.0   -212.0  \n",
       "11746ec8-1a75-4250-8d9f-07eccb5998eb     -3.0     -3.0  \n",
       "11cc56fe-19da-4042-9c6a-6bf4baaf7162    -12.0    -12.0  \n",
       "...                                       ...      ...  \n",
       "ec0a1e13-85c0-4f7c-b8e4-ee0b559a4115    -10.0    -10.0  \n",
       "f15aa97e-4d5e-4cb2-8537-3b641b7ce6b6  -1353.0  -1459.0  \n",
       "f46c196b-a311-4f1e-a565-caf6d6065cd5    -10.0    -10.0  \n",
       "fc1688fc-f318-4deb-8914-e9f14ea06f1b    -18.0    -18.0  \n",
       "fcd2c327-2aac-4606-90eb-785be7328325   -897.0   -897.0  \n",
       "\n",
       "[74 rows x 6 columns]"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Clientes con todas sus deudas negativas\n",
    "deudas[deudas<0].dropna(how='any')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Observamos que existen 74 clientes con todas sus deudas negativas, en donde no tiene sentido realizar la imputación, por lo tanto, eliminaremos dichos registros para la construcción del modelo."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Pagos mensuales"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    1.440000e+05\n",
       "mean     5.225306e+03\n",
       "std      1.712472e+04\n",
       "min      0.000000e+00\n",
       "25%      3.900000e+02\n",
       "50%      1.874000e+03\n",
       "75%      4.516000e+03\n",
       "max      1.684259e+06\n",
       "Name: PAGO_MES, dtype: float64"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Medidas estadísticas básicas\n",
    "transacciones[\"PAGO_MES\"].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Graficar un histograma y boxplot para conocer la distribución de la variable\n",
    "fig, ax = grafico_distribucion(transacciones[\"PAGO_MES\"])\n",
    "plt.xlabel(\"\\nMonto de pago ($)\")\n",
    "fig.suptitle(\"Distribución de la variable monto de pago en todo el semestre\")\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "Obsermos la presencia de datos **inconsistentes** correspondientes a pagos mayores al cupo de la tarjeta, lo que no nos permite entender claramente los gráficos de dsitribución."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "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>MONTH</th>\n",
       "      <th>t-0</th>\n",
       "      <th>t-1</th>\n",
       "      <th>t-2</th>\n",
       "      <th>t-3</th>\n",
       "      <th>t-4</th>\n",
       "      <th>t-5</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CLIENT_ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>00004d47-2e99-461f-b6bf-59802fc72a8c</th>\n",
       "      <td>2600</td>\n",
       "      <td>3009</td>\n",
       "      <td>2200</td>\n",
       "      <td>2000</td>\n",
       "      <td>2000</td>\n",
       "      <td>2832</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>000af878-0c8d-4d31-901a-d4b5c584e08f</th>\n",
       "      <td>3000</td>\n",
       "      <td>1000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>000c3e2f-013e-4894-81aa-aa96357a1cf2</th>\n",
       "      <td>1000</td>\n",
       "      <td>390</td>\n",
       "      <td>390</td>\n",
       "      <td>0</td>\n",
       "      <td>780</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>000e7e42-2d80-441c-9ed5-df0e9dbb4c02</th>\n",
       "      <td>6655</td>\n",
       "      <td>6359</td>\n",
       "      <td>18543</td>\n",
       "      <td>6022</td>\n",
       "      <td>1334</td>\n",
       "      <td>3864</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0011d46a-bf48-482d-aab0-526fe8b6e254</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ffeeee1b-ad11-4bca-b461-2955c872a158</th>\n",
       "      <td>2000</td>\n",
       "      <td>2000</td>\n",
       "      <td>3293</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fff43052-7f1f-473f-8a56-99fff7ed0018</th>\n",
       "      <td>6500</td>\n",
       "      <td>2515</td>\n",
       "      <td>0</td>\n",
       "      <td>8000</td>\n",
       "      <td>0</td>\n",
       "      <td>1000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fff71109-4aae-49d5-9233-3efd3c7f45b1</th>\n",
       "      <td>10000</td>\n",
       "      <td>8000</td>\n",
       "      <td>12000</td>\n",
       "      <td>50000</td>\n",
       "      <td>28000</td>\n",
       "      <td>60000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fff95170-fb6d-4561-91b2-94d81f4eebb2</th>\n",
       "      <td>1183</td>\n",
       "      <td>2736</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fffa4fb3-4495-4622-8cce-2652ec71e2e6</th>\n",
       "      <td>1790</td>\n",
       "      <td>1229</td>\n",
       "      <td>519</td>\n",
       "      <td>687</td>\n",
       "      <td>557</td>\n",
       "      <td>546</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>24000 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "MONTH                                   t-0   t-1    t-2    t-3    t-4    t-5\n",
       "CLIENT_ID                                                                    \n",
       "00004d47-2e99-461f-b6bf-59802fc72a8c   2600  3009   2200   2000   2000   2832\n",
       "000af878-0c8d-4d31-901a-d4b5c584e08f   3000  1000      0      0      0      0\n",
       "000c3e2f-013e-4894-81aa-aa96357a1cf2   1000   390    390      0    780      0\n",
       "000e7e42-2d80-441c-9ed5-df0e9dbb4c02   6655  6359  18543   6022   1334   3864\n",
       "0011d46a-bf48-482d-aab0-526fe8b6e254      0     0      0      0      0      0\n",
       "...                                     ...   ...    ...    ...    ...    ...\n",
       "ffeeee1b-ad11-4bca-b461-2955c872a158   2000  2000   3293      0      0      0\n",
       "fff43052-7f1f-473f-8a56-99fff7ed0018   6500  2515      0   8000      0   1000\n",
       "fff71109-4aae-49d5-9233-3efd3c7f45b1  10000  8000  12000  50000  28000  60000\n",
       "fff95170-fb6d-4561-91b2-94d81f4eebb2   1183  2736      0      0      0      0\n",
       "fffa4fb3-4495-4622-8cce-2652ec71e2e6   1790  1229    519    687    557    546\n",
       "\n",
       "[24000 rows x 6 columns]"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Tabla pivote de pagos\n",
    "pagos = pd.pivot(transacciones, index=\"CLIENT_ID\", columns=\"MONTH\", values=\"PAGO_MES\")\n",
    "pagos"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Registros monetarios inconsistentes"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Existen registros de pagos y deudas en la base de datos de transacciones que superan el monto del cupo de la tarjeta de crédito de cada cliente, por lo que se consideran como datos inconsistentes. A continuación realizaremos un análisis de la cantidad de clientes que presentan este comportamiento.\n",
    "\n",
    "Para ello construimos una tabla con la información monetaria, es decir, pagos, deudas y cupo."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "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>Cupo_TC</th>\n",
       "      <th>DEUDA_MES_t-0</th>\n",
       "      <th>DEUDA_MES_t-1</th>\n",
       "      <th>DEUDA_MES_t-2</th>\n",
       "      <th>DEUDA_MES_t-3</th>\n",
       "      <th>DEUDA_MES_t-4</th>\n",
       "      <th>DEUDA_MES_t-5</th>\n",
       "      <th>PAGO_MES_t-0</th>\n",
       "      <th>PAGO_MES_t-1</th>\n",
       "      <th>PAGO_MES_t-2</th>\n",
       "      <th>PAGO_MES_t-3</th>\n",
       "      <th>PAGO_MES_t-4</th>\n",
       "      <th>PAGO_MES_t-5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>270000</td>\n",
       "      <td>322249</td>\n",
       "      <td>298700</td>\n",
       "      <td>308133</td>\n",
       "      <td>263516</td>\n",
       "      <td>94062</td>\n",
       "      <td>92618</td>\n",
       "      <td>10000</td>\n",
       "      <td>13000</td>\n",
       "      <td>12000</td>\n",
       "      <td>5000</td>\n",
       "      <td>4000</td>\n",
       "      <td>5000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>80000</td>\n",
       "      <td>42635</td>\n",
       "      <td>39508</td>\n",
       "      <td>36060</td>\n",
       "      <td>30470</td>\n",
       "      <td>18306</td>\n",
       "      <td>7432</td>\n",
       "      <td>1623</td>\n",
       "      <td>1740</td>\n",
       "      <td>1200</td>\n",
       "      <td>366</td>\n",
       "      <td>7432</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>210000</td>\n",
       "      <td>143225</td>\n",
       "      <td>142996</td>\n",
       "      <td>144415</td>\n",
       "      <td>148190</td>\n",
       "      <td>147993</td>\n",
       "      <td>131629</td>\n",
       "      <td>6483</td>\n",
       "      <td>5082</td>\n",
       "      <td>6000</td>\n",
       "      <td>5900</td>\n",
       "      <td>4544</td>\n",
       "      <td>4700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>230000</td>\n",
       "      <td>54798</td>\n",
       "      <td>11632</td>\n",
       "      <td>4987</td>\n",
       "      <td>6222</td>\n",
       "      <td>15121</td>\n",
       "      <td>17425</td>\n",
       "      <td>11632</td>\n",
       "      <td>4987</td>\n",
       "      <td>6222</td>\n",
       "      <td>15121</td>\n",
       "      <td>17425</td>\n",
       "      <td>17007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>130000</td>\n",
       "      <td>129689</td>\n",
       "      <td>107070</td>\n",
       "      <td>103942</td>\n",
       "      <td>94623</td>\n",
       "      <td>90556</td>\n",
       "      <td>92394</td>\n",
       "      <td>5000</td>\n",
       "      <td>4400</td>\n",
       "      <td>5000</td>\n",
       "      <td>4000</td>\n",
       "      <td>3500</td>\n",
       "      <td>4000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Cupo_TC  DEUDA_MES_t-0  DEUDA_MES_t-1  DEUDA_MES_t-2  DEUDA_MES_t-3  \\\n",
       "0   270000         322249         298700         308133         263516   \n",
       "1    80000          42635          39508          36060          30470   \n",
       "2   210000         143225         142996         144415         148190   \n",
       "3   230000          54798          11632           4987           6222   \n",
       "4   130000         129689         107070         103942          94623   \n",
       "\n",
       "   DEUDA_MES_t-4  DEUDA_MES_t-5  PAGO_MES_t-0  PAGO_MES_t-1  PAGO_MES_t-2  \\\n",
       "0          94062          92618         10000         13000         12000   \n",
       "1          18306           7432          1623          1740          1200   \n",
       "2         147993         131629          6483          5082          6000   \n",
       "3          15121          17425         11632          4987          6222   \n",
       "4          90556          92394          5000          4400          5000   \n",
       "\n",
       "   PAGO_MES_t-3  PAGO_MES_t-4  PAGO_MES_t-5  \n",
       "0          5000          4000          5000  \n",
       "1           366          7432             0  \n",
       "2          5900          4544          4700  \n",
       "3         15121         17425         17007  \n",
       "4          4000          3500          4000  "
      ]
     },
     "execution_count": 157,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Transponer deudas y pagos de la tabla transacciones\n",
    "money_df = pd.pivot(transacciones, index=\"CLIENT_ID\", columns=\"MONTH\", values=[\"DEUDA_MES\", \"PAGO_MES\"])\n",
    "# Eliminar el multíndice generado por la transposición\n",
    "money_df.columns = money_df.columns.to_flat_index()\n",
    "money_df.columns = [ f\"{var}_{time_tag}\" for (var, time_tag) in money_df.columns ]\n",
    "money_df = money_df.reset_index()\n",
    "# Traer el cupo de la tabla Clientes\n",
    "money_df = clientes[[\"CLIENT_ID\",\"Cupo_TC\"]].merge(money_df, on=\"CLIENT_ID\")\n",
    "# Eliminar el ID\n",
    "money_df = money_df.drop(columns=[\"CLIENT_ID\"])\n",
    "money_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "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>Cupo_TC</th>\n",
       "      <th>DEUDA_MES_t-0</th>\n",
       "      <th>DEUDA_MES_t-1</th>\n",
       "      <th>DEUDA_MES_t-2</th>\n",
       "      <th>DEUDA_MES_t-3</th>\n",
       "      <th>DEUDA_MES_t-4</th>\n",
       "      <th>DEUDA_MES_t-5</th>\n",
       "      <th>PAGO_MES_t-0</th>\n",
       "      <th>PAGO_MES_t-1</th>\n",
       "      <th>PAGO_MES_t-2</th>\n",
       "      <th>PAGO_MES_t-3</th>\n",
       "      <th>PAGO_MES_t-4</th>\n",
       "      <th>PAGO_MES_t-5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>270000</td>\n",
       "      <td>322249</td>\n",
       "      <td>298700</td>\n",
       "      <td>308133</td>\n",
       "      <td>263516</td>\n",
       "      <td>94062</td>\n",
       "      <td>92618</td>\n",
       "      <td>10000</td>\n",
       "      <td>13000</td>\n",
       "      <td>12000</td>\n",
       "      <td>5000</td>\n",
       "      <td>4000</td>\n",
       "      <td>5000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>20000</td>\n",
       "      <td>19149</td>\n",
       "      <td>20221</td>\n",
       "      <td>19812</td>\n",
       "      <td>10717</td>\n",
       "      <td>13534</td>\n",
       "      <td>15298</td>\n",
       "      <td>1392</td>\n",
       "      <td>1201</td>\n",
       "      <td>1000</td>\n",
       "      <td>3000</td>\n",
       "      <td>2000</td>\n",
       "      <td>399</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>50000</td>\n",
       "      <td>44470</td>\n",
       "      <td>51192</td>\n",
       "      <td>48506</td>\n",
       "      <td>18524</td>\n",
       "      <td>19239</td>\n",
       "      <td>19283</td>\n",
       "      <td>7800</td>\n",
       "      <td>173</td>\n",
       "      <td>1000</td>\n",
       "      <td>1000</td>\n",
       "      <td>1000</td>\n",
       "      <td>1000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>60000</td>\n",
       "      <td>60745</td>\n",
       "      <td>61085</td>\n",
       "      <td>56863</td>\n",
       "      <td>18473</td>\n",
       "      <td>15434</td>\n",
       "      <td>7301</td>\n",
       "      <td>2613</td>\n",
       "      <td>5900</td>\n",
       "      <td>1000</td>\n",
       "      <td>428</td>\n",
       "      <td>1000</td>\n",
       "      <td>2000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>20000</td>\n",
       "      <td>20571</td>\n",
       "      <td>19089</td>\n",
       "      <td>19658</td>\n",
       "      <td>19453</td>\n",
       "      <td>19108</td>\n",
       "      <td>18868</td>\n",
       "      <td>1323</td>\n",
       "      <td>1600</td>\n",
       "      <td>830</td>\n",
       "      <td>700</td>\n",
       "      <td>674</td>\n",
       "      <td>376</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Cupo_TC  DEUDA_MES_t-0  DEUDA_MES_t-1  DEUDA_MES_t-2  DEUDA_MES_t-3  \\\n",
       "0    270000         322249         298700         308133         263516   \n",
       "7     20000          19149          20221          19812          10717   \n",
       "19    50000          44470          51192          48506          18524   \n",
       "37    60000          60745          61085          56863          18473   \n",
       "43    20000          20571          19089          19658          19453   \n",
       "\n",
       "    DEUDA_MES_t-4  DEUDA_MES_t-5  PAGO_MES_t-0  PAGO_MES_t-1  PAGO_MES_t-2  \\\n",
       "0           94062          92618         10000         13000         12000   \n",
       "7           13534          15298          1392          1201          1000   \n",
       "19          19239          19283          7800           173          1000   \n",
       "37          15434           7301          2613          5900          1000   \n",
       "43          19108          18868          1323          1600           830   \n",
       "\n",
       "    PAGO_MES_t-3  PAGO_MES_t-4  PAGO_MES_t-5  \n",
       "0           5000          4000          5000  \n",
       "7           3000          2000           399  \n",
       "19          1000          1000          1000  \n",
       "37           428          1000          2000  \n",
       "43           700           674           376  "
      ]
     },
     "execution_count": 169,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Hallar los registros en los que por lo mens un pago o una deuda supera al cupo\n",
    "condition = money_df.iloc[:, 1:].gt(money_df['Cupo_TC'], axis=0).any(axis=1)\n",
    "money_incosistente = money_df[condition]\n",
    "money_incosistente.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Hay 3201 clientes con registros monetarios inconsistentes\n"
     ]
    }
   ],
   "source": [
    "print(f\"Hay {len(money_incosistente)} clientes con registros monetarios inconsistentes\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Variable objetivo"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "La variable objetivo proviene de la tabla de etiquetas. Cada cliente se encuentra etiquetado con $0$ o $1$, donde $0$ significa que el pago el próximo mes será puntual y $1$ quiere decir que el cliente no va a pagar su deuda de la tarjeta de crédito el próximo mes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Proporción \n",
    "sns.countplot(etiquetas, x=\"default payment next month\")\n",
    "plt.ylabel(\"Cantidad de clientes\")\n",
    "plt.xlabel(\"Impago próximo mes\")\n",
    "plt.title(\"Cantidad de impagos para el próximo mes\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "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>cantidad</th>\n",
       "      <th>proporcion</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>default payment next month</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>18715</td>\n",
       "      <td>0.779792</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5285</td>\n",
       "      <td>0.220208</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                            cantidad  proporcion\n",
       "default payment next month                      \n",
       "0                              18715    0.779792\n",
       "1                               5285    0.220208"
      ]
     },
     "execution_count": 180,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "next_pay_summary = etiquetas.groupby(\"default payment next month\").count()\n",
    "next_pay_summary = next_pay_summary.rename(columns={\"CLIENT_ID\":\"cantidad\"})\n",
    "next_pay_summary[\"proporcion\"] = next_pay_summary[\"cantidad\"] /   next_pay_summary[\"cantidad\"].sum()\n",
    "next_pay_summary"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Se presenta un desbalance de clase, ya que el 78% de los deudores pagarán cumplido el siguiente mes y el 22% presentarán mora. Este desequilibrio de clases representa una gran dificultad a la hora construir el modelo predictivo, por tal razón, se deben usar técnicas para manejarlo."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Resumen"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "| **Campo**                            | **Descripción**                                                                                                                                | **Comentario**                                                                           |\n",
    "|--------------------------------------|------------------------------------------------------------------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------|\n",
    "| CLIENT_ID                            | Identificador único para cada cliente.                                                                                |    Se usa para unir tablas                                                                                      |\n",
    "| Cupo_TC (int)                        | Cupo inicial de la tarjeta de crédito.                                                                                                         | No presenta datos inconsistentes                                                         |\n",
    "| SEX (nominal)                        | Sexo biológico del cliente.  <br>Categorías: Masculino , Femenino                                                                                  | No presenta datos inconsistentes                                                         |\n",
    "| EDUCATION (nominal)                  | Nivel de escolaridad del cliente <br>Categorías: colegio, universidad, posgrado, otro                                                              | Valores inconsistentes \"0\",\"5\",\"6\"                                                       |\n",
    "| MARRIAGE (nominal)                   | Estado civil del cliente <br>Categorías: soltero_a, casado, otro                                                                                   | Valores inconsistentes  \"0\"                                                              |\n",
    "| AGE (int)                            | Edad del cliente                                                                                                                               | No presenta valores inconsistentes                                                       |\n",
    "| RETRASO_PAGO_ESTADO (int)            | Variable numérica ordinal <br>Serie de tiempo mensual de 6 observaciones para cada cliente <br>Valores: -2, -1 , 0, 1, 2, 3, 4, 5, 6, 7, 8             | No presenta valores inconsistentes                                                       |\n",
    "| PAGO_MES (int)                       | Variable entera, representa el monto pagado en un mes determinado <br>Serie de tiempo mensual de 6 observaciones para cada cliente                 | Pagos inconsistentes que  superan el valor del cupo                                      |\n",
    "| DEUDA_MES (int)                      | Variable entera, representa el monto que un cliente  adeuda en un mes determinado <br> Serie de tiempo mensual de 6 observaciones para cada cliente | - Montos inconsistentes que  superan el valor del cupo <br>- Montos negativos inconsistentes |\n",
    "| default payment next month (binaria) | Variable objetivo binaria, 1 para impago.                                                                                                      | Existe un gran desbalance de clase                                                       |"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Se eliminarán todos los registros que contengan por lo menos un valor inconsistente en alguna columna. Se descarta la imputación de datos en lugar de valores inconsistentes ya que estos registros representan un pequeño porcentaje de la muestra.\n",
    "- Existe un importante desbalance de clase importante en la variable objetivo, para ellos se propone usar tres técnicas distintas: Validación cruzada estratificada, imputación de datos (oversampling) y eliminación de datos (undersampling)."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "venv",
   "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": "3.11.3"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}