Notebook-taller-bloque1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Bloque 1: Naturaleza de las series de tiempo\n",
    "\n",
    "Melanie Oyarzún W.\n",
    "\n",
    "\n",
    "**Resultado de aprendizaje esperado:**\n",
    "\n",
    "Identificar datos de series temporales, sus particularidades y riesgos, en el contexto de posibles aplicaciones profesionales.\n",
    "\n",
    "**Bibliografía recomendada:**\n",
    "\n",
    "Stock & Watson, C.14 [link](https://www.dropbox.com/s/d09qcat340jy5q2/stock%20y%20watson%20-%20Introduccion%20a%20la%20Econometria.pdf?dl=0) ; Wooldridge, c.12 [link](https://www.dropbox.com/s/wrh0tuna27qp747/Wooldridge__Introductory_Econometrics_2nd_Ed_Solutions.pdf?dl=0), Gujarati, c.12  [link](https://www.dropbox.com/s/rrvth8vsa1c5gts/Gujarati-cap%2012.pdf?dl=0)\n",
    "\n",
    "**Material de apoyo***\n",
    "\n",
    "El taller cuenta con un repositorio  [https://github.com/melanieoyarzun/taller_IDS2021](https://github.com/melanieoyarzun/taller_IDS2021) en el cual está dispoinible todo el material.\n",
    "\n",
    "**Slides**\n",
    "\n",
    "Estan disponibles en el repositorio o directamente en el link \n",
    "\n",
    "---"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 1000x600 with 0 Axes>"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 1000x600 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Paquetes y settings\n",
    "\n",
    "from dateutil.parser import parse \n",
    "import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import watermark\n",
    "\n",
    "\n",
    "#%matplotlib inline\n",
    "%load_ext watermark\n",
    "\n",
    "# setting de graficos\n",
    "\n",
    "plt.figure(figsize=(5,3), dpi= 200, facecolor='w', edgecolor='k')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Python implementation: CPython\n",
      "Python version       : 3.9.7\n",
      "IPython version      : 7.29.0\n",
      "\n",
      "Compiler    : Clang 10.0.0 \n",
      "OS          : Darwin\n",
      "Release     : 20.6.0\n",
      "Machine     : x86_64\n",
      "Processor   : i386\n",
      "CPU cores   : 4\n",
      "Architecture: 64bit\n",
      "\n",
      "Git hash: aef3c6ddbb2a0445b60074d46581d039a1946002\n",
      "\n",
      "seaborn   : 0.11.2\n",
      "pandas    : 1.3.4\n",
      "watermark : 2.2.0\n",
      "sys       : 3.9.7 (default, Sep 16 2021, 08:50:36) \n",
      "[Clang 10.0.0 ]\n",
      "numpy     : 1.20.3\n",
      "matplotlib: 3.4.3\n",
      "\n"
     ]
    }
   ],
   "source": [
    "%watermark -n -v -m -g -iv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Tipos de datos\n",
    "\n",
    "Los datos que analizamos con modelos se pueden categorizar en tres grandes tipos: \n",
    "\n",
    "- corte transversal \n",
    "- series de tiempo (o longitudinal)\n",
    "- panel\n",
    "\n",
    "## Corte transversal\n",
    "Cuando se empieza a estudiar modelamiento de datos, la aplicación natural de estos es en lo que llamamos datos de corte transversal. En estos, la **unidad de observación** son individuos separados unos de otros. \n",
    "\n",
    "Usualmente, cada fila reprsenta un individuo (personas, familias, empresas) en un momento del tiempo específico (mismo año, mismo mes, etc). Es análogo a una *fotografía* de do grupo.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Un ejemplo de esto es la encuesta Casen [(http://observatorio.ministeriodesarrollosocial.gob.cl/encuesta-casen)](http://observatorio.ministeriodesarrollosocial.gob.cl/encuesta-casen) o Encuesta de Caracterización Socioeconómica Nacional.\n",
    "\n",
    "Esta es una encuesta de caracter nacional realizada a hogares, en la cual se busca levantar información para conocer la situación de los hogares y de la población, especialmente de aquella en situación de pobreza y de aquellos grupos definidos como prioritarios y permite evaluar impacto de política social. Es representativa a nivel regional.\n",
    "\n",
    "Es realizada por el Ministerio de Desarrollo Social desde el año 1990 con una periodicidad; bianual o trianual. Hasta ahora, las encuestas aplicadas corresponden a los años 1990, 1992, 1994, 1996, 1998, 2000, 2003, 2006, 2009, 2011, 2013, 2105 y 2017."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>folio</th>\n",
       "      <th>o</th>\n",
       "      <th>id_persona</th>\n",
       "      <th>region</th>\n",
       "      <th>comuna</th>\n",
       "      <th>zona</th>\n",
       "      <th>expr</th>\n",
       "      <th>edad</th>\n",
       "      <th>sexo</th>\n",
       "      <th>tot_per</th>\n",
       "      <th>...</th>\n",
       "      <th>esc2</th>\n",
       "      <th>educ</th>\n",
       "      <th>o1</th>\n",
       "      <th>yaut</th>\n",
       "      <th>yauth</th>\n",
       "      <th>yautcor</th>\n",
       "      <th>yautcorh</th>\n",
       "      <th>ytrabajocor</th>\n",
       "      <th>ytrabajocorh</th>\n",
       "      <th>yae</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.101100e+11</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>Región de Tarapacá</td>\n",
       "      <td>Iquique</td>\n",
       "      <td>Urbano</td>\n",
       "      <td>67</td>\n",
       "      <td>34</td>\n",
       "      <td>Mujer</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Media humanista completa</td>\n",
       "      <td>No</td>\n",
       "      <td>220000.0</td>\n",
       "      <td>300000</td>\n",
       "      <td>220000.0</td>\n",
       "      <td>300000</td>\n",
       "      <td>150000.0</td>\n",
       "      <td>150000.0</td>\n",
       "      <td>240586.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.101100e+11</td>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>Región de Tarapacá</td>\n",
       "      <td>Iquique</td>\n",
       "      <td>Urbano</td>\n",
       "      <td>67</td>\n",
       "      <td>4</td>\n",
       "      <td>Mujer</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Sin educación formal</td>\n",
       "      <td>NaN</td>\n",
       "      <td>80000.0</td>\n",
       "      <td>300000</td>\n",
       "      <td>80000.0</td>\n",
       "      <td>300000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>150000.0</td>\n",
       "      <td>240586.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.101100e+11</td>\n",
       "      <td>2</td>\n",
       "      <td>31</td>\n",
       "      <td>Región de Tarapacá</td>\n",
       "      <td>Iquique</td>\n",
       "      <td>Urbano</td>\n",
       "      <td>67</td>\n",
       "      <td>5</td>\n",
       "      <td>Mujer</td>\n",
       "      <td>3</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Básica incompleta</td>\n",
       "      <td>NaN</td>\n",
       "      <td>25000.0</td>\n",
       "      <td>941583</td>\n",
       "      <td>25000.0</td>\n",
       "      <td>941583</td>\n",
       "      <td>NaN</td>\n",
       "      <td>891583.0</td>\n",
       "      <td>439170.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.101100e+11</td>\n",
       "      <td>1</td>\n",
       "      <td>32</td>\n",
       "      <td>Región de Tarapacá</td>\n",
       "      <td>Iquique</td>\n",
       "      <td>Urbano</td>\n",
       "      <td>67</td>\n",
       "      <td>45</td>\n",
       "      <td>Hombre</td>\n",
       "      <td>3</td>\n",
       "      <td>...</td>\n",
       "      <td>15.0</td>\n",
       "      <td>Técnico nivel superior incompleta</td>\n",
       "      <td>Sí</td>\n",
       "      <td>889500.0</td>\n",
       "      <td>941583</td>\n",
       "      <td>889500.0</td>\n",
       "      <td>941583</td>\n",
       "      <td>889500.0</td>\n",
       "      <td>891583.0</td>\n",
       "      <td>439170.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.101100e+11</td>\n",
       "      <td>3</td>\n",
       "      <td>30</td>\n",
       "      <td>Región de Tarapacá</td>\n",
       "      <td>Iquique</td>\n",
       "      <td>Urbano</td>\n",
       "      <td>67</td>\n",
       "      <td>19</td>\n",
       "      <td>Mujer</td>\n",
       "      <td>3</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>No sabe</td>\n",
       "      <td>No</td>\n",
       "      <td>27083.0</td>\n",
       "      <td>941583</td>\n",
       "      <td>27083.0</td>\n",
       "      <td>941583</td>\n",
       "      <td>2083.0</td>\n",
       "      <td>891583.0</td>\n",
       "      <td>439170.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.101100e+11</td>\n",
       "      <td>2</td>\n",
       "      <td>117</td>\n",
       "      <td>Región de Tarapacá</td>\n",
       "      <td>Iquique</td>\n",
       "      <td>Urbano</td>\n",
       "      <td>67</td>\n",
       "      <td>57</td>\n",
       "      <td>Hombre</td>\n",
       "      <td>3</td>\n",
       "      <td>...</td>\n",
       "      <td>16.0</td>\n",
       "      <td>Técnico nivel superior completo</td>\n",
       "      <td>Sí</td>\n",
       "      <td>3000000.0</td>\n",
       "      <td>3004167</td>\n",
       "      <td>3000000.0</td>\n",
       "      <td>3004167</td>\n",
       "      <td>3000000.0</td>\n",
       "      <td>3004167.0</td>\n",
       "      <td>1522295.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1.101100e+11</td>\n",
       "      <td>3</td>\n",
       "      <td>118</td>\n",
       "      <td>Región de Tarapacá</td>\n",
       "      <td>Iquique</td>\n",
       "      <td>Urbano</td>\n",
       "      <td>67</td>\n",
       "      <td>20</td>\n",
       "      <td>Hombre</td>\n",
       "      <td>3</td>\n",
       "      <td>...</td>\n",
       "      <td>14.0</td>\n",
       "      <td>Técnico nivel superior incompleta</td>\n",
       "      <td>No</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3004167</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3004167</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3004167.0</td>\n",
       "      <td>1522295.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1.101100e+11</td>\n",
       "      <td>1</td>\n",
       "      <td>116</td>\n",
       "      <td>Región de Tarapacá</td>\n",
       "      <td>Iquique</td>\n",
       "      <td>Urbano</td>\n",
       "      <td>67</td>\n",
       "      <td>56</td>\n",
       "      <td>Mujer</td>\n",
       "      <td>3</td>\n",
       "      <td>...</td>\n",
       "      <td>15.0</td>\n",
       "      <td>Técnico nivel superior completo</td>\n",
       "      <td>No</td>\n",
       "      <td>4167.0</td>\n",
       "      <td>3004167</td>\n",
       "      <td>4167.0</td>\n",
       "      <td>3004167</td>\n",
       "      <td>4167.0</td>\n",
       "      <td>3004167.0</td>\n",
       "      <td>1522295.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1.101100e+11</td>\n",
       "      <td>1</td>\n",
       "      <td>2209</td>\n",
       "      <td>Región de Tarapacá</td>\n",
       "      <td>Iquique</td>\n",
       "      <td>Urbano</td>\n",
       "      <td>67</td>\n",
       "      <td>77</td>\n",
       "      <td>Hombre</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>10.0</td>\n",
       "      <td>Media técnica profesional incompleta</td>\n",
       "      <td>No</td>\n",
       "      <td>1250.0</td>\n",
       "      <td>128250</td>\n",
       "      <td>1250.0</td>\n",
       "      <td>128250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>375653.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1.101100e+11</td>\n",
       "      <td>2</td>\n",
       "      <td>2210</td>\n",
       "      <td>Región de Tarapacá</td>\n",
       "      <td>Iquique</td>\n",
       "      <td>Urbano</td>\n",
       "      <td>67</td>\n",
       "      <td>60</td>\n",
       "      <td>Mujer</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Media humanista completa</td>\n",
       "      <td>No</td>\n",
       "      <td>127000.0</td>\n",
       "      <td>128250</td>\n",
       "      <td>127000.0</td>\n",
       "      <td>128250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>375653.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 22 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          folio  o  id_persona              region   comuna    zona  expr  \\\n",
       "0  1.101100e+11  1           5  Región de Tarapacá  Iquique  Urbano    67   \n",
       "1  1.101100e+11  2           6  Región de Tarapacá  Iquique  Urbano    67   \n",
       "2  1.101100e+11  2          31  Región de Tarapacá  Iquique  Urbano    67   \n",
       "3  1.101100e+11  1          32  Región de Tarapacá  Iquique  Urbano    67   \n",
       "4  1.101100e+11  3          30  Región de Tarapacá  Iquique  Urbano    67   \n",
       "5  1.101100e+11  2         117  Región de Tarapacá  Iquique  Urbano    67   \n",
       "6  1.101100e+11  3         118  Región de Tarapacá  Iquique  Urbano    67   \n",
       "7  1.101100e+11  1         116  Región de Tarapacá  Iquique  Urbano    67   \n",
       "8  1.101100e+11  1        2209  Región de Tarapacá  Iquique  Urbano    67   \n",
       "9  1.101100e+11  2        2210  Región de Tarapacá  Iquique  Urbano    67   \n",
       "\n",
       "   edad    sexo  tot_per  ...  esc2                                  educ  \\\n",
       "0    34   Mujer        2  ...  12.0              Media humanista completa   \n",
       "1     4   Mujer        2  ...   NaN                  Sin educación formal   \n",
       "2     5   Mujer        3  ...   NaN                     Básica incompleta   \n",
       "3    45  Hombre        3  ...  15.0     Técnico nivel superior incompleta   \n",
       "4    19   Mujer        3  ...   NaN                               No sabe   \n",
       "5    57  Hombre        3  ...  16.0       Técnico nivel superior completo   \n",
       "6    20  Hombre        3  ...  14.0     Técnico nivel superior incompleta   \n",
       "7    56   Mujer        3  ...  15.0       Técnico nivel superior completo   \n",
       "8    77  Hombre        2  ...  10.0  Media técnica profesional incompleta   \n",
       "9    60   Mujer        2  ...  12.0              Media humanista completa   \n",
       "\n",
       "    o1       yaut    yauth    yautcor  yautcorh  ytrabajocor  ytrabajocorh  \\\n",
       "0   No   220000.0   300000   220000.0    300000     150000.0      150000.0   \n",
       "1  NaN    80000.0   300000    80000.0    300000          NaN      150000.0   \n",
       "2  NaN    25000.0   941583    25000.0    941583          NaN      891583.0   \n",
       "3   Sí   889500.0   941583   889500.0    941583     889500.0      891583.0   \n",
       "4   No    27083.0   941583    27083.0    941583       2083.0      891583.0   \n",
       "5   Sí  3000000.0  3004167  3000000.0   3004167    3000000.0     3004167.0   \n",
       "6   No        NaN  3004167        NaN   3004167          NaN     3004167.0   \n",
       "7   No     4167.0  3004167     4167.0   3004167       4167.0     3004167.0   \n",
       "8   No     1250.0   128250     1250.0    128250          NaN           0.0   \n",
       "9   No   127000.0   128250   127000.0    128250          NaN           0.0   \n",
       "\n",
       "         yae  \n",
       "0   240586.0  \n",
       "1   240586.0  \n",
       "2   439170.0  \n",
       "3   439170.0  \n",
       "4   439170.0  \n",
       "5  1522295.0  \n",
       "6  1522295.0  \n",
       "7  1522295.0  \n",
       "8   375653.0  \n",
       "9   375653.0  \n",
       "\n",
       "[10 rows x 22 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_casen2020= pd.read_stata(\"Data/casen_2020_ingresos.dta\")\n",
    "df_casen2020.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 185437 entries, 0 to 185436\n",
      "Data columns (total 22 columns):\n",
      " #   Column        Non-Null Count   Dtype   \n",
      "---  ------        --------------   -----   \n",
      " 0   folio         185437 non-null  float64 \n",
      " 1   o             185437 non-null  int8    \n",
      " 2   id_persona    185437 non-null  int16   \n",
      " 3   region        185437 non-null  category\n",
      " 4   comuna        185437 non-null  category\n",
      " 5   zona          185437 non-null  category\n",
      " 6   expr          185437 non-null  int16   \n",
      " 7   edad          185437 non-null  int16   \n",
      " 8   sexo          185437 non-null  category\n",
      " 9   tot_per       185437 non-null  int8    \n",
      " 10  ecivil        153892 non-null  category\n",
      " 11  esc           148886 non-null  float64 \n",
      " 12  esc2          148886 non-null  float64 \n",
      " 13  educ          185437 non-null  category\n",
      " 14  o1            151315 non-null  category\n",
      " 15  yaut          95399 non-null   float64 \n",
      " 16  yauth         185437 non-null  int32   \n",
      " 17  yautcor       102165 non-null  float64 \n",
      " 18  yautcorh      185437 non-null  int32   \n",
      " 19  ytrabajocor   73509 non-null   float32 \n",
      " 20  ytrabajocorh  185437 non-null  float32 \n",
      " 21  yae           185339 non-null  float64 \n",
      "dtypes: category(7), float32(2), float64(6), int16(3), int32(2), int8(2)\n",
      "memory usage: 15.6 MB\n"
     ]
    }
   ],
   "source": [
    "df_casen2020.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='esc', ylabel='yaut'>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "sns.scatterplot(data=df_casen2020, x=\"esc\", y=\"yaut\", alpha=0.5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Series de tiempo\n",
    "\n",
    "En cambio, un tipo diferente de datos son las **Series Temporales** o tambien longitudinales.\n",
    "\n",
    "En este tipo de información, se tiene diferentes momentos del tiempo (día, semana, mes, año, etc.) para una misma unidad de análisis (individuo, país, empresa, etc)\n",
    "\n",
    "Ejemplos típicos de series de tiempo son: \n",
    "- Datos macroeconómicos (PIB. Inglación, empleo, etc.)\n",
    "- Financieros (precios de acciones)\n",
    "- Empresariales (Ventas, costos, etc...)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>date</th>\n",
       "      <th>open</th>\n",
       "      <th>high</th>\n",
       "      <th>low</th>\n",
       "      <th>close</th>\n",
       "      <th>adjclose</th>\n",
       "      <th>volume</th>\n",
       "      <th>ticker</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2009-11-30</td>\n",
       "      <td>143.619995</td>\n",
       "      <td>145.910004</td>\n",
       "      <td>135.110001</td>\n",
       "      <td>137.580002</td>\n",
       "      <td>137.580002</td>\n",
       "      <td>31350900</td>\n",
       "      <td>AMZN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2009-12-07</td>\n",
       "      <td>138.000000</td>\n",
       "      <td>139.000000</td>\n",
       "      <td>129.820007</td>\n",
       "      <td>134.149994</td>\n",
       "      <td>134.149994</td>\n",
       "      <td>47863000</td>\n",
       "      <td>AMZN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2009-12-14</td>\n",
       "      <td>132.500000</td>\n",
       "      <td>132.610001</td>\n",
       "      <td>125.650002</td>\n",
       "      <td>128.479996</td>\n",
       "      <td>128.479996</td>\n",
       "      <td>45794900</td>\n",
       "      <td>AMZN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2009-12-21</td>\n",
       "      <td>130.479996</td>\n",
       "      <td>139.699997</td>\n",
       "      <td>130.190002</td>\n",
       "      <td>138.470001</td>\n",
       "      <td>138.470001</td>\n",
       "      <td>32406000</td>\n",
       "      <td>AMZN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2009-12-28</td>\n",
       "      <td>139.750000</td>\n",
       "      <td>142.580002</td>\n",
       "      <td>134.520004</td>\n",
       "      <td>134.520004</td>\n",
       "      <td>134.520004</td>\n",
       "      <td>28600700</td>\n",
       "      <td>AMZN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>518</th>\n",
       "      <td>2019-11-04</td>\n",
       "      <td>1801.010010</td>\n",
       "      <td>1815.060059</td>\n",
       "      <td>1774.040039</td>\n",
       "      <td>1785.880005</td>\n",
       "      <td>1785.880005</td>\n",
       "      <td>11461600</td>\n",
       "      <td>AMZN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>519</th>\n",
       "      <td>2019-11-11</td>\n",
       "      <td>1778.000000</td>\n",
       "      <td>1786.219971</td>\n",
       "      <td>1732.859985</td>\n",
       "      <td>1739.489990</td>\n",
       "      <td>1739.489990</td>\n",
       "      <td>13165500</td>\n",
       "      <td>AMZN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>520</th>\n",
       "      <td>2019-11-18</td>\n",
       "      <td>1738.300049</td>\n",
       "      <td>1762.520020</td>\n",
       "      <td>1722.709961</td>\n",
       "      <td>1745.719971</td>\n",
       "      <td>1745.719971</td>\n",
       "      <td>13042300</td>\n",
       "      <td>AMZN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>521</th>\n",
       "      <td>2019-11-25</td>\n",
       "      <td>1753.250000</td>\n",
       "      <td>1824.689941</td>\n",
       "      <td>1753.239990</td>\n",
       "      <td>1800.800049</td>\n",
       "      <td>1800.800049</td>\n",
       "      <td>11616400</td>\n",
       "      <td>AMZN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>522</th>\n",
       "      <td>2019-12-02</td>\n",
       "      <td>1804.400024</td>\n",
       "      <td>1805.550049</td>\n",
       "      <td>1747.229980</td>\n",
       "      <td>1769.959961</td>\n",
       "      <td>1769.959961</td>\n",
       "      <td>7306500</td>\n",
       "      <td>AMZN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>523 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          date         open         high          low        close  \\\n",
       "0   2009-11-30   143.619995   145.910004   135.110001   137.580002   \n",
       "1   2009-12-07   138.000000   139.000000   129.820007   134.149994   \n",
       "2   2009-12-14   132.500000   132.610001   125.650002   128.479996   \n",
       "3   2009-12-21   130.479996   139.699997   130.190002   138.470001   \n",
       "4   2009-12-28   139.750000   142.580002   134.520004   134.520004   \n",
       "..         ...          ...          ...          ...          ...   \n",
       "518 2019-11-04  1801.010010  1815.060059  1774.040039  1785.880005   \n",
       "519 2019-11-11  1778.000000  1786.219971  1732.859985  1739.489990   \n",
       "520 2019-11-18  1738.300049  1762.520020  1722.709961  1745.719971   \n",
       "521 2019-11-25  1753.250000  1824.689941  1753.239990  1800.800049   \n",
       "522 2019-12-02  1804.400024  1805.550049  1747.229980  1769.959961   \n",
       "\n",
       "        adjclose    volume ticker  \n",
       "0     137.580002  31350900   AMZN  \n",
       "1     134.149994  47863000   AMZN  \n",
       "2     128.479996  45794900   AMZN  \n",
       "3     138.470001  32406000   AMZN  \n",
       "4     134.520004  28600700   AMZN  \n",
       "..           ...       ...    ...  \n",
       "518  1785.880005  11461600   AMZN  \n",
       "519  1739.489990  13165500   AMZN  \n",
       "520  1745.719971  13042300   AMZN  \n",
       "521  1800.800049  11616400   AMZN  \n",
       "522  1769.959961   7306500   AMZN  \n",
       "\n",
       "[523 rows x 8 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Ejemplo datos de acciones\n",
    "\n",
    "#Usemos la api de Yahoo finance: pip install yahoo_fin y pip install requests_html\n",
    "\n",
    "from yahoo_fin.stock_info import get_data\n",
    "\n",
    "amazon_weekly= get_data(\"amzn\", start_date=\"12/04/2009\", end_date=\"12/04/2019\", index_as_date = False, interval=\"1wk\")\n",
    "amazon_weekly\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='date', ylabel='open'>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.scatterplot(data=amazon_weekly, y=\"open\", x=\"date\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Uso series de tiempo\n",
    "\n",
    "La información temporal(de series de tiempo) permite responder cómo una variable (o más) responde a cambios a través del tiempo.\n",
    "- ¿Cuál es el efecto causal dinámico de Xt sobre Yt? \n",
    "- ¿Cuál es la mejor predicción del valor de Y en el futuro?\n",
    "  \n",
    "**PERO su uso sin cuidado puede llevarnos a conclusiones MUY equiviocadas**\n",
    "\n",
    "\n",
    "![Ejemplo idotamente peligroso](slide_figuras/spurius.png)\n",
    "\n",
    "\n",
    "(Fuente: [https://www.tylervigen.com/spurious-correlations](https://www.tylervigen.com/spurious-correlations))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Trabajar con series de tiempo no es para nada sencillo: **necesitaremos técnicas nuevas.**\n",
    "\n",
    "¿Por qué? \n",
    "\n",
    "Necesitaremos modelos sofisticados de errores y técnicas para estimarlos, por ciertas dieferencias en la naturaleza del **proceso generador de datos**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Panel\n",
    "\n",
    "El último tipo de datos, es lo que se llama un panel. Este es la combinacion de los dos tipos anteriores, donde para un conjunto de individuos tenemos varias observaciones en el tiempo.\n",
    "\n",
    "Suele ser el tipo de datos más completo, pero tabien más dificil de conseguir. Además, enfrenta los problemas típicos de ambos tipos anteriores, dependiendo si es un panel corto o largo.\n",
    "\n",
    "Una alternativa a etsos son los llamados pooled cross section.\n",
    "\n",
    "Ela análisis de este tipo de datos escapa a los objetivos del taller, pero una introducción pueden revisarla en Stock y Watson Cap. 10.\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Corte transversal vs series de tiempo\n",
    "\n",
    "## Muestras en corte transversal\n",
    "\n",
    "En corte transversal solemos trabajar con **MUESTRAS**\n",
    "\n",
    "![Corte transversal](slide_figuras/datos_ct.png)\n",
    "\n",
    "Estimamos modelos de la forma:\n",
    "$$ y_{i}=\\beta_{0}+\\beta_{1}x_{1i}+\\dots+\\beta_{1}x_{ki}+u_{i} $$\n",
    "\n",
    "Estos modelos representan una correlación marginal en las observaciones entre y y x's  (escalada por la varianza de x) y para que podamos interpretarlas causalmente tenemos varias condiciones o supuetsos que de sedeb cumplir.\n",
    "\n",
    "Uno de estos, es el **supuesto de exogeneidad**: $$  E[u_{i}|X_{i}]=0 $$\n",
    "\n",
    "Sin embargo, esta forma de ver este supuesto es una **simplificación** ya que, como estamos en una uestra aleatoria, no tenemos que verificar que los efectos cruzados tambien sean exogenos:\n",
    "\n",
    "$$  E[u_{i}|X_{j}]=0 $$\n",
    "\n",
    "* Esto se daba por cumplido, como consecuencia de que era una muestra aletaoria.\n",
    "* Lo cual, generalente, ocurre en corte transversal por lo cual cada observación es i.i.d.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "\n",
    "## La imposibilidad de la muestra en series de tiempo\n",
    "\n",
    "En serie de tiempo, nuestro universo es un proceso estocástico:\n",
    "\n",
    "![Corte transversal](slide_figuras/datos_st.png)\n",
    "\n",
    "Estimamos modelos de la forma:\n",
    "\n",
    "$$ y_{t}=\\beta_{0}+\\beta_{1}x_{1t}+\\dots+\\beta_{1}x_{kt}+u_{t} $$\n",
    "\n",
    "* Estamos en presencia de un **proceso estocástico** \n",
    "* En cada momento se observa un posible resultado (o realización) del proceso estocásticos.\n",
    "\n",
    "Esto tiene vaias implicancias:\n",
    "\n",
    "* NO SON INDEPENDIENTES ya que por construcción, viene del mismo proceso.\n",
    "* Por lo tanto, el supuesto de exogeneidad **$E[u_{i}|X_{i}]=0 $** NO ES SUFICIENTE\n",
    "* Requerimos su versión más exigente: (Exogeneidad estricta)\n",
    "$$E[u_{t}|X_{s}]=0 \\qquad \\forall s $$\n",
    "\n",
    "* Este supuesto, generalmente NO SE CUMPLE.\n",
    "* Si se cumpliese, seguiruíamos operando como siempre con modelos de regresión multiple estándar."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**¿Qué hacer entonces?**\n",
    "\n",
    "Reconocer los datos como procesos estocásticos e incluir sus particularidades en la modelación. De eso se tratarán los dos siguientes bloques\n",
    "- Bloque 2: Peculiaridades de los procesos estocasticos y su exploración en la data\n",
    "- Bloque 3: Modelamiento\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Conceptos propios de las series de tiempo\n",
    "\n",
    "Dado que una serie de tiempo tiene un orden específico, que en si mismo es importante. En base a este orden se suelen crear nuevos indicadores o variables. Revisemos los más comunes:\n",
    "\n",
    "## Notación y transformaciones:\n",
    "\n",
    "**Notación**\n",
    "\n",
    "* Variables de series de tiempo se denominan con sub-índice t para indicar el perído en el tiempo:   $y_t$\n",
    "* El total de periodos se suele referir como T.\n",
    "\n",
    "### Rezagos \n",
    "\n",
    "* Un **rezago** es el valor de la variable en períodos anteriores:   \n",
    "  * Primer rezago: $y_{t-1}$ es el valor 1 período anterior\n",
    "  * Segundo rezago: $y_{t-2}$ es el valor 2 períodos atrás\n",
    "  * j-ésimo rezago:  $y_{t-j}$ es el valor j períodos atrás\n",
    "\n",
    "### Diferencias\n",
    "\n",
    "Una diferencia corresponde al cambio en una variable entre dos periodos específicos y se usa la notación $\\Delta$\n",
    "\n",
    "* Primera diferencia: $$\\Delta y_t = y_t- y_{t-1}$$\n",
    "\n",
    "### Tasas de crecimiento\n",
    "\n",
    "Si calculamos la primera diferencia el logaritmo natural, podemos obtener incorporar la tasa de crecimiento en una regresión:\n",
    "\n",
    "* Primera diferencia en logs:\n",
    "  $$ \\Delta ln(y_{t})=ln(y_{t})-ln(y_{t-1}) $$\n",
    "\n",
    "* Cambio porcentual de $y_t$  entre $t -1$ y $t\\approx  100\\times \\Delta ln(y_{t})$\n",
    "\n",
    "## Tipos de modelos\n",
    "\n",
    " ### Modelo estático \n",
    "\n",
    " Se modela la relación contemporánea entre dos variables, i.e., relación en el mismo momento en el tiempo\n",
    "\n",
    "$$y_{t}=\\beta_{0}+\\beta_{1}z_{t}+u_{t}$$\n",
    "\n",
    "Ejemplos:\n",
    "* Modelo Simple:\n",
    "$$ inflación_{t}=\\beta_{0}+\\beta_{1}desempleo_{t}+u_{t}$$\n",
    "\n",
    "* Modelo múltiple:\n",
    "$$ homicidio_{t}=\\beta_{0}+\\beta_{1}condena_{t}+\\beta_{2}desmepleo_{t}+\\beta_{3}hombres_{t}+u_{t} $$\n",
    "\n",
    " ### Modelo dinámico\n",
    "\n",
    " Se incluyen efectos temporales, que se piensan que tienen que ver con **tendencias**, **inercia** o **estacionalidades**\n",
    "\n",
    " $$ inflación_{t}=\\beta_{0}+\\beta_{1}desempleo_{t}+\\beta_{2} inflación_{t-1} +u_{t}$$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Pandas y series de tiempo\n",
    "\n",
    "* Pandas fue desarrollado originalmente para trabajar con datos financieros.\n",
    "* Como las series temporales son un tipo de datos comunes encontrados en aplicaiones de finanzas, naturalmente pandas tiene muy buen soporte para la mayoria de operaciones comunes.\n",
    "\n",
    "\n",
    "* `pd.to_datetime()` : conviere una serie o valor en un timestamp\n",
    "  * Este formato permite un mejor manejo de las series."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "#pandas remote data access support for calls to the World Bank Indicators API\n",
    "from pandas_datareader import data, wb #instalar conda install pandas-datareader  o  pip installpandas-datareader"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>id</th>\n",
       "      <th>name</th>\n",
       "      <th>unit</th>\n",
       "      <th>source</th>\n",
       "      <th>sourceNote</th>\n",
       "      <th>sourceOrganization</th>\n",
       "      <th>topics</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>716</th>\n",
       "      <td>6.0.GDPpc_constant</td>\n",
       "      <td>GDP per capita, PPP (constant 2011 internation...</td>\n",
       "      <td></td>\n",
       "      <td>LAC Equity Lab</td>\n",
       "      <td>GDP per capita based on purchasing power parit...</td>\n",
       "      <td>b'World Development Indicators (World Bank)'</td>\n",
       "      <td>Economy &amp; Growth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10384</th>\n",
       "      <td>NY.GDP.PCAP.KD</td>\n",
       "      <td>GDP per capita (constant 2015 US$)</td>\n",
       "      <td></td>\n",
       "      <td>World Development Indicators</td>\n",
       "      <td>GDP per capita is gross domestic product divid...</td>\n",
       "      <td>b'World Bank national accounts data, and OECD ...</td>\n",
       "      <td>Economy &amp; Growth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10386</th>\n",
       "      <td>NY.GDP.PCAP.KN</td>\n",
       "      <td>GDP per capita (constant LCU)</td>\n",
       "      <td></td>\n",
       "      <td>World Development Indicators</td>\n",
       "      <td>GDP per capita is gross domestic product divid...</td>\n",
       "      <td>b'World Bank national accounts data, and OECD ...</td>\n",
       "      <td>Economy &amp; Growth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10388</th>\n",
       "      <td>NY.GDP.PCAP.PP.KD</td>\n",
       "      <td>GDP per capita, PPP (constant 2017 internation...</td>\n",
       "      <td></td>\n",
       "      <td>World Development Indicators</td>\n",
       "      <td>GDP per capita based on purchasing power parit...</td>\n",
       "      <td>b'International Comparison Program, World Bank...</td>\n",
       "      <td>Economy &amp; Growth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10389</th>\n",
       "      <td>NY.GDP.PCAP.PP.KD.87</td>\n",
       "      <td>GDP per capita, PPP (constant 1987 internation...</td>\n",
       "      <td></td>\n",
       "      <td>WDI Database Archives</td>\n",
       "      <td></td>\n",
       "      <td>b''</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         id  \\\n",
       "716      6.0.GDPpc_constant   \n",
       "10384        NY.GDP.PCAP.KD   \n",
       "10386        NY.GDP.PCAP.KN   \n",
       "10388     NY.GDP.PCAP.PP.KD   \n",
       "10389  NY.GDP.PCAP.PP.KD.87   \n",
       "\n",
       "                                                    name unit  \\\n",
       "716    GDP per capita, PPP (constant 2011 internation...        \n",
       "10384                 GDP per capita (constant 2015 US$)        \n",
       "10386                      GDP per capita (constant LCU)        \n",
       "10388  GDP per capita, PPP (constant 2017 internation...        \n",
       "10389  GDP per capita, PPP (constant 1987 internation...        \n",
       "\n",
       "                             source  \\\n",
       "716                  LAC Equity Lab   \n",
       "10384  World Development Indicators   \n",
       "10386  World Development Indicators   \n",
       "10388  World Development Indicators   \n",
       "10389         WDI Database Archives   \n",
       "\n",
       "                                              sourceNote  \\\n",
       "716    GDP per capita based on purchasing power parit...   \n",
       "10384  GDP per capita is gross domestic product divid...   \n",
       "10386  GDP per capita is gross domestic product divid...   \n",
       "10388  GDP per capita based on purchasing power parit...   \n",
       "10389                                                      \n",
       "\n",
       "                                      sourceOrganization            topics  \n",
       "716         b'World Development Indicators (World Bank)'  Economy & Growth  \n",
       "10384  b'World Bank national accounts data, and OECD ...  Economy & Growth  \n",
       "10386  b'World Bank national accounts data, and OECD ...  Economy & Growth  \n",
       "10388  b'International Comparison Program, World Bank...  Economy & Growth  \n",
       "10389                                                b''                    "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Revisemos que indicadores hay disponibles. En este caso revisare de PIB (GDP en ingés), pero se pueden explorar muchas más opciones.\n",
    "\n",
    "wb.search('gdp.*capita.*const')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "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>iso3c</th>\n",
       "      <th>iso2c</th>\n",
       "      <th>name</th>\n",
       "      <th>region</th>\n",
       "      <th>adminregion</th>\n",
       "      <th>incomeLevel</th>\n",
       "      <th>lendingType</th>\n",
       "      <th>capitalCity</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ABW</td>\n",
       "      <td>AW</td>\n",
       "      <td>Aruba</td>\n",
       "      <td>Latin America &amp; Caribbean</td>\n",
       "      <td></td>\n",
       "      <td>High income</td>\n",
       "      <td>Not classified</td>\n",
       "      <td>Oranjestad</td>\n",
       "      <td>-70.0167</td>\n",
       "      <td>12.5167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>AFE</td>\n",
       "      <td>ZH</td>\n",
       "      <td>Africa Eastern and Southern</td>\n",
       "      <td>Aggregates</td>\n",
       "      <td></td>\n",
       "      <td>Aggregates</td>\n",
       "      <td>Aggregates</td>\n",
       "      <td></td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>AFG</td>\n",
       "      <td>AF</td>\n",
       "      <td>Afghanistan</td>\n",
       "      <td>South Asia</td>\n",
       "      <td>South Asia</td>\n",
       "      <td>Low income</td>\n",
       "      <td>IDA</td>\n",
       "      <td>Kabul</td>\n",
       "      <td>69.1761</td>\n",
       "      <td>34.5228</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>AFR</td>\n",
       "      <td>A9</td>\n",
       "      <td>Africa</td>\n",
       "      <td>Aggregates</td>\n",
       "      <td></td>\n",
       "      <td>Aggregates</td>\n",
       "      <td>Aggregates</td>\n",
       "      <td></td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>AFW</td>\n",
       "      <td>ZI</td>\n",
       "      <td>Africa Western and Central</td>\n",
       "      <td>Aggregates</td>\n",
       "      <td></td>\n",
       "      <td>Aggregates</td>\n",
       "      <td>Aggregates</td>\n",
       "      <td></td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  iso3c iso2c                         name                      region  \\\n",
       "0   ABW    AW                        Aruba  Latin America & Caribbean    \n",
       "1   AFE    ZH  Africa Eastern and Southern                  Aggregates   \n",
       "2   AFG    AF                  Afghanistan                  South Asia   \n",
       "3   AFR    A9                       Africa                  Aggregates   \n",
       "4   AFW    ZI   Africa Western and Central                  Aggregates   \n",
       "\n",
       "  adminregion  incomeLevel     lendingType capitalCity  longitude  latitude  \n",
       "0              High income  Not classified  Oranjestad   -70.0167   12.5167  \n",
       "1               Aggregates      Aggregates                    NaN       NaN  \n",
       "2  South Asia   Low income             IDA       Kabul    69.1761   34.5228  \n",
       "3               Aggregates      Aggregates                    NaN       NaN  \n",
       "4               Aggregates      Aggregates                    NaN       NaN  "
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Obtengamos la lista de paises disponibles\n",
    "countries=wb.get_countries()\n",
    "\n",
    "#Preview primeras filas lista de paises\n",
    "countries[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "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>iso3c</th>\n",
       "      <th>iso2c</th>\n",
       "      <th>name</th>\n",
       "      <th>region</th>\n",
       "      <th>adminregion</th>\n",
       "      <th>incomeLevel</th>\n",
       "      <th>lendingType</th>\n",
       "      <th>capitalCity</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>CHL</td>\n",
       "      <td>CL</td>\n",
       "      <td>Chile</td>\n",
       "      <td>Latin America &amp; Caribbean</td>\n",
       "      <td></td>\n",
       "      <td>High income</td>\n",
       "      <td>IBRD</td>\n",
       "      <td>Santiago</td>\n",
       "      <td>-70.6475</td>\n",
       "      <td>-33.475</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   iso3c iso2c   name                      region adminregion  incomeLevel  \\\n",
       "49   CHL    CL  Chile  Latin America & Caribbean               High income   \n",
       "\n",
       "   lendingType capitalCity  longitude  latitude  \n",
       "49        IBRD    Santiago   -70.6475   -33.475  "
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#sabemos que queremos Chile, asi que busquemos su info\n",
    "\n",
    "countries[ countries['name'] == 'Chile' ]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "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></th>\n",
       "      <th>NY.GDP.PCAP.KD</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>country</th>\n",
       "      <th>year</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"10\" valign=\"top\">Chile</th>\n",
       "      <th>2020</th>\n",
       "      <td>12954.405354</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019</th>\n",
       "      <td>13866.955843</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018</th>\n",
       "      <td>13901.018309</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017</th>\n",
       "      <td>13590.991346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016</th>\n",
       "      <td>13624.678549</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015</th>\n",
       "      <td>13574.169238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014</th>\n",
       "      <td>13425.683275</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013</th>\n",
       "      <td>13333.346834</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2012</th>\n",
       "      <td>12941.029999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2011</th>\n",
       "      <td>12406.413571</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              NY.GDP.PCAP.KD\n",
       "country year                \n",
       "Chile   2020    12954.405354\n",
       "        2019    13866.955843\n",
       "        2018    13901.018309\n",
       "        2017    13590.991346\n",
       "        2016    13624.678549\n",
       "        2015    13574.169238\n",
       "        2014    13425.683275\n",
       "        2013    13333.346834\n",
       "        2012    12941.029999\n",
       "        2011    12406.413571"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Descarguemos la data desde la API del banco muncial a un dataframe\n",
    "\n",
    "df_GPDpc_Chile = wb.download(\n",
    "                    #Use the indicator attribute to identify which indicator or indicators to download\n",
    "                    indicator='NY.GDP.PCAP.KD',\n",
    "                    #Use the country attribute to identify the countries you want data for\n",
    "                    country=['CL'],\n",
    "                    #Identify the first year for which you want the data, as an integer or a string\n",
    "                    start='1980',\n",
    "                    #Identify the last year for which you want the data, as an integer or a string\n",
    "                    end=2020,\n",
    "                )\n",
    "\n",
    "#Veamos el data frame\n",
    "df_GPDpc_Chile.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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></th>\n",
       "      <th>NY.GDP.PCAP.PP.KD</th>\n",
       "      <th>NY.GDP.PCAP.KD</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>country</th>\n",
       "      <th>year</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">Chile</th>\n",
       "      <th>2010</th>\n",
       "      <td>21262.466986</td>\n",
       "      <td>11809.141627</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2009</th>\n",
       "      <td>20298.250925</td>\n",
       "      <td>11273.617502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2008</th>\n",
       "      <td>20840.404214</td>\n",
       "      <td>11574.728609</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              NY.GDP.PCAP.PP.KD  NY.GDP.PCAP.KD\n",
       "country year                                   \n",
       "Chile   2010       21262.466986    11809.141627\n",
       "        2009       20298.250925    11273.617502\n",
       "        2008       20840.404214    11574.728609"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "wb.download( indicator=['NY.GDP.PCAP.PP.KD','NY.GDP.PCAP.KD'], country=['CL'], start=2008, end=2010 )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='year', ylabel='NY.GDP.PCAP.KD'>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.scatterplot(data=df_GPDpc_Chile, x=\"year\", y=\"NY.GDP.PCAP.KD\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax = df_GPDpc_Chile.plot(legend=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Ejemplos\n",
    "\n",
    "## 1. PIB de EEUU\n",
    "\n",
    "Empecemos con un ejemplo muy clásico de series de tiempo, con datos del PIB de Estados Unidos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 290 entries, 0 to 289\n",
      "Data columns (total 2 columns):\n",
      " #   Column  Non-Null Count  Dtype         \n",
      "---  ------  --------------  -----         \n",
      " 0   DATE    290 non-null    datetime64[ns]\n",
      " 1   GDP     290 non-null    float64       \n",
      "dtypes: datetime64[ns](1), float64(1)\n",
      "memory usage: 4.7 KB\n"
     ]
    }
   ],
   "source": [
    "GDP = pd.read_csv('data/GDP.csv', parse_dates=['DATE'])\n",
    "GDP.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "DatetimeIndex: 290 entries, 1947-01-01 to 2019-04-01\n",
      "Data columns (total 1 columns):\n",
      " #   Column  Non-Null Count  Dtype  \n",
      "---  ------  --------------  -----  \n",
      " 0   GDP     290 non-null    float64\n",
      "dtypes: float64(1)\n",
      "memory usage: 4.5 KB\n"
     ]
    }
   ],
   "source": [
    "GDP.set_index('DATE', inplace=True)\n",
    "GDP.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "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>GDP</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DATE</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1947-01-01</th>\n",
       "      <td>2033.061</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1947-04-01</th>\n",
       "      <td>2027.639</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1947-07-01</th>\n",
       "      <td>2023.452</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1947-10-01</th>\n",
       "      <td>2055.103</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1948-01-01</th>\n",
       "      <td>2086.017</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 GDP\n",
       "DATE                \n",
       "1947-01-01  2033.061\n",
       "1947-04-01  2027.639\n",
       "1947-07-01  2023.452\n",
       "1947-10-01  2055.103\n",
       "1948-01-01  2086.017"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "GDP.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'GDP ($\\\\$B$)')"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax = GDP['1980':].plot(legend=False)\n",
    "ax.set_ylabel(r'GDP ($\\$B$)')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Observamos claramente una **tendencia** al alza."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Mortalidad por influenza\n",
    "\n",
    "Vemos un segundo ejemplo "
   ]
  },
  {
   "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>Year</th>\n",
       "      <th>Week</th>\n",
       "      <th>Percent of Deaths Due to Pneumonia and Influenza</th>\n",
       "      <th>Expected</th>\n",
       "      <th>Threshold</th>\n",
       "      <th>All Deaths</th>\n",
       "      <th>Pneumonia Deaths</th>\n",
       "      <th>Influenza Deaths</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2012</td>\n",
       "      <td>1</td>\n",
       "      <td>8.479120</td>\n",
       "      <td>8.15718</td>\n",
       "      <td>8.49104</td>\n",
       "      <td>51102</td>\n",
       "      <td>4323</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2012</td>\n",
       "      <td>2</td>\n",
       "      <td>8.343472</td>\n",
       "      <td>8.22181</td>\n",
       "      <td>8.55556</td>\n",
       "      <td>50962</td>\n",
       "      <td>4245</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2012</td>\n",
       "      <td>3</td>\n",
       "      <td>8.370908</td>\n",
       "      <td>8.27534</td>\n",
       "      <td>8.60898</td>\n",
       "      <td>51010</td>\n",
       "      <td>4261</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2012</td>\n",
       "      <td>4</td>\n",
       "      <td>8.448458</td>\n",
       "      <td>8.31696</td>\n",
       "      <td>8.65049</td>\n",
       "      <td>50163</td>\n",
       "      <td>4227</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2012</td>\n",
       "      <td>5</td>\n",
       "      <td>8.140332</td>\n",
       "      <td>8.34602</td>\n",
       "      <td>8.67945</td>\n",
       "      <td>49568</td>\n",
       "      <td>4026</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Year  Week  Percent of Deaths Due to Pneumonia and Influenza  Expected  \\\n",
       "0  2012     1                                          8.479120   8.15718   \n",
       "1  2012     2                                          8.343472   8.22181   \n",
       "2  2012     3                                          8.370908   8.27534   \n",
       "3  2012     4                                          8.448458   8.31696   \n",
       "4  2012     5                                          8.140332   8.34602   \n",
       "\n",
       "   Threshold  All Deaths  Pneumonia Deaths  Influenza Deaths  \n",
       "0    8.49104       51102              4323                10  \n",
       "1    8.55556       50962              4245                 7  \n",
       "2    8.60898       51010              4261                 9  \n",
       "3    8.65049       50163              4227                11  \n",
       "4    8.67945       49568              4026                 9  "
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ILI = pd.read_csv('data/CDC.csv')\n",
    "ILI.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "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>Year</th>\n",
       "      <th>Week</th>\n",
       "      <th>Percent of Deaths Due to Pneumonia and Influenza</th>\n",
       "      <th>Expected</th>\n",
       "      <th>Threshold</th>\n",
       "      <th>All Deaths</th>\n",
       "      <th>Pneumonia Deaths</th>\n",
       "      <th>Influenza Deaths</th>\n",
       "      <th>date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2012</td>\n",
       "      <td>1</td>\n",
       "      <td>8.479120</td>\n",
       "      <td>8.15718</td>\n",
       "      <td>8.49104</td>\n",
       "      <td>51102</td>\n",
       "      <td>4323</td>\n",
       "      <td>10</td>\n",
       "      <td>2012.019231</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2012</td>\n",
       "      <td>2</td>\n",
       "      <td>8.343472</td>\n",
       "      <td>8.22181</td>\n",
       "      <td>8.55556</td>\n",
       "      <td>50962</td>\n",
       "      <td>4245</td>\n",
       "      <td>7</td>\n",
       "      <td>2012.038462</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2012</td>\n",
       "      <td>3</td>\n",
       "      <td>8.370908</td>\n",
       "      <td>8.27534</td>\n",
       "      <td>8.60898</td>\n",
       "      <td>51010</td>\n",
       "      <td>4261</td>\n",
       "      <td>9</td>\n",
       "      <td>2012.057692</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2012</td>\n",
       "      <td>4</td>\n",
       "      <td>8.448458</td>\n",
       "      <td>8.31696</td>\n",
       "      <td>8.65049</td>\n",
       "      <td>50163</td>\n",
       "      <td>4227</td>\n",
       "      <td>11</td>\n",
       "      <td>2012.076923</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2012</td>\n",
       "      <td>5</td>\n",
       "      <td>8.140332</td>\n",
       "      <td>8.34602</td>\n",
       "      <td>8.67945</td>\n",
       "      <td>49568</td>\n",
       "      <td>4026</td>\n",
       "      <td>9</td>\n",
       "      <td>2012.096154</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Year  Week  Percent of Deaths Due to Pneumonia and Influenza  Expected  \\\n",
       "0  2012     1                                          8.479120   8.15718   \n",
       "1  2012     2                                          8.343472   8.22181   \n",
       "2  2012     3                                          8.370908   8.27534   \n",
       "3  2012     4                                          8.448458   8.31696   \n",
       "4  2012     5                                          8.140332   8.34602   \n",
       "\n",
       "   Threshold  All Deaths  Pneumonia Deaths  Influenza Deaths         date  \n",
       "0    8.49104       51102              4323                10  2012.019231  \n",
       "1    8.55556       50962              4245                 7  2012.038462  \n",
       "2    8.60898       51010              4261                 9  2012.057692  \n",
       "3    8.65049       50163              4227                11  2012.076923  \n",
       "4    8.67945       49568              4026                 9  2012.096154  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ILI['date'] = ILI['Year']+ILI['Week']/52.\n",
    "ILI.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, '% Mortality')"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ILI.plot(x='date', y=['Percent of Deaths Due to Pneumonia and Influenza', 'Expected', 'Threshold'])\n",
    "ax = plt.gca()\n",
    "ax.legend(['Mortality', 'Expected', 'Threshold'])\n",
    "ax.set_xlabel('Date')\n",
    "ax.set_ylabel('% Mortality')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "El comportamiento es muy diferente, podemos ver una **estacionalidad**.\n",
    "\n",
    "Es decir hay un cambio periodico asociado al tiempo de la variable."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Transformaciones\n",
    "\n",
    "En base a estos conceptos, entonces, es muy comun realizar transformaciones a las series de tiempo ya sea para obtener rezagos, diferencias, tasas de crecimiento, etc.\n",
    "Ilustraremos varios de estos, con el dataset del DOw-Jonses indsutrial Average.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'Date')"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "DJIA = pd.read_csv('data/DJIA.csv', parse_dates=['DATE'], na_values='.').dropna()\n",
    "DJIA.plot(x='DATE', legend=False)\n",
    "ax = plt.gca()\n",
    "ax.set_ylabel('DJIA')\n",
    "ax.set_xlabel('Date')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "\n",
    "## Diferencias\n",
    "\n",
    "Una manera de remover tendencias de un dataset es diferenciandolo. Como nuestros datasets son discretos, usaremos diferencias finitas."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "def differentiate(values, d=1):\n",
    "    # First value is required so that we can recover the original values with np.cumsum\n",
    "    x = np.concatenate([[values[0]], values[1:]-values[:-1]])\n",
    "\n",
    "    if d == 1:\n",
    "        return x\n",
    "    else:    \n",
    "        return difference(x, d - 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "values = DJIA['DJIA'].values\n",
    "differences = differentiate(values)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Como podemos ver, el grafico se ve mucho más estacionario."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Differences')"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(DJIA['DATE'].iloc[1:], differences[1:])\n",
    "plt.xlabel('Date')\n",
    "plt.ylabel('Differences')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Para recuperar la data original, basta con integrar los puntos diferenciados. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "def integrate(values, d=1):\n",
    "    x = np.cumsum(values)\n",
    "    \n",
    "    if d == 1:\n",
    "        return x\n",
    "    else:\n",
    "        return integrate(x, d-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "rebuilt = integrate(differences)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Y un chequeo rapido para ver que son los mismos valores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(rebuilt-values)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Windowing\n",
    "\n",
    "Muchas veces queremos calcular variables de cuerta cantidad en una ventana de datos.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "def rolling(x, order):\n",
    "    npoints = x.shape[0]\n",
    "    running = []\n",
    "    \n",
    "    for i in range(npoints-order+1):\n",
    "        running.append(x[i:i+order])\n",
    "        \n",
    "    return np.array(running)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0,  1,  2,  3,  4,  5],\n",
       "       [ 1,  2,  3,  4,  5,  6],\n",
       "       [ 2,  3,  4,  5,  6,  7],\n",
       "       [ 3,  4,  5,  6,  7,  8],\n",
       "       [ 4,  5,  6,  7,  8,  9],\n",
       "       [ 5,  6,  7,  8,  9, 10]])"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#And a simple example\n",
    "values = np.arange(11)\n",
    "values\n",
    "rolling(values, 6)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "Como tenemos un array con todas las ventanas, esto facilita el calculo de medias moviles o otras variables que requieran calculo iterativo en ciertos intervalos."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.5, 1.5, 2.5, 3.5, 4.5, 5.5, 6.5, 7.5, 8.5, 9.5])"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "rolling(values, 2)\n",
    "rolling(values, 2).mean(axis=1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 1,  2,  3,  4,  5,  6,  7,  8,  9, 10])"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Or the running maximum, etc.\n",
    "\n",
    "rolling(values, 2).max(axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Exponential Smoothing\n",
    "Another form of smoothing a noise time series is called exponential smoothing. This is equivalent to an exponentially weighted running average where past values get exponentially reduced.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def ES(values, alpha= 0.05):\n",
    "    N = len(values)\n",
    "    S = [values[0]*alpha]\n",
    "    \n",
    "    for i in range(1, N):\n",
    "        S.append(alpha*values[i]+(1-alpha)*S[-1])\n",
    "        \n",
    "    return np.array(S)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Podemos ver rapidamente, que mientras menor es el valor del alpha, menos ruidosa (o más suave) es el resultado."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fa4f922cac0>"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "smooth = []\n",
    "smooth.append(ES(differences[1:], 0.01))\n",
    "smooth.append(ES(differences[1:], 0.1))\n",
    "smooth.append(ES(differences[1:], 0.5))\n",
    "plt.plot(DJIA['DATE'].iloc[1:100], differences[1:100], label='Differences')\n",
    "plt.plot(DJIA['DATE'].iloc[1:100], smooth[2][:99], label=r'$\\alpha=0.5$')\n",
    "plt.plot(DJIA['DATE'].iloc[1:100], smooth[1][:99], label=r'$\\alpha=0.1$')\n",
    "plt.plot(DJIA['DATE'].iloc[1:100], smooth[0][:99], label=r'$\\alpha=0.01$')\n",
    "plt.xlabel('Date')\n",
    "plt.ylabel('Differences')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Missing Data\n",
    "\n",
    "Desafortunadamente, los datos no siempre están limpios o completos, lo que nos obliga a lidiar datos faltantes. Aquí ilustramos varios enfoques para introducir valores perdidos. Comenzamos generando un conjunto de datos con valores perdidos.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = np.linspace(-np.pi, np.pi, 100)\n",
    "y = np.cos(x)\n",
    "y_missing = y.copy()\n",
    "y_missing[40:55] = np.nan\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is simply a cosine function with a few missing values at the peak.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fa509eed1f0>]"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(x, y, '*')\n",
    "plt.plot(x, y_missing)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "Quizás la estrategia más común es simplemente mantener el último valor 'bueno' conocido y usarlo para completar los puntos de datos faltantes. Este enfoque no puede lidiar con los valores faltantes al comienzo del conjunto de datos. \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "def ffill(y):\n",
    "    y0 = y.copy()\n",
    "    N = len(y0)\n",
    "    \n",
    "    current = None\n",
    "    for i in range(1, N):\n",
    "        if np.isnan(y0[i]):\n",
    "            y0[i] = current\n",
    "        else:\n",
    "            current = y0[i]\n",
    "    \n",
    "    return y0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "Naturalmente, el enfoque opuesto también es común cuando usamos el siguiente valor bueno. De esta manera podemos manejar fácilmente los valores perdidos iniciales, pero no podemos hacer nada con los valores perdidos al final de la serie de tiempo.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def bfill(y):\n",
    "    y0 = y.copy()\n",
    "    N = len(y0)\n",
    "    \n",
    "    current = None\n",
    "    for i in range(N-1, 0, -1):\n",
    "        if np.isnan(y0[i]):\n",
    "            y0[i] = current\n",
    "        else:\n",
    "            current = y0[i]\n",
    "    \n",
    "    return y0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "Back-fill y Forward-fill son enfoques simples pero poderosos para lidiar con los datos faltantes. Sin embargo, a menudo queremos tener más cuidado con el valor que atribuimos. Un enfoque común es interpolar entre el valor anterior y el siguiente y conectarlos con una línea recta.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "def interpolate(y):\n",
    "    y0 = y.copy()\n",
    "    N = len(y0)\n",
    "    \n",
    "    pos = 0\n",
    "    while pos < N:\n",
    "        if np.isnan(y0[pos]):\n",
    "            count = 0\n",
    "            \n",
    "            while np.isnan(y0[pos+count]):\n",
    "                count += 1\n",
    "            \n",
    "            current = y0[pos-1]\n",
    "            future = y0[pos+count]\n",
    "            slope = (future-current)/count\n",
    "            \n",
    "            y0[pos:pos+count] = current + np.arange(1, count+1)*slope\n",
    "            \n",
    "            pos += count\n",
    "        else:\n",
    "            pos += 1\n",
    "            \n",
    "    return y0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "La imputación de datos (el cálculo de los valores perdidos esperados) es un gran subcampo de estadísticas con una amplia gama de técnicas y enfoques. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "y_bfill = bfill(y_missing)\n",
    "y_ffill = ffill(y_missing)\n",
    "y_inter = interpolate(y_missing)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Y graficamos para ver las diferencias."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fa4f9e5bc10>"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "\n",
    "plt.plot(x, y_bfill, label='back fill')\n",
    "plt.plot(x, y_ffill, label='forward fill')\n",
    "plt.plot(x, y_inter, label='interpolate')\n",
    "plt.plot(x, y_missing, label='Data')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Resampling\n",
    "\n",
    "En muchos casos, también necesitamos cambiar la frecuencia a la que estamos operando. Por ejemplo, nuestro conjunto de datos DJIA tiene valores al final del día, pero podríamos estar interesados en puntos de datos semanales o mensuales. El remuestreo es una serie de técnicas diseñadas para lidiar con esta situación y es similar en espíritu a las técnicas de ventanas que vimos anteriormente. La principal diferencia es que en lugar de simplemente mover la ventana en un paso fijo, cada ventana corresponde a nuestro período de interés.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "mapping = DJIA['DATE'].dt.year\n",
    "values = DJIA['DJIA'].values\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "En el caso más simple, simplemente calculamos cuál es la ventana correcta para cada punto de datos y la agregamos en consecuencia."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def groupBy(values, mapping, func = None):\n",
    "    agg = {}\n",
    "    pos = {}\n",
    "    \n",
    "    for i in range(values.shape[0]):\n",
    "        key = mapping.iloc[i]\n",
    "        \n",
    "        if key not in agg:\n",
    "            agg[key] = []\n",
    "        \n",
    "        pos[key] = i\n",
    "        \n",
    "        if not np.isnan(values[i]):\n",
    "            agg[key].append(values[i])\n",
    "        \n",
    "    order = sorted(agg.keys())\n",
    "    \n",
    "    if func is not None:\n",
    "        for key in agg:\n",
    "            agg[key] = func(np.array(agg[key]).astype('float'))\n",
    "            \n",
    "    return agg, pos\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Naturalmente, esta función groupBy es útil no solo para remuestrear sino también para una amplia gama de análisis estadísticos. Además de un mapeo, también debemos especificar qué función de agregación queremos usar. ¿Nos interesa el valor medio? ¿el maximo? ¿Desviación Estándar?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "agg, pos = groupBy(values, mapping, np.mean)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Aca simplemente calculamos la media por año"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{2009: 10232.334385964914,\n",
       " 2010: 10668.589087301589,\n",
       " 2011: 11957.570238095239,\n",
       " 2012: 12965.28744,\n",
       " 2013: 15009.523492063492,\n",
       " 2014: 16777.690912698414,\n",
       " 2015: 17587.029166666664,\n",
       " 2016: 17927.107341269842,\n",
       " 2017: 21750.20374501992,\n",
       " 2018: 25046.85734939759,\n",
       " 2019: 26000.832040816327}"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "agg"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Como nuestra función groupBy también devuelve las posiciones de índice de la última vez que se vio cada bin, podemos comparar fácilmente los datos originales con los muestreados nuevamente."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'Date')"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "aggregated = []\n",
    "\n",
    "for key in pos:\n",
    "    aggregated.append([pos[key], agg[key]])\n",
    "\n",
    "aggregated = np.array(aggregated)\n",
    "aggregated\n",
    "plt.plot(DJIA['DATE'], DJIA['DJIA'])\n",
    "ax = plt.gca()\n",
    "ax.plot(DJIA.set_index('DATE').index[aggregated.T[0].astype('int')], aggregated.T[1])\n",
    "ax.plot(DJIA.set_index('DATE').index[aggregated.T[0].astype('int')], aggregated.T[1], 'ro', markersize=10,)\n",
    "ax.set_ylabel('DJIA')\n",
    "ax.set_xlabel('Date')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Jackknife estimators\n",
    "\n",
    "Finalmente, en muchos casos queremos estimar cantidades estadísticas de series de tiempo. Un ejemplo obvio podría ser estimar la media móvil de una serie sin tendencia para verificar si de hecho está lo suficientemente cerca de cero para ser considerada estacionaria.\n",
    "\n",
    "El estimador JackKnife nos permite obtener no solo el valor esperado en cuestión, sino también una medida de su varianza. Esto se logra mediante el uso de un enfoque de omisión para calcular N estimaciones de nuestra métrica. A partir de esta población de estimación, podemos obtener los promedios como la mejor estimación posible y la desviación estándar como una medida de las barras de error involucradas.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "def jackknife(x, func, variance = False):\n",
    "    N = len(x)\n",
    "    pos = np.arange(N)\n",
    "    values = [func(x[pos != i]) for i in pos]\n",
    "    jack = np.sum(values)/N\n",
    "    \n",
    "    if variance:\n",
    "        values = [np.power(func(x[pos != i]) - jack, 2.0) for i in pos]\n",
    "        var = (N-1)/N * np.sum(values)\n",
    "        return jack, var\n",
    "    else:\n",
    "        return jack"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.8994140828534198\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(1.8992558510791335, 0.023063797890560296)"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x = np.random.normal(0, 2, 100)\n",
    "print(x.std())\n",
    "jackknife(x, np.std, True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "Con jackknife obtenemos no solo una estimación del valor sino también una medida del error"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Bootstrapping\n",
    "\n",
    "Otra técnica común para estimar propiedades estadísticas se conoce como bootstrapping y está estrechamente relacionada con Jackknife. En este enfoque, simplemente tomamos muestras (con reemplazo) de la población original para obtener una medida de cuánta variabilidad se puede esperar.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "def bootstrapping(x, n_samples, func=np.mean):\n",
    "    y = x.copy()\n",
    "    N = len(y)\n",
    "    population = []\n",
    "    \n",
    "    for i in range(n_samples):\n",
    "        population.append(func(np.random.choice(y, N, replace=True)))\n",
    "        \n",
    "    return np.array(population)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "Podemos generar fácilmente un histograma de las muestras bostrappeadas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "def histogram(values, n_bins=100):\n",
    "    xmax = values.max()\n",
    "    xmin = values.min()\n",
    "    delta  = (xmax-xmin)/n_bins\n",
    "    \n",
    "    counts = np.zeros(n_bins+1, dtype='int')\n",
    "    \n",
    "    for value in values:\n",
    "        val_bin = np.around((value-xmin)/delta).astype('int')\n",
    "        counts[val_bin] += 1.0\n",
    "    \n",
    "    bins = xmin+delta*np.arange(n_bins+1)\n",
    "    \n",
    "    return bins, counts/values.shape[0]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "  \n",
    "x = np.random.normal(0, 2, size=100)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "boot = bootstrapping(x, 1000)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-0.22862039068812348"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "x.mean()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.7922436343752959"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x.std()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "bins, counts = histogram(boot)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.LineCollection at 0x7fa52adc44c0>"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(bins, counts)\n",
    "plt.vlines(x=boot.mean(), ymin=0, ymax=counts.max(), label='mean', color='g')\n",
    "plt.vlines(x=boot.mean()+boot.std(), ymin=0, ymax=counts.max(), label='std', linestyles='--', color='r')\n",
    "plt.vlines(x=boot.mean()-boot.std(), ymin=0, ymax=counts.max(), label='std', linestyles='--', color='r')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Actividad Bloque 1:\n",
    "\n",
    "1. Siga este ejemplo practico de importar datos desde la API del Banco Mundial y preparar la base para su análisis de series de tiempo.\n",
    "2. Importe usted la serie de GDP total Y Percapita para otro país serie desde la API del Banco mundial, muestre sus principales características y realice un grafico. \n",
    "3. Obtenga las primeras diferencias y compare, ¿pareciera haber tendencias?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "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>id</th>\n",
       "      <th>name</th>\n",
       "      <th>unit</th>\n",
       "      <th>source</th>\n",
       "      <th>sourceNote</th>\n",
       "      <th>sourceOrganization</th>\n",
       "      <th>topics</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>711</th>\n",
       "      <td>5.51.01.10.gdp</td>\n",
       "      <td>Per capita GDP growth</td>\n",
       "      <td></td>\n",
       "      <td>Statistical Capacity Indicators</td>\n",
       "      <td>GDP per capita is the sum of gross value added...</td>\n",
       "      <td>b'World Development Indicator (WDI) databank. ...</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>713</th>\n",
       "      <td>6.0.GDP_current</td>\n",
       "      <td>GDP (current $)</td>\n",
       "      <td></td>\n",
       "      <td>LAC Equity Lab</td>\n",
       "      <td>GDP is the sum of gross value added by all res...</td>\n",
       "      <td>b'World Development Indicators (World Bank)'</td>\n",
       "      <td>Economy &amp; Growth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>714</th>\n",
       "      <td>6.0.GDP_growth</td>\n",
       "      <td>GDP growth (annual %)</td>\n",
       "      <td></td>\n",
       "      <td>LAC Equity Lab</td>\n",
       "      <td>Annual percentage growth rate of GDP at market...</td>\n",
       "      <td>b'World Development Indicators (World Bank)'</td>\n",
       "      <td>Economy &amp; Growth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>715</th>\n",
       "      <td>6.0.GDP_usd</td>\n",
       "      <td>GDP (constant 2005 $)</td>\n",
       "      <td></td>\n",
       "      <td>LAC Equity Lab</td>\n",
       "      <td>GDP is the sum of gross value added by all res...</td>\n",
       "      <td>b'World Development Indicators (World Bank)'</td>\n",
       "      <td>Economy &amp; Growth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>716</th>\n",
       "      <td>6.0.GDPpc_constant</td>\n",
       "      <td>GDP per capita, PPP (constant 2011 internation...</td>\n",
       "      <td></td>\n",
       "      <td>LAC Equity Lab</td>\n",
       "      <td>GDP per capita based on purchasing power parit...</td>\n",
       "      <td>b'World Development Indicators (World Bank)'</td>\n",
       "      <td>Economy &amp; Growth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16297</th>\n",
       "      <td>UIS.XUNIT.GDPCAP.23.FSGOV</td>\n",
       "      <td>Initial government funding per secondary stude...</td>\n",
       "      <td></td>\n",
       "      <td>Education Statistics</td>\n",
       "      <td>Total general (local, regional and central, cu...</td>\n",
       "      <td>b'UNESCO Institute for Statistics'</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16298</th>\n",
       "      <td>UIS.XUNIT.GDPCAP.23.FSHH</td>\n",
       "      <td>Initial household funding per secondary studen...</td>\n",
       "      <td></td>\n",
       "      <td>Education Statistics</td>\n",
       "      <td>Total payments of households (pupils, students...</td>\n",
       "      <td>b'UNESCO Institute for Statistics'</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16299</th>\n",
       "      <td>UIS.XUNIT.GDPCAP.3.FSGOV</td>\n",
       "      <td>Initial government funding per upper secondary...</td>\n",
       "      <td></td>\n",
       "      <td>Education Statistics</td>\n",
       "      <td>Total general (local, regional and central, cu...</td>\n",
       "      <td>b'UNESCO Institute for Statistics'</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16300</th>\n",
       "      <td>UIS.XUNIT.GDPCAP.5T8.FSGOV</td>\n",
       "      <td>Initial government funding per tertiary studen...</td>\n",
       "      <td></td>\n",
       "      <td>Education Statistics</td>\n",
       "      <td>Total general (local, regional and central, cu...</td>\n",
       "      <td>b'UNESCO Institute for Statistics'</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16301</th>\n",
       "      <td>UIS.XUNIT.GDPCAP.5T8.FSHH</td>\n",
       "      <td>Initial household funding per tertiary student...</td>\n",
       "      <td></td>\n",
       "      <td>Education Statistics</td>\n",
       "      <td>Total payments of households (pupils, students...</td>\n",
       "      <td>b'UNESCO Institute for Statistics'</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>542 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                               id  \\\n",
       "711                5.51.01.10.gdp   \n",
       "713               6.0.GDP_current   \n",
       "714                6.0.GDP_growth   \n",
       "715                   6.0.GDP_usd   \n",
       "716            6.0.GDPpc_constant   \n",
       "...                           ...   \n",
       "16297   UIS.XUNIT.GDPCAP.23.FSGOV   \n",
       "16298    UIS.XUNIT.GDPCAP.23.FSHH   \n",
       "16299    UIS.XUNIT.GDPCAP.3.FSGOV   \n",
       "16300  UIS.XUNIT.GDPCAP.5T8.FSGOV   \n",
       "16301   UIS.XUNIT.GDPCAP.5T8.FSHH   \n",
       "\n",
       "                                                    name unit  \\\n",
       "711                                Per capita GDP growth        \n",
       "713                                      GDP (current $)        \n",
       "714                                GDP growth (annual %)        \n",
       "715                                GDP (constant 2005 $)        \n",
       "716    GDP per capita, PPP (constant 2011 internation...        \n",
       "...                                                  ...  ...   \n",
       "16297  Initial government funding per secondary stude...        \n",
       "16298  Initial household funding per secondary studen...        \n",
       "16299  Initial government funding per upper secondary...        \n",
       "16300  Initial government funding per tertiary studen...        \n",
       "16301  Initial household funding per tertiary student...        \n",
       "\n",
       "                                source  \\\n",
       "711    Statistical Capacity Indicators   \n",
       "713                     LAC Equity Lab   \n",
       "714                     LAC Equity Lab   \n",
       "715                     LAC Equity Lab   \n",
       "716                     LAC Equity Lab   \n",
       "...                                ...   \n",
       "16297             Education Statistics   \n",
       "16298             Education Statistics   \n",
       "16299             Education Statistics   \n",
       "16300             Education Statistics   \n",
       "16301             Education Statistics   \n",
       "\n",
       "                                              sourceNote  \\\n",
       "711    GDP per capita is the sum of gross value added...   \n",
       "713    GDP is the sum of gross value added by all res...   \n",
       "714    Annual percentage growth rate of GDP at market...   \n",
       "715    GDP is the sum of gross value added by all res...   \n",
       "716    GDP per capita based on purchasing power parit...   \n",
       "...                                                  ...   \n",
       "16297  Total general (local, regional and central, cu...   \n",
       "16298  Total payments of households (pupils, students...   \n",
       "16299  Total general (local, regional and central, cu...   \n",
       "16300  Total general (local, regional and central, cu...   \n",
       "16301  Total payments of households (pupils, students...   \n",
       "\n",
       "                                      sourceOrganization            topics  \n",
       "711    b'World Development Indicator (WDI) databank. ...                    \n",
       "713         b'World Development Indicators (World Bank)'  Economy & Growth  \n",
       "714         b'World Development Indicators (World Bank)'  Economy & Growth  \n",
       "715         b'World Development Indicators (World Bank)'  Economy & Growth  \n",
       "716         b'World Development Indicators (World Bank)'  Economy & Growth  \n",
       "...                                                  ...               ...  \n",
       "16297                 b'UNESCO Institute for Statistics'                    \n",
       "16298                 b'UNESCO Institute for Statistics'                    \n",
       "16299                 b'UNESCO Institute for Statistics'                    \n",
       "16300                 b'UNESCO Institute for Statistics'                    \n",
       "16301                 b'UNESCO Institute for Statistics'                    \n",
       "\n",
       "[542 rows x 7 columns]"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#pandas remote data access support for calls to the World Bank Indicators API\n",
    "from pandas_datareader import data, wb #instalar conda install pandas-datareader  o  pip installpandas-datareader\n",
    "\n",
    "#Revisemos que indicadores hay disponibles. En este caso revisare de PIB (GDP en ingés), pero se pueden explorar muchas más opciones.\n",
    "\n",
    "wb.search('gdp')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Obtengamos la lista de paises disponibles\n",
    "countries=wb.get_countries()\n",
    "\n",
    "#Preview primeras filas lista de paises\n",
    "countries[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#sabemos que queremos Chile, asi que busquemos su info\n",
    "\n",
    "countries[ countries['name'] == 'Chile' ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Descarguemos la data desde la API del banco muncial a un dataframe\n",
    "\n",
    "df_GPDpc_Chile = wb.download(\n",
    "                    #Use the indicator attribute to identify which indicator or indicators to download\n",
    "                    indicator='NY.GDP.PCAP.KD',\n",
    "                    #Use the country attribute to identify the countries you want data for\n",
    "                    country=['CL'],\n",
    "                    #Identify the first year for which you want the data, as an integer or a string\n",
    "                    start='1980',\n",
    "                    #Identify the last year for which you want the data, as an integer or a string\n",
    "                    end=2020\n",
    "                )\n",
    "\n",
    "#Veamos el data frame\n",
    "df_GPDpc_Chile.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "wb.download( indicator=['NY.GDP.PCAP.PP.KD','NY.GDP.PCAP.KD'], country=['CL'], start=2008, end=2010 )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sns.scatterplot(data=df_GPDpc_Chile, x=\"year\", y=\"NY.GDP.PCAP.KD\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "aee8b7b246df8f9039afb4144a1f6fd8d2ca17a180786b69acc140d282b71a49"
  },
  "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}