{ "cells": [ { "cell_type": "markdown", "id": "5c8decec", "metadata": {}, "source": [ "# Propósito del Modelo\n", "El objetivo es hacer análisis que involucran múltiples variables de diferentes bases de datos combinadas." ] }, { "cell_type": "code", "execution_count": 1, "id": "ff4e01c3", "metadata": {}, "outputs": [ { "data": { "text/plain": [ "( 1960 1961 1962 \\\n", " Country Name Country Code \n", " Aruba ABW -0.201874 -0.200982 -0.200327 \n", " Africa Eastern and Southern AFE -0.094914 -0.098259 -0.094392 \n", " Afghanistan AFG -0.201874 -0.200982 -0.200327 \n", " Africa Western and Central AFW -0.135881 -0.136426 -0.136892 \n", " Angola AGO -0.201874 -0.200982 -0.200327 \n", " \n", " 1963 1964 1965 \\\n", " Country Name Country Code \n", " Aruba ABW -0.199863 -0.200426 -0.202832 \n", " Africa Eastern and Southern AFE -0.094446 -0.097204 -0.095992 \n", " Afghanistan AFG -0.199863 -0.200426 -0.202832 \n", " Africa Western and Central AFW -0.135198 -0.136542 -0.136252 \n", " Angola AGO -0.199863 -0.200426 -0.202832 \n", " \n", " 1966 1967 1968 \\\n", " Country Name Country Code \n", " Aruba ABW -0.202705 -0.202662 -0.203255 \n", " Africa Eastern and Southern AFE -0.097136 -0.094766 -0.097880 \n", " Afghanistan AFG -0.202705 -0.202662 -0.203255 \n", " Africa Western and Central AFW -0.141240 -0.150569 -0.154342 \n", " Angola AGO -0.202705 -0.202662 -0.203255 \n", " \n", " 1969 ... 2013 2014 \\\n", " Country Name Country Code ... \n", " Aruba ABW -0.204786 ... -0.288371 -0.289320 \n", " Africa Eastern and Southern AFE -0.099362 ... -0.173739 -0.173601 \n", " Afghanistan AFG -0.204786 ... -0.286345 -0.287284 \n", " Africa Western and Central AFW -0.149334 ... -0.195458 -0.193785 \n", " Angola AGO -0.204786 ... -0.277472 -0.278209 \n", " \n", " 2015 2016 2017 \\\n", " Country Name Country Code \n", " Aruba ABW -0.289823 -0.290696 -0.291427 \n", " Africa Eastern and Southern AFE -0.174254 -0.175796 -0.177555 \n", " Afghanistan AFG -0.287813 -0.288695 -0.289457 \n", " Africa Western and Central AFW -0.194550 -0.197907 -0.199728 \n", " Angola AGO -0.278940 -0.280400 -0.281517 \n", " \n", " 2018 2019 2020 \\\n", " Country Name Country Code \n", " Aruba ABW -0.292036 -0.292556 -0.292998 \n", " Africa Eastern and Southern AFE -0.179122 -0.180347 -0.180619 \n", " Afghanistan AFG -0.290112 -0.290586 -0.290938 \n", " Africa Western and Central AFW -0.200799 -0.200851 -0.199375 \n", " Angola AGO -0.282588 -0.283415 -0.284049 \n", " \n", " 2021 2022 \n", " Country Name Country Code \n", " Aruba ABW -0.293201 -0.294042 \n", " Africa Eastern and Southern AFE -0.183041 -0.183463 \n", " Afghanistan AFG -0.291793 -0.287261 \n", " Africa Western and Central AFW -0.201723 -0.201925 \n", " Angola AGO -0.284755 -0.285617 \n", " \n", " [5 rows x 63 columns],\n", " 1960 1961 1962 \\\n", " Country Name Country Code \n", " Aruba ABW -0.084868 -0.08484 -0.083201 \n", " Africa Eastern and Southern AFE -0.084868 -0.08484 -0.083201 \n", " Afghanistan AFG -0.084868 -0.08484 -0.083201 \n", " Africa Western and Central AFW -0.084868 -0.08484 -0.083201 \n", " Angola AGO -0.084868 -0.08484 -0.083201 \n", " \n", " 1963 1964 1965 \\\n", " Country Name Country Code \n", " Aruba ABW -0.082048 -0.080368 -0.074607 \n", " Africa Eastern and Southern AFE -0.082048 -0.080368 -0.074607 \n", " Afghanistan AFG -0.082048 -0.080368 -0.074607 \n", " Africa Western and Central AFW -0.082048 -0.080368 -0.074607 \n", " Angola AGO -0.082048 -0.080368 -0.074607 \n", " \n", " 1966 1967 1968 \\\n", " Country Name Country Code \n", " Aruba ABW -0.075705 -0.073737 -0.072911 \n", " Africa Eastern and Southern AFE -0.075705 -0.073737 -0.072911 \n", " Afghanistan AFG -0.075705 -0.073737 -0.072911 \n", " Africa Western and Central AFW -0.075705 -0.073737 -0.072911 \n", " Angola AGO -0.075705 -0.073737 -0.072911 \n", " \n", " 1969 ... 2013 2014 \\\n", " Country Name Country Code ... \n", " Aruba ABW -0.071835 ... -0.105875 -0.107196 \n", " Africa Eastern and Southern AFE -0.071835 ... -0.105875 -0.107196 \n", " Afghanistan AFG -0.071835 ... -0.105875 -0.107196 \n", " Africa Western and Central AFW -0.071835 ... -0.105875 -0.107196 \n", " Angola AGO -0.071835 ... -0.104616 -0.106276 \n", " \n", " 2015 2016 2017 \\\n", " Country Name Country Code \n", " Aruba ABW -0.105721 -0.104515 -0.105551 \n", " Africa Eastern and Southern AFE -0.105721 -0.104515 -0.105551 \n", " Afghanistan AFG -0.105721 -0.104515 -0.105551 \n", " Africa Western and Central AFW -0.105721 -0.104515 -0.105551 \n", " Angola AGO -0.105172 -0.104014 -0.105000 \n", " \n", " 2018 2019 2020 \\\n", " Country Name Country Code \n", " Aruba ABW -0.107831 -0.111312 -0.111963 \n", " Africa Eastern and Southern AFE -0.107831 -0.111312 -0.111963 \n", " Afghanistan AFG -0.107831 -0.111312 -0.111067 \n", " Africa Western and Central AFW -0.107831 -0.111312 -0.111963 \n", " Angola AGO -0.107297 -0.110756 -0.111903 \n", " \n", " 2021 2022 \n", " Country Name Country Code \n", " Aruba ABW -0.110932 -0.111682 \n", " Africa Eastern and Southern AFE -0.110932 -0.111682 \n", " Afghanistan AFG -0.110692 -0.111682 \n", " Africa Western and Central AFW -0.110932 -0.111682 \n", " Angola AGO -0.111045 -0.111847 \n", " \n", " [5 rows x 63 columns],\n", " 1960 1961 1962 \\\n", " Country Name Country Code \n", " Aruba ABW -0.146825 -0.158595 -0.153408 \n", " Africa Eastern and Southern AFE -0.146825 -0.158595 -0.153408 \n", " Afghanistan AFG -0.146825 -0.158595 -0.153408 \n", " Africa Western and Central AFW -0.146825 -0.158595 -0.153408 \n", " Angola AGO -0.146825 -0.158595 -0.153408 \n", " \n", " 1963 1964 1965 \\\n", " Country Name Country Code \n", " Aruba ABW -0.158273 -0.158734 -0.162173 \n", " Africa Eastern and Southern AFE -0.158273 -0.158734 -0.162173 \n", " Afghanistan AFG -0.158273 -0.158734 -0.162173 \n", " Africa Western and Central AFW -0.158273 -0.158734 -0.162173 \n", " Angola AGO -0.158273 -0.158734 -0.162173 \n", " \n", " 1966 1967 1968 \\\n", " Country Name Country Code \n", " Aruba ABW -0.158459 -0.162375 -0.1682 \n", " Africa Eastern and Southern AFE -0.158459 -0.162375 -0.1682 \n", " Afghanistan AFG -0.158459 -0.162375 -0.1682 \n", " Africa Western and Central AFW -0.158459 -0.162375 -0.1682 \n", " Angola AGO -0.158459 -0.162375 -0.1682 \n", " \n", " 1969 ... 2013 2014 \\\n", " Country Name Country Code ... \n", " Aruba ABW -0.184395 ... -0.203738 -0.202930 \n", " Africa Eastern and Southern AFE -0.184395 ... -0.071547 -0.081470 \n", " Afghanistan AFG -0.184395 ... -0.203738 -0.202930 \n", " Africa Western and Central AFW -0.184395 ... -0.203738 -0.202930 \n", " Angola AGO -0.184395 ... -0.194796 -0.193997 \n", " \n", " 2015 2016 2017 \\\n", " Country Name Country Code \n", " Aruba ABW -0.285333 -0.202682 -0.202571 \n", " Africa Eastern and Southern AFE -0.195116 -0.092256 -0.090919 \n", " Afghanistan AFG -0.285333 -0.202682 -0.202571 \n", " Africa Western and Central AFW -0.285333 -0.202682 -0.202571 \n", " Angola AGO -0.280513 -0.198833 -0.198563 \n", " \n", " 2018 2019 2020 \\\n", " Country Name Country Code \n", " Aruba ABW -0.202676 -0.202183 -0.201677 \n", " Africa Eastern and Southern AFE -0.093306 -0.094302 -0.094236 \n", " Afghanistan AFG -0.202676 -0.202183 -0.201677 \n", " Africa Western and Central AFW -0.202676 -0.202183 -0.201677 \n", " Angola AGO -0.200401 -0.201239 -0.199575 \n", " \n", " 2021 2022 \n", " Country Name Country Code \n", " Aruba ABW -0.201440 -0.201034 \n", " Africa Eastern and Southern AFE -0.093547 -0.090401 \n", " Afghanistan AFG -0.201440 -0.201034 \n", " Africa Western and Central AFW -0.201440 -0.201034 \n", " Angola AGO -0.201274 -0.203768 \n", " \n", " [5 rows x 63 columns],\n", " 1960 1961 1962 \\\n", " Country Name Country Code \n", " Aruba ABW -0.090809 -0.091864 -0.09344 \n", " Africa Eastern and Southern AFE -0.090809 -0.091864 -0.09344 \n", " Afghanistan AFG -0.090809 -0.091864 -0.09344 \n", " Africa Western and Central AFW -0.090809 -0.091864 -0.09344 \n", " Angola AGO -0.090809 -0.091864 -0.09344 \n", " \n", " 1963 1964 1965 \\\n", " Country Name Country Code \n", " Aruba ABW -0.093163 -0.094031 -0.092187 \n", " Africa Eastern and Southern AFE -0.093163 -0.094031 -0.092187 \n", " Afghanistan AFG -0.093163 -0.094031 -0.092187 \n", " Africa Western and Central AFW -0.093163 -0.094031 -0.092187 \n", " Angola AGO -0.093163 -0.094031 -0.092187 \n", " \n", " 1966 1967 1968 \\\n", " Country Name Country Code \n", " Aruba ABW -0.090953 -0.089691 -0.090635 \n", " Africa Eastern and Southern AFE -0.090953 -0.089691 -0.090635 \n", " Afghanistan AFG -0.090953 -0.089691 -0.090635 \n", " Africa Western and Central AFW -0.090953 -0.089691 -0.090635 \n", " Angola AGO -0.090953 -0.089691 -0.090635 \n", " \n", " 1969 ... 2013 2014 \\\n", " Country Name Country Code ... \n", " Aruba ABW -0.089882 ... -0.113248 -0.114293 \n", " Africa Eastern and Southern AFE -0.089882 ... -0.113248 -0.114293 \n", " Afghanistan AFG -0.089882 ... -0.113248 -0.114293 \n", " Africa Western and Central AFW -0.089882 ... -0.113248 -0.114293 \n", " Angola AGO -0.089882 ... -0.112900 -0.113945 \n", " \n", " 2015 2016 2017 \\\n", " Country Name Country Code \n", " Aruba ABW -0.115194 -0.114932 -0.115308 \n", " Africa Eastern and Southern AFE -0.115194 -0.114932 -0.115308 \n", " Afghanistan AFG -0.115194 -0.114932 -0.115308 \n", " Africa Western and Central AFW -0.115194 -0.114932 -0.115308 \n", " Angola AGO -0.114668 -0.114465 -0.114863 \n", " \n", " 2018 2019 2020 \\\n", " Country Name Country Code \n", " Aruba ABW -0.116877 -0.119049 -0.118598 \n", " Africa Eastern and Southern AFE -0.116877 -0.119049 -0.118598 \n", " Afghanistan AFG -0.116877 -0.119049 -0.117498 \n", " Africa Western and Central AFW -0.116877 -0.119049 -0.118598 \n", " Angola AGO -0.116601 -0.118793 -0.118486 \n", " \n", " 2021 2022 \n", " Country Name Country Code \n", " Aruba ABW -0.118736 -0.117705 \n", " Africa Eastern and Southern AFE -0.118736 -0.117705 \n", " Afghanistan AFG -0.118341 -0.117705 \n", " Africa Western and Central AFW -0.118736 -0.117705 \n", " Angola AGO -0.118741 -0.118102 \n", " \n", " [5 rows x 63 columns],\n", " 1960 1961 1962 \\\n", " Country Name Country Code \n", " Aruba ABW -0.162675 -0.152507 -0.175345 \n", " Africa Eastern and Southern AFE 2.381755 2.202534 2.959211 \n", " Afghanistan AFG -0.162675 -0.152507 -0.175345 \n", " Africa Western and Central AFW -0.162675 -0.152507 -0.175345 \n", " Angola AGO -0.162675 -0.109791 -0.142530 \n", " \n", " 1963 1964 1965 \\\n", " Country Name Country Code \n", " Aruba ABW -0.199606 -0.207111 -0.257548 \n", " Africa Eastern and Southern AFE 2.810270 2.750210 3.317384 \n", " Afghanistan AFG -0.199606 -0.207111 -0.257548 \n", " Africa Western and Central AFW -0.199606 -0.207111 -0.257548 \n", " Angola AGO -0.165347 -0.144329 -0.250614 \n", " \n", " 1966 1967 1968 \\\n", " Country Name Country Code \n", " Aruba ABW -0.288931 -0.277482 -0.278466 \n", " Africa Eastern and Southern AFE 4.608648 3.940742 3.916744 \n", " Afghanistan AFG -0.288931 -0.277482 -0.278466 \n", " Africa Western and Central AFW -0.288931 -0.277482 0.066949 \n", " Angola AGO -0.261803 -0.277482 -0.278466 \n", " \n", " 1969 ... 2011 2012 \\\n", " Country Name Country Code ... \n", " Aruba ABW -0.270550 ... -0.325080 -0.311868 \n", " Africa Eastern and Southern AFE 4.802118 ... -0.325080 -0.311868 \n", " Afghanistan AFG -0.270550 ... -0.325080 -0.311868 \n", " Africa Western and Central AFW 0.209348 ... 0.935966 0.632917 \n", " Angola AGO -0.270550 ... 0.266913 0.170396 \n", " \n", " 2013 2014 2015 \\\n", " Country Name Country Code \n", " Aruba ABW -0.306099 -0.309704 -0.250354 \n", " Africa Eastern and Southern AFE -0.306099 -0.309704 -0.250354 \n", " Afghanistan AFG -0.306099 -0.309704 -0.250354 \n", " Africa Western and Central AFW 0.353438 0.164271 -0.175233 \n", " Angola AGO 0.224552 0.035088 -0.153382 \n", " \n", " 2016 2017 2018 \\\n", " Country Name Country Code \n", " Aruba ABW -0.252749 -0.257753 -0.291725 \n", " Africa Eastern and Southern AFE -0.252749 -0.257753 -0.291725 \n", " Afghanistan AFG -0.252749 -0.257753 -0.291725 \n", " Africa Western and Central AFW -0.252749 -0.133328 0.042448 \n", " Angola AGO -0.115071 -0.036075 0.042644 \n", " \n", " 2019 2020 \n", " Country Name Country Code \n", " Aruba ABW -0.281673 -0.251303 \n", " Africa Eastern and Southern AFE -0.281673 -0.251303 \n", " Afghanistan AFG -0.281673 -0.251303 \n", " Africa Western and Central AFW -0.051936 -0.239934 \n", " Angola AGO -0.088127 -0.217208 \n", " \n", " [5 rows x 61 columns])" ] }, "execution_count": 1, "metadata": {}, "output_type": "execute_result" } ], "source": [ "import pandas as pd\n", "import numpy as np\n", "import statsmodels.formula.api as smf\n", "import statsmodels.api as sm\n", "import matplotlib.pyplot as plt\n", "import seaborn as sns\n", "\n", "main_gdp = pd.read_csv('main_gdp.csv', index_col=['Country Name', 'Country Code'])\n", "main_government = pd.read_csv('main_government.csv', index_col=['Country Name', 'Country Code'])\n", "main_investments = pd.read_csv('main_investments.csv', index_col=['Country Name', 'Country Code'])\n", "main_consumption = pd.read_csv('main_consumption.csv', index_col=['Country Name', 'Country Code'])\n", "main_trade = pd.read_csv('main_trade.csv', index_col=['Country Name', 'Country Code'])\n", "\n", "(main_gdp.head(), main_government.head(), main_investments.head(), main_consumption.head(), main_trade.head())" ] }, { "cell_type": "markdown", "id": "5c1b62dd", "metadata": {}, "source": [ "#### Unir múltiples DataFrames al mismo tiempo con sufijos especificados" ] }, { "cell_type": "code", "execution_count": 2, "id": "f35b32d2", "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
GDP_calculated_1960GDP_calculated_1961GDP_calculated_1962GDP_calculated_1963GDP_calculated_1964GDP_calculated_1965GDP_calculated_1966GDP_calculated_1967GDP_calculated_1968GDP_calculated_1969...GDP_calculated_2013GDP_calculated_2014GDP_calculated_2015GDP_calculated_2016GDP_calculated_2017GDP_calculated_2018GDP_calculated_2019GDP_calculated_2020GDP_calculated_2021GDP_calculated_2022
Country NameCountry Code
ArubaABW-0.687051-0.688787-0.705721-0.732952-0.740670-0.789346-0.816753-0.805947-0.813468-0.821448...-1.017331-1.023442-1.046426-0.965575-0.972609-1.011146-1.006774-0.976539-0.925748-0.925497
Africa Eastern and SouthernAFE1.9643401.7689772.5347702.3823402.3198732.8924254.1863963.5201733.4871174.356643...-0.770508-0.786264-0.840639-0.740249-0.747085-0.788861-0.786684-0.756719-0.599803-0.593651
AfghanistanAFG-0.687051-0.688787-0.705721-0.732952-0.740670-0.789346-0.816753-0.805947-0.813468-0.821448...-1.015306-1.021407-1.044415-0.963573-0.970639-1.009221-1.004804-0.972484-0.923705-0.918717
Africa Western and CentralAFW-0.621058-0.624231-0.642287-0.668287-0.676786-0.722767-0.755287-0.753854-0.419140-0.286099...-0.264882-0.453932-0.876033-0.872786-0.756485-0.585735-0.685331-0.871547-0.834270-0.833380
AngolaAGO-0.687051-0.646070-0.672906-0.698693-0.677888-0.782412-0.789625-0.805947-0.813468-0.821448...-0.465233-0.657339-0.932675-0.812782-0.736018-0.664243-0.802329-0.931221-0.917089-0.923103
\n", "

5 rows × 63 columns

\n", "
" ], "text/plain": [ " GDP_calculated_1960 \\\n", "Country Name Country Code \n", "Aruba ABW -0.687051 \n", "Africa Eastern and Southern AFE 1.964340 \n", "Afghanistan AFG -0.687051 \n", "Africa Western and Central AFW -0.621058 \n", "Angola AGO -0.687051 \n", "\n", " GDP_calculated_1961 \\\n", "Country Name Country Code \n", "Aruba ABW -0.688787 \n", "Africa Eastern and Southern AFE 1.768977 \n", "Afghanistan AFG -0.688787 \n", "Africa Western and Central AFW -0.624231 \n", "Angola AGO -0.646070 \n", "\n", " GDP_calculated_1962 \\\n", "Country Name Country Code \n", "Aruba ABW -0.705721 \n", "Africa Eastern and Southern AFE 2.534770 \n", "Afghanistan AFG -0.705721 \n", "Africa Western and Central AFW -0.642287 \n", "Angola AGO -0.672906 \n", "\n", " GDP_calculated_1963 \\\n", "Country Name Country Code \n", "Aruba ABW -0.732952 \n", "Africa Eastern and Southern AFE 2.382340 \n", "Afghanistan AFG -0.732952 \n", "Africa Western and Central AFW -0.668287 \n", "Angola AGO -0.698693 \n", "\n", " GDP_calculated_1964 \\\n", "Country Name Country Code \n", "Aruba ABW -0.740670 \n", "Africa Eastern and Southern AFE 2.319873 \n", "Afghanistan AFG -0.740670 \n", "Africa Western and Central AFW -0.676786 \n", "Angola AGO -0.677888 \n", "\n", " GDP_calculated_1965 \\\n", "Country Name Country Code \n", "Aruba ABW -0.789346 \n", "Africa Eastern and Southern AFE 2.892425 \n", "Afghanistan AFG -0.789346 \n", "Africa Western and Central AFW -0.722767 \n", "Angola AGO -0.782412 \n", "\n", " GDP_calculated_1966 \\\n", "Country Name Country Code \n", "Aruba ABW -0.816753 \n", "Africa Eastern and Southern AFE 4.186396 \n", "Afghanistan AFG -0.816753 \n", "Africa Western and Central AFW -0.755287 \n", "Angola AGO -0.789625 \n", "\n", " GDP_calculated_1967 \\\n", "Country Name Country Code \n", "Aruba ABW -0.805947 \n", "Africa Eastern and Southern AFE 3.520173 \n", "Afghanistan AFG -0.805947 \n", "Africa Western and Central AFW -0.753854 \n", "Angola AGO -0.805947 \n", "\n", " GDP_calculated_1968 \\\n", "Country Name Country Code \n", "Aruba ABW -0.813468 \n", "Africa Eastern and Southern AFE 3.487117 \n", "Afghanistan AFG -0.813468 \n", "Africa Western and Central AFW -0.419140 \n", "Angola AGO -0.813468 \n", "\n", " GDP_calculated_1969 ... \\\n", "Country Name Country Code ... \n", "Aruba ABW -0.821448 ... \n", "Africa Eastern and Southern AFE 4.356643 ... \n", "Afghanistan AFG -0.821448 ... \n", "Africa Western and Central AFW -0.286099 ... \n", "Angola AGO -0.821448 ... \n", "\n", " GDP_calculated_2013 \\\n", "Country Name Country Code \n", "Aruba ABW -1.017331 \n", "Africa Eastern and Southern AFE -0.770508 \n", "Afghanistan AFG -1.015306 \n", "Africa Western and Central AFW -0.264882 \n", "Angola AGO -0.465233 \n", "\n", " GDP_calculated_2014 \\\n", "Country Name Country Code \n", "Aruba ABW -1.023442 \n", "Africa Eastern and Southern AFE -0.786264 \n", "Afghanistan AFG -1.021407 \n", "Africa Western and Central AFW -0.453932 \n", "Angola AGO -0.657339 \n", "\n", " GDP_calculated_2015 \\\n", "Country Name Country Code \n", "Aruba ABW -1.046426 \n", "Africa Eastern and Southern AFE -0.840639 \n", "Afghanistan AFG -1.044415 \n", "Africa Western and Central AFW -0.876033 \n", "Angola AGO -0.932675 \n", "\n", " GDP_calculated_2016 \\\n", "Country Name Country Code \n", "Aruba ABW -0.965575 \n", "Africa Eastern and Southern AFE -0.740249 \n", "Afghanistan AFG -0.963573 \n", "Africa Western and Central AFW -0.872786 \n", "Angola AGO -0.812782 \n", "\n", " GDP_calculated_2017 \\\n", "Country Name Country Code \n", "Aruba ABW -0.972609 \n", "Africa Eastern and Southern AFE -0.747085 \n", "Afghanistan AFG -0.970639 \n", "Africa Western and Central AFW -0.756485 \n", "Angola AGO -0.736018 \n", "\n", " GDP_calculated_2018 \\\n", "Country Name Country Code \n", "Aruba ABW -1.011146 \n", "Africa Eastern and Southern AFE -0.788861 \n", "Afghanistan AFG -1.009221 \n", "Africa Western and Central AFW -0.585735 \n", "Angola AGO -0.664243 \n", "\n", " GDP_calculated_2019 \\\n", "Country Name Country Code \n", "Aruba ABW -1.006774 \n", "Africa Eastern and Southern AFE -0.786684 \n", "Afghanistan AFG -1.004804 \n", "Africa Western and Central AFW -0.685331 \n", "Angola AGO -0.802329 \n", "\n", " GDP_calculated_2020 \\\n", "Country Name Country Code \n", "Aruba ABW -0.976539 \n", "Africa Eastern and Southern AFE -0.756719 \n", "Afghanistan AFG -0.972484 \n", "Africa Western and Central AFW -0.871547 \n", "Angola AGO -0.931221 \n", "\n", " GDP_calculated_2021 \\\n", "Country Name Country Code \n", "Aruba ABW -0.925748 \n", "Africa Eastern and Southern AFE -0.599803 \n", "Afghanistan AFG -0.923705 \n", "Africa Western and Central AFW -0.834270 \n", "Angola AGO -0.917089 \n", "\n", " GDP_calculated_2022 \n", "Country Name Country Code \n", "Aruba ABW -0.925497 \n", "Africa Eastern and Southern AFE -0.593651 \n", "Afghanistan AFG -0.918717 \n", "Africa Western and Central AFW -0.833380 \n", "Angola AGO -0.923103 \n", "\n", "[5 rows x 63 columns]" ] }, "execution_count": 2, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Unir los DataFrames uno por uno\n", "df_combined = main_gdp.join(main_government, lsuffix='_gdp', rsuffix='_gov')\n", "df_combined = df_combined.join(main_investments, rsuffix='_inv')\n", "df_combined = df_combined.join(main_consumption, rsuffix='_con')\n", "df_combined = df_combined.join(main_trade, rsuffix='_trade')\n", "\n", "\n", "# Corregir la fórmula para calcular el GDP correctamente usando los sufijos adecuados para cada año\n", "for year in range(1960, 2023):\n", " gdp_col = f'{year}_gdp' if f'{year}_gdp' in df_combined.columns else str(year)\n", " gov_col = f'{year}_gov' if f'{year}_gov' in df_combined.columns else str(year)\n", " inv_col = f'{year}_inv' if f'{year}_inv' in df_combined.columns else str(year)\n", " con_col = f'{year}_con' if f'{year}_con' in df_combined.columns else str(year)\n", " trade_col = f'{year}_trade' if f'{year}_trade' in df_combined.columns else str(year)\n", "\n", " df_combined[f'GDP_calculated_{year}'] = (df_combined[gdp_col] +\n", " df_combined[gov_col] +\n", " df_combined[inv_col] +\n", " df_combined[con_col] +\n", " df_combined[trade_col])\n", "\n", "df_combined[[f'GDP_calculated_{year}' for year in range(1960, 2023)]].head()" ] }, { "cell_type": "code", "execution_count": 3, "id": "8d4ef2f3", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "MultiIndex: 239 entries, ('Aruba', 'ABW') to ('Zimbabwe', 'ZWE')\n", "Columns: 376 entries, 1960_gdp to GDP_calculated_2022\n", "dtypes: float64(376)\n", "memory usage: 731.2+ KB\n" ] }, { "data": { "text/plain": [ "( 1960_gdp 1961_gdp 1962_gdp 1963_gdp 1964_gdp 1965_gdp \\\n", " count 239.000000 239.000000 239.000000 239.000000 239.000000 239.000000 \n", " mean -0.081045 -0.080586 -0.079957 -0.079715 -0.080026 -0.081028 \n", " std 0.811894 0.811040 0.811016 0.810860 0.810938 0.811302 \n", " min -0.224510 -0.223332 -0.222338 -0.222336 -0.222801 -0.221246 \n", " 25% -0.202029 -0.205308 -0.204461 -0.204545 -0.204212 -0.207263 \n", " 50% -0.201874 -0.200982 -0.200327 -0.199863 -0.200426 -0.202832 \n", " 75% -0.201874 -0.200982 -0.200327 -0.199863 -0.200426 -0.202832 \n", " max 8.925325 8.916576 8.903261 8.903658 8.909846 8.905935 \n", " \n", " 1966_gdp 1967_gdp 1968_gdp 1969_gdp ... \\\n", " count 239.000000 239.000000 239.000000 239.000000 ... \n", " mean -0.080735 -0.080437 -0.081089 -0.082057 ... \n", " std 0.811470 0.810851 0.810718 0.810380 ... \n", " min -0.220064 -0.219031 -0.219934 -0.221195 ... \n", " 25% -0.208238 -0.209050 -0.210544 -0.212014 ... \n", " 50% -0.202705 -0.202662 -0.203255 -0.204786 ... \n", " 75% -0.202705 -0.202662 -0.203255 -0.204786 ... \n", " max 8.893228 8.892912 8.894431 8.920468 ... \n", " \n", " GDP_calculated_2013 GDP_calculated_2014 GDP_calculated_2015 \\\n", " count 239.000000 239.000000 239.000000 \n", " mean -0.213543 -0.243189 -0.328689 \n", " std 2.892561 2.864930 2.860354 \n", " min -1.030556 -1.036704 -1.056181 \n", " 25% -1.017694 -1.023788 -1.047886 \n", " 50% -1.010545 -1.016673 -1.039728 \n", " 75% -0.773237 -0.793074 -0.879409 \n", " max 28.144082 27.835159 27.572665 \n", " \n", " GDP_calculated_2016 GDP_calculated_2017 GDP_calculated_2018 \\\n", " count 239.000000 239.000000 239.000000 \n", " mean -0.247591 -0.239064 -0.249890 \n", " std 2.856636 2.864796 2.849596 \n", " min -0.978107 -0.985575 -1.024731 \n", " 25% -0.965917 -0.973220 -1.012258 \n", " 50% -0.959720 -0.966302 -1.004253 \n", " 75% -0.808664 -0.790391 -0.808545 \n", " max 27.837297 27.701692 27.367289 \n", " \n", " GDP_calculated_2019 GDP_calculated_2020 GDP_calculated_2021 \\\n", " count 239.000000 239.000000 239.000000 \n", " mean -0.257321 -0.264348 -0.200992 \n", " std 2.831007 2.824620 3.298784 \n", " min -1.021194 -0.990254 -0.954674 \n", " 25% -1.007834 -0.976800 -0.933824 \n", " 50% -0.999458 -0.970490 -0.921164 \n", " 75% -0.807603 -0.831460 -0.816629 \n", " max 26.262588 26.264463 31.407289 \n", " \n", " GDP_calculated_2022 \n", " count 239.000000 \n", " mean -0.202149 \n", " std 3.301215 \n", " min -0.962321 \n", " 25% -0.936761 \n", " 50% -0.918836 \n", " 75% -0.826804 \n", " max 31.527472 \n", " \n", " [8 rows x 376 columns],\n", " None,\n", " (239, 376),\n", " Index(['1960_gdp', '1961_gdp', '1962_gdp', '1963_gdp', '1964_gdp', '1965_gdp',\n", " '1966_gdp', '1967_gdp', '1968_gdp', '1969_gdp',\n", " ...\n", " 'GDP_calculated_2013', 'GDP_calculated_2014', 'GDP_calculated_2015',\n", " 'GDP_calculated_2016', 'GDP_calculated_2017', 'GDP_calculated_2018',\n", " 'GDP_calculated_2019', 'GDP_calculated_2020', 'GDP_calculated_2021',\n", " 'GDP_calculated_2022'],\n", " dtype='object', length=376),\n", " MultiIndex([( 'Aruba', 'ABW'),\n", " ('Africa Eastern and Southern', 'AFE'),\n", " ( 'Afghanistan', 'AFG'),\n", " ( 'Africa Western and Central', 'AFW'),\n", " ( 'Angola', 'AGO'),\n", " ( 'Albania', 'ALB'),\n", " ( 'Andorra', 'AND'),\n", " ( 'Arab World', 'ARB'),\n", " ( 'United Arab Emirates', 'ARE'),\n", " ( 'Argentina', 'ARG'),\n", " ...\n", " ( 'Virgin Islands (U.S.)', 'VIR'),\n", " ( 'Viet Nam', 'VNM'),\n", " ( 'Vanuatu', 'VUT'),\n", " ( 'World', 'WLD'),\n", " ( 'Samoa', 'WSM'),\n", " ( 'Kosovo', 'XKX'),\n", " ( 'Yemen, Rep.', 'YEM'),\n", " ( 'South Africa', 'ZAF'),\n", " ( 'Zambia', 'ZMB'),\n", " ( 'Zimbabwe', 'ZWE')],\n", " names=['Country Name', 'Country Code'], length=239),\n", " 1960_gdp float64\n", " 1961_gdp float64\n", " 1962_gdp float64\n", " 1963_gdp float64\n", " 1964_gdp float64\n", " ... \n", " GDP_calculated_2018 float64\n", " GDP_calculated_2019 float64\n", " GDP_calculated_2020 float64\n", " GDP_calculated_2021 float64\n", " GDP_calculated_2022 float64\n", " Length: 376, dtype: object,\n", " 1960_gdp 0\n", " 1961_gdp 0\n", " 1962_gdp 0\n", " 1963_gdp 0\n", " 1964_gdp 0\n", " ..\n", " GDP_calculated_2018 0\n", " GDP_calculated_2019 0\n", " GDP_calculated_2020 0\n", " GDP_calculated_2021 0\n", " GDP_calculated_2022 0\n", " Length: 376, dtype: int64)" ] }, "execution_count": 3, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_combined.describe(), df_combined.info(), df_combined.shape, df_combined.columns, df_combined.index, df_combined.dtypes, df_combined.isnull().sum()" ] }, { "cell_type": "markdown", "id": "3646af4e", "metadata": {}, "source": [ "## Prueba ADF " ] }, { "cell_type": "code", "execution_count": 4, "id": "8e6598de", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " Augmented Dickey-Fuller Test on \"GDP_calculated_2021\" \n", " -----------------------------------------------\n", " Null Hypothesis: Data has unit root. Non-Stationary.\n", " Significance Level = 0.05\n", " Test Statistic = -15.8577\n", " No. Lags Chosen = 0\n", " Critical value 1% = -3.458\n", " Critical value 5% = -2.874\n", " Critical value 10% = -2.573\n", " => P-Value = 0.0. Rejecting Null Hypothesis.\n", " => Series is Stationary.\n" ] } ], "source": [ "from statsmodels.tsa.stattools import adfuller\n", "\n", "def test_stationarity(series, signif=0.05, name='', verbose=False):\n", " r = adfuller(series, autolag='AIC')\n", " output = {'test_statistic': round(r[0], 4), 'pvalue': round(r[1], 4), 'n_lags': r[2], 'n_obs': r[3]}\n", " p_value = output['pvalue'] \n", " def adjust(val, length= 6): return str(val).ljust(length)\n", "\n", " # Imprimir Resultados de la Prueba\n", " if verbose:\n", " print(f' Augmented Dickey-Fuller Test on \"{name}\"', \"\\n \", '-'*47)\n", " print(f' Null Hypothesis: Data has unit root. Non-Stationary.')\n", " print(f' Significance Level = {signif}')\n", " print(f' Test Statistic = {output[\"test_statistic\"]}')\n", " print(f' No. Lags Chosen = {output[\"n_lags\"]}')\n", "\n", " for key, val in r[4].items():\n", " print(f' Critical value {adjust(key)} = {round(val, 3)}')\n", "\n", " if p_value <= signif:\n", " print(f\" => P-Value = {p_value}. Rejecting Null Hypothesis.\")\n", " print(f\" => Series is Stationary.\")\n", " else:\n", " print(f\" => P-Value = {p_value}. Weak evidence to reject the Null Hypothesis.\")\n", " print(f\" => Series is Non-Stationary.\")\n", " \n", " return output\n", "\n", "# Aplicar al año 2021\n", "series = df_combined['GDP_calculated_2021']\n", "result = test_stationarity(series, name='GDP_calculated_2021', verbose=True)" ] }, { "cell_type": "markdown", "id": "28192e55", "metadata": {}, "source": [ "## Análisis de Componentes Temporales" ] }, { "cell_type": "code", "execution_count": 5, "id": "98bab1cc", "metadata": { "scrolled": false }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "def decompose_time_series(series, title):\n", " if len(series) > 2: \n", " decomposition = sm.tsa.seasonal_decompose(series, model='additive', period=1) \n", " # Crear gráficos de los componentes\n", " fig, ax = plt.subplots(4, 1, figsize=(16, 10), sharex=True)\n", " series.plot(ax=ax[0], color='b', title=title)\n", " ax[0].set_ylabel('Original')\n", " decomposition.trend.plot(ax=ax[1], color='r')\n", " ax[1].set_ylabel('Trend')\n", " decomposition.seasonal.plot(ax=ax[2], color='g')\n", " ax[2].set_ylabel('Seasonal')\n", " decomposition.resid.plot(ax=ax[3], color='k')\n", " ax[3].set_ylabel('Residual')\n", " plt.tight_layout()\n", " plt.show()\n", " else:\n", " print(f\"Not enough data to decompose {title}\")\n", "\n", "decompose_time_series(df_combined['GDP_calculated_2021'], 'Countries GDP 2021')" ] }, { "cell_type": "code", "execution_count": null, "id": "602d60bb", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "markdown", "id": "b574d8d2", "metadata": {}, "source": [ "### Homocedasticidad" ] }, { "cell_type": "code", "execution_count": 6, "id": "f976fa64", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "C:\\Users\\Oscar Murgueytio\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\patsy\\util.py:672: FutureWarning: is_categorical_dtype is deprecated and will be removed in a future version. Use isinstance(dtype, CategoricalDtype) instead\n", " return _pandas_is_categorical_dtype(dt)\n", "C:\\Users\\Oscar Murgueytio\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\patsy\\util.py:672: FutureWarning: is_categorical_dtype is deprecated and will be removed in a future version. Use isinstance(dtype, CategoricalDtype) instead\n", " return _pandas_is_categorical_dtype(dt)\n", "C:\\Users\\Oscar Murgueytio\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\patsy\\util.py:672: FutureWarning: is_categorical_dtype is deprecated and will be removed in a future version. Use isinstance(dtype, CategoricalDtype) instead\n", " return _pandas_is_categorical_dtype(dt)\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "p-value de la prueba de Breusch-Pagan: 0.00017441609247072733\n" ] }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "import statsmodels.api as sm\n", "from statsmodels.stats.diagnostic import het_breuschpagan\n", "from statsmodels.formula.api import ols\n", "import matplotlib.pyplot as plt\n", "\n", "modelo = 'GDP_calculated_2022 ~ GDP_calculated_2021'\n", "\n", "model = ols(modelo, data=df_combined).fit()\n", "\n", "residuos = model.resid\n", "\n", "# Prueba de Breusch-Pagan\n", "_, pvalue, _, _ = het_breuschpagan(residuos, model.model.exog)\n", "print(f\"p-value de la prueba de Breusch-Pagan: {pvalue}\")\n", "\n", "plt.figure(figsize=(20, 8))\n", "plt.scatter(model.fittedvalues, residuos)\n", "plt.axhline(0, color='red', linestyle='--')\n", "plt.title('Residuos vs Valores Ajustados')\n", "plt.xlabel('Valores Ajustados')\n", "plt.ylabel('Residuos')\n", "plt.show()\n", "\n", "# Si el p-value es menor a 0.05, existe evidencia de heterocedasticidad." ] }, { "cell_type": "markdown", "id": "ac981b36", "metadata": {}, "source": [ "### Weighted Least Squares (WLS)" ] }, { "cell_type": "code", "execution_count": 7, "id": "ab351c19", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " WLS Regression Results \n", "===============================================================================\n", "Dep. Variable: GDP_calculated_2021 R-squared: 1.000\n", "Model: WLS Adj. R-squared: 1.000\n", "Method: Least Squares F-statistic: 2.752e+06\n", "Date: Sun, 12 May 2024 Prob (F-statistic): 0.00\n", "Time: 20:58:08 Log-Likelihood: 733.57\n", "No. Observations: 239 AIC: -1457.\n", "Df Residuals: 234 BIC: -1440.\n", "Df Model: 4 \n", "Covariance Type: nonrobust \n", "==============================================================================\n", " coef std err t P>|t| [0.025 0.975]\n", "------------------------------------------------------------------------------\n", "const -0.0637 0.004 -16.682 0.000 -0.071 -0.056\n", "2020_gov 1.1296 0.009 122.174 0.000 1.111 1.148\n", "2020 2.8250 0.005 579.645 0.000 2.815 2.835\n", "2020_con 0.8689 0.007 128.030 0.000 0.856 0.882\n", "2020_trade 0.2523 0.016 15.870 0.000 0.221 0.284\n", "==============================================================================\n", "Omnibus: 48.340 Durbin-Watson: 1.650\n", "Prob(Omnibus): 0.000 Jarque-Bera (JB): 396.541\n", "Skew: -0.460 Prob(JB): 7.80e-87\n", "Kurtosis: 9.243 Cond. No. 72.8\n", "==============================================================================\n", "\n", "Notes:\n", "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n" ] } ], "source": [ "import statsmodels.api as sm\n", "\n", "X = df_combined[['2020_gov', '2020', '2020_con', '2020_trade']] # G + I + C + T\n", "y = df_combined['GDP_calculated_2021'] \n", "\n", "# Agregar una constante al modelo\n", "X = sm.add_constant(X)\n", "\n", "# Modelo OLS para obtener los residuos\n", "model_ols = sm.OLS(y, X).fit()\n", "residuos = model_ols.resid\n", "\n", "# Calcular los pesos como el inverso de los residuos al cuadrado\n", "pesos = 1.0 / (residuos ** 2)\n", "\n", "# Aplicar WLS con los pesos obtenidos\n", "model_wls = sm.WLS(y, X, weights=pesos).fit()\n", "\n", "print(model_wls.summary())" ] }, { "cell_type": "markdown", "id": "4740eaf2", "metadata": {}, "source": [ "### Normalidad de los Residuos" ] }, { "cell_type": "code", "execution_count": 8, "id": "715a37f2", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Shapiro-Wilk Test:\n", "Statistic: 0.2418427268818486\n", "p-value: 4.245462941888389e-30\n" ] } ], "source": [ "import scipy.stats as stats\n", "\n", "residuos = model_wls.resid\n", "\n", "# Prueba de Shapiro-Wilk\n", "shapiro_test = stats.shapiro(residuos)\n", "\n", "print(\"Shapiro-Wilk Test:\")\n", "print(\"Statistic:\", shapiro_test.statistic)\n", "print(\"p-value:\", shapiro_test.pvalue)" ] }, { "cell_type": "markdown", "id": "21ce23e6", "metadata": {}, "source": [ "### Ausencia de Multicolinealidad" ] }, { "cell_type": "code", "execution_count": 9, "id": "f7dd008c", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " feature VIF\n", "0 const 1.037174\n", "1 2020_gov 16.410593\n", "2 2020 1.004569\n", "3 2020_con 16.409976\n", "4 2020_trade 1.004691\n" ] } ], "source": [ "from statsmodels.stats.outliers_influence import variance_inflation_factor\n", "\n", "# VIF para cada variable en el modelo\n", "vif_data = pd.DataFrame()\n", "vif_data[\"feature\"] = X.columns\n", "vif_data[\"VIF\"] = [variance_inflation_factor(X.values, i) for i in range(X.shape[1])]\n", "\n", "print(vif_data)" ] }, { "cell_type": "markdown", "id": "0fed85b3", "metadata": {}, "source": [ "### Regresión Robusta" ] }, { "cell_type": "code", "execution_count": 10, "id": "d3cb156e", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " Robust linear Model Regression Results \n", "===============================================================================\n", "Dep. Variable: GDP_calculated_2021 No. Observations: 239\n", "Model: RLM Df Residuals: 234\n", "Method: IRLS Df Model: 4\n", "Norm: HuberT \n", "Scale Est.: mad \n", "Cov Type: H1 \n", "Date: Sun, 12 May 2024 \n", "Time: 20:59:27 \n", "No. Iterations: 50 \n", "==============================================================================\n", " coef std err z P>|z| [0.025 0.975]\n", "------------------------------------------------------------------------------\n", "const -0.0776 0.001 -153.867 0.000 -0.079 -0.077\n", "2020_gov 1.1230 0.002 590.301 0.000 1.119 1.127\n", "2020 2.8207 0.001 5098.021 0.000 2.820 2.822\n", "2020_con 0.8751 0.002 460.001 0.000 0.871 0.879\n", "2020_trade 0.1813 0.001 229.744 0.000 0.180 0.183\n", "==============================================================================\n", "\n", "If the model instance has been used for another fit with different fit parameters, then the fit options might not be the correct ones anymore .\n" ] }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "import statsmodels.api as sm\n", "\n", "X = df_combined[['2020_gov', '2020', '2020_con', '2020_trade']] # Predictores \n", "y = df_combined['GDP_calculated_2021'] # La variable objetivo\n", "\n", "# Agrega una constante al modelo\n", "X = sm.add_constant(X)\n", "\n", "# Crear y ajustar un modelo de regresión lineal robusta\n", "model_robust = sm.RLM(y, X).fit()\n", "\n", "# Resumen del modelo\n", "print(model_robust.summary())\n", "\n", "# Visualizar los residuos y verificar la mejora\n", "import matplotlib.pyplot as plt\n", "\n", "plt.figure(figsize=(20, 8))\n", "plt.scatter(model_robust.fittedvalues, model_robust.resid)\n", "plt.axhline(y=0, color='r', linestyle='--')\n", "plt.xlabel('Valores Ajustados')\n", "plt.ylabel('Residuos')\n", "plt.title('Residuos vs Valores Ajustados en Regresión Robusta')\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "c622c224", "metadata": {}, "source": [ "### Diagnóstico Visual de los Residuos" ] }, { "cell_type": "code", "execution_count": 11, "id": "651dcb0e", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Shapiro-Wilk Test - Statistic: 0.33540911798629547, p-value: 1.3335239077225291e-28\n", "Breusch-Pagan Test - p-value: 4.210590527338177e-14\n", " feature VIF\n", "0 const 1.037174\n", "1 2020_gov 16.410593\n", "2 2020 1.004569\n", "3 2020_con 16.409976\n", "4 2020_trade 1.004691\n", " OLS Regression Results \n", "===============================================================================\n", "Dep. Variable: GDP_calculated_2021 R-squared: 0.997\n", "Model: OLS Adj. R-squared: 0.997\n", "Method: Least Squares F-statistic: 1.725e+04\n", "Date: Sun, 12 May 2024 Prob (F-statistic): 8.71e-288\n", "Time: 21:01:57 Log-Likelihood: 56.088\n", "No. Observations: 239 AIC: -102.2\n", "Df Residuals: 234 BIC: -84.79\n", "Df Model: 4 \n", "Covariance Type: nonrobust \n", "==============================================================================\n", " coef std err t P>|t| [0.025 0.975]\n", "------------------------------------------------------------------------------\n", "const -0.0314 0.013 -2.464 0.014 -0.056 -0.006\n", "2020_gov 1.1131 0.048 23.157 0.000 1.018 1.208\n", "2020 2.8144 0.014 201.307 0.000 2.787 2.842\n", "2020_con 0.8821 0.048 18.350 0.000 0.787 0.977\n", "2020_trade 0.4075 0.020 20.440 0.000 0.368 0.447\n", "==============================================================================\n", "Omnibus: 269.675 Durbin-Watson: 2.034\n", "Prob(Omnibus): 0.000 Jarque-Bera (JB): 47503.892\n", "Skew: 4.021 Prob(JB): 0.00\n", "Kurtosis: 71.597 Cond. No. 7.98\n", "==============================================================================\n", "\n", "Notes:\n", "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n" ] }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "import statsmodels.api as sm\n", "from scipy.stats import shapiro\n", "import statsmodels.stats.diagnostic as sms\n", "from statsmodels.stats.outliers_influence import variance_inflation_factor\n", "\n", "# Prueba de Normalidad de los Residuos con Shapiro-Wilk\n", "stat, p_value = shapiro(model_ols.resid)\n", "print(f'Shapiro-Wilk Test - Statistic: {stat}, p-value: {p_value}')\n", "\n", "# Verificación de Homocedasticidad con Breusch-Pagan\n", "_, p_value, _, _ = sms.het_breuschpagan(model_ols.resid, model_ols.model.exog)\n", "print(f'Breusch-Pagan Test - p-value: {p_value}')\n", "\n", "# Calculando VIF para cada predictor\n", "vif_data = pd.DataFrame()\n", "vif_data[\"feature\"] = X.columns\n", "vif_data[\"VIF\"] = [variance_inflation_factor(X.values, i) for i in range(X.shape[1])]\n", "print(vif_data)\n", "\n", "# Resumen del modelo para ajuste\n", "print(model_ols.summary())\n", "\n", "# Diagnóstico visual de residuos\n", "plt.figure(figsize=(20, 8))\n", "plt.scatter(model_ols.fittedvalues, model_ols.resid)\n", "plt.axhline(y=0, color='red', linestyle='--')\n", "plt.xlabel('Valores Ajustados')\n", "plt.ylabel('Residuos')\n", "plt.title('Residuos vs Valores Ajustados')\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": null, "id": "9aa78e11", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "markdown", "id": "6e06a067", "metadata": {}, "source": [ "## Guardar los DataFrames como CSV" ] }, { "cell_type": "code", "execution_count": 12, "id": "def5e36d", "metadata": {}, "outputs": [], "source": [ "# Guardar el DataFrame como CSV\n", "df_combined.to_csv('df_combined.csv', index=False)" ] }, { "cell_type": "code", "execution_count": null, "id": "3cf71b29", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": 13, "id": "be01121b", "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/plain": [ "array(['1960_gdp', '1961_gdp', '1962_gdp', '1963_gdp', '1964_gdp',\n", " '1965_gdp', '1966_gdp', '1967_gdp', '1968_gdp', '1969_gdp',\n", " '1970_gdp', '1971_gdp', '1972_gdp', '1973_gdp', '1974_gdp',\n", " '1975_gdp', '1976_gdp', '1977_gdp', '1978_gdp', '1979_gdp',\n", " '1980_gdp', '1981_gdp', '1982_gdp', '1983_gdp', '1984_gdp',\n", " '1985_gdp', '1986_gdp', '1987_gdp', '1988_gdp', '1989_gdp',\n", " '1990_gdp', '1991_gdp', '1992_gdp', '1993_gdp', '1994_gdp',\n", " '1995_gdp', '1996_gdp', '1997_gdp', '1998_gdp', '1999_gdp',\n", " '2000_gdp', '2001_gdp', '2002_gdp', '2003_gdp', '2004_gdp',\n", " '2005_gdp', '2006_gdp', '2007_gdp', '2008_gdp', '2009_gdp',\n", " '2010_gdp', '2011_gdp', '2012_gdp', '2013_gdp', '2014_gdp',\n", " '2015_gdp', '2016_gdp', '2017_gdp', '2018_gdp', '2019_gdp',\n", " '2020_gdp', '2021_gdp', '2022_gdp', '1960_gov', '1961_gov',\n", " '1962_gov', '1963_gov', '1964_gov', '1965_gov', '1966_gov',\n", " '1967_gov', '1968_gov', '1969_gov', '1970_gov', '1971_gov',\n", " '1972_gov', '1973_gov', '1974_gov', '1975_gov', '1976_gov',\n", " '1977_gov', '1978_gov', '1979_gov', '1980_gov', '1981_gov',\n", " '1982_gov', '1983_gov', '1984_gov', '1985_gov', '1986_gov',\n", " '1987_gov', '1988_gov', '1989_gov', '1990_gov', '1991_gov',\n", " '1992_gov', '1993_gov', '1994_gov', '1995_gov', '1996_gov',\n", " '1997_gov', '1998_gov', '1999_gov', '2000_gov', '2001_gov',\n", " '2002_gov', '2003_gov', '2004_gov', '2005_gov', '2006_gov',\n", " '2007_gov', '2008_gov', '2009_gov', '2010_gov', '2011_gov',\n", " '2012_gov', '2013_gov', '2014_gov', '2015_gov', '2016_gov',\n", " '2017_gov', '2018_gov', '2019_gov', '2020_gov', '2021_gov',\n", " '2022_gov', '1960', '1961', '1962', '1963', '1964', '1965', '1966',\n", " '1967', '1968', '1969', '1970', '1971', '1972', '1973', '1974',\n", " '1975', '1976', '1977', '1978', '1979', '1980', '1981', '1982',\n", " '1983', '1984', '1985', '1986', '1987', '1988', '1989', '1990',\n", " '1991', '1992', '1993', '1994', '1995', '1996', '1997', '1998',\n", " '1999', '2000', '2001', '2002', '2003', '2004', '2005', '2006',\n", " '2007', '2008', '2009', '2010', '2011', '2012', '2013', '2014',\n", " '2015', '2016', '2017', '2018', '2019', '2020', '2021', '2022',\n", " '1960_con', '1961_con', '1962_con', '1963_con', '1964_con',\n", " '1965_con', '1966_con', '1967_con', '1968_con', '1969_con',\n", " '1970_con', '1971_con', '1972_con', '1973_con', '1974_con',\n", " '1975_con', '1976_con', '1977_con', '1978_con', '1979_con',\n", " '1980_con', '1981_con', '1982_con', '1983_con', '1984_con',\n", " '1985_con', '1986_con', '1987_con', '1988_con', '1989_con',\n", " '1990_con', '1991_con', '1992_con', '1993_con', '1994_con',\n", " '1995_con', '1996_con', '1997_con', '1998_con', '1999_con',\n", " '2000_con', '2001_con', '2002_con', '2003_con', '2004_con',\n", " '2005_con', '2006_con', '2007_con', '2008_con', '2009_con',\n", " '2010_con', '2011_con', '2012_con', '2013_con', '2014_con',\n", " '2015_con', '2016_con', '2017_con', '2018_con', '2019_con',\n", " '2020_con', '2021_con', '2022_con', '1960_trade', '1961_trade',\n", " '1962_trade', '1963_trade', '1964_trade', '1965_trade',\n", " '1966_trade', '1967_trade', '1968_trade', '1969_trade',\n", " '1970_trade', '1971_trade', '1972_trade', '1973_trade',\n", " '1974_trade', '1975_trade', '1976_trade', '1977_trade',\n", " '1978_trade', '1979_trade', '1980_trade', '1981_trade',\n", " '1982_trade', '1983_trade', '1984_trade', '1985_trade',\n", " '1986_trade', '1987_trade', '1988_trade', '1989_trade',\n", " '1990_trade', '1991_trade', '1992_trade', '1993_trade',\n", " '1994_trade', '1995_trade', '1996_trade', '1997_trade',\n", " '1998_trade', '1999_trade', '2000_trade', '2001_trade',\n", " '2002_trade', '2003_trade', '2004_trade', '2005_trade',\n", " '2006_trade', '2007_trade', '2008_trade', '2009_trade',\n", " '2010_trade', '2011_trade', '2012_trade', '2013_trade',\n", " '2014_trade', '2015_trade', '2016_trade', '2017_trade',\n", " '2018_trade', '2019_trade', '2020_trade', 'GDP_calculated_1960',\n", " 'GDP_calculated_1961', 'GDP_calculated_1962',\n", " 'GDP_calculated_1963', 'GDP_calculated_1964',\n", " 'GDP_calculated_1965', 'GDP_calculated_1966',\n", " 'GDP_calculated_1967', 'GDP_calculated_1968',\n", " 'GDP_calculated_1969', 'GDP_calculated_1970',\n", " 'GDP_calculated_1971', 'GDP_calculated_1972',\n", " 'GDP_calculated_1973', 'GDP_calculated_1974',\n", " 'GDP_calculated_1975', 'GDP_calculated_1976',\n", " 'GDP_calculated_1977', 'GDP_calculated_1978',\n", " 'GDP_calculated_1979', 'GDP_calculated_1980',\n", " 'GDP_calculated_1981', 'GDP_calculated_1982',\n", " 'GDP_calculated_1983', 'GDP_calculated_1984',\n", " 'GDP_calculated_1985', 'GDP_calculated_1986',\n", " 'GDP_calculated_1987', 'GDP_calculated_1988',\n", " 'GDP_calculated_1989', 'GDP_calculated_1990',\n", " 'GDP_calculated_1991', 'GDP_calculated_1992',\n", " 'GDP_calculated_1993', 'GDP_calculated_1994',\n", " 'GDP_calculated_1995', 'GDP_calculated_1996',\n", " 'GDP_calculated_1997', 'GDP_calculated_1998',\n", " 'GDP_calculated_1999', 'GDP_calculated_2000',\n", " 'GDP_calculated_2001', 'GDP_calculated_2002',\n", " 'GDP_calculated_2003', 'GDP_calculated_2004',\n", " 'GDP_calculated_2005', 'GDP_calculated_2006',\n", " 'GDP_calculated_2007', 'GDP_calculated_2008',\n", " 'GDP_calculated_2009', 'GDP_calculated_2010',\n", " 'GDP_calculated_2011', 'GDP_calculated_2012',\n", " 'GDP_calculated_2013', 'GDP_calculated_2014',\n", " 'GDP_calculated_2015', 'GDP_calculated_2016',\n", " 'GDP_calculated_2017', 'GDP_calculated_2018',\n", " 'GDP_calculated_2019', 'GDP_calculated_2020',\n", " 'GDP_calculated_2021', 'GDP_calculated_2022'], dtype=object)" ] }, "execution_count": 13, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Detalles de las columnas\n", "df_combined.columns.values[:376]" ] }, { "cell_type": "code", "execution_count": 14, "id": "3b00e818", "metadata": {}, "outputs": [ { "data": { "text/plain": [ "array([[-0.20187388, -0.20098171, -0.20032654, ..., -0.97653872,\n", " -0.92574843, -0.92549705],\n", " [-0.09491353, -0.09825918, -0.09439167, ..., -0.75671896,\n", " -0.59980343, -0.59365105],\n", " [-0.20187388, -0.20098171, -0.20032654, ..., -0.97248442,\n", " -0.9237053 , -0.9187167 ],\n", " ...,\n", " [-0.16960274, -0.16843671, -0.16711353, ..., -0.79129927,\n", " -0.85280902, -0.85958157],\n", " [-0.22148338, -0.22037191, -0.21961394, ..., -0.95118224,\n", " -0.92343036, -0.92314575],\n", " [-0.22099433, -0.2197217 , -0.21887512, ..., -0.98686413,\n", " -0.95003866, -0.95714594]])" ] }, "execution_count": 14, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_combined.values[:376]" ] }, { "cell_type": "code", "execution_count": 15, "id": "1a49bd46", "metadata": {}, "outputs": [ { "data": { "text/plain": [ "1960_gdp 0\n", "1961_gdp 0\n", "1962_gdp 0\n", "1963_gdp 0\n", "1964_gdp 0\n", " ..\n", "GDP_calculated_2018 0\n", "GDP_calculated_2019 0\n", "GDP_calculated_2020 0\n", "GDP_calculated_2021 0\n", "GDP_calculated_2022 0\n", "Length: 376, dtype: int64" ] }, "execution_count": 15, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_combined.isnull().sum()" ] }, { "cell_type": "code", "execution_count": 16, "id": "0be90489", "metadata": {}, "outputs": [ { "data": { "text/plain": [ "1960_gdp 0\n", "1961_gdp 0\n", "1962_gdp 0\n", "1963_gdp 0\n", "1964_gdp 0\n", " ..\n", "GDP_calculated_2018 0\n", "GDP_calculated_2019 0\n", "GDP_calculated_2020 0\n", "GDP_calculated_2021 0\n", "GDP_calculated_2022 0\n", "Length: 376, dtype: int64" ] }, "execution_count": 16, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_combined.iloc[:376].isnull().sum()" ] }, { "cell_type": "code", "execution_count": 17, "id": "d6c2aacf", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "1960_gdp 0\n", "1961_gdp 0\n", "1962_gdp 0\n", "1963_gdp 0\n", "1964_gdp 0\n", " ..\n", "GDP_calculated_2018 0\n", "GDP_calculated_2019 0\n", "GDP_calculated_2020 0\n", "GDP_calculated_2021 0\n", "GDP_calculated_2022 0\n", "Length: 376, dtype: int64\n", "1960_gdp float64\n", "1961_gdp float64\n", "1962_gdp float64\n", "1963_gdp float64\n", "1964_gdp float64\n", " ... \n", "GDP_calculated_2018 float64\n", "GDP_calculated_2019 float64\n", "GDP_calculated_2020 float64\n", "GDP_calculated_2021 float64\n", "GDP_calculated_2022 float64\n", "Length: 376, dtype: object\n" ] } ], "source": [ "# Verificar NaNs en las primeras 376 filas y los tipos de datos\n", "print(df_combined.iloc[:376].isnull().sum())\n", "print(df_combined.iloc[:376].dtypes)" ] }, { "cell_type": "code", "execution_count": null, "id": "8387b391", "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "Python 3 (ipykernel)", "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": "3.10.8" } }, "nbformat": 4, "nbformat_minor": 5 }