{ "cells": [ { "cell_type": "markdown", "id": "hindu-assistant", "metadata": {}, "source": [ "# Identificación, tratamiento y manejo de Valores faltantes o Missing Values\n", "" ] }, { "cell_type": "markdown", "id": "graduate-trigger", "metadata": {}, "source": [ "## En este notebook aprenderás a:\n", "
\n", "\n", "
" ] }, { "cell_type": "markdown", "id": "hired-royalty", "metadata": {}, "source": [ "## Antes de empezar" ] }, { "cell_type": "markdown", "id": "noted-herald", "metadata": {}, "source": [ "En este notebook se han utilizado los siguientes paquetes que puedes instalar utilizando las líneas de código anexas en la siguiente celda:\n", "- pandas\n", "- numpy\n", "- missingno\n" ] }, { "cell_type": "code", "execution_count": 1, "id": "cross-savannah", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:30.987265Z", "start_time": "2022-02-16T20:11:30.983293Z" } }, "outputs": [], "source": [ "# !pip install pandas --user\n", "# !pip install numpy --user\n", "# !pip install scikit-learn --user\n", "# !pip install missingno --user" ] }, { "cell_type": "markdown", "id": "returning-retention", "metadata": {}, "source": [ "# Ejemplo 1" ] }, { "cell_type": "markdown", "id": "skilled-relaxation", "metadata": {}, "source": [ "## Leyendo el DataFrame\n", "\n", "Comentábamos en la parte de teoría que en ocasiones los Missing Values nos pueden venir dados como:\n", "- Un valor ausente en la tabla\n", "- Un valor por defecto que quiere indicar Missing Value\n", "\n", "🎁 Te dejo por aquí una lista de valores frequentes *(treat_NaNs)* que por defecto podrían representar valores faltantes" ] }, { "cell_type": "markdown", "id": "economic-course", "metadata": {}, "source": [ "Si leemos la primera hoja de cálculo del archivo excel con nombre \"dataset_notebook_demo.xlsx\" sin utilizar esa lista observa lo que sucede:" ] }, { "cell_type": "code", "execution_count": 2, "id": "tamil-wilderness", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:32.145321Z", "start_time": "2022-02-16T20:11:30.998262Z" } }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
AñoGéneroNúmero Ventas
02014Aventuras400.0
1-Bélico80.0
2-Biografías200.0
3-Novela Romántica350.0
4-Poesía80.0
52015Aventuras500.0
6-Bélico150.0
7-Biografías200.0
8-Novela Romántica300.0
9-Poesía120.0
102016Aventuras700.0
11-Bélico200.0
12-Biografías200.0
13-Novela Romántica300.0
14-Poesía150.0
152017Aventuras700.0
16-Bélico200.0
17-BiografíasNaN
18-Novela Romántica350.0
19-Poesía100.0
202018Aventuras880.0
21-Bélico180.0
22-Biografías190.0
23-Novela Romántica100.0
24-Poesía95.0
252019Aventuras200.0
26-Bélico200.0
27-Biografías300.0
28-Novela Romántica120.0
29-Poesía120.0
302020Aventuras1000.0
31-Bélico400.0
32-Biografías350.0
33-Novela RománticaNaN
34-Poesía400.0
352021Aventuras900.0
36-Bélico350.0
37-Biografías250.0
38-Novela Romántica200.0
39-Poesía100.0
\n", "
" ], "text/plain": [ " Año Género Número Ventas\n", "0 2014 Aventuras 400.0\n", "1 - Bélico 80.0\n", "2 - Biografías 200.0\n", "3 - Novela Romántica 350.0\n", "4 - Poesía 80.0\n", "5 2015 Aventuras 500.0\n", "6 - Bélico 150.0\n", "7 - Biografías 200.0\n", "8 - Novela Romántica 300.0\n", "9 - Poesía 120.0\n", "10 2016 Aventuras 700.0\n", "11 - Bélico 200.0\n", "12 - Biografías 200.0\n", "13 - Novela Romántica 300.0\n", "14 - Poesía 150.0\n", "15 2017 Aventuras 700.0\n", "16 - Bélico 200.0\n", "17 - Biografías NaN\n", "18 - Novela Romántica 350.0\n", "19 - Poesía 100.0\n", "20 2018 Aventuras 880.0\n", "21 - Bélico 180.0\n", "22 - Biografías 190.0\n", "23 - Novela Romántica 100.0\n", "24 - Poesía 95.0\n", "25 2019 Aventuras 200.0\n", "26 - Bélico 200.0\n", "27 - Biografías 300.0\n", "28 - Novela Romántica 120.0\n", "29 - Poesía 120.0\n", "30 2020 Aventuras 1000.0\n", "31 - Bélico 400.0\n", "32 - Biografías 350.0\n", "33 - Novela Romántica NaN\n", "34 - Poesía 400.0\n", "35 2021 Aventuras 900.0\n", "36 - Bélico 350.0\n", "37 - Biografías 250.0\n", "38 - Novela Romántica 200.0\n", "39 - Poesía 100.0" ] }, "execution_count": 2, "metadata": {}, "output_type": "execute_result" } ], "source": [ "import pandas as pd\n", "ruta_dataset = \"dataset_notebook_demo.xlsx\"\n", "df= pd.read_excel(ruta_dataset)\n", "df" ] }, { "cell_type": "markdown", "id": "patent-falls", "metadata": {}, "source": [ "Tenemos un valor \"-\" que representa un valor faltante, pero no se detecta como tal si no como un caracter.\n", "\n", "Si ahora por el contrario, empleamos la lista que te comentaba para que reconozca ese caracter como un valor faltante:" ] }, { "cell_type": "code", "execution_count": 3, "id": "blessed-joyce", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:32.209295Z", "start_time": "2022-02-16T20:11:32.148293Z" } }, "outputs": [], "source": [ "treat_NaNs = [\" \",\"\",\".\",\"-\",\"._\",\",\",\";\",\":\",\"''\",\"'\",\"/\",\"?\",\"!\",\"[]\",\"#\",\n", "\"´\",\"&\",\"$\",\"()\",\"{}\",\"ç\",\"Ç\",\"`\",\"^\",\"*\",\"+\",\"|\",\"%\",\"n/a\",\"N/A\",\"--\",\"NA\",\"na\"]\n", "\n", "ruta_dataset = \"dataset_notebook_demo.xlsx\"\n", "df = pd.read_excel(ruta_dataset, na_values=treat_NaNs)" ] }, { "cell_type": "code", "execution_count": 4, "id": "uniform-spice", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:32.223300Z", "start_time": "2022-02-16T20:11:32.212278Z" } }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
AñoGéneroNúmero Ventas
02014.0Aventuras400.0
1NaNBélico80.0
2NaNBiografías200.0
3NaNNovela Romántica350.0
4NaNPoesía80.0
\n", "
" ], "text/plain": [ " Año Género Número Ventas\n", "0 2014.0 Aventuras 400.0\n", "1 NaN Bélico 80.0\n", "2 NaN Biografías 200.0\n", "3 NaN Novela Romántica 350.0\n", "4 NaN Poesía 80.0" ] }, "execution_count": 4, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.head()" ] }, { "cell_type": "markdown", "id": "genuine-minimum", "metadata": {}, "source": [ "comprobamos como ahora el caracter \"-\" es reconocido como un valor NaN" ] }, { "cell_type": "markdown", "id": "further-facing", "metadata": {}, "source": [ "## Describiendo el DataFrame\n", "Uno de los primeros pasos cuando llega a nuestras manos un set de datos es **explorarlo**.\n", "\n", "Y digo uno de los primeros porque el más importante es conseguir abrir el archivo. Ese aspecto daría para\n", "otra tarea completa dedicada a a la correcta conexión y lectura de información proveniente de diferentes orígenes de datos.\n", "\n", "No obstante, para poder empezar a trabajar con un conjunto de datos cualesquiera es importante **analizarlos previamente**.\n", "\n", "Los pasos más habituales suelen ser:\n", "- Ver las primeras y últimas filas de nuestro DataFrame\n", "- Intentar realizar una descripción rápida\n", "- Ver valores únicos, tipos de datos de los campos, y analizar los missing values\n", "\n", "Vamos allá❗" ] }, { "cell_type": "markdown", "id": "adequate-harvard", "metadata": {}, "source": [ "🔵 Este dataset representa el número de libros vendidos por género y por año por un comercio local." ] }, { "cell_type": "code", "execution_count": 5, "id": "blank-indicator", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:32.240266Z", "start_time": "2022-02-16T20:11:32.227272Z" } }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
AñoGéneroNúmero Ventas
02014.0Aventuras400.0
1NaNBélico80.0
2NaNBiografías200.0
3NaNNovela Romántica350.0
4NaNPoesía80.0
52015.0Aventuras500.0
6NaNBélico150.0
7NaNBiografías200.0
8NaNNovela Romántica300.0
9NaNPoesía120.0
\n", "
" ], "text/plain": [ " Año Género Número Ventas\n", "0 2014.0 Aventuras 400.0\n", "1 NaN Bélico 80.0\n", "2 NaN Biografías 200.0\n", "3 NaN Novela Romántica 350.0\n", "4 NaN Poesía 80.0\n", "5 2015.0 Aventuras 500.0\n", "6 NaN Bélico 150.0\n", "7 NaN Biografías 200.0\n", "8 NaN Novela Romántica 300.0\n", "9 NaN Poesía 120.0" ] }, "execution_count": 5, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# visualizando las n primeras filas del dataset\n", "df.head(10)\n", "\n", "# visualizando las n últimas filas del dataset\n", "# df.tail(2)" ] }, { "cell_type": "code", "execution_count": 6, "id": "atlantic-shuttle", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:32.254265Z", "start_time": "2022-02-16T20:11:32.243280Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "RangeIndex: 40 entries, 0 to 39\n", "Data columns (total 3 columns):\n", " # Column Non-Null Count Dtype \n", "--- ------ -------------- ----- \n", " 0 Año 8 non-null float64\n", " 1 Género 40 non-null object \n", " 2 Número Ventas 38 non-null float64\n", "dtypes: float64(2), object(1)\n", "memory usage: 4.1 KB\n" ] } ], "source": [ "# el método info nos da información inicial sobre el dataset, representando las columnas, \n", "# el número de valores NO MISSSING y el tipo de columna. \n", "df.info(memory_usage='deep') #deep nos indica cuánto ocupa el dataset en memoria" ] }, { "cell_type": "markdown", "id": "rising-tennis", "metadata": {}, "source": [ "Personalmente, creo que puede estar bien para hacerse una idea de la cantidad de valores existentes siempre y cuando sepamos la forma del dataset *(df.shape)*. Pero yo no lo utilizo mucho 😉" ] }, { "cell_type": "markdown", "id": "negative-heating", "metadata": {}, "source": [ "A la hora de obtener un *overview* de la cantidad de valores faltantes que tiene nuestro conjunto de datos, podemos utilizar indistintamente los métodos `.isna()` o `.isnull()`" ] }, { "cell_type": "code", "execution_count": 7, "id": "behavioral-angle", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:32.281268Z", "start_time": "2022-02-16T20:11:32.263262Z" } }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
AñoGéneroNúmero Ventas
0FalseFalseFalse
1TrueFalseFalse
2TrueFalseFalse
3TrueFalseFalse
4TrueFalseFalse
\n", "
" ], "text/plain": [ " Año Género Número Ventas\n", "0 False False False\n", "1 True False False\n", "2 True False False\n", "3 True False False\n", "4 True False False" ] }, "execution_count": 7, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.isna().head()" ] }, { "cell_type": "markdown", "id": "quiet-dragon", "metadata": {}, "source": [ "Estarás de acuerdo conmigo en que esto no es muy intuitivo, ¿verdad?. 🤔🤔\n", "Si utilizamos los métodos `df.isna().sum()` o `df.isnull().sum()` obtendremos un recuento de los valores faltantes que tenemos por cada columna.\n", "\n", "💭 Si me preguntas por mis preferencias, personalmente tengo tendencia a utilizar el segundo método." ] }, { "cell_type": "code", "execution_count": 8, "id": "soviet-argentina", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:32.295309Z", "start_time": "2022-02-16T20:11:32.284262Z" } }, "outputs": [ { "data": { "text/plain": [ "Año 32\n", "Género 0\n", "Número Ventas 2\n", "dtype: int64" ] }, "execution_count": 8, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.isnull().sum()" ] }, { "cell_type": "markdown", "id": "purple-support", "metadata": {}, "source": [ "Si quieres ver el número de Missing Values por cada fila:" ] }, { "cell_type": "code", "execution_count": 9, "id": "vocational-madonna", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:32.308268Z", "start_time": "2022-02-16T20:11:32.297268Z" } }, "outputs": [ { "data": { "text/plain": [ "0 0\n", "1 1\n", "2 1\n", "3 1\n", "4 1\n", "dtype: int64" ] }, "execution_count": 9, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.isnull().sum(axis=1).head()" ] }, { "cell_type": "markdown", "id": "double-funds", "metadata": {}, "source": [ "¡Ya empezamos a hacernos una idea de cómo de completo está nuestro conjunto de datos!\n", "Pero podemos ir un pasito más allá y ver qué porcentaje del total suponen esos valores faltantes para cada columna.\n", "Te dejo por aquí una función que puedes reutilizar en tu día a día 😊\n", "\n", "🔥 Te animo a que construyas tu propia función para obtener el porcentaje de completitud de cada fila." ] }, { "cell_type": "code", "execution_count": 10, "id": "occupational-variation", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:32.317264Z", "start_time": "2022-02-16T20:11:32.312266Z" } }, "outputs": [], "source": [ "def porcentaje_missings(data):\n", " m_per = data.isnull().sum() * 100 / len(data)\n", " percent_missing = round(m_per,2)\n", " missing_value_df = pd.DataFrame({'percent_missing (%)': percent_missing})\n", " return missing_value_df " ] }, { "cell_type": "code", "execution_count": 11, "id": "private-swaziland", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:32.325268Z", "start_time": "2022-02-16T20:11:32.320293Z" } }, "outputs": [], "source": [ "#El huequecito para tu función :P" ] }, { "cell_type": "code", "execution_count": 12, "id": "saving-tonight", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:34.438458Z", "start_time": "2022-02-16T20:11:32.327273Z" } }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "C:\\Users\\Borja\\Anaconda3\\lib\\site-packages\\pandas\\compat\\_optional.py:123: UserWarning: Pandas requires version '2.6.8' or newer of 'numexpr' (version '2.6.2' currently installed).\n", " warnings.warn(msg, UserWarning)\n" ] }, { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
percent_missing (%)
Año80.0
Género0.0
Número Ventas5.0
\n", "
" ], "text/plain": [ " percent_missing (%)\n", "Año 80.0\n", "Género 0.0\n", "Número Ventas 5.0" ] }, "execution_count": 12, "metadata": {}, "output_type": "execute_result" } ], "source": [ "porcentaje_missings(df)" ] }, { "cell_type": "markdown", "id": "spiritual-prague", "metadata": { "ExecuteTime": { "end_time": "2022-02-13T18:41:03.082025Z", "start_time": "2022-02-13T18:41:03.076043Z" } }, "source": [ "Si todavía te quedas con ganas de obtener una visualización gráfica sobre la cantidad de missings que tienen tus datos,\n", "\n", "¡Échale un vistazo a las siguientes líneas de código!" ] }, { "cell_type": "code", "execution_count": 13, "id": "previous-artist", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:35.214508Z", "start_time": "2022-02-16T20:11:34.441463Z" } }, "outputs": [], "source": [ "import missingno as msno" ] }, { "cell_type": "code", "execution_count": 14, "id": "handed-administrator", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:35.712485Z", "start_time": "2022-02-16T20:11:35.216457Z" } }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "msno.matrix(df,figsize=(8,5), fontsize=12); #en blanco, cada missing value\n", "msno.bar(df, figsize=(8,5), fontsize=12); " ] }, { "cell_type": "markdown", "id": "equal-saudi", "metadata": {}, "source": [ "Finalmente, si quieres ver el número total de Missing Values que hay repartidos por todo tu dataset..." ] }, { "cell_type": "code", "execution_count": 15, "id": "external-watershed", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:35.723490Z", "start_time": "2022-02-16T20:11:35.714460Z" } }, "outputs": [ { "data": { "text/plain": [ "34" ] }, "execution_count": 15, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.isnull().sum().sum() \n", "#la primera expresión sum() suma a lo largo de cada columna y como lo que \n", "#obtenemos es una pd.Series, al aplicar el segundo sum() vuelve a sumar sobre la columna\n", "#que contiene el número total de missing values por columna representando, ahora sí,\n", "#el número total de missing values del dataset." ] }, { "cell_type": "markdown", "id": "thousand-spyware", "metadata": { "ExecuteTime": { "end_time": "2022-02-13T19:27:16.505889Z", "start_time": "2022-02-13T19:27:16.501890Z" } }, "source": [ "## Empecemos a rellenar missing Values" ] }, { "cell_type": "markdown", "id": "experimental-cotton", "metadata": {}, "source": [ "Analizando un poco el dataset, vemos que se trata de un conjunto de datos donde la columna *Año* solo indica el valor al lado del primer género.\n", "Existe un método habitualmente utilizado y que no se suele mencionar como una estrategia disponible, pero que en este caso es realmente útil: `.ffill()`" ] }, { "cell_type": "code", "execution_count": 16, "id": "amazing-comment", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:35.731461Z", "start_time": "2022-02-16T20:11:35.726459Z" } }, "outputs": [], "source": [ "df[\"Año\"] = df[\"Año\"].ffill() #de manera análoga también existe el método .bfill()" ] }, { "cell_type": "code", "execution_count": 17, "id": "secondary-editing", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:35.750491Z", "start_time": "2022-02-16T20:11:35.733458Z" } }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
AñoGéneroNúmero Ventas
02014.0Aventuras400.0
12014.0Bélico80.0
22014.0Biografías200.0
32014.0Novela Romántica350.0
42014.0Poesía80.0
52015.0Aventuras500.0
62015.0Bélico150.0
72015.0Biografías200.0
\n", "
" ], "text/plain": [ " Año Género Número Ventas\n", "0 2014.0 Aventuras 400.0\n", "1 2014.0 Bélico 80.0\n", "2 2014.0 Biografías 200.0\n", "3 2014.0 Novela Romántica 350.0\n", "4 2014.0 Poesía 80.0\n", "5 2015.0 Aventuras 500.0\n", "6 2015.0 Bélico 150.0\n", "7 2015.0 Biografías 200.0" ] }, "execution_count": 17, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.head(8)" ] }, { "cell_type": "markdown", "id": "frequent-replication", "metadata": {}, "source": [ "Como puedes apreciar, hasta que no encuentra otro valor rellena todas las filas hacia adelante con el primer valor encontrado." ] }, { "cell_type": "code", "execution_count": 18, "id": "unique-court", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:35.763492Z", "start_time": "2022-02-16T20:11:35.753459Z" } }, "outputs": [ { "data": { "text/plain": [ "Año 0\n", "Género 0\n", "Número Ventas 2\n", "dtype: int64" ] }, "execution_count": 18, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.isnull().sum() " ] }, { "cell_type": "markdown", "id": "julian-blanket", "metadata": {}, "source": [ "En estos momentos nos quedan dos Missings. ¿Cómo podemos localizarlos y únicamente ver los registros que corresponden a las observaciones que contienen valores missing en el campo *Número Ventas*?\n", "\n", "Utilizando la máscara *booleana* que se crea con el método `pd.isnull()`" ] }, { "cell_type": "code", "execution_count": 19, "id": "guilty-truck", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:35.781486Z", "start_time": "2022-02-16T20:11:35.766459Z" } }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
AñoGéneroNúmero Ventas
172017.0BiografíasNaN
332020.0Novela RománticaNaN
\n", "
" ], "text/plain": [ " Año Género Número Ventas\n", "17 2017.0 Biografías NaN\n", "33 2020.0 Novela Romántica NaN" ] }, "execution_count": 19, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df[pd.isnull(df[\"Número Ventas\"])]" ] }, { "cell_type": "markdown", "id": "annual-alpha", "metadata": {}, "source": [ "Vemos que, por el motivo que fuera, no hay dato para el género de Biografías en 2017 y tampoco para el género Novela Romántica en 2020.\n", "\n", "En este momento, podríamos ver alguna medida estadística que nos de información sobre la columna *Número Ventas*." ] }, { "cell_type": "code", "execution_count": 20, "id": "reasonable-breakdown", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:35.798459Z", "start_time": "2022-02-16T20:11:35.784461Z" } }, "outputs": [ { "data": { "text/plain": [ "count 38.000000\n", "mean 305.657895\n", "std 237.964826\n", "min 80.000000\n", "25% 150.000000\n", "50% 200.000000\n", "75% 350.000000\n", "max 1000.000000\n", "Name: Número Ventas, dtype: float64" ] }, "execution_count": 20, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df[\"Número Ventas\"].describe()" ] }, { "cell_type": "markdown", "id": "arbitrary-interview", "metadata": {}, "source": [ "Ahora podríamos asumir la media como un valor para rellenar los valores faltantes y ya tendríamos todos nuestro set de datos completo." ] }, { "cell_type": "code", "execution_count": 21, "id": "similar-omaha", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:35.807479Z", "start_time": "2022-02-16T20:11:35.800457Z" } }, "outputs": [], "source": [ "#Voy a hacer una copia del dataset solo para ver cómo aplicamos esa estrategia.\n", "df_estrategia = df.copy()\n", "df_estrategia[\"Número Ventas\"] = df_estrategia[\"Número Ventas\"].fillna(df_estrategia[\"Número Ventas\"].mean())" ] }, { "cell_type": "code", "execution_count": 22, "id": "sexual-active", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:35.819464Z", "start_time": "2022-02-16T20:11:35.810457Z" } }, "outputs": [ { "data": { "text/plain": [ "Año 0\n", "Género 0\n", "Número Ventas 0\n", "dtype: int64" ] }, "execution_count": 22, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_estrategia.isnull().sum() " ] }, { "cell_type": "markdown", "id": "ruled-witch", "metadata": {}, "source": [ "En lugar de la **media**, podríamos haber imputado la **mediana** (que siempre es más conservadora) o **un valor aleatorio escogido entre el mínimo de libros vendidos y el máximo**. En fin, muchas estragias diferentes con las que ya tendríamos un dataset completo y sin valores faltantes. \n", "\n", "No obstante, en este caso, **podemos aprovecharnos del contexto que rodea a los datos** para imputar los valores faltantes, yendo por dos caminos diferentes:\n", "\n", "- Podemos utilizar algún método de imputación teniendo en cuenta el número de ventas pero por Género\n", "- Podemos ver qué sucede cada año y tomar una decisión.\n", "\n", "Voy a coger el dataset original y voy a analizar cuántos libros se han vendido cada año (Sin tener en cuenta el género)" ] }, { "cell_type": "code", "execution_count": 23, "id": "pursuant-suite", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:35.842459Z", "start_time": "2022-02-16T20:11:35.821459Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "El número total de libros vendidos en 2014.0 es: 1110.0\n", "El número total de libros vendidos en 2015.0 es: 1270.0\n", "El número total de libros vendidos en 2016.0 es: 1550.0\n", "El número total de libros vendidos en 2017.0 es: 1350.0\n", "El número total de libros vendidos en 2018.0 es: 1445.0\n", "El número total de libros vendidos en 2019.0 es: 940.0\n", "El número total de libros vendidos en 2020.0 es: 2150.0\n", "El número total de libros vendidos en 2021.0 es: 1800.0\n" ] } ], "source": [ "for anio in df[\"Año\"].unique():\n", " n_ventas = df[df[\"Año\"] == anio][\"Número Ventas\"].sum()\n", " print(f\"El número total de libros vendidos en {anio} es: {n_ventas}\")" ] }, { "cell_type": "markdown", "id": "collect-avatar", "metadata": {}, "source": [ "A través del análisis anterior lo que vemos es que, en 2020 (año de de confinamiento), se vendieron más libros que el resto de años. Pero no obtenemos mucha más información." ] }, { "cell_type": "code", "execution_count": 24, "id": "continental-restriction", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:35.853457Z", "start_time": "2022-02-16T20:11:35.845485Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\t El número mínimo: 350.0\n", "\t El número máximo: 1000.0\n" ] } ], "source": [ "n_min = df[df[\"Año\"] == 2020][\"Número Ventas\"].min()\n", "n_max = df[df[\"Año\"] == 2020][\"Número Ventas\"].max() \n", "print(f\"\\t El número mínimo: {n_min}\")\n", "print(f\"\\t El número máximo: {n_max}\")" ] }, { "cell_type": "markdown", "id": "labeled-cambodia", "metadata": { "ExecuteTime": { "end_time": "2022-02-13T20:16:22.032258Z", "start_time": "2022-02-13T20:16:22.025247Z" } }, "source": [ "Además, si vemos el valor mínimo y el valor máximo de números vendidos para 2020, vemos que hay bastante diferencia.\n", "\n", "Veamos si podemos hacer algo teniendo en cuenta el número de libros vendidos por año y género:" ] }, { "cell_type": "code", "execution_count": 25, "id": "educated-private", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:35.895459Z", "start_time": "2022-02-16T20:11:35.855459Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "El número total de libros vendidos para el género Biografías es:\n", "\tEn 2014.0 : 200.0\n", "\tEn 2015.0 : 200.0\n", "\tEn 2016.0 : 200.0\n", "\tEn 2017.0 : 0.0\n", "\tEn 2018.0 : 190.0\n", "\tEn 2019.0 : 300.0\n", "\tEn 2020.0 : 350.0\n", "\tEn 2021.0 : 250.0\n", "La media de libros de Biografías vendidos por año es: 241.42857142857142\n", "La mediana de libros de Biografías vendidos por año es: 200.0\n", "********************\n", "\n", "El número total de libros vendidos para el género Novela Romántica es:\n", "\tEn 2014.0 : 350.0\n", "\tEn 2015.0 : 300.0\n", "\tEn 2016.0 : 300.0\n", "\tEn 2017.0 : 350.0\n", "\tEn 2018.0 : 100.0\n", "\tEn 2019.0 : 120.0\n", "\tEn 2020.0 : 0.0\n", "\tEn 2021.0 : 200.0\n", "La media de libros de Novela Romántica vendidos por año es: 245.71428571428572\n", "La mediana de libros de Novela Romántica vendidos por año es: 300.0\n", "********************\n", "\n" ] } ], "source": [ "generos = [\"Biografías\", \"Novela Romántica\"]\n", "for genero in generos: \n", " print(f\"El número total de libros vendidos para el género {genero} es:\")\n", " for anio in df[\"Año\"].unique():\n", " n_ventas = df[(df[\"Año\"] == anio) & (df[\"Género\"] == genero)][\"Número Ventas\"].sum()\n", " print(f\"\\tEn {anio} : {n_ventas}\")\n", " print(f\"La media de libros de {genero} vendidos por año es: \" , df[df[\"Género\"] == genero][\"Número Ventas\"].mean());\n", " print(f\"La mediana de libros de {genero} vendidos por año es: \" , df[df[\"Género\"] == genero][\"Número Ventas\"].median());\n", " print(\"*\"*20)\n", " print()" ] }, { "cell_type": "markdown", "id": "crazy-webcam", "metadata": {}, "source": [ "En este caso no encuentro un criterio claro para imputar en ningún caso.\n", "Por buscar algún razonamiento, imputaremos la mediana para los libros de Biografías ya que en 2018 baja el número de ventas.\n", "\n", "🛑 **¡Un momento! Me acaba de llamar José, el responsable de la tienda, diciéndome que antes de perder los datos para el número de libros vendidos del género de Novela Romántica, recuerda que en 2020 se vendieron 310 Novelas Románticas.**\n", "\n", "¡Bueno!¡Pues algo es algo! Imputemos ambos valores." ] }, { "cell_type": "code", "execution_count": 26, "id": "korean-transformation", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:35.905488Z", "start_time": "2022-02-16T20:11:35.898462Z" } }, "outputs": [], "source": [ "condicion = (pd.isnull(df[\"Número Ventas\"])) & (df[\"Género\"]==\"Biografías\")\n", "\n", "df.loc[condicion, \"Número Ventas\" ] = df[df[\"Género\"] == \"Biografías\"][\"Número Ventas\"].median()" ] }, { "cell_type": "code", "execution_count": 27, "id": "portable-karma", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:35.915491Z", "start_time": "2022-02-16T20:11:35.908458Z" } }, "outputs": [], "source": [ "#como ya solo queda un missing utilizaré el método fillna, solo para que\n", "#veas casos donde no me apetece \"comerme la cabeza\" filtrando el dataset\n", "#y aprovecho lo que sé que hace cada método a mi favor.\n", "\n", "df[\"Número Ventas\"] = df[\"Número Ventas\"].fillna(310)" ] }, { "cell_type": "code", "execution_count": 28, "id": "useful-admission", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:35.932479Z", "start_time": "2022-02-16T20:11:35.925485Z" } }, "outputs": [ { "data": { "text/plain": [ "Año 0\n", "Género 0\n", "Número Ventas 0\n", "dtype: int64" ] }, "execution_count": 28, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.isnull().sum()" ] }, { "cell_type": "markdown", "id": "behavioral-judges", "metadata": {}, "source": [ "**¡Reto conseguido!** En este punto ya tendríamos un dataset preparado para el análisis." ] }, { "cell_type": "markdown", "id": "collective-tuition", "metadata": {}, "source": [ "# Ejemplo 2\n", "\n", "> Este dataset contiene información personal de personas ficticias como el nombre, apellidos, edad, estado civil o si se ha vacunado o no." ] }, { "cell_type": "code", "execution_count": 29, "id": "geological-latin", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.011458Z", "start_time": "2022-02-16T20:11:35.936458Z" } }, "outputs": [], "source": [ "#leemos el dataset\n", "\n", "ruta_dataset = \"dataset_notebook_demo.xlsx\"\n", "df2 = pd.read_excel(ruta_dataset, na_values=treat_NaNs, sheet_name=\"ej2\")" ] }, { "cell_type": "code", "execution_count": 30, "id": "incident-exposure", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.020459Z", "start_time": "2022-02-16T20:11:36.013458Z" } }, "outputs": [ { "data": { "text/plain": [ "(69, 12)" ] }, "execution_count": 30, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df2.shape #miramos su forma" ] }, { "cell_type": "markdown", "id": "black-redhead", "metadata": {}, "source": [ "Podemos utilizar la siguiente expresión para simplemente saber si tenemos valores faltantes en cada columna.\n", "Es otro recurso que puedes utilizar. Dependiendo de lo que en el momento te venga a la cabeza o te apetezca aplicar.\n", "\n", "Permíteme anticiparme y comentarte que el método `df.isnull().sum()` será tu gran aliado en la mayoría de ocasiones." ] }, { "cell_type": "code", "execution_count": 31, "id": "aware-boxing", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.033462Z", "start_time": "2022-02-16T20:11:36.022459Z" } }, "outputs": [ { "data": { "text/plain": [ "Nombre True\n", "Apellido 1 True\n", "Apellido 2 True\n", "Sexo True\n", "Municipio True\n", "Provincia True\n", "NIF True\n", "Edad True\n", "Hijos True\n", "Ingresos True\n", "Estado Civil True\n", "Vacunadx True\n", "dtype: bool" ] }, "execution_count": 31, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df2.isna().any() #todas las columnas tienen al menos un NaN. Mala suerte XD!" ] }, { "cell_type": "code", "execution_count": 32, "id": "editorial-productivity", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.052458Z", "start_time": "2022-02-16T20:11:36.036463Z" } }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
percent_missing (%)
Nombre1.45
Apellido 18.70
Apellido 25.80
Sexo10.14
Municipio10.14
Provincia8.70
NIF66.67
Edad5.80
Hijos8.70
Ingresos11.59
Estado Civil7.25
Vacunadx5.80
\n", "
" ], "text/plain": [ " percent_missing (%)\n", "Nombre 1.45\n", "Apellido 1 8.70\n", "Apellido 2 5.80\n", "Sexo 10.14\n", "Municipio 10.14\n", "Provincia 8.70\n", "NIF 66.67\n", "Edad 5.80\n", "Hijos 8.70\n", "Ingresos 11.59\n", "Estado Civil 7.25\n", "Vacunadx 5.80" ] }, "execution_count": 32, "metadata": {}, "output_type": "execute_result" } ], "source": [ "porcentaje_missings(df2) #reutilizamos la función del primer ejemplo para\n", "#sacar el porcentaje de missings por columna" ] }, { "cell_type": "markdown", "id": "instant-connection", "metadata": {}, "source": [ "Observamos que la columna del NIF contiene un alto porcentaje de valores faltantes. \n", "\n", "Dado que no hemos definido ningún objetivo de análisis para este conjunto de datos y puesto que nuestra meta principal es aprender a tratar los valores faltantes, eliminaremos esta columna utilizando el método `.drop()`" ] }, { "cell_type": "code", "execution_count": 33, "id": "featured-mountain", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.060476Z", "start_time": "2022-02-16T20:11:36.054465Z" } }, "outputs": [], "source": [ "df2 = df2.drop(columns=[\"NIF\"])" ] }, { "cell_type": "markdown", "id": "united-jesus", "metadata": { "ExecuteTime": { "end_time": "2022-02-14T19:19:36.650465Z", "start_time": "2022-02-14T19:19:36.642487Z" } }, "source": [ "En estos momentos voy a hacer una copia del dataset `df2` que llamaré `df2Copia` y que utilizaremos más adelante." ] }, { "cell_type": "code", "execution_count": 34, "id": "searching-northern", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.069460Z", "start_time": "2022-02-16T20:11:36.062461Z" } }, "outputs": [], "source": [ "df2Copia = df2.copy()" ] }, { "cell_type": "markdown", "id": "beautiful-disclosure", "metadata": {}, "source": [ "Vamos a construir una función que analice de manera rápida (no muy estética, pero práctica) cuáles son las proporciones de las columnas categóricas:" ] }, { "cell_type": "code", "execution_count": 35, "id": "qualified-flood", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.079486Z", "start_time": "2022-02-16T20:11:36.071460Z" }, "scrolled": true }, "outputs": [], "source": [ "def proporciones(data):\n", " columns = [\"Sexo\", \"Hijos\",\"Estado Civil\", \"Vacunadx\"]\n", " for col in columns:\n", " print(f\"La columna {col} tiene las siguientes proporciones: \")\n", " print(dict(round(data[col].value_counts(normalize=True)*100,2)))" ] }, { "cell_type": "code", "execution_count": 36, "id": "appreciated-demand", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.098458Z", "start_time": "2022-02-16T20:11:36.081457Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "La columna Sexo tiene las siguientes proporciones: \n", "{'H': 50.0, 'M': 50.0}\n", "La columna Hijos tiene las siguientes proporciones: \n", "{'NO': 58.73, 'SÍ': 41.27}\n", "La columna Estado Civil tiene las siguientes proporciones: \n", "{'Solter/x': 65.62, 'Casad/x': 20.31, 'Divorciad/x': 14.06}\n", "La columna Vacunadx tiene las siguientes proporciones: \n", "{'No': 64.62, 'SI': 35.38}\n" ] } ], "source": [ "proporciones(df2)" ] }, { "cell_type": "markdown", "id": "encouraging-sleeping", "metadata": {}, "source": [ "> **Atención**❗ : Más tarde haremos referencia a estas proporciones obtenidas, con el dataset `df2Copia`" ] }, { "cell_type": "code", "execution_count": 37, "id": "abroad-victory", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.354467Z", "start_time": "2022-02-16T20:11:36.100474Z" } }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "msno.matrix(df2,figsize=(8,5), fontsize=12); #en blanco, cada missing value" ] }, { "cell_type": "markdown", "id": "heated-geology", "metadata": {}, "source": [ "Vemos que tenemos algunas filas con valores faltantes.\n", "\n", "Consideraremos que si un registro no contiene sus dos apellidos no es un registro válido." ] }, { "cell_type": "code", "execution_count": 38, "id": "greek-lucas", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.365473Z", "start_time": "2022-02-16T20:11:36.358461Z" } }, "outputs": [], "source": [ "df2 = df2[(pd.notnull(df2[\"Apellido 1\"])) & (pd.notnull(df2[\"Apellido 2\"]))]" ] }, { "cell_type": "code", "execution_count": 39, "id": "aboriginal-organization", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.616459Z", "start_time": "2022-02-16T20:11:36.368459Z" } }, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAf8AAAFpCAYAAACSziHgAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8+yak3AAAACXBIWXMAAAsTAAALEwEAmpwYAAA3K0lEQVR4nO3deZxWdfn/8dc1M7IrLii4ggqCYIJLJKmA+5rhllpumRb0MyVLSC0tNXcLt1SUJNdSE9EUxVRCzcAFcgG+4b6lqRmKggt8fn9cn8OcuR0Q5D7n3MN5Px+P+zFzn/vM3Nd93+c+12c/FkJAREREyqOu6ABEREQkX0r+IiIiJaPkLyIiUjJK/iIiIiWj5C8iIlIySv4iIiIlo+QvIiItnpk1FB1DS6LkLyIiLZqZrQ/81Mw2KTqWlkLJX0REWrqvAIcCh5jZxkUH0xKomURERFokM+sFvBhCuNvMFgKnAPVm9ocQwvMFh1fTVPMXEZGW6izgcIAQwj3ABcAOwBFqAVgy09r+IiLSEplZXQhhYRzsVx9C+NjMhgA/AR4E1AKwGKr5i4hIi2Nm9SGEhfHuecDbZtY2hHA7cCGNLQAbFRVjLVPNX0REWiQzM6AT8D9gArA+0C+EMC+2ABwPPAlcEkJ4qaAwa5Jq/iIi0lJdDNwUQvgU2A14BXgq1QJwOdAb+LC4EGuTav7yOWZmIR4Y6d8LjKcu1bwnIiXU3LnIzLoCfwLOCCHcZWZtgDuA9YCtQwgfmVmHEMLcAkKuaar5SxOxHy2YWWsz6xJ/L+w4Sfr1zKy9mR1sZsPMrLeZrVRUTCKSHzNbGWAxlZD3gWnAlnGf+cA+wAfA5LjPRzmE2eIo+csisWS9wMw6AI8A48ysR0y+uR8rqXhWwfvtvgkci0/v2SHveEQkX2a2LnB+Mm3P3Flmtr+ZrRxCeA+4DTjRzL4GiwoAA4H94v1MWg2LrBRVQ4sOXqor1vIbgCuABcBTwMVmtkkRBYCkBQIYB/w1hHAI8FXgE+CIPGMRkULUAX9OTddrADYEDgEeM7OdgSeAs4EDYotlXQjh4xDCK1kGlhQqzGyimW2R5XNlQSv8SaXOeC37d8CnwA/wAsBxIYR/xWb4BTnG8xW8ae9UgNiHdwleG1jJN4XPcoynxdBYCWmJ0n37IYRXgVfj9jOAh0IIh8S+/V8CP8Ob+NfCKyxtQggfZx1fjC2YWVugAzA/y+fMgpK/NBFCeN3Mbg4hvAaLDvRj8ALAj0MIM+P2leII26zjedzMLsK/4ImPgFWBdiGEOTGevAslNSV5/WbWHj8JfqLELy1NxWDjY/BWvpnAi8AmwEZm1iaEcAfwMzPbEugG/Bbogp8X5mQQ10ohhE8rzzNxSuG7wN4xzhZT6Fazv3xOkvjj71OB0cDrwAVmtraZjcJbBDKVKmFPCiF8Ymb18aF5wEepxH86/uUrpXiyWWBmXwHuA24HppvZHma2agHx2JLuizSnIvHfAnwLqAfeCSG8DYzAKwEHmVnSn/9kCOE2YGugawjh5Qzi6gJcZGabx+9ZnZmNM7NbzOw8oBWwehyblNkYg2rTVD9ZrIov4+Z4C8BQ4GVg0zxq/ouJaxPg+hBCfzO7AR8H0LtMzf+VLS9xYNRjwG+Au4EjgUHApSGE6/KqjZhZQ/I5lL01Rr4cM/sDfn7pn9qWtGx1A07Cm9r/FFsAso5nEPAjvOBxfghhhpntj4892ADYEdgIeBhoB/wHuCGE8OesY1seSv6yRBUFgMfwJuXtQgifFXVyjyeAccA7wNrAFs01ya2ozGwn/ITzC/CahpltD5wQQtg3td+JeAvNNiGEd3KIKzlB1wHXAG3xcSPDY81NalAtNVObX6XvcuDwEMKrFYXJhnje6QacCHQFfhdCuDuHuAYC38VbIk4PITyXemxP4Hp8hsEWePfD3SGEZ7OOa3mo2b/ElqZ5NpX4v4OXtrePX8CGaifaZWgurgP6Aq3xpTw/zSKeGrYmcF08YSel9zbADmbWJ9kphHA+vuzprlkHlOp6qMNHX68B3IzXjG5siaOhy8Aa19HoZmYDio4HWBmvUSddfota81K/fwKcCcwG/pllMGbWKj73ZOAloCdwqpn1TO32GDADeC6EcF0I4fxaT/yg5F9a1riYz5rmi+Z0A5a0cM5koE8q0Va1iX1Z4gkhvIBftWunVEGkNE3+IYQ/hhBmmdl6wOlmtmoI4T58ZbPDzVc9S7yHFwCqzsw2iuMtoLEQ8ivgXyGEvUMIt+InxR2By81sc/X/145Uga0vMAnYPem3LlADsApesCcZ55McN2a2HfDdEMK/gRNDCK9nGUwca9RgZlPxVQNfAzYFTjGzTeNu7+HXFNgxy1iqTck/B6mBaultuZ4E088Xm/IXxH78Kfj62P8AzjWzbZv7+xDCq7GGYNVItF82nuS9DCH8NquCSK2zxvUWtgO2AU4yXw/hRmBj4AozOzGOh1gDuDeDGAzYDHgamqy+Nhm4Ku5zPb7yWit8KtYYfOqm1ID4fd4QvyDOeSGE00II75tPXysqpkeBvwNXm9nqSWtS6vjaDugcx7zk9b0/Fp89c0wIYX/gNLz5/yQz6xnjmAjUfG0/Tck/Y6l+UDOzQWbWz8zWi7XczxUKMtQhFU8ws9WBG4CLQgg7A9/DB6t838x6LO6fpL6EhcRT2bRfpsSfJP1U/+wteELthc95vh8fBzAZ73t8E/haPP6qeqzF4+DuEMIt8dg+Lm6/D5gca2gbAYPjZ3YPXkP6bzXjkOXWC7g3hPA7M1vJzMYCN5vZpQXGdCk+VuR6M1s7NC6mMxwYDlye82DjT/FBzklF5W68wPQNvIKyLnBsFjMNsqTkn6FUjbYOeBRfheoc4A4z2zavPmrzubDPmlnvpCCCl1zfxBMuIYS78EFa6+L9WqWJpyWItZ+FZtbdzI41sx8APUIIfwSuBfoAZwCvhRDOBg4LIfwkq/EQsdCWFLwGAkeb2QXgTaXA6vjYhJVirBsCB4XUNFLJXzMtjp8BPWOy/wfe3D4G+IGZ7ZN3fNFE4DK82+//zOwhM7sDvzzvniGuNZKFxRSSX8VXD/xqqvLzR/yaAi/jZeFMFxbKgpJ/hlIHymhgVgjh63hpcTV8zmpe3sDnf99iZr1iXK3xJtldYVFy+Qfep7VTyeKpOak+TksnfuAZoD9wHL7K4cjgU4quA7oDvzWz9qHpQiTVHp/RkGrN6h1C+Bve8tDPzH4Td/sbfnnVR/DVGU8Nvg67FCTVyraGmfUys7ViS80ofCnvi0IIhwS/FO74jGNpdnBvcqyHEMYBewLDgJuA3wMDQwhPZhhT+rgeaGbbmlm74NMJRwF3mtkuZrYa3jK5EPh1COGNrGLKklb4y4A1TklJptC0w2ux4KXq/wL7m9nawKch+2lY/8Gby84GxpvZN4Iv1XsmMMLM3gshTIj7tsEX9ClTPLVobeCNWDAKZrY+vpDJSSGE35rZWsAewMFm9iLeBbAGPigps6uYxWP6s9ia9RAwzcxOxZtB64BhZnZ+COFEM9sL2Ap4WTX+YqVaIfviLUWfAJ+Y2bP4VMyPLF4p08xuwlfTy2QKnTWdvrcS0DrES+7GQm5SAPiU2BKYtYrj+hG8W/JjoM7MBoUQRprZh/i4mmeAHsA3Qgj/ySO+TIQQdKviDahLfuKl1e2AP+NXpLsEL2G3ivucgi/GYhnG0xB/9gROwGvdT+G1RMNrZe/iV8a6F5ie/E0Z4qnFGzAAr8l3xLtD6vFBUAvxEc7Jfqvgy5qOTR1zydoddRnGZ3jT7B/Sz4UXcg/AW3UuL/p91O1zn9taeOI6Oh5be+DrZUzEC9l98ILBI8BK8W/qqxxD+vx4J74a5fP4hbq6LOHvMjtHpp8DeAC4Nt7vgK8l8hSwetzWCx/Z37noz3N5b2r2r6JYul4Ym7B+iX+BHsYHO43D+6s2Dz595Dh8tbyHQjyqshC8NLsJPuagHrgaXyf7L8BGIYTT8UtfPozXHreOf5PJYMRai6dGzQVOC758cZvgTfgH4TMhFi3iE0J4Hz+2NjSzVYLXlkJyHGYY3/p4q+EJ6Y0hhI/w6YbXAOuaWecMY5Bltwq+7v3t8di6Bx8g+gmwa/C56UnzetXGiphfKZSK43ISfsGun+Oze04Avmtm7ZoZl0CW58iUXsC/8fMy+LiDF/CxSJPNbI0QwqzgM5/eyiGebBVd+lgRb8AFeOm5X2rbj/HlIS8AxuL9oVtk8NxDgAHx96QWeDK+FGayz/p40n0G6N7M/6haab/W4mlJN3ylsGuBveP9dYDn8BaRnvgKejfirSSZ14ySzxAfm/EasGbclrRkrY0vswywctHvX9lvyfcm9bM3fvGZ5DOqwwvgE0i1KCWPVSmGLniLZ6/U8bMu3jrUkNpvCD6wblDe70/Fto3jzzPxq5s2AAfjrW7/rNb7Ugs31fyrzMzWwS/v2I/UQLUQwm/xC1U8izctDQohTKvi8yal5d3wC98Q4lGMN6O3N7OV4/ZX8UE0PYBHremiMITqlPZrKp6WoqKFI7lC3+FmtmvwgUUD8cLSDPwEBXBoCF7jzyCeJuOCgnsS+D98ShbBR/eDt06cYn7VtQ+QwljjFOM+wK/MrHMIYQY+Qn18qqVoAd7SNC/996F6LUd98HEoPzWzHvEcYPj1OPaNsVrwQYaTSLVsZcmaTsEeYGbrmFnrEMLz5uscbAacEXxswnrA4cBeVXxfCqcBf8vJKhaZCSG8YWbn4k1pR5rZGyGEm+JjExb3f6qgD/BMCGFYjGsjvF/qUby22Bsff5DE8De81jgTr8Wt6PHUvNQJaRO80DQabyn6PjDUzAghTDSzXfD+0m1CCNvGv20bQpi32H/+5eJJD4Iai7dcvRtCOBVfW/1cM5uEN/P3wBdDGRhCaHHXNl+RVAzu+yu+6NJ8gBDCt81sPDDTzO7DB4l2Aw7JIpYQwv2xTHo4XjA8M4TwnJldCexqZi+GEB6Pu78PZN6cnnp/6vDpjavhffv3mNklIYT/mtn7+Bojg4GjgK+GFWzQqmr+X4KZbWJm+8KiPuw6M/uNmV1pZj/C5zefiddmjzezgzKOZx1gopmdk9r8E+AG8/UE7scTyQ1m9mPzC1GMxpvkzg9VXgSm1uKpVemaeuqEtBnet98Ob1Z/Fk+ur+Jzr3cJvqTpN4A144mcDBJ/sua74f34a+OFsu+a2dhY+/8uPs95L7w1YrsQwlPVjCPGovPUMoitQKsCVwKnhBBODiHMMbP+5ivSfRMYiQ9kexjvfqz6uBprXBf/fnxQ36b4qnid8e6sdsAvzOw8MzsZbxm9vZoxNBPToguV4QOu/4nPbLiJxnX72+LfuWfwiwdtF0KYlWVchSi636El3vCT3kLggHj/ceBWvG/rBrxprS++rOlIYBawf0axtIs/98ebYn+VemxMjGW7eP8ovH/4AXyAXTKit2r9xbUWTy3fqBjdjI8ufgj4aTP79sXnGk8C+sdt6+A1lpurFQ+wRsW2E4ErUvc3wmdo3FCxXybjMvALAz0ArFX059WSbvjgvnvw2Red8QLlQ/hFl45vZv+sPr+V4nNfFb/jT+CDCteMn+338Na/S4HNc3hfknFH1+NTGfdMXj9wKD4z6xzimJXknLQi3goPoKXe8EulfoYP5BuV2r4+cC4+fWY1fHDL8UC3DGLYFrgQ2DDeH4KPTv11ap+x+HS5JOG2wqf1JF+Cqk2jq7V4avmG93neG0/SyYCslfFm2mTQUQPeP5q8N33i8Vaf+j9rJ/svZzxt8ULswNS2vePJ+iXiVKe4vUfcdk9O79VMYCrQqejPLcZTc4VTmhmIhg+snRGPsyvitguA3+QY14n4jKbk/jfxqz3+HlgnbmtoLv4qx1Ffcf/XNM42aJt8rsC347n7bFJTZ1fEW+EBtLRb+iDFp4QsjAmuc2r7VvFEulW8n0lCA76Dl+5/DXSN25pLuNfEeHYmVZKt9oFda/FUxNZszSbrk84S4tmYWCCksbWkPT7t8fiKfVeOx1qrL3o9yxlTl+RziIWBBnwQ32TgPKB9at9eMbGsm8dnhq9zMJWCWwBoLKh1Ygnz0guKqQe+eNbP8Aso1cfjrFdq3xuA0TnGdjxwTXJcxZ9H4dd5uAVfojqvWAz4Rer+ycC/8NaRdql9DszyuK6VW+EBtKRb6ktmqW1H4AWAwyr2fRjYJ4eYvoWvIXDOFyTc8cCYssVT8bnV4cuFHgN8K/V4YdN38IVXJgB7xPtD8Zrawal9ro3vaS6FI3zd8qvxQkc9vhDVbfEzbZvaL9Mm0YqYBuI1tYnEKYYFfFbJAjWb4/3B+wOrFHXsVMTWBx8sdy3e/fgBMCw+1hALAbfjs40y+dxofurct/CZBJultrXBC5SjiLX/nN6j3vgU6+tS234ZCwD7kSrcluFWeAAt5VaRQHbF11dPVqv7ET4l60RgEN4l8HaS/DKKJ90CcUj8Ylcm3Ofw6Sqf+5sVPZ7m4sPHG9wJ3IW3PNyRejzX5j0aa0E98P7Ov+Er+7XDVzmcga8VcTfwGDmMh8BrPR2Br+HjNc7HxyEkBYBbYmJpm1UMi/ncnovv0SXx9ycorgDQE1+e+6Qlfa55Hkfx8/k9qbEiwP+AX8bfO+DjlG4ku5X7GlLxDMAvNZ1suxhfons7vED5fbyAm2krTuX5BS8EbRO/+zektv8CHzuTeWWtlm6FB9CSbqkEMgsvud5C4wIn/w9vAZiJN3tXfQGfdBzpn/H3A/BR2emE+03gQ+D7lX+7IsdTEVuSZH8KjIu/t8anNz1DXBo3x2MoKUSmm/C74n2xf8enFIGPQB6Kj+pP/ibT8RDxuP1F/H3bmGgvoLEAMBRfdji3pne8K2lixbZ/4F0AuRYAYmI7idiCFZPJGXi/8Y8KOo46xZ930jgAeRpxES08Ea9F06Wfq3oc0XTJ3r8DT+ML5DyKr9tPPA+8jQ9YfSPL82Mzn9m3gFVTn1lSAEi3AIykCmNnWtKt8ABq/UbTpsf9iP1lwPZ47XYiTQsA7wKrZR0PXgP5ET6lsHc8yPfAm9PPpjHhbk9GI3lrMZ7mPrd4fyQ+0Ch9EtwWr11vlNOxlDzvZjGWP+DjHlrhA/fOj/Hs+EWvJ6P4huI1oOSzGoAXAM6lsQugYx7vVSqmQ/FWiNVS29bBL170ODkPAsRb9f6D116n4S0zo+Pn1jvn46gP3oz9NbzgdjKedK9N7Xsr8JPKv80iJjyxJ+vir4a3PkwlNqfj4xC6AWtn/P6kKyEH4l2ww4ldNHgB4Bt418iNeR4/tXTT/NklSC26Umdme+NToZJLSk7Bm4s+xlfMahNCuAxfnva9rGKK8fTGD+iueCI5HTg3+CJCN+K1xpFm1iWE8FDIcN58rcVTEZfF9Rfq8abatfBkEeJuL9A4oj5T1ng51bXxJv7/w0+Qx+HjEN4DfoNPx7rCzLasfD1Vjudz3/0QwhX4dKxDzaxV8AWZDsEvBDMCWBh8TfhMLOaYeBLvk909zr8GT77X411tK2cdj5m1NbPWACGEK4Hf4cf5dSGEPfE1LOaRw3EUYwhmtiZeeL0ghDAF7xr6Ad6ydkKM+3o82V6U/tvlee6KtSnSi8Rtgq+Lf0y8PwqYjb8vD5lZhxDC0yGEl0II/16eGL4gvoYQ16cws61DCLfg0/cGAseY2arBF2VL1jhYx8zWzSqemlZ06aNWbzRtypqGN+d/gJdue6Qe64N3AYxL/12GcRk+Xe5X8X49fjI8PbXPYXgzW15XwqqZeJKY4s8tgemp7Y/ic8a3xEfWH4Ev8pFL8zHev3808LPUthPx1pHj8YFQ6wE/JJ+avsXn2j21bRhwX8V+W9HMNReqHEu6z3gvfIxIx7jtDHycxk/wJPNDvMad2QCt1Pf/K/E7PwGv3Q8kNmWn9r0mPpbXZ9YeX3DpNXzJ2eSxg/Cutil4IW4y2fXxb5/8X3x6XDsap6ieFc+ZDfG7vxAvnGT6/afpOfsxfIxIMoNlBD5o9RR8eu0R+MDW0l6DovAAav2G18wui7/vGxPd1cCmcVs9vnLV+jnGdAdxPjbe9HlT/L0HsEnFvnkUAGoiHj4/wKcL3r84ILXtrnhynIIX6LbM+L1Jzwz5VTwRjqPpRU1GxG0/J/vpfOl4tsW7Ht7AB2UlF2B6hNTAzIzfn7qKk/aTeMvIO3jtfuf42PF4v/azeK3tq1m+N/H+BvgV3UbiBaCr8D7tA2KsW+AFt2k0JtlMCv/QtL8e2Ce+R+dX7LcuPg2zZ+p9rXYf/6B4HB+Jt2AlawgYXoC9g8aLUZ2IFw42yON4is95D3B95eeBd0veg7dIvJH1d7/Wb4UHUMu3eEKcDpya2rYXPp3mKlLTVzKMobnFOy7Fm7KmAVeltv8JOLZM8TQTi+HXuD8yniAnEFfES+3TI564c5mnTaplAe8qmgdsW7HPGWTcOrK4JIDXbv8YT4zj8ULK3Vm+P1S0tsTPbQKNA9U64IWzSalEYngzdtXH1AC7AD+Pv9eltv2pYr8z8QFt7fBum4NpTMhZreeRxNMJb8LvFu/vjV9DZGTlvou7X8WYjgY+BR5OPxc+fuVavKD2G7y1tGdWx1Ezca0Wj9214/16mhYA1sILvblV1mr1VngAtXSrPPECu+O1oJtpulDGHnhN7ZJ4sGc1iCYZTNctnoi+Ge/3wwcaPkfjYMOxeK07sxHhtRZPc58bPp1odDz5PIbXUJ7GW2vOwJuMV83xmOoQj5V0oSg5KW7f3OvI4niiae36Zryw9jhxvYOYzDbEl2B+DR8jkUl3CD7b4kVSi83grWfn0FiDHo3Pxrg2fge/neXxgzcDb1qxfd/4PmxYsX1GcuyntmW1PG7yndsovj9v4deZ7xq3fwMfdzQiq/enudeIz8KYjI+7OKTiGNsfHyT6F6Bv1nGlYqoDuuPrQXw9bksKZquR4+WCW8Kt8ABq5UbTZtg2NE5R2QYfhHVO+uQQk19mC1SkEkFvvHnvfry5ago+2GhvvK/xZXxE7wNk1L9Xi/E097lVxhtvF8REcwh+8Y4J5LuqWHu83/Ne4Lep7Rfio6F3au59zuqYion0Ovxy0yPxgYbHVey3FRmvcIb3nb9LaqlZ4vQvfOGVp+Lv2+BjSC4hg5kGwNZAm9T99fF+4dZ4wf6P+ADW9Aqef6WZGRkZvld98Hnyp+CtaQ/ihdlu8fG98ULuoRnGkCTROrxVLTkfHEa87HRq3554v3pui0BVbL8Mn4bdLbXtOErex/+596noAGrhBk0W8PkL3vT5fDzI2+JTaR7Cp9Pk0dSffLFWx/s9f5B6bDw+zzmZtrINPkgsk/69WoxnMZ/bdXit9S68WT8peOxDnM+LNwFmUatubmWzDjSuGd4Bn2t8P00LAGOoGFyXZWx4zX5yxeMH4jXHbbKMo+I5k89mG7wFJL3yYz1e698n3h+Gdyt1rnIM9fiUwX+SqtnHZPYgcFo8rg7AB4rdgY9kvx5fZCjzwX2pmC4Hrkzd3w8vYI8mFtLwFq+suh3S37Mn4vuxaeq8cDReADgKL7jNJINWo+a+uzGmn+MLY30bv5DQFvF88AhewD0bL2j2y+szawm3wgOolVs8GUzG177eOB5M9xBHrQM74AONTiM1KCvDeFaPX+7ngcEVjz2enAxo2uSd5Qp+NRVP+jnwucS3AIPxcRof0rhYziC8NSLr1cQ6xBNOd3yVvFvwptFkzfCV8UT7OE3HkGRa048/W+GDnbrgNcjd0p9PjPXkrD+r+FxJ4m/A56Xfitdaz0ntcxc+IOu8+Fn2qnIMzSWRLniLX6uYxO6IScXwlpBz4/t0GTm0aFXEdjFwZsW2E/G1Dq4gtZIoVSgALO6YBO6j6cI4nWgs9B+JV5AeBrbO+P1Ir88/He/jvx0vlFyMn8vXxRdk+jM+TqJPHp9VS7oVHkChL95PPofF39eKB3e71OMH47Xar8X7g6jykr1LOoHEL/gsfF7xqqntPwIuzeg9qal4FhNHHxovuTkAuCv12Pl4IS1JbF3xGvfqGcf0m5jEkvUOLotJ7AAaFzlZEx+H8CYwPG4zqlRIomnBK0n8dXhXx214v/4YvAVrk9S+1ybxZPC+pGNqk3rNj8TnHYgvwPIBTa+OeRk+cDOTy7zSODDtmHj/CeLSxXjt8Wh83MipNCb79EyMTGvZFdt+GN+f9Ge2Nl45uTqex6pSEKn4vE6icVZTUrhNCtW/w1tJX6Dx/Lgm0CGL9yUV0zrxHHQkPlNlfOqxe/C1Bn5PY7dtbi00Le1WeACFvXBYAy/NP4wPUGmP1zL2rNjvDuDyjGNpj68OuAoVrQrxxHgnPkp8vVRM15Qlnornvxbv99wy3h+El/6Tk/lT8fcG4pKr5DCPH+/nnIAPlvsnXgA4B+/r35fGLoAr8S6AqraKVJy0e+E12ZXxgtmo1GN74YXc0XjN9gR8EFnVR2RXxHQV8L34e2882aYvErQ9MJ+4Hn3cluVYkc54IeiR+Nyb49PWzqGxAHAM3q11Jo0DWTO/PgbepH4KPusiKTBdhrfa7IhPQbwKL8gdj09/XO4LDFV8XncBk1L3V8Nr13fiNe2pePfaOODWrN6TZmJsHV/75XhXR9e4/Zr43f8WPsbnT6TGcujWzHtZdAC5vdDmm/o2xEvNf8dH8P84Hsz9UvtcQOqCGVnEFU/EC/FR2Jfx+dHFP8VbIF7FE9z9VCw0sqLGU/G8t+FNixsTaxjxBHQnnmT/mdr3eHxRn6rX+CuTUnzPVo1JYnA8Ec3CCwBnx9iujyfUR2nsQ82itnZ9PAlOx2tCC/GBhenBqtvhrTj34rMy+mZxHKV+Hw88nrq/Vjx+9ktta4UXnhaS3xoDyeC+Kfj4g83wbq3KAsBteAtAluNX0i1V7+ODU6fhiWxjvCn7LOAlvHvyseQYwmdFbLqcz5/+vG4nNYUvtX1jvKK0TyreE/ACbR6zepLvTS+8D//4eP/o5LsfP7NJeCvFCn9Z3uV6P4sOIPcXDF+puL8+3o//N7yZ8Yx4gv4FPnjlbTJesxvYE/gMb8a6DpiLt0rsk9rniBjXr2m8mEcmYw9qLZ74vw8GplRsax9vV+KDjA4F+uKtE++Q4TQjvBn0G6RGoOOtJdPiCehGvEbWNcZ+Nl6oymwxGLzg+hA+4HI43kd9Ej5a/koqCsB4oSXrEdnjKhMJ3lf8pxhT99T2n+NL1GYyLxxPlCtVbLP4Wc3ECwCb4lNWz4rHVhv8mgd3ksGgyBhT0qqwbnzeoanHb8bX709WFd0Ab7VMku9P8KmHa1QpnpuBJyu2dcULi61T29rEbf/N8nsWn2uN1O/J6z4fuCj+Ppw4kBYvCNxOhtdXWVFuhQeQ64v1ZrKF+CCQ62Ki6BxP5D/Fa0qnxpP1X+LJOpM+x4q4WsXYzsWbq38QT4zz8ZHOu9B4LfpxeAtFZk3ZtRZPjGkY8SIc+IC6r+O168fxkv7T8cT5AJ70+mYYSx3etL8QLyCNorF59hzitCcaV6TrXvH3WczIuBpviVkpte0hvCC3Hl6bvDz1WB4rPyYxpWccdMdbRnbFC4834OMlTgXmELuSMoilHm+B+A5Np/btE9+ba/AWgG3xAsBcYFhqv4eJ44OqfBzdHmPqiFc45lMx+JLGAkC69aZr3H8uVVqpDm+ReRMfIJcM5FsfT/AjUvuthreYPkHGV+eLx8V44OCK7QfhFbPueKH/lXgueh+N6l+697boAHJ9sf7FnoePkL0b76v+N43TQX4eT0jbFhDbSXiTY5JEzsP7Yh/GR6tfGLf/GE9wP8ryBF6D8WwfT3QX4QOzXsO7ZPbDW2j+hjdLtieHvr54LL0Tj5u/xGPp13gz+mWp/SYBt+UQzyC8z3qPeL9zvD8k3t8kxntd1rEsIab14ud2dry/Nl7ovgpvCeibcTy74LXk5NK3ySI+O+CJ+BS8O2JHPLkmzcyd8ZaBLMZF7IIXEL+JtzxcHZ+rT8V+fyVePyTeXwtPgNWeCbFZjOdcoD9eeKssjHTEB49WdfrlEuJJCoY30nQ1w6uAMfH3A/FC1CZZx7Si3AoPIPcX7P2dT+F9/GvhJf9kyc5ZeG1uSkwiuVwYJ/X703gXxPF4gu0ZT0o9aTrS+P+RXQ2ppuKpiO1gvKZ/bnICj9t3wcdtVKXpcxni2R5vAdgFL4QMxfsi59J03EjmUx7j8wyKiWNY+qSdSmKbxs8xl2WNK2IaWhFT0ny7Iz6IK69pc4PwAsD5eOLfOfXY+nhh7urUtuS965hhTINjTN/EpxVegheoN6/Yr3Lp3qxWFv1K/MzmA+dVPPbDGFu7LJ57CTH1xAv5T8Tv3DC8UHATOV2Ya0W7FR5AIS/aS/qzgANT21rTOBAq1zmhNE7LOgGvnf2L1MVoUvtlvr5ALcbzRc+Jtzo8RAH9fPFYeoY4eA0fjLQNFQsK5VgAGBgLH39MPzeNYw0y7eNf2pji9uPw5tpc11mP3/OPaGZgIV4haG7KZNZXpEsXAPrjXUn3UzFnPsfjqEc8rs+m8cp4w/AaeKZN/UuIqR4fT3MW3t32Hl5ZO7WIeFr6rfAACnvhXgOYTZzmV3Q8MaYN8Kb1c+P9XC6B21LiaSa+9WICeY8C+/niiTs5lgp/j/AuiRl4s3a7iscKiS/GNDNVSPoBXvMu5MpqeKvNDHwdhs91ExXxPqUKAPsAX8W7kC4p8Dj6Cl5JOgPvEvlvUYm/8jPBZ9fsi8+gyXRA9op6KzyAQl+8FwCexdd9z3yq2lLGdALetJ15M3pLjCcV16Z4n9/0IhN/Kp6kefs7FFC7Xkw8ybGdewvNEmJ6Cp8e+m5Rib9WP7MY02C8u+2geIznUtNfQjyb4astzi0y8afiqZyxUhPHdku8JU1apWVmO+NNWzuGED6ogXh64td3PzrUwIdTa/EkzKyBeJGhEMIbRccDYGa7AkeFEA4uOhaovWMbwMx2wAdu7R5C+GcNxFNTnxmAme2Gzxj5TrxfF0JYWGA8GwEhhPBiUTEsjplZLZ2XWpLSJ38AM2sXQvio6DgSyQFd9Je+VuOpZbV2Mqq1YxvAzNqEEOYXHUei1j4zqM2YZMWi5C9SZTpxtzy1+JnVYkyy4lDyFxERKZm6ogMQERGRfCn5i4iIlExmyd/MDjCzS8zsITN738yCmV2f1fOJiIjI0mnI8H//HL9wzlx8Pe9eGT6XiIiILKUsm/1/jF9MZBV8WUgRERGpAZnV/EMIDya/m1lWTyMiIiLLKMtm/6oYPHhwzcxFHDVqFADDhw8vNI60Woup1uIBxbQ0ai0eqL2Yai0eUEzLYtKkSctbC61aLnr//fc5+eSTmTFjBscddxxDhgyp3CXzGnPNJ38REZEVxVtvvcXIkSN54403OO200xg0aFAhcSj5i4iI5OCFF15g5MiRzJs3j/PPP5++ffsWFouSv2Ri0qRJRYfwObUW0/Tp04sOQURy8sorr3DcccfRtm1bLr74YjbaaKNC49EiPyIiIhn75JNP+PTTT2nTpg1t27YtOhwlfxERkax1796dCy+8kDlz5nDssccye/bsQuNR8hcREcnBZpttxiWXXEJDQwPDhw/niSeeKCwWJX8REZGcdO3alUsvvZTOnTvzs5/9jAceeKCQODIb8GdmQ4Ah8W6X+HOAmY2Nv78TQvhpVs8vxRo8eHDRISxSi/OOk5hEpHzWXHNNLr74Yo488kiuuOIKdtxxx9xjyHK0fz/giIptG8UbwMuAkr+IiJTOCy+8wLvvvsv3vve9Qp4/s2b/EMIvQwi2hFu3rJ5bRESkVi1cuJDLL7+cTp06ccABBxQSg/r8RUREcvTggw8ya9Ysjj76aNq0aVNIDFrkRzKhBXVERJp344030q1bN3bZZZfCYlDNX0REJEfrrLMO//nPf3j//fcLi0HJX0REJEfHHHMM8+fPZ+zYsYXFoOQvIiKSow022IB99tmHO++8k1deeaWQGNTnL5moxXn+IiK14ogjjmDixImMHj2aM888M/fnV81fREQkZ6ussgqdO3fm9ddfL+T5lfxFRERydt999/Hiiy/y7W9/u5DnV/IXERHJ0fz58xkzZgw9e/Zkp512KiQGJX8REZEc3Xrrrbz99tsMGzaMurpi0rAG/IkUpF+/fjW3GJKIZG/ChAn06tWLvn37FhaDav4iIiI5GjBgAP/617948cUXC4tByV9ERCRHhx12GO3atePKK68sLAYlfxERkRx17NiRww47jClTpvD4448XEoP6/EUKMn36dIYPH150GIAWQhLJ25AhQxg3bhyjR49m6623zv35VfMXERHJWatWrRg8eDAvvfRSIc+v5C8iIlKAoqb5gZK/iIhI6Sj5i4iIlIySv4iISMko+YuIiJSMkr+IiEjJKPmLiIiUjJK/iIhIAebOnYuZFfLcSv4iIiI5e/vtt7n33nvZbrvtCnl+JX8REZGcjRkzhoULF3LMMccU8vxK/iIiIjmaPXs2EydOZL/99qNLly6FxKAL+4iILINJkyYVHYK0cFdddRUrr7wyhx56aGExqOYvIiKSozlz5tCxY0fatGlTWAxK/iIiIjk6/PDDefXVV7nrrrsKi0HJX0REJEdf//rX6du3L2PHjuXDDz8sJAb1+YuILIPBgwcXHcIio0aNKjoE+RLMjGHDhjF06FBuvPHGQkb8q+YvIiKSs549ezJo0CDGjx9fyPMr+YuIiBRg3XXX5ZNPPinkuZX8RURESkbJX0REpGSU/EVERApQVJM/KPmLiIjkbu7cuUycOJHNN9+8kOdX8hcREcnZ9ddfzwcffMDQoUMLeX4lfxERkRz9+9//5rbbbmO33Xaje/fuhcSgRX5ERJaBLuwjy+vqq6+mrq6Oo446qrAYVPMXERHJ0fPPP896661Hp06dCotByV9ERCRHBx54IM8//zyTJ08uLAYlfxERkRztvvvubLjhhowePbqw6X7q8xcRWQa6sI8sr/r6eoYOHcrIkSMZP348Bx54YO4xqOYvIiKSs/79+7PVVltx0003FfL8Sv4iIiIF6NmzJ3Pnzi3kuZX8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFREQKEEIo7LkbCntmkRz169ePSZMmFR3G59RaTLUWD9ReTLUWDyimLJjZScBZwGUhhGOr8T/nzZvH008/zZNPPsm0adN47rnnaNu2bTX+9TJT8hcREUkxs22A7wNPLc//+fjjj3n22WeZPn06Tz75JLNmzWLBggU0NDTQu3dvDj/8cLbddtvqBL2MlPxFREQiM+sI3AAcBZzW3D4LFixg6tSpzJ49mx49etC/f3/q6+v57LPPmDlzJtOnT2fatGk888wzfPrpp9TV1dGzZ08OOuggttxyS/r06UObNm1yfV2VlPxFREQajQZuDSE8aGafS/4LFixgxIgRzJw5k/nz59OqVSs6derEOuusw9NPP838+fMB6N69O0OGDGGLLbZg8803p3379nm/jiVS8hcREQHM7BigO3Do4vaZOnUqM2fOZN68eYA37b/++uvMmzeP3XffnS222IK+ffvSsWPHnKL+cpT8pRSmT5/O8OHDiw5jkVGjRgHUTEy1Fg/UXky1Fg8opmXxRQMQzawnPsBvuxDCp4vbb/bs2Ytq92lz5szh5ZdfZo011mD11Venffv2NDTUboqt3chERETyMwDoBDxrZsm2emCgmQ2dP38+rVu3pkePHrRp02ZRzR+gVatWbL311rz11luMGTMGgDZt2tC3b1+22GIL+vXrR/fu3amvr8/5JS2ekr+IiAjcDjxese0aYDZwVqtWrZ4G6N+/P5tuuikzZszg448/pnXr1vTu3ZvTTz+d+vp65syZs2jA37Rp05gyZQoAHTp0oF+/fvTr148tt9ySbt26kSpk5E7JX0RESi+E8D/gf+ltZvYh8N8QwjPJtvr6es477zymTp3Kc889R/fu3ReN9gfo2LEjgwYNYtCgQQC88847TQoDDz/8MACrrbYa/fr1Y4cddmD77bfP5TWmKfmLiIgsg/r6egYMGMCAAQO+cN9OnTqx8847s/POOwPw5ptvLioITJ06lUcffZQJEyZkHfLnaHlfERGRZoQQBldrdb9Ely5d2GOPPTj55JPZa6+9WLBgQTX//VJT8hcRESkZJX8REZGSUfIXEREpmWVO/ma2k5mNM7M3zexjM3vDzO41sz1T+6xvZr8zsykV+z1kZt81s5Wq+zJERERkaS1T8jez84C/AlsDdwAXAncBawKDU7tuDHwHmIPPnbwQuBPoCvweuNfMNNNARESkAEudgOOaxycCfwC+H0L4pOLxdG3+78BqIYSFzewzEdgB2A+4+UvGLSIiIl/SUtX8zaw18GvgFZpJ/ADptZBDCJ9UJv7UPrfHuz2+TMAiIiJlYGYDzewOM3vdzIKZHVnx+H6x2/3t+Pjgpf3fS9vsvwvetH8bsNDM9jKzkWZ2vJl98SoHjYHWA8nYgKeW9u9ERERKqAPwDHA8MK+Zx9vjLe0nLOs/Xtpm/6/Gn/OBacBm6QfNbDJwQAjh7YrtnYBjAcMLD7vgl0u8MYRw57IGKyIisiJYsGABr7/+OgsWLODRRx9tskRwIoRwN3A3gJmNrfwfIYTr4mOdlvX5l7bmv1b8eSIQgO2BlYHN8T78gcAtzfxdJ+A04FRgGD4Q8ALgyGUNVEREZEWwYMECRowYwSOPPMLChQs544wzGDFiRK6r/S1t8k/2+wzYJ4TwcAhhbgjhaWBf4DVgUGUXQAhhVgjB8BaGrsCPge8Dk81s9aq8AhERkRZk6tSpzJw5k88++wyAefPmMWPGDKZOnZpbDBZC+OKdzM4FRgD/CCF8ro/fzK4GvgcMDyFc9AX/62DgJuCyaq+ZLCIiUut22GGHXwC/pGkFfCFw2oMPPnhmc39jZnOBY0MIY5t5rBPwNrBDCGHS0sSwtH3+/xd//m8xj78Xf7Zdiv+VXL5o8FI+t4iIyArjwQcfPAM4o8gYlrbZ/368r7+3mTX3N8kAwBeX4n+tG39+tpTPLSIiIlW0VMk/hPAyvkLfBviUg0XMbFdgN7xV4J64bcs4rY+KfTsASbfAXV86ahERkRWcmXUws35m1g/P1xvE+xvEx1ePjyUV8O7x8S5f+L+Xps8/Psl6+HzC9fGWgGnAhsAQvFXg4BDCn+O+twPbxv1fAT6Kf7cHsGrcvlsIYe5SPbmIiEjJxEV7HmzmoT+EEI6Mi/5c08zjvwoh/HKJ/3tpk38MZE182t4+wNrA+8BDwNkhhKmp/fYCDgH6A52Bdvi4gKfwJX1/H0JQs7+IiEgBlin5i4iISMu3zJf0FRERkZZNyV9ERKRklPxFRERKRslfRESkZJT8RURESkbJX0REpGSU/EVEREpGyV9ERKRklPxFRERKRslfRESkZP4/miUOGPjH8AgAAAAASUVORK5CYII=\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "msno.matrix(df2,figsize=(8,5), fontsize=12); #en blanco, cada missing value" ] }, { "cell_type": "markdown", "id": "respected-programmer", "metadata": {}, "source": [ "Filtremos nuestros datos para ver qué pinta tienen las filas que al menos contienen un valor faltante:" ] }, { "cell_type": "code", "execution_count": 40, "id": "unauthorized-landscape", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.642459Z", "start_time": "2022-02-16T20:11:36.619459Z" } }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
NombreApellido 1Apellido 2SexoMunicipioProvinciaEdadHijosIngresosEstado CivilVacunadx
1Ana BelénCarrascoDimitrovaNaNNaNZaragoza22.0NO1000.0Solter/xSI
3MaríaGallegoMartínezNaNSevillaSevilla45.0NO1500.0Solter/xNo
5ManuelGarcíaMuñozHMadridNaN59.0SÍ1825.0Divorciad/xNo
10JuanSanzHernándezHRenteríaGuipúzcoa22.0SÍNaNSolter/xSI
13José LuisRomeroBustoHSan SebastiánGuipúzcoa45.0SÍNaNSolter/xNo
62María CarmenRomanoAbellánMNaNNaNNaNNaNNaNNaNNaN
\n", "
" ], "text/plain": [ " Nombre Apellido 1 Apellido 2 Sexo Municipio Provincia Edad \\\n", "1 Ana Belén Carrasco Dimitrova NaN NaN Zaragoza 22.0 \n", "3 María Gallego Martínez NaN Sevilla Sevilla 45.0 \n", "5 Manuel García Muñoz H Madrid NaN 59.0 \n", "10 Juan Sanz Hernández H Rentería Guipúzcoa 22.0 \n", "13 José Luis Romero Busto H San Sebastián Guipúzcoa 45.0 \n", "62 María Carmen Romano Abellán M NaN NaN NaN \n", "\n", " Hijos Ingresos Estado Civil Vacunadx \n", "1 NO 1000.0 Solter/x SI \n", "3 NO 1500.0 Solter/x No \n", "5 SÍ 1825.0 Divorciad/x No \n", "10 SÍ NaN Solter/x SI \n", "13 SÍ NaN Solter/x No \n", "62 NaN NaN NaN NaN " ] }, "execution_count": 40, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df2[df2.isnull().any(axis=\"columns\")]" ] }, { "cell_type": "markdown", "id": "other-application", "metadata": {}, "source": [ "Dado que la columna Sexo contiene 2 NaN:" ] }, { "cell_type": "code", "execution_count": 41, "id": "ceramic-virus", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.654456Z", "start_time": "2022-02-16T20:11:36.645459Z" } }, "outputs": [ { "data": { "text/plain": [ "2" ] }, "execution_count": 41, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df2[\"Sexo\"].isnull().sum()" ] }, { "cell_type": "markdown", "id": "instant-escape", "metadata": {}, "source": [ "Y analizando los datos asumimos que Ana Belén y María son de género Femenino, imputaremos la categoría \"M\" en la columna Sexo sin necesidad de hacerlo seleccionando los índices. Para ello, utilizaremos en esta ocasión el método `.fillna()`" ] }, { "cell_type": "code", "execution_count": 42, "id": "beautiful-editor", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.663459Z", "start_time": "2022-02-16T20:11:36.657460Z" } }, "outputs": [], "source": [ "df2[\"Sexo\"] = df2[\"Sexo\"].fillna(\"M\")" ] }, { "cell_type": "markdown", "id": "alpha-preliminary", "metadata": {}, "source": [ "Otro registro que podemos eliminar por completo es el que tiene índice 62 por contar con demasiados Missing Values. Asumiremos ese riesgo! 😋\n", "\n", "La siguiente operación funcionará porque pandas no pierde el índice hasta que no lo reseteemos nosotrxs." ] }, { "cell_type": "code", "execution_count": 43, "id": "personalized-emerald", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.673458Z", "start_time": "2022-02-16T20:11:36.666472Z" } }, "outputs": [], "source": [ "df2 = df2.drop(62)" ] }, { "cell_type": "markdown", "id": "oriented-olympus", "metadata": {}, "source": [ "Para poder disponer del registro 1 (Ana Belén Carrasco) imputaremos Zaragoza al municipio.\n", "\n", "De forma análoga, para disponer del registro 5 (Manuel García) imputaremos Madrid a la provincia (Madrid = CC.AA. uniprovincial)" ] }, { "cell_type": "code", "execution_count": 44, "id": "legislative-malta", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.684466Z", "start_time": "2022-02-16T20:11:36.677463Z" } }, "outputs": [], "source": [ "df2.loc[1,\"Municipio\"] = \"Zaragoza\"\n", "df2.loc[5,\"Provincia\"] = \"Madrid\"" ] }, { "cell_type": "markdown", "id": "posted-gazette", "metadata": {}, "source": [ "Nos queda trabajar con los valores faltantes correspondientes a ingresos:" ] }, { "cell_type": "code", "execution_count": 45, "id": "uniform-bobby", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.712457Z", "start_time": "2022-02-16T20:11:36.686458Z" } }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
NombreApellido 1Apellido 2SexoMunicipioProvinciaEdadHijosIngresosEstado CivilVacunadx
10JuanSanzHernándezHRenteríaGuipúzcoa22.0SÍNaNSolter/xSI
13José LuisRomeroBustoHSan SebastiánGuipúzcoa45.0SÍNaNSolter/xNo
\n", "
" ], "text/plain": [ " Nombre Apellido 1 Apellido 2 Sexo Municipio Provincia Edad \\\n", "10 Juan Sanz Hernández H Rentería Guipúzcoa 22.0 \n", "13 José Luis Romero Busto H San Sebastián Guipúzcoa 45.0 \n", "\n", " Hijos Ingresos Estado Civil Vacunadx \n", "10 SÍ NaN Solter/x SI \n", "13 SÍ NaN Solter/x No " ] }, "execution_count": 45, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df2[df2.isnull().any(axis=\"columns\")]" ] }, { "cell_type": "markdown", "id": "focused-brass", "metadata": {}, "source": [ "En este caso imputaremos el valor más habitual de Ingresos para las personas de 22 años." ] }, { "cell_type": "code", "execution_count": 46, "id": "timely-privacy", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.725463Z", "start_time": "2022-02-16T20:11:36.715459Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "El valor más frecuente de ingresos para una persona de 22 años es: 1000.0\n" ] } ], "source": [ "#extremos la moda de los ingresos de todos los registros donde la persona tenga 22 años\n", "val = df2[df2[\"Edad\"]==22][\"Ingresos\"].mode()[0]\n", "print(f\"El valor más frecuente de ingresos para una persona de 22 años es: {val}\")\n", "\n", "#imputamos ese valor al registro 10.\n", "df2.loc[10,\"Ingresos\"] = val" ] }, { "cell_type": "markdown", "id": "civil-journal", "metadata": {}, "source": [ "Finalmente, para la persona de 45 años imputaremos la mediana en su rango de edad." ] }, { "cell_type": "code", "execution_count": 47, "id": "secondary-visibility", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.738456Z", "start_time": "2022-02-16T20:11:36.729461Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "El valor más frecuente de ingresos para una persona de 45 años es: 1500.0\n" ] } ], "source": [ "#extremos la mediana de los ingresos de todos los registros donde la persona tenga 45 años\n", "val = df2[df2[\"Edad\"]==45][\"Ingresos\"].median()\n", "print(f\"El valor más frecuente de ingresos para una persona de 45 años es: {val}\")\n", "\n", "#imputamos ese valor al registro 13.\n", "df2.loc[13,\"Ingresos\"] = val" ] }, { "cell_type": "markdown", "id": "appointed-middle", "metadata": {}, "source": [ "En estos momentos ya dispondríamos de un dataset apto para el análisis y modelado." ] }, { "cell_type": "code", "execution_count": 48, "id": "weighted-reggae", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.750459Z", "start_time": "2022-02-16T20:11:36.741460Z" } }, "outputs": [ { "data": { "text/plain": [ "Nombre 0\n", "Apellido 1 0\n", "Apellido 2 0\n", "Sexo 0\n", "Municipio 0\n", "Provincia 0\n", "Edad 0\n", "Hijos 0\n", "Ingresos 0\n", "Estado Civil 0\n", "Vacunadx 0\n", "dtype: int64" ] }, "execution_count": 48, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df2.isnull().sum()" ] }, { "cell_type": "markdown", "id": "minor-talent", "metadata": {}, "source": [ "Recuperando la copia del dataset `df2` llamada `df2Copia` como dataset de referencia ahora, también podríamos haber optado por eliminar todas las filas que contengan algún valor nulo con el método `.dropna(axis=1)`\n", "- `df.dropna(axis=\"index\")` elimina todas las filas que tengan como mínimo un missing value\n", "\n", "Para las columnas:\n", "- `df.dropna(axis=\"columns\")` elimina todas las columnas que tengan como mínimo un missing value\n" ] }, { "cell_type": "code", "execution_count": 49, "id": "anonymous-cooling", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.767459Z", "start_time": "2022-02-16T20:11:36.754466Z" } }, "outputs": [ { "data": { "text/plain": [ "Nombre 0\n", "Apellido 1 0\n", "Apellido 2 0\n", "Sexo 0\n", "Municipio 0\n", "Provincia 0\n", "Edad 0\n", "Hijos 0\n", "Ingresos 0\n", "Estado Civil 0\n", "Vacunadx 0\n", "dtype: int64" ] }, "execution_count": 49, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df2Copia = df2Copia.dropna(axis=\"index\")\n", "df2Copia.isnull().sum()" ] }, { "cell_type": "markdown", "id": "fiscal-criminal", "metadata": {}, "source": [ "Podemos analizar si para las columnas categóricas, el hecho de haber eliminado las filas sin ton ni son afecta a las proporciones de las mismas:" ] }, { "cell_type": "code", "execution_count": 50, "id": "satellite-albany", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.784465Z", "start_time": "2022-02-16T20:11:36.770469Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "La columna Sexo tiene las siguientes proporciones: \n", "{'M': 52.63, 'H': 47.37}\n", "La columna Hijos tiene las siguientes proporciones: \n", "{'NO': 59.65, 'SÍ': 40.35}\n", "La columna Estado Civil tiene las siguientes proporciones: \n", "{'Solter/x': 64.91, 'Casad/x': 22.81, 'Divorciad/x': 12.28}\n", "La columna Vacunadx tiene las siguientes proporciones: \n", "{'No': 63.16, 'SI': 36.84}\n" ] } ], "source": [ "proporciones(df2Copia)" ] }, { "cell_type": "markdown", "id": "invisible-oregon", "metadata": {}, "source": [ "Si comparamos con las proporciones originales, vemos que en ocasiones podemos sacrificar unos cuantos registros para ganar tiempo, sin que ello perjudique gravemente la muestra de datos con la que contamos para el análisis/entrenamiento del modelo." ] }, { "cell_type": "markdown", "id": "concrete-vessel", "metadata": {}, "source": [ "
Bonus!
" ] }, { "cell_type": "markdown", "id": "entitled-extra", "metadata": {}, "source": [ "También me gustaría dejarte por aquí algunas otras píldoras que pueden servirte de gran ayuda.\n", "\n", "\n", "
1- Puedes hacer que pandas interprete los valores infinitos `np.inf` como `np.nan`
" ] }, { "cell_type": "code", "execution_count": 51, "id": "absent-generic", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.790460Z", "start_time": "2022-02-16T20:11:36.786465Z" } }, "outputs": [], "source": [ "pd.set_option('use_inf_as_na', True)" ] }, { "cell_type": "code", "execution_count": 52, "id": "trying-miracle", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.815459Z", "start_time": "2022-02-16T20:11:36.793458Z" }, "scrolled": true }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
AñoGéneroNúmero Ventas
02014.0NaN400.0
12014.0Bélico80.0
\n", "
" ], "text/plain": [ " Año Género Número Ventas\n", "0 2014.0 NaN 400.0\n", "1 2014.0 Bélico 80.0" ] }, "execution_count": 52, "metadata": {}, "output_type": "execute_result" } ], "source": [ "import numpy as np\n", "#coloco a propósito un valor infinito en la primera celda del campo Género.\n", "\n", "df.iloc[0,1] = np.inf \n", "df.head(2) #como puedes ver lo interpreta como NaN" ] }, { "cell_type": "markdown", "id": "unsigned-rider", "metadata": {}, "source": [ "\n", "
2- También puedes utilizar la función .replace() para reemplazar un valor cualquiera (también un missing value) por el valor que desees.
" ] }, { "cell_type": "code", "execution_count": 53, "id": "straight-content", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.822464Z", "start_time": "2022-02-16T20:11:36.817458Z" } }, "outputs": [], "source": [ "df[\"Género\"] = df[\"Género\"].replace(np.nan,\"Aventuras\") #vuelvo a restaurar el missing" ] }, { "cell_type": "code", "execution_count": 54, "id": "conditional-breach", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.839457Z", "start_time": "2022-02-16T20:11:36.825460Z" } }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
AñoGéneroNúmero Ventas
02014.0Aventuras400.0
12014.0Bélico80.0
\n", "
" ], "text/plain": [ " Año Género Número Ventas\n", "0 2014.0 Aventuras 400.0\n", "1 2014.0 Bélico 80.0" ] }, "execution_count": 54, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.head(2)" ] }, { "cell_type": "markdown", "id": "consecutive-liability", "metadata": {}, "source": [ "\n", "
3- Pandas cuenta con la opción .interpolate() para rellenar los missings que hay entre dos puntos utilizando diferentes estrategias:\n", "
    \n", "
  • Lineal
    • \n", "
  • Cúbica
    • \n", "
  • Cuadrática
    • \n", "
  • etc.
    • \n", "\n", "
\n", "
\n", "\n", "Te dejo por aquí la docu de pandas\n" ] }, { "cell_type": "code", "execution_count": 55, "id": "committed-syndicate", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:36.922489Z", "start_time": "2022-02-16T20:11:36.842458Z" } }, "outputs": [], "source": [ "ruta_dataset = \"dataset_notebook_demo.xlsx\"\n", "df3 = pd.read_excel(ruta_dataset, na_values=treat_NaNs, sheet_name=\"ejinterpolacion\")" ] }, { "cell_type": "code", "execution_count": 56, "id": "hollywood-basin", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:37.428462Z", "start_time": "2022-02-16T20:11:36.925462Z" } }, "outputs": [ { "data": { "text/plain": [ "" ] }, "execution_count": 56, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "df3.index = df3.fecha #colocamos la fecha como índice y eliminamos dicha columna\n", "df3[[\"d1\"]].plot(marker='o', figsize=(10,5), linewidth=2, title=\"La señal original. Con sus Missing y todo.\")\n", "df3[[\"d1\",\"d2\"]].plot(marker='o', figsize=(10,5), linewidth=2, title=\"La señal perfecta deseada.\")" ] }, { "cell_type": "markdown", "id": "stopped-account", "metadata": { "ExecuteTime": { "end_time": "2022-02-15T20:04:09.894929Z", "start_time": "2022-02-15T20:04:09.886935Z" } }, "source": [ "¿Pero, y si interpolamos linealmente...**nos perdemos mucho**?" ] }, { "cell_type": "code", "execution_count": 57, "id": "searching-patio", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:37.650484Z", "start_time": "2022-02-16T20:11:37.430460Z" } }, "outputs": [ { "data": { "text/plain": [ "" ] }, "execution_count": 57, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "df3[\"Señal interpolada\"] = df3[\"d1\"].interpolate(\"linear\")\n", "df3[[\"d2\",\"Señal interpolada\"]].plot(marker='o', figsize=(10,5), linewidth=2,title=\"Señal interpolada vs. Señal esperada.\")" ] }, { "cell_type": "markdown", "id": "damaged-tampa", "metadata": {}, "source": [ "Como te comentaba en anteriores ocasiones, dependerá del caso. Pero a veces las estrategias más simples son suficientes como para permitirnos continuar con nuestra labor." ] }, { "cell_type": "markdown", "id": "green-mentor", "metadata": {}, "source": [ "\n", "
4- utilizar las componentes de estacionalidad y tendencia en una serie temporal con Missing Values para interpolar y reemplazar los valores faltantes.\n", "
\n", "\n", "> 🔵 Puedes volver aquí cuando quieras, si aún no tienes los conocimientos sobre Time Series." ] }, { "cell_type": "markdown", "id": "demanding-sailing", "metadata": {}, "source": [ "🔵 El conjunto de datos AirPassengers proporciona el recuento mensual de los pasajeros de una aerolínea estadounidense, desde 1949 hasta 1960." ] }, { "cell_type": "code", "execution_count": 58, "id": "irish-verse", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:37.710494Z", "start_time": "2022-02-16T20:11:37.653462Z" } }, "outputs": [], "source": [ "ruta_dataset = \"dataset_notebook_demo.xlsx\"\n", "df4 = pd.read_excel(ruta_dataset, na_values=treat_NaNs, sheet_name=\"ejAirPassengers\")" ] }, { "cell_type": "code", "execution_count": 59, "id": "proved-contrast", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:37.780502Z", "start_time": "2022-02-16T20:11:37.713474Z" } }, "outputs": [], "source": [ "from statsmodels.tsa.seasonal import seasonal_decompose\n", "\n", "df4[\"Month\"] = pd.to_datetime(df4[\"Month\"]) #convertimos a tipo fecha la variable Month.\n", "df4.set_index('Month', inplace=True) #la colocamos como índice del dataset.\n", "index = df4[pd.isnull(df4[\"#Passengers\"])].index #creamos una máscara con los índices donde #Passengers es null.\n", "#para poder detectar la estacionalidad de la serie y la tendencia tomamos la decisión\n", "#de interpolar linealmente la serie temporal.\n", "df4['#Passengers1']= df4['#Passengers'].interpolate()" ] }, { "cell_type": "code", "execution_count": 60, "id": "ethical-reach", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:38.237461Z", "start_time": "2022-02-16T20:11:37.782458Z" } }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "analysis = df4[['#Passengers1']].copy()\n", "\n", "#descomponemos la serie en tendencia, estacionalidad y residuo\n", "decompose_result_mult = seasonal_decompose(analysis, model=\"multiplicative\")\n", "\n", "trend = decompose_result_mult.trend\n", "seasonal = decompose_result_mult.seasonal\n", "residual = decompose_result_mult.resid\n", "\n", "decompose_result_mult.plot();" ] }, { "cell_type": "code", "execution_count": 61, "id": "valid-notification", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:38.248463Z", "start_time": "2022-02-16T20:11:38.240462Z" } }, "outputs": [], "source": [ "#dado que tenemos delante una serie temporal con tendencia ascendente\n", "#y una estacionalidad marcada, podemos interpolar de la siguiente manera:\n", "df4.loc[index, \"#Passengers\"] = trend.loc[index]*seasonal.loc[index]" ] }, { "cell_type": "code", "execution_count": 62, "id": "sought-fashion", "metadata": { "ExecuteTime": { "end_time": "2022-02-16T20:11:38.471495Z", "start_time": "2022-02-16T20:11:38.251456Z" } }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "df4[\"#Passengers\"].plot(figsize = (10,6)); #Señal ajustada\n", "df4[\"#Passengers1\"].plot(figsize = (10,6)); #Señal interpolada linealmente" ] }, { "cell_type": "markdown", "id": "smaller-flush", "metadata": {}, "source": [ "Nada mal, ¿Verdad?😊" ] }, { "cell_type": "markdown", "id": "protecting-dallas", "metadata": {}, "source": [ "\n", "
5- No te ciñas a pandas y explora otros métodos de imputación pertenecientes a otras librerías como scikit-learn, fancyimpute, miceforest,etc.)
\n", " \n", "Te dejo un par de enlaces que considero que explican bastante bien la imputación múltiple iterativa e imputación KNN:\n", "- Imputación KNN\n", "- Imputación Iterativa\n" ] }, { "cell_type": "markdown", "id": "clinical-watts", "metadata": {}, "source": [ "
\n", "¡Hasta la próxima pequeñx gran Egger! 🐣 Ya estás un pasito más cerca de lograr tus metas!\n", "
" ] } ], "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": "3.7.7" }, "toc": { "base_numbering": 1, "nav_menu": {}, "number_sections": true, "sideBar": true, "skip_h1_title": false, "title_cell": "Table of Contents", "title_sidebar": "Contents", "toc_cell": false, "toc_position": {}, "toc_section_display": true, "toc_window_display": false }, "varInspector": { "cols": { "lenName": 16, "lenType": 16, "lenVar": 40 }, "kernels_config": { "python": { "delete_cmd_postfix": "", "delete_cmd_prefix": "del ", "library": "var_list.py", "varRefreshCmd": "print(var_dic_list())" }, "r": { "delete_cmd_postfix": ") ", "delete_cmd_prefix": "rm(", "library": "var_list.r", "varRefreshCmd": "cat(var_dic_list()) " } }, "types_to_exclude": [ "module", "function", "builtin_function_or_method", "instance", "_Feature" ], "window_display": false } }, "nbformat": 4, "nbformat_minor": 5 }