pandas-tutorial.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Pandas\n",
    "\n",
    "Libreria de código abierto y extensión de NumPy para manipulación a análisis de datos en Python.\n",
    "Permite de forma rápida la limpieza, preparación y análisis de los datos."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Instalación en Jupyter\n",
    "\n",
    "* Con anaconda:\n",
    "```bash\n",
    "conda install pandas\n",
    "```\n",
    "* Sin anaconda\n",
    "```bash\n",
    "pip install pandas\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Series"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np # a partir de ahora utilizaremos np (seudonimo) en vez de numpy \n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "a    1\n",
      "b    2\n",
      "c    3\n",
      "dtype: int64\n",
      "0    1\n",
      "1    3\n",
      "2    5\n",
      "3    7\n",
      "4    9\n",
      "dtype: int64\n",
      "venta_1    1\n",
      "venta_2    4\n",
      "venta_3    7\n",
      "dtype: int64\n",
      "4\n",
      "venta_1     2\n",
      "venta_2     8\n",
      "venta_3    14\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "etiquetas = ['a','b','c']\n",
    "datos =[1,2,3]\n",
    "\n",
    "print(pd.Series(data=datos, index=etiquetas))\n",
    "\n",
    "# si creamos una serie de datos con arrange, el index es por defecto: 0,1,2,3,4...\n",
    "datos = np.arange(1,10,2)\n",
    "print(pd.Series(datos))\n",
    "\n",
    "# Para acceder a los datos\n",
    "data = np.arange(1,10,3)\n",
    "labels = ['venta_1','venta_2','venta_3']\n",
    "serie = pd.Series(data,labels)\n",
    "\n",
    "# imprimir serie \n",
    "print(serie)\n",
    "# accedemos a dato 1\n",
    "print(serie['venta_2'])\n",
    "\n",
    "# Podemos realizar operaciones sobre estas series. Por ejemplo, vamos a sumarlas:\n",
    "serie_suma = serie + serie\n",
    "print(serie_suma)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Dataframes:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "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>columna_1</th>\n",
       "      <th>columna_2</th>\n",
       "      <th>columna_3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>fila_1</th>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fila_2</th>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fila_3</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        columna_1  columna_2  columna_3\n",
       "fila_1          3          3          8\n",
       "fila_2          3          4          1\n",
       "fila_3          4          1          1"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "# Los dataframes nos permite crear tablas bonitas\n",
    "filas = ['fila_1', 'fila_2', 'fila_3']\n",
    "columnas = ['columna_1','columna_2','columna_3']\n",
    "datos = np.random.randint(1,10,9)\n",
    "matriz = datos.reshape(3,3) # convertimos el array en una matriz 3x3\n",
    "dataframe = pd.DataFrame(matriz, filas, columnas) # creamos la tabla\n",
    "dataframe"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dataframes: seleccionar datos de una tabla"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2, 3)"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "# Los dataframes nos permite crear tablas bonitas\n",
    "filas = ['fila_1', 'fila_2', 'fila_3']\n",
    "columnas = ['columna_1','columna_2','columna_3']\n",
    "datos = np.random.randint(1,10,9)\n",
    "matriz = datos.reshape(3,3) # convertimos el array en una matriz 3x3\n",
    "dataframe = pd.DataFrame(matriz, filas, columnas) # creamos la tabla\n",
    "\n",
    "# Seleccionamos primera fila\n",
    "dataframe.loc['fila_1']\n",
    "\n",
    "# Si queremos seleecionar varias filas tendriamos que hacer una lista de valores\n",
    "dataframe.loc[['fila_1', 'fila_2']]\n",
    "\n",
    "# Si queremos seleccionar una columna\n",
    "dataframe['columna_1']\n",
    "\n",
    "# Si queremos selecionar un elemento de la table\n",
    "dataframe.loc['fila_1']['columna_1']\n",
    "\n",
    "# Añadimos una nueva columna con la suma de las otras 3\n",
    "dataframe['suma'] = dataframe['columna_1'] + dataframe['columna_2'] + dataframe['columna_3']\n",
    "\n",
    "# Eliminar una columna en su representación (aunque no la elimina del dataframe)\n",
    "dataframe.drop('suma', axis=1)\n",
    "# Eliminar una columna del dataframe\n",
    "dataframe.drop('suma', axis=1, inplace=True)\n",
    "\n",
    "# Eliminar una fila\n",
    "dataframe.drop('fila_3', inplace=True)\n",
    "\n",
    "# Conocer número de filas y columnas\n",
    "dataframe.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dataframes: selección de datos por condición"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "          infectados  fallecidos  curados\n",
      "Madrid         20000        1000     1500\n",
      "Cataluña       15000         800     1300\n"
     ]
    },
    {
     "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>fallecidos</th>\n",
       "      <th>curados</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Madrid</th>\n",
       "      <td>1000</td>\n",
       "      <td>1500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Cataluña</th>\n",
       "      <td>800</td>\n",
       "      <td>1300</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          fallecidos  curados\n",
       "Madrid          1000     1500\n",
       "Cataluña         800     1300"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "filas = ['Madrid', 'Cataluña', 'Andalucía']\n",
    "columnas = ['infectados', 'fallecidos', 'curados']\n",
    "datos = [[20000, 1000, 1500], [15000, 800, 1300], [10000, 300, 600]]\n",
    "tabla = pd.DataFrame(datos, filas, columnas)\n",
    "\n",
    "# selección por condición: Los que no la cumplan, valdrán Null\n",
    "condicion = tabla >1000\n",
    "tabla[condicion]\n",
    "\n",
    "# selección por condición en una columna\n",
    "condicion = tabla['infectados'] > 12000 # se pueden concatenar codiciones con & o | (and, or)\n",
    "print(tabla[condicion]) # eliminaria la fila que no cumpla con la condición de la columna\n",
    "\n",
    "# selección por condicion de varias columnas\n",
    "condicion = tabla['infectados'] > 12000 # se pueden concatenar codiciones con & o | (and, or)\n",
    "tabla[condicion]['fallecidos'] # aplica la condicion a una columna\n",
    "tabla[condicion][ ['fallecidos', 'curados'] ] # aplica la condicion a un grupo de columnas"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dataframes: modificación de filas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>infectados</th>\n",
       "      <th>fallecidos</th>\n",
       "      <th>curados</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fake_news</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>20000</td>\n",
       "      <td>1000</td>\n",
       "      <td>1500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>15000</td>\n",
       "      <td>800</td>\n",
       "      <td>1300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>10000</td>\n",
       "      <td>300</td>\n",
       "      <td>600</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           infectados  fallecidos  curados\n",
       "fake_news                                 \n",
       "30              20000        1000     1500\n",
       "40              15000         800     1300\n",
       "70              10000         300      600"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "filas = 'Madrid Cataluña Andalucía'.split()\n",
    "columnas = 'infectados fallecidos curados'.split()\n",
    "datos = [[20000, 1000, 1500], [15000, 800, 1300], [10000, 300, 600]]\n",
    "tabla = pd.DataFrame(datos, filas, columnas)\n",
    "\n",
    "# Añadimos una nueva columna\n",
    "tabla['fake_news'] = '30 40 70'.split()\n",
    "\n",
    "# Seleccionamos una columna como índice\n",
    "tabla.set_index('fake_news')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Tratamientos de datos nulos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "    Columna_1  Columna_2  Columna_3\n",
      "0       23.0         67        NaN\n",
      "1       45.0         23       73.0\n",
      "2        NaN         89        NaN\n",
      "\n",
      "    Columna_1  Columna_2  Columna_3\n",
      "1       45.0         23       73.0\n",
      "\n",
      "    Columna_2\n",
      "0         67\n",
      "1         23\n",
      "2         89\n",
      "\n",
      "    Columna_1  Columna_2  Columna_3\n",
      "0       23.0         67      100.0\n",
      "1       45.0         23       73.0\n",
      "2      100.0         89      100.0\n",
      "\n",
      " Columna_1    34.000000\n",
      "Columna_2    59.666667\n",
      "Columna_3    73.000000\n",
      "dtype: float64\n",
      "\n",
      "    Columna_1  Columna_2  Columna_3\n",
      "0       23.0         67       73.0\n",
      "1       45.0         23       73.0\n",
      "2       34.0         89       73.0\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "# creamos una tabla a partir de un diccionario\n",
    "diccionario = {'Columna_1': [23,45,np.nan], 'Columna_2': [67,23,89], 'Columna_3': [np.nan,73,np.nan]}\n",
    "tabla = pd.DataFrame(diccionario)\n",
    "print(\"\\n\",tabla)\n",
    "\n",
    "# vamos a eliminar las filas que contengan valores nulos (solo lo muestra)\n",
    "print(\"\\n\",tabla.dropna())\n",
    "\n",
    "# vamos a eliminar las columnas que contengan valores nulos (solo lo muestra)\n",
    "print(\"\\n\",tabla.dropna(axis=1)) # si quieremos aplicarlo sobre la tabla ponemos el parámetro: inplace=True\n",
    "\n",
    "# ahora vamos a rellenar los valores nulos con un valor\n",
    "print(\"\\n\", tabla.fillna(value=100))\n",
    "\n",
    "# ahora vamos a obtener los valores medios de cada columna\n",
    "print(\"\\n\", tabla.mean())\n",
    "\n",
    "# vamos a rellenar los valores nulos con los valores medios\n",
    "print(\"\\n\", tabla.fillna(value=tabla.mean()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Agrupación por columnas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   dias vendedores  ventas\n",
      "0  dia1      Paula       2\n",
      "1  dia1      Maria       5\n",
      "2  dia2     Josefa       5\n",
      "3  dia3     Pepito       8\n",
      "4  dia3     Manolo       3\n",
      "\n",
      "       ventas\n",
      "dias        \n",
      "dia1     3.5\n",
      "dia2     5.0\n",
      "dia3     5.5\n",
      "\n",
      "       ventas\n",
      "dias        \n",
      "dia1       7\n",
      "dia2       5\n",
      "dia3      11\n",
      "\n",
      "      ventas                                          \n",
      "      count mean       std  min   25%  50%   75%  max\n",
      "dias                                                 \n",
      "dia1    2.0  3.5  2.121320  2.0  2.75  3.5  4.25  5.0\n",
      "dia2    1.0  5.0       NaN  5.0  5.00  5.0  5.00  5.0\n",
      "dia3    2.0  5.5  3.535534  3.0  4.25  5.5  6.75  8.0\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "# Creamos un diccionario a partir del cual se creará la tabla\n",
    "diccionario = {'dias':['dia1','dia1','dia2','dia3','dia3'], 'vendedores':['Paula','Maria','Josefa','Pepito','Manolo'], 'ventas':[2,5,5,8,3]}\n",
    "tabla = pd.DataFrame(diccionario)\n",
    "print(tabla)\n",
    "\n",
    "# Operamos con los datos de aquellas filas que se repiten, agrupandolos de esta manera\n",
    "agrupacion = tabla.groupby('dias') \n",
    "\n",
    "print(\"\\n\",agrupacion.mean())     # media de los valores que se han agrupado\n",
    "print(\"\\n\",agrupacion.sum())      # suma de los valores que se han agrupado\n",
    "print(\"\\n\",agrupacion.describe()) # datos estadísticos de los valores que se han agrupado (std: desviación típica )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Combinar dataframes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    A   B   C\n",
      "0   1   4   7\n",
      "1   2   5   8\n",
      "2   3   6   9\n",
      "0  14  64  27\n",
      "1  45  51  18\n",
      "2  23  64  69\n",
      "   A  B  C   A   B   C\n",
      "0  1  4  7  14  64  27\n",
      "1  2  5  8  45  51  18\n",
      "2  3  6  9  23  64  69\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "diccionario1 = {'A':[1,2,3], 'B':[4,5,6], 'C':[7,8,9]}\n",
    "diccionario2 = {'A':[14,45,23], 'B':[64,51,64], 'C':[27,18,69]}\n",
    "tabla1 = pd.DataFrame(diccionario1)\n",
    "tabla2 = pd.DataFrame(diccionario2)\n",
    "\n",
    "# Vamos a añadir las filas de la tabla 2 a la tabla 1\n",
    "tabla = pd.concat([tabla1, tabla2])\n",
    "print(tabla)\n",
    "\n",
    "# vamos a añadir las columnas de la tabla 2 a la tabla 1\n",
    "tabla = pd.concat([tabla1, tabla2], axis=1)\n",
    "print(tabla)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Merge en Dataframes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>clave</th>\n",
       "      <th>D</th>\n",
       "      <th>E</th>\n",
       "      <th>F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>7</td>\n",
       "      <td>c1</td>\n",
       "      <td>14</td>\n",
       "      <td>64</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>8</td>\n",
       "      <td>c2</td>\n",
       "      <td>45</td>\n",
       "      <td>51</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>9</td>\n",
       "      <td>c3</td>\n",
       "      <td>23</td>\n",
       "      <td>64</td>\n",
       "      <td>69</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   A  B  C clave   D   E   F\n",
       "0  1  4  7    c1  14  64  27\n",
       "1  2  5  8    c2  45  51  18\n",
       "2  3  6  9    c3  23  64  69"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "diccionario1 = {'A':[1,2,3], 'B':[4,5,6], 'C':[7,8,9], 'clave':['c1','c2','c3']}\n",
    "diccionario2 = {'D':[14,45,23], 'E':[64,51,64], 'F':[27,18,69],'clave':['c1','c2','c3']}\n",
    "\n",
    "tabla1 = pd.DataFrame(diccionario1)\n",
    "tabla2 = pd.DataFrame(diccionario2)\n",
    "\n",
    "# vamos a realizar un merge de la tabla a través de la clave\n",
    "pd.merge(tabla1, tabla2, on='clave')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Join en Dataframes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>E</th>\n",
       "      <th>F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>i1</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>7</td>\n",
       "      <td>14</td>\n",
       "      <td>64</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>i2</th>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>8</td>\n",
       "      <td>45</td>\n",
       "      <td>51</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>i3</th>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>9</td>\n",
       "      <td>23</td>\n",
       "      <td>64</td>\n",
       "      <td>69</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    A  B  C   D   E   F\n",
       "i1  1  4  7  14  64  27\n",
       "i2  2  5  8  45  51  18\n",
       "i3  3  6  9  23  64  69"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "diccionario1 = {'A':[1,2,3], 'B':[4,5,6], 'C':[7,8,9]}\n",
    "diccionario2 = {'D':[14,45,23], 'E':[64,51,64], 'F':[27,18,69]}\n",
    "\n",
    "# en este caso vamos a añadir un index a cada tabla\n",
    "tabla1 = pd.DataFrame(diccionario1, index=['i1','i2','i3'])\n",
    "tabla2 = pd.DataFrame(diccionario2, index=['i1','i2','i3'])\n",
    "\n",
    "# junta las tablas por indice\n",
    "tabla1.join(tabla2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Operaciones con Dataframes "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    A  B   C\n",
      "i1  1  4   7\n",
      "i2  2  5   8\n",
      "i3  3  6   9\n",
      "i4  2  5  10\n",
      "i5  3  2  11\n",
      "i6  4  1  12\n",
      "\n",
      "Elementos no repetidos de la columna A:\n",
      "[1 2 3 4]\n",
      "\n",
      "Numero de valores únicos que hay:\n",
      "4\n",
      "\n",
      "Veces que se repite un valor de la columna A:\n",
      "3    2\n",
      "2    2\n",
      "4    1\n",
      "1    1\n",
      "Name: A, dtype: int64\n",
      "\n",
      "Columna A duplicada por 2:\n",
      "i1    2\n",
      "i2    4\n",
      "i3    6\n",
      "i4    4\n",
      "i5    6\n",
      "i6    8\n",
      "Name: A, dtype: int64\n",
      "\n",
      "Borrar columna B:\n",
      "    A  B\n",
      "i1  1  4\n",
      "i2  2  5\n",
      "i3  3  6\n",
      "i4  2  5\n",
      "i5  3  2\n",
      "i6  4  1\n",
      "\n",
      "Borrar fila i2:\n",
      "    A  B\n",
      "i1  1  4\n",
      "i3  3  6\n",
      "i4  2  5\n",
      "i5  3  2\n",
      "i6  4  1\n",
      "\n",
      "Ordenar los elementos de la tabla atendiendo a la columna B (de menor a mayor):\n",
      "    A  B\n",
      "i6  4  1\n",
      "i5  3  2\n",
      "i1  1  4\n",
      "i4  2  5\n",
      "i3  3  6\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "diccionario = {'A':[1,2,3,2,3,4], 'B':[4,5,6,5,2,1], 'C':[7,8,9,10,11,12]}\n",
    "\n",
    "# en este caso vamos a añadir un index a cada tabla\n",
    "tabla = pd.DataFrame(diccionario, index=['i1','i2','i3','i4','i5','i6'])\n",
    "\n",
    "print(tabla)\n",
    "\n",
    "# vamos a quedarnos solo con los elementos no repetidos de la columna A\n",
    "print(\"\\nElementos no repetidos de la columna A:\")\n",
    "print(tabla['A'].unique())\n",
    "\n",
    "# numero de valores únicos que hay\n",
    "print(\"\\nNumero de valores únicos que hay:\")\n",
    "print(tabla['A'].nunique())\n",
    "\n",
    "# imprimimos cuantas veces se repite un valor\n",
    "print(\"\\nVeces que se repite un valor de la columna A:\")\n",
    "print(tabla['A'].value_counts())\n",
    "\n",
    "# aplicar una funcion a una columna\n",
    "def duplicar(x):\n",
    "    return x*2\n",
    "\n",
    "print(\"\\nColumna A duplicada por 2:\")\n",
    "print(tabla['A'].apply(duplicar)) # esto es equivalente a: tabla['A'].apply(lambda x: x*2)\n",
    "\n",
    "print(\"\\nBorrar columna B:\")\n",
    "tabla.drop('C', axis=1,inplace=True)\n",
    "print(tabla) \n",
    "\n",
    "print(\"\\nBorrar fila i2:\")\n",
    "tabla.drop('i2',inplace=True)\n",
    "print(tabla) \n",
    "\n",
    "print(\"\\nOrdenar los elementos de la tabla atendiendo a la columna B (de menor a mayor):\")\n",
    "print(tabla.sort_values('B'))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Leer xlsx: un fichero excel con Pandas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "# se pueden utilizar los comandos 'pwd' y 'ls' para mostoja1rar rutas y directorios \n",
    "\n",
    "dataframe = pd.read_excel('fichero.xlsx')\n",
    "\n",
    "dataframe['e'] = [1,2,3,4,5] # añadimos una nueva columna\n",
    "\n",
    "dataframe.to_excel('fichero.xlsx', sheet_name='Hoja1') # lo guardamos en un excel"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Leer html con Pandas "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tipo de datos obtenidos de la url:  <class 'list'>\n"
     ]
    },
    {
     "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>País, Estado o territorio</th>\n",
       "      <th colspan=\"2\" halign=\"left\">Casos</th>\n",
       "      <th colspan=\"3\" halign=\"left\">Fallecidos</th>\n",
       "      <th colspan=\"2\" halign=\"left\">Recuperados</th>\n",
       "      <th>Ref.</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>País, Estado o territorio</th>\n",
       "      <th>Totales</th>\n",
       "      <th>CxMhab.</th>\n",
       "      <th>Totales</th>\n",
       "      <th>%</th>\n",
       "      <th>FxMhab.</th>\n",
       "      <th>Totales</th>\n",
       "      <th>%</th>\n",
       "      <th>Ref.</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>202 territorios</th>\n",
       "      <th>752 747</th>\n",
       "      <th>n/a</th>\n",
       "      <th>36 226</th>\n",
       "      <th>4.81 %</th>\n",
       "      <th>n/a</th>\n",
       "      <th>158 705</th>\n",
       "      <th>21.08 %</th>\n",
       "      <th>Ref.</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Estados Unidos(1)</td>\n",
       "      <td>145 131</td>\n",
       "      <td>438.94</td>\n",
       "      <td>2608</td>\n",
       "      <td>1.8 %</td>\n",
       "      <td>7.55</td>\n",
       "      <td>4767</td>\n",
       "      <td>3.28 %</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Italia</td>\n",
       "      <td>101 739</td>\n",
       "      <td>1693.11</td>\n",
       "      <td>11 591</td>\n",
       "      <td>11.39 %</td>\n",
       "      <td>192.89</td>\n",
       "      <td>14 620</td>\n",
       "      <td>14.37 %</td>\n",
       "      <td>[151]​</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>España</td>\n",
       "      <td>85 195</td>\n",
       "      <td>1805.36</td>\n",
       "      <td>7340</td>\n",
       "      <td>9.16 %</td>\n",
       "      <td>155.54</td>\n",
       "      <td>16 780</td>\n",
       "      <td>20.95 %</td>\n",
       "      <td>[152]​</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>China</td>\n",
       "      <td>81 470</td>\n",
       "      <td>57.99</td>\n",
       "      <td>3304</td>\n",
       "      <td>4.06 %</td>\n",
       "      <td>2.35</td>\n",
       "      <td>75 770</td>\n",
       "      <td>93 %</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Alemania</td>\n",
       "      <td>63 929</td>\n",
       "      <td>768.01</td>\n",
       "      <td>560</td>\n",
       "      <td>0.9 %</td>\n",
       "      <td>6.73</td>\n",
       "      <td>9211</td>\n",
       "      <td>14.41 %</td>\n",
       "      <td>[153]​</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Irán</td>\n",
       "      <td>41 495</td>\n",
       "      <td>497.72</td>\n",
       "      <td>2757</td>\n",
       "      <td>6.64 %</td>\n",
       "      <td>33.07</td>\n",
       "      <td>13 911</td>\n",
       "      <td>33.52 %</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Francia</td>\n",
       "      <td>40 174</td>\n",
       "      <td>619.2</td>\n",
       "      <td>2606</td>\n",
       "      <td>6.49 %</td>\n",
       "      <td>40.17</td>\n",
       "      <td>7202</td>\n",
       "      <td>17.93 %</td>\n",
       "      <td>[154]​</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Reino Unido</td>\n",
       "      <td>22 141</td>\n",
       "      <td>330.31</td>\n",
       "      <td>1408</td>\n",
       "      <td>6.36 %</td>\n",
       "      <td>21.01</td>\n",
       "      <td>140</td>\n",
       "      <td>0.63 %</td>\n",
       "      <td>[155]​</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Suiza</td>\n",
       "      <td>15 526</td>\n",
       "      <td>1805.35</td>\n",
       "      <td>333</td>\n",
       "      <td>2.14 %</td>\n",
       "      <td>38.72</td>\n",
       "      <td>1823</td>\n",
       "      <td>12.1 %</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Bélgica</td>\n",
       "      <td>11 899</td>\n",
       "      <td>1032.9</td>\n",
       "      <td>513</td>\n",
       "      <td>4.31 %</td>\n",
       "      <td>44.53</td>\n",
       "      <td>1359</td>\n",
       "      <td>11.42 %</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  País, Estado o territorio    Casos          Fallecidos                   \\\n",
       "  País, Estado o territorio  Totales  CxMhab.    Totales        % FxMhab.   \n",
       "            202 territorios  752 747      n/a     36 226   4.81 %     n/a   \n",
       "0         Estados Unidos(1)  145 131   438.94       2608    1.8 %    7.55   \n",
       "1                    Italia  101 739  1693.11     11 591  11.39 %  192.89   \n",
       "2                    España   85 195  1805.36       7340   9.16 %  155.54   \n",
       "3                     China   81 470    57.99       3304   4.06 %    2.35   \n",
       "4                  Alemania   63 929   768.01        560    0.9 %    6.73   \n",
       "5                      Irán   41 495   497.72       2757   6.64 %   33.07   \n",
       "6                   Francia   40 174    619.2       2606   6.49 %   40.17   \n",
       "7               Reino Unido   22 141   330.31       1408   6.36 %   21.01   \n",
       "8                     Suiza   15 526  1805.35        333   2.14 %   38.72   \n",
       "9                   Bélgica   11 899   1032.9        513   4.31 %   44.53   \n",
       "\n",
       "  Recuperados             Ref.  \n",
       "      Totales        %    Ref.  \n",
       "      158 705  21.08 %    Ref.  \n",
       "0        4767   3.28 %     NaN  \n",
       "1      14 620  14.37 %  [151]​  \n",
       "2      16 780  20.95 %  [152]​  \n",
       "3      75 770     93 %     NaN  \n",
       "4        9211  14.41 %  [153]​  \n",
       "5      13 911  33.52 %     NaN  \n",
       "6        7202  17.93 %  [154]​  \n",
       "7         140   0.63 %  [155]​  \n",
       "8        1823   12.1 %     NaN  \n",
       "9        1359  11.42 %     NaN  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "url = 'https://es.wikipedia.org/wiki/Pandemia_de_enfermedad_por_coronavirus_de_2019-2020'\n",
    "datos = pd.read_html(url)\n",
    "print('Tipo de datos obtenidos de la url: ',type(datos))\n",
    "paises_afectados = datos[5] # la tabla 5 es sobre los paises afectados\n",
    "paises_afectados.head(10)   # cogemos los 10 primeros paises "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Grabar DataFrame es una tabla SQL"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Instalación de los paquetes necesarios:\n",
    "Vamos a cargar un motor de SQL en memoria y para ello necesitaremos instalar la librerías siguientes:\n",
    "\n",
    "\n",
    "* Con Anaconda:\n",
    "```bash\n",
    "conda install sqlalchemy\n",
    "```\n",
    "* Sin Anaconda:\n",
    "```bash\n",
    "pip install sqlalchemy\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   A  B  C\n",
      "0  1  4  7\n",
      "1  2  5  8\n",
      "2  3  6  9\n"
     ]
    },
    {
     "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   A  B  C\n",
       "0  1  4  7\n",
       "1  2  5  8\n",
       "2  3  6  9"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "from sqlalchemy import create_engine\n",
    "\n",
    "\n",
    "diccionario = {'A':[1,2,3], 'B':[4,5,6], 'C':[7,8,9]}\n",
    "dataframe = pd.DataFrame(diccionario)\n",
    "\n",
    "print(dataframe)\n",
    "\n",
    "# cargamos en memoria un motor de sql\n",
    "engine = create_engine('sqlite:///:memory:')\n",
    "\n",
    "dataframe.to_sql('tabla',engine,index=False) # decimos que el index no lo pase\n",
    "\n",
    "datos_leidos_bd = pd.read_sql('tabla', con=engine) # utilizamos una conexion llamado engine\n",
    "\n",
    "datos_leidos_bd"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Gráficos con Pandas "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    a  b  c  d\n",
      "0   1  1  1  4\n",
      "1   3  0  1  0\n",
      "2   1  1  1  1\n",
      "3   3  1  3  0\n",
      "4   0  3  0  1\n",
      "5   2  4  4  1\n",
      "6   2  1  1  3\n",
      "7   2  3  3  1\n",
      "8   0  1  4  1\n",
      "9   1  0  1  3\n",
      "10  4  0  0  1\n",
      "11  4  4  1  1\n",
      "12  3  3  0  3\n",
      "13  2  1  3  4\n",
      "14  1  1  3  4\n",
      "15  4  0  0  1\n",
      "16  0  3  3  2\n",
      "17  2  0  4  1\n",
      "18  2  3  1  2\n",
      "19  3  1  1  3\n",
      "AxesSubplot(0.125,0.125;0.775x0.755)\n",
      "AxesSubplot(0.125,0.125;0.775x0.755)\n",
      "AxesSubplot(0.125,0.125;0.775x0.755)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fb7b20e92d0>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "# creamos una matriz de 2x4 con numeros aleatorios de 0 a 5\n",
    "matriz = np.random.randint(5, size=(20,4)) \n",
    "dataframe = pd.DataFrame(matriz, columns=['a','b','c','d']) # agregamos el nombre de las columnas que compondrán la tabla\n",
    "print(dataframe)\n",
    "\n",
    "# creamos un histograma\n",
    "print(dataframe['b'].hist())\n",
    "\n",
    "# creamos una gráfica con los valores de las columnas\n",
    "print(dataframe.plot.area(alpha=0.3)) # alpha = valor de transparencia\n",
    "\n",
    "# gráfico de barras\n",
    "print(dataframe.plot.bar()) # stacked=True si queremos acumular valores en las barras.\n",
    "\n",
    "# grafico tipo scatter\n",
    "dataframe.plot.scatter(x='a',y='b',c='c',cmap='coolwarm') # identificamos las columnas según la escala de grises\n",
    "# 'coolwarm' sería una paleta de colores distinta (más cálida)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Ejercicio 1: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "matemáticas    9\n",
      "fisica         9\n",
      "historia       8\n",
      "dtype: int64\n",
      "La nota de física que has sacado es  9\n"
     ]
    }
   ],
   "source": [
    "import numpy as mp\n",
    "import pandas as pd\n",
    "\n",
    "asignaturas = ['matemáticas', 'fisica', 'historia']\n",
    "notas = [9,9,8]\n",
    "serie = pd.Series(index=asignaturas, data=notas)\n",
    "print(serie)\n",
    "nota_fisica = serie['fisica']\n",
    "print('La nota de física que has sacado es ',nota_fisica)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Ejercicio 2: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                oficina_1  oficina_2  oficina_3\n",
      "ventas_enero          457        442        288\n",
      "ventas_febrero        492        313        413\n",
      "ventas_marzo          357        488        188\n",
      "\n",
      "Las ventas de enero para la oficina_1 son:  457\n"
     ]
    }
   ],
   "source": [
    "import numpy as mp\n",
    "import pandas as pd\n",
    "\n",
    "filas = ['ventas_enero','ventas_febrero','ventas_marzo']\n",
    "columnas =['oficina_1','oficina_2','oficina_3']\n",
    "\n",
    "\n",
    "ventas_enero = np.random.randint(100,500,3)\n",
    "ventas_febrero = np.random.randint(100,500,3)\n",
    "ventas_marzo = np.random.randint(100,500,3)\n",
    "\n",
    "ventas = [ventas_enero,ventas_febrero,ventas_marzo]\n",
    "\n",
    "dataframe = pd.DataFrame(index=filas, columns=columnas, data=ventas)\n",
    "print(dataframe)\n",
    "\n",
    "print(\"\\nLas ventas de enero para la oficina_1 son: \",dataframe['oficina_1']['ventas_enero'])\n",
    "# tambien se puede poner como : dataframe.loc['ventas_enero']['oficina_1']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}