Skip to content

Instantly share code, notes, and snippets.

@pr0nstar

pr0nstar/covid.alt.ipynb

Created December 30, 2021 15:46
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save pr0nstar/d8109d562ed409ecaec4f60176d1a538 to your computer and use it in GitHub Desktop.
Save pr0nstar/d8109d562ed409ecaec4f60176d1a538 to your computer and use it in GitHub Desktop.
Download ZIP
Display the source blob
Display the rendered blob
Raw
covid.alt.ipynb
{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"id": "accompanied-myanmar",
"metadata": {},
"outputs": [],
"source": [
"import warnings\n",
"import datetime\n",
"import unidecode\n",
"\n",
"import numpy as np\n",
"import pandas as pd\n",
"import seaborn as sns\n",
"\n",
"import sklearn.linear_model\n",
"import statsmodels.formula.api as smf\n",
"\n",
"from matplotlib import pyplot\n",
"\n",
"warnings.simplefilter(action='ignore', category=pd.errors.PerformanceWarning)"
]
},
{
"cell_type": "markdown",
"id": "convenient-vienna",
"metadata": {},
"source": [
"### Exceso mortalidad"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "another-palmer",
"metadata": {},
"outputs": [],
"source": [
"mortality_df = pd.read_csv(\n",
" 'https://github.com/pr0nstar/covid19-data/raw/master/raw/mortality/south.america.subnational.mortality.csv'\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "surrounded-malaysia",
"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></th>\n",
" <th>country_name</th>\n",
" <th>adm1_name</th>\n",
" <th>frequency</th>\n",
" <th>deaths</th>\n",
" </tr>\n",
" <tr>\n",
" <th>iso_code</th>\n",
" <th>adm1_isocode</th>\n",
" <th>date</th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>BO</th>\n",
" <th>BO-B</th>\n",
" <th>2019-01-01</th>\n",
" <td>Bolivia</td>\n",
" <td>Beni</td>\n",
" <td>monthly</td>\n",
" <td>131</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" country_name adm1_name frequency deaths\n",
"iso_code adm1_isocode date \n",
"BO BO-B 2019-01-01 Bolivia Beni monthly 131"
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"mortality_df['date'] = pd.to_datetime(mortality_df['date'])\n",
"\n",
"geo_codes = {'PY-{}'.format(_):'PY-0{}'.format(_) for _ in range(1, 10)}\n",
"\n",
"mortality_df['adm1_isocode'] = mortality_df['adm1_isocode'].replace(geo_codes)\n",
"mortality_df = mortality_df.set_index(['iso_code', 'adm1_isocode', 'date'])\n",
"mortality_df = mortality_df.sort_index()\n",
"\n",
"mortality_df.head(1)"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "consecutive-spain",
"metadata": {},
"outputs": [],
"source": [
"location_df = mortality_df.reset_index()[['country_name', 'adm1_name', 'adm1_isocode']]\n",
"location_df = location_df.drop_duplicates()"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "trying-ethics",
"metadata": {},
"outputs": [],
"source": [
"df_co = mortality_df.loc[['CO']].reset_index(level='date')\n",
"df_co['date'] = df_co['date'] - pd.DateOffset(days=1)\n",
"df_co = df_co.set_index('date', append=True)\n",
"\n",
"mortality_df = mortality_df.drop('CO', level='iso_code')\n",
"mortality_df = pd.concat([mortality_df, df_co])"
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "accomplished-ordinance",
"metadata": {},
"outputs": [],
"source": [
"# https://en.wikipedia.org/wiki/2017_Mocoa_landslide\n",
"mortality_df.loc[('CO', 'CO-PUT', '2017-03-26'), 'deaths'] = 24"
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "tribal-problem",
"metadata": {},
"outputs": [],
"source": [
"#### Drop Peru Lambayaque\n",
"mortality_df = mortality_df.drop(('PE', 'PE-LAM'))\n",
"\n",
"#### Drop Peru Ayacucho 2017\n",
"df_pe_lam = mortality_df.xs(('PE', 'PE-AYA'), drop_level=False).reset_index(level='date')\n",
"df_pe_lam = df_pe_lam[df_pe_lam['date'] < '2018'].set_index('date', append=True)\n",
"\n",
"mortality_df = mortality_df.drop(df_pe_lam.index)"
]
},
{
"cell_type": "code",
"execution_count": 10,
"id": "equivalent-management",
"metadata": {},
"outputs": [],
"source": [
"#### Drop Brasil 2018\n",
"df_br = mortality_df.loc[['BR']].reset_index(level='date')\n",
"df_br = df_br[df_br['date'] < '2019'].set_index('date', append=True)\n",
"\n",
"df = mortality_df.drop(df_br.index)"
]
},
{
"cell_type": "code",
"execution_count": 11,
"id": "annual-aerospace",
"metadata": {},
"outputs": [],
"source": [
"# Monthly\n",
"mortality_df = mortality_df.groupby([\n",
" 'adm1_isocode',\n",
" pd.Grouper(level='date', freq='M')\n",
"])['deaths'].sum().unstack(level=0)"
]
},
{
"cell_type": "code",
"execution_count": 12,
"id": "arabic-solution",
"metadata": {},
"outputs": [],
"source": [
"# Right censoring (two months)\n",
"mortality_df = mortality_df.iloc[:-2]\n",
"# 2016:\n",
"mortality_df = mortality_df.loc['2016':]"
]
},
{
"cell_type": "code",
"execution_count": 13,
"id": "southwest-titanium",
"metadata": {},
"outputs": [],
"source": [
"def get_base(df):\n",
" base_df = df.loc[:'2019']\n",
" \n",
" base_df = base_df.to_frame().set_index([\n",
" base_df.index.year, base_df.index.month\n",
" ])[df.name]\n",
"\n",
" base_years = base_df.index.get_level_values(0).values.reshape(-1 ,1)\n",
" base_years_n = len(np.unique(base_years))\n",
" \n",
" onehot = np.concatenate([np.eye(12) for _ in range(base_years_n)])\n",
" predictors = np.concatenate((base_years, onehot), axis=1)\n",
"\n",
" reg = sklearn.linear_model.LinearRegression(fit_intercept=False).fit(\n",
" predictors, base_df.values\n",
" )\n",
" \n",
" baseline = reg.predict(\n",
" np.concatenate((np.ones((12, 1)) * 2020, np.eye(12)), axis=1)\n",
" )\n",
" \n",
" fbase_df = pd.DataFrame([])\n",
" for year in df.loc['2020':].index.year.unique():\n",
" indexed_base_df = pd.Series(baseline)\n",
" indexed_base_df.index = pd.date_range(\n",
" start=str(year), periods=12, freq='M'\n",
" )\n",
" fbase_df = pd.concat([fbase_df, indexed_base_df])\n",
"\n",
" return fbase_df[0].rename(df.name)[:df.index[-1]]\n"
]
},
{
"cell_type": "code",
"execution_count": 14,
"id": "disciplinary-ancient",
"metadata": {},
"outputs": [],
"source": [
"mortality_base_df = mortality_df.apply(lambda _: get_base(_.dropna()))"
]
},
{
"cell_type": "code",
"execution_count": 15,
"id": "declared-defense",
"metadata": {},
"outputs": [],
"source": [
"mortality_total_diff = (mortality_df - mortality_base_df).loc['2020-03':].sum()"
]
},
{
"cell_type": "code",
"execution_count": 16,
"id": "intense-yahoo",
"metadata": {},
"outputs": [],
"source": [
"mortality_total_expected = mortality_base_df.apply(\n",
" lambda _: _['2020-03':mortality_df[_.name].last_valid_index()].sum()\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 17,
"id": "rolled-effort",
"metadata": {},
"outputs": [],
"source": [
"mortality_prate = (mortality_total_diff / mortality_total_expected)"
]
},
{
"cell_type": "code",
"execution_count": 18,
"id": "dimensional-barrel",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"adm1_isocode\n",
"BO-B 0.476628\n",
"BO-C 0.482366\n",
"BO-H 0.369276\n",
"BO-L 0.704266\n",
"BO-N 0.859537\n",
" ... \n",
"UY-SA -0.110480\n",
"UY-SJ 0.020264\n",
"UY-SO -0.009402\n",
"UY-TA -0.091659\n",
"UY-TT -0.037525\n",
"Length: 170, dtype: float64"
]
},
"execution_count": 18,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"mortality_prate"
]
},
{
"cell_type": "markdown",
"id": "massive-guidance",
"metadata": {},
"source": [
"### Elevacion"
]
},
{
"cell_type": "code",
"execution_count": 19,
"id": "matched-setup",
"metadata": {},
"outputs": [],
"source": [
"elevation_df = pd.read_csv(\n",
" 'https://github.com/daliagachc/cov-alt/raw/master/data/city_data.csv',\n",
" na_values='QuantityMagnitude[Missing[\"NotAvailable\"]]'\n",
")\n",
"elevation_df = elevation_df.dropna()\n",
"elevation_df.columns = [_.lower() for _ in elevation_df.columns]"
]
},
{
"cell_type": "code",
"execution_count": 20,
"id": "documentary-liechtenstein",
"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>city</th>\n",
" <th>population</th>\n",
" <th>elevation</th>\n",
" <th>administrativedivision</th>\n",
" <th>country</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>Santa Cruz</td>\n",
" <td>1453549</td>\n",
" <td>420.0</td>\n",
" <td>Santa Cruz, Bolivia</td>\n",
" <td>Bolivia</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>El Alto</td>\n",
" <td>903080</td>\n",
" <td>4100.0</td>\n",
" <td>La Paz, Bolivia</td>\n",
" <td>Bolivia</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>La Paz</td>\n",
" <td>757184</td>\n",
" <td>3700.0</td>\n",
" <td>La Paz, Bolivia</td>\n",
" <td>Bolivia</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>Cochabamba</td>\n",
" <td>632013</td>\n",
" <td>2600.0</td>\n",
" <td>Cochabamba, Bolivia</td>\n",
" <td>Bolivia</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>Sucre</td>\n",
" <td>300000</td>\n",
" <td>2800.0</td>\n",
" <td>Chuquisaca, Bolivia</td>\n",
" <td>Bolivia</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" city population elevation administrativedivision country\n",
"0 Santa Cruz 1453549 420.0 Santa Cruz, Bolivia Bolivia\n",
"1 El Alto 903080 4100.0 La Paz, Bolivia Bolivia\n",
"2 La Paz 757184 3700.0 La Paz, Bolivia Bolivia\n",
"3 Cochabamba 632013 2600.0 Cochabamba, Bolivia Bolivia\n",
"4 Sucre 300000 2800.0 Chuquisaca, Bolivia Bolivia"
]
},
"execution_count": 20,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"elevation_df.head()"
]
},
{
"cell_type": "code",
"execution_count": 21,
"id": "median-origin",
"metadata": {},
"outputs": [],
"source": [
"elevation_adm1_map = {\n",
" 'Distrito Federal': 'Federal District',\n",
" 'Rio Grande Do Sul': 'Rio Grande do Sul',\n",
" 'Bio Bio': 'Biobio',\n",
" 'Metropolitana': 'Santiago Metropolitan',\n",
" 'Distrito Capital': 'Bogota D.C.',\n",
" 'Morona Santiago': 'Morona-Santiago',\n",
" 'Zamora Chinchipe': 'Zamora-Chinchipe',\n",
" 'Canendiyu': 'Canindeyu',\n",
"}"
]
},
{
"cell_type": "code",
"execution_count": 22,
"id": "executed-delta",
"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>elevation_name</th>\n",
" <th>adm1_name</th>\n",
" <th>country_name</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>Santa Cruz, Bolivia</td>\n",
" <td>Santa Cruz</td>\n",
" <td>Bolivia</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>La Paz, Bolivia</td>\n",
" <td>La Paz</td>\n",
" <td>Bolivia</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>Cochabamba, Bolivia</td>\n",
" <td>Cochabamba</td>\n",
" <td>Bolivia</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>Chuquisaca, Bolivia</td>\n",
" <td>Chuquisaca</td>\n",
" <td>Bolivia</td>\n",
" </tr>\n",
" <tr>\n",
" <th>5</th>\n",
" <td>Oruro, Bolivia</td>\n",
" <td>Oruro</td>\n",
" <td>Bolivia</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" elevation_name adm1_name country_name\n",
"0 Santa Cruz, Bolivia Santa Cruz Bolivia\n",
"1 La Paz, Bolivia La Paz Bolivia\n",
"3 Cochabamba, Bolivia Cochabamba Bolivia\n",
"4 Chuquisaca, Bolivia Chuquisaca Bolivia\n",
"5 Oruro, Bolivia Oruro Bolivia"
]
},
"execution_count": 22,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"elevation_locations = elevation_df['administrativedivision'].str.split(',').apply(pd.Series)\n",
"elevation_locations = pd.concat([\n",
" elevation_df['administrativedivision'],\n",
" elevation_locations\n",
"], axis=1)\n",
"elevation_locations.columns = ['elevation_name', 'adm1_name', 'country_name']\n",
"\n",
"elevation_locations = elevation_locations.drop_duplicates()\n",
"elevation_locations = elevation_locations.applymap(lambda _: _.strip())\n",
"\n",
"elevation_locations['adm1_name'] = elevation_locations['adm1_name'].apply(unidecode.unidecode)\n",
"elevation_locations['adm1_name'] = elevation_locations['adm1_name'].replace(elevation_adm1_map)\n",
"\n",
"elevation_locations.head()"
]
},
{
"cell_type": "code",
"execution_count": 23,
"id": "sophisticated-culture",
"metadata": {},
"outputs": [],
"source": [
"elevation_locations = pd.concat([\n",
" location_df.set_index(['country_name', 'adm1_name']),\n",
" elevation_locations.set_index(['country_name', 'adm1_name'])\n",
"], axis=1)"
]
},
{
"cell_type": "code",
"execution_count": 24,
"id": "controlled-front",
"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>adm1_isocode</th>\n",
" <th>elevation_name</th>\n",
" </tr>\n",
" <tr>\n",
" <th>country_name</th>\n",
" <th>adm1_name</th>\n",
" <th></th>\n",
" <th></th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th rowspan=\"4\" valign=\"top\">Chile</th>\n",
" <th>Arica y Parinacota</th>\n",
" <td>CL-AP</td>\n",
" <td>NaN</td>\n",
" </tr>\n",
" <tr>\n",
" <th>Libertador</th>\n",
" <td>NaN</td>\n",
" <td>Libertador, Chile</td>\n",
" </tr>\n",
" <tr>\n",
" <th>O'Higgins</th>\n",
" <td>CL-LI</td>\n",
" <td>NaN</td>\n",
" </tr>\n",
" <tr>\n",
" <th>Ñuble</th>\n",
" <td>CL-NB</td>\n",
" <td>NaN</td>\n",
" </tr>\n",
" <tr>\n",
" <th rowspan=\"2\" valign=\"top\">Ecuador</th>\n",
" <th>Santa Elena</th>\n",
" <td>EC-SE</td>\n",
" <td>NaN</td>\n",
" </tr>\n",
" <tr>\n",
" <th>Santo Domingo de los Tsachilas</th>\n",
" <td>EC-SD</td>\n",
" <td>NaN</td>\n",
" </tr>\n",
" <tr>\n",
" <th rowspan=\"2\" valign=\"top\">Peru</th>\n",
" <th>Callao</th>\n",
" <td>PE-CAL</td>\n",
" <td>NaN</td>\n",
" </tr>\n",
" <tr>\n",
" <th>Lima Provincias</th>\n",
" <td>NaN</td>\n",
" <td>Lima Provincias, Peru</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" adm1_isocode \\\n",
"country_name adm1_name \n",
"Chile Arica y Parinacota CL-AP \n",
" Libertador NaN \n",
" O'Higgins CL-LI \n",
" Ñuble CL-NB \n",
"Ecuador Santa Elena EC-SE \n",
" Santo Domingo de los Tsachilas EC-SD \n",
"Peru Callao PE-CAL \n",
" Lima Provincias NaN \n",
"\n",
" elevation_name \n",
"country_name adm1_name \n",
"Chile Arica y Parinacota NaN \n",
" Libertador Libertador, Chile \n",
" O'Higgins NaN \n",
" Ñuble NaN \n",
"Ecuador Santa Elena NaN \n",
" Santo Domingo de los Tsachilas NaN \n",
"Peru Callao NaN \n",
" Lima Provincias Lima Provincias, Peru "
]
},
"execution_count": 24,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"elevation_locations[elevation_locations.isna().any(axis=1)]"
]
},
{
"cell_type": "code",
"execution_count": 25,
"id": "revised-answer",
"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>elevation_arithmetic</th>\n",
" <th>elevation_geometric</th>\n",
" <th>population</th>\n",
" </tr>\n",
" <tr>\n",
" <th>administrativedivision</th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>Acre, Brazil</th>\n",
" <td>149.997</td>\n",
" <td>148.901</td>\n",
" <td>531527</td>\n",
" </tr>\n",
" <tr>\n",
" <th>Aisen, Chile</th>\n",
" <td>280.000</td>\n",
" <td>280.000</td>\n",
" <td>57818</td>\n",
" </tr>\n",
" <tr>\n",
" <th>Alagoas, Brazil</th>\n",
" <td>98.868</td>\n",
" <td>43.424</td>\n",
" <td>2352063</td>\n",
" </tr>\n",
" <tr>\n",
" <th>Alto Paraguay, Paraguay</th>\n",
" <td>94.214</td>\n",
" <td>88.817</td>\n",
" <td>13120</td>\n",
" </tr>\n",
" <tr>\n",
" <th>Alto Parana, Paraguay</th>\n",
" <td>222.197</td>\n",
" <td>221.085</td>\n",
" <td>550795</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" elevation_arithmetic elevation_geometric population\n",
"administrativedivision \n",
"Acre, Brazil 149.997 148.901 531527\n",
"Aisen, Chile 280.000 280.000 57818\n",
"Alagoas, Brazil 98.868 43.424 2352063\n",
"Alto Paraguay, Paraguay 94.214 88.817 13120\n",
"Alto Parana, Paraguay 222.197 221.085 550795"
]
},
"execution_count": 25,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Population weighted\n",
"\n",
"population_weighted_elevation = elevation_df.groupby('administrativedivision').apply(\n",
" lambda _: (_['population'] * _['elevation']).sum() / _['population'].sum()\n",
").round(3)\n",
"\n",
"elevation_df_ = elevation_df.groupby('administrativedivision').apply(\n",
" lambda _: (_['population'] * np.log(_['elevation'] + 1e-6)).sum() / _['population'].sum()\n",
")\n",
"elevation_df_ = np.exp(elevation_df_).round(3)\n",
"\n",
"population_weighted_elevation = pd.concat([\n",
" population_weighted_elevation.rename('elevation_arithmetic'),\n",
" elevation_df_.rename('elevation_geometric'),\n",
" elevation_df.groupby('administrativedivision')['population'].sum()\n",
"], axis=1)\n",
"\n",
"population_weighted_elevation.head()"
]
},
{
"cell_type": "code",
"execution_count": 26,
"id": "august-ceiling",
"metadata": {},
"outputs": [],
"source": [
"population_weighted_elevation.index = population_weighted_elevation.index.to_series().replace(\n",
" elevation_locations.set_index('elevation_name')['adm1_isocode']\n",
")\n",
"population_weighted_elevation = pd.concat([\n",
" population_weighted_elevation[~population_weighted_elevation.index.isna()],\n",
" mortality_prate.rename('prate'),\n",
" location_df.set_index('adm1_isocode')\n",
"], axis=1).dropna()"
]
},
{
"cell_type": "code",
"execution_count": 27,
"id": "dramatic-banana",
"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>elevation_arithmetic</th>\n",
" <th>elevation_geometric</th>\n",
" <th>population</th>\n",
" <th>prate</th>\n",
" <th>country_name</th>\n",
" <th>adm1_name</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>BR-AC</th>\n",
" <td>149.997</td>\n",
" <td>148.901</td>\n",
" <td>531527.0</td>\n",
" <td>0.376842</td>\n",
" <td>Brazil</td>\n",
" <td>Acre</td>\n",
" </tr>\n",
" <tr>\n",
" <th>CL-AI</th>\n",
" <td>280.000</td>\n",
" <td>280.000</td>\n",
" <td>57818.0</td>\n",
" <td>0.112842</td>\n",
" <td>Chile</td>\n",
" <td>Aisen</td>\n",
" </tr>\n",
" <tr>\n",
" <th>BR-AL</th>\n",
" <td>98.868</td>\n",
" <td>43.424</td>\n",
" <td>2352063.0</td>\n",
" <td>0.058929</td>\n",
" <td>Brazil</td>\n",
" <td>Alagoas</td>\n",
" </tr>\n",
" <tr>\n",
" <th>PY-16</th>\n",
" <td>94.214</td>\n",
" <td>88.817</td>\n",
" <td>13120.0</td>\n",
" <td>0.424879</td>\n",
" <td>Paraguay</td>\n",
" <td>Alto Paraguay</td>\n",
" </tr>\n",
" <tr>\n",
" <th>PY-10</th>\n",
" <td>222.197</td>\n",
" <td>221.085</td>\n",
" <td>550795.0</td>\n",
" <td>0.573519</td>\n",
" <td>Paraguay</td>\n",
" <td>Alto Parana</td>\n",
" </tr>\n",
" <tr>\n",
" <th>...</th>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" <td>...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>CO-VAC</th>\n",
" <td>904.354</td>\n",
" <td>554.585</td>\n",
" <td>4539294.0</td>\n",
" <td>0.368710</td>\n",
" <td>Colombia</td>\n",
" <td>Valle del Cauca</td>\n",
" </tr>\n",
" <tr>\n",
" <th>CL-VS</th>\n",
" <td>164.757</td>\n",
" <td>85.328</td>\n",
" <td>1376910.0</td>\n",
" <td>0.123665</td>\n",
" <td>Chile</td>\n",
" <td>Valparaiso</td>\n",
" </tr>\n",
" <tr>\n",
" <th>CO-VAU</th>\n",
" <td>183.742</td>\n",
" <td>182.452</td>\n",
" <td>16374.0</td>\n",
" <td>-0.070932</td>\n",
" <td>Colombia</td>\n",
" <td>Vaupes</td>\n",
" </tr>\n",
" <tr>\n",
" <th>CO-VID</th>\n",
" <td>81.406</td>\n",
" <td>76.166</td>\n",
" <td>18502.0</td>\n",
" <td>0.143237</td>\n",
" <td>Colombia</td>\n",
" <td>Vichada</td>\n",
" </tr>\n",
" <tr>\n",
" <th>EC-Z</th>\n",
" <td>1163.784</td>\n",
" <td>1144.693</td>\n",
" <td>19594.0</td>\n",
" <td>0.521158</td>\n",
" <td>Ecuador</td>\n",
" <td>Zamora-Chinchipe</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"<p>164 rows × 6 columns</p>\n",
"</div>"
],
"text/plain": [
" elevation_arithmetic elevation_geometric population prate \\\n",
"BR-AC 149.997 148.901 531527.0 0.376842 \n",
"CL-AI 280.000 280.000 57818.0 0.112842 \n",
"BR-AL 98.868 43.424 2352063.0 0.058929 \n",
"PY-16 94.214 88.817 13120.0 0.424879 \n",
"PY-10 222.197 221.085 550795.0 0.573519 \n",
"... ... ... ... ... \n",
"CO-VAC 904.354 554.585 4539294.0 0.368710 \n",
"CL-VS 164.757 85.328 1376910.0 0.123665 \n",
"CO-VAU 183.742 182.452 16374.0 -0.070932 \n",
"CO-VID 81.406 76.166 18502.0 0.143237 \n",
"EC-Z 1163.784 1144.693 19594.0 0.521158 \n",
"\n",
" country_name adm1_name \n",
"BR-AC Brazil Acre \n",
"CL-AI Chile Aisen \n",
"BR-AL Brazil Alagoas \n",
"PY-16 Paraguay Alto Paraguay \n",
"PY-10 Paraguay Alto Parana \n",
"... ... ... \n",
"CO-VAC Colombia Valle del Cauca \n",
"CL-VS Chile Valparaiso \n",
"CO-VAU Colombia Vaupes \n",
"CO-VID Colombia Vichada \n",
"EC-Z Ecuador Zamora-Chinchipe \n",
"\n",
"[164 rows x 6 columns]"
]
},
"execution_count": 27,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"population_weighted_elevation"
]
},
{
"cell_type": "markdown",
"id": "cathedral-ordinance",
"metadata": {},
"source": [
"### Exceso/Elevacion"
]
},
{
"cell_type": "code",
"execution_count": 28,
"id": "fluid-account",
"metadata": {},
"outputs": [],
"source": [
"PAISES_ANDINOS = [\n",
" 'Bolivia', 'Peru', 'Ecuador', 'Colombia'\n",
"]"
]
},
{
"cell_type": "code",
"execution_count": 29,
"id": "social-storage",
"metadata": {},
"outputs": [],
"source": [
"paises_andinos_df = population_weighted_elevation[\n",
" population_weighted_elevation['country_name'].isin(PAISES_ANDINOS)\n",
"]"
]
},
{
"cell_type": "code",
"execution_count": 30,
"id": "numerical-bolivia",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"<AxesSubplot:xlabel='elevation_geometric', ylabel='prate'>"
]
},
"execution_count": 30,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 1417.32x708.661 with 2 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"fig, axs = pyplot.subplots(figsize=(50/2.54, 25/2.54), ncols=2)\n",
"\n",
"sns.scatterplot(data=paises_andinos_df, x='elevation_arithmetic', y='prate', size='population', ax=axs[0])\n",
"sns.scatterplot(data=paises_andinos_df, x='elevation_geometric', y='prate', size='population', ax=axs[1])\n"
]
},
{
"cell_type": "code",
"execution_count": 34,
"id": "adequate-testimony",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<table class=\"simpletable\">\n",
"<caption>OLS Regression Results</caption>\n",
"<tr>\n",
" <th>Dep. Variable:</th> <td>prate</td> <th> R-squared: </th> <td> 0.001</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Model:</th> <td>OLS</td> <th> Adj. R-squared: </th> <td> -0.010</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Method:</th> <td>Least Squares</td> <th> F-statistic: </th> <td> 0.1077</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Date:</th> <td>Thu, 30 Dec 2021</td> <th> Prob (F-statistic):</th> <td> 0.744</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Time:</th> <td>11:34:23</td> <th> Log-Likelihood: </th> <td> -1.8934</td>\n",
"</tr>\n",
"<tr>\n",
" <th>No. Observations:</th> <td> 87</td> <th> AIC: </th> <td> 7.787</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Df Residuals:</th> <td> 85</td> <th> BIC: </th> <td> 12.72</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Df Model:</th> <td> 1</td> <th> </th> <td> </td> \n",
"</tr>\n",
"<tr>\n",
" <th>Covariance Type:</th> <td>nonrobust</td> <th> </th> <td> </td> \n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <td></td> <th>coef</th> <th>std err</th> <th>t</th> <th>P>|t|</th> <th>[0.025</th> <th>0.975]</th> \n",
"</tr>\n",
"<tr>\n",
" <th>Intercept</th> <td> 0.4728</td> <td> 0.038</td> <td> 12.341</td> <td> 0.000</td> <td> 0.397</td> <td> 0.549</td>\n",
"</tr>\n",
"<tr>\n",
" <th>elevation_arithmetic</th> <td> 7.415e-06</td> <td> 2.26e-05</td> <td> 0.328</td> <td> 0.744</td> <td>-3.75e-05</td> <td> 5.23e-05</td>\n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <th>Omnibus:</th> <td>52.035</td> <th> Durbin-Watson: </th> <td> 1.863</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Prob(Omnibus):</th> <td> 0.000</td> <th> Jarque-Bera (JB): </th> <td> 187.955</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Skew:</th> <td> 1.976</td> <th> Prob(JB): </th> <td>1.54e-41</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Kurtosis:</th> <td> 9.020</td> <th> Cond. No. </th> <td>2.42e+03</td>\n",
"</tr>\n",
"</table><br/><br/>Notes:<br/>[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.<br/>[2] The condition number is large, 2.42e+03. This might indicate that there are<br/>strong multicollinearity or other numerical problems."
],
"text/plain": [
"<class 'statsmodels.iolib.summary.Summary'>\n",
"\"\"\"\n",
" OLS Regression Results \n",
"==============================================================================\n",
"Dep. Variable: prate R-squared: 0.001\n",
"Model: OLS Adj. R-squared: -0.010\n",
"Method: Least Squares F-statistic: 0.1077\n",
"Date: Thu, 30 Dec 2021 Prob (F-statistic): 0.744\n",
"Time: 11:34:23 Log-Likelihood: -1.8934\n",
"No. Observations: 87 AIC: 7.787\n",
"Df Residuals: 85 BIC: 12.72\n",
"Df Model: 1 \n",
"Covariance Type: nonrobust \n",
"========================================================================================\n",
" coef std err t P>|t| [0.025 0.975]\n",
"----------------------------------------------------------------------------------------\n",
"Intercept 0.4728 0.038 12.341 0.000 0.397 0.549\n",
"elevation_arithmetic 7.415e-06 2.26e-05 0.328 0.744 -3.75e-05 5.23e-05\n",
"==============================================================================\n",
"Omnibus: 52.035 Durbin-Watson: 1.863\n",
"Prob(Omnibus): 0.000 Jarque-Bera (JB): 187.955\n",
"Skew: 1.976 Prob(JB): 1.54e-41\n",
"Kurtosis: 9.020 Cond. No. 2.42e+03\n",
"==============================================================================\n",
"\n",
"Notes:\n",
"[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
"[2] The condition number is large, 2.42e+03. This might indicate that there are\n",
"strong multicollinearity or other numerical problems.\n",
"\"\"\""
]
},
"execution_count": 34,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"smf.ols('prate ~ elevation_arithmetic', data=paises_andinos_df).fit().summary()"
]
},
{
"cell_type": "code",
"execution_count": 35,
"id": "generous-ownership",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<table class=\"simpletable\">\n",
"<caption>OLS Regression Results</caption>\n",
"<tr>\n",
" <th>Dep. Variable:</th> <td>prate</td> <th> R-squared: </th> <td> 0.002</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Model:</th> <td>OLS</td> <th> Adj. R-squared: </th> <td> -0.010</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Method:</th> <td>Least Squares</td> <th> F-statistic: </th> <td> 0.1704</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Date:</th> <td>Thu, 30 Dec 2021</td> <th> Prob (F-statistic):</th> <td> 0.681</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Time:</th> <td>11:34:23</td> <th> Log-Likelihood: </th> <td> -1.8614</td>\n",
"</tr>\n",
"<tr>\n",
" <th>No. Observations:</th> <td> 87</td> <th> AIC: </th> <td> 7.723</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Df Residuals:</th> <td> 85</td> <th> BIC: </th> <td> 12.65</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Df Model:</th> <td> 1</td> <th> </th> <td> </td> \n",
"</tr>\n",
"<tr>\n",
" <th>Covariance Type:</th> <td>nonrobust</td> <th> </th> <td> </td> \n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <td></td> <th>coef</th> <th>std err</th> <th>t</th> <th>P>|t|</th> <th>[0.025</th> <th>0.975]</th> \n",
"</tr>\n",
"<tr>\n",
" <th>Intercept</th> <td> 0.4715</td> <td> 0.037</td> <td> 12.917</td> <td> 0.000</td> <td> 0.399</td> <td> 0.544</td>\n",
"</tr>\n",
"<tr>\n",
" <th>elevation_geometric</th> <td> 9.362e-06</td> <td> 2.27e-05</td> <td> 0.413</td> <td> 0.681</td> <td>-3.57e-05</td> <td> 5.45e-05</td>\n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <th>Omnibus:</th> <td>51.982</td> <th> Durbin-Watson: </th> <td> 1.871</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Prob(Omnibus):</th> <td> 0.000</td> <th> Jarque-Bera (JB): </th> <td> 187.208</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Skew:</th> <td> 1.975</td> <th> Prob(JB): </th> <td>2.23e-41</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Kurtosis:</th> <td> 9.004</td> <th> Cond. No. </th> <td>2.19e+03</td>\n",
"</tr>\n",
"</table><br/><br/>Notes:<br/>[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.<br/>[2] The condition number is large, 2.19e+03. This might indicate that there are<br/>strong multicollinearity or other numerical problems."
],
"text/plain": [
"<class 'statsmodels.iolib.summary.Summary'>\n",
"\"\"\"\n",
" OLS Regression Results \n",
"==============================================================================\n",
"Dep. Variable: prate R-squared: 0.002\n",
"Model: OLS Adj. R-squared: -0.010\n",
"Method: Least Squares F-statistic: 0.1704\n",
"Date: Thu, 30 Dec 2021 Prob (F-statistic): 0.681\n",
"Time: 11:34:23 Log-Likelihood: -1.8614\n",
"No. Observations: 87 AIC: 7.723\n",
"Df Residuals: 85 BIC: 12.65\n",
"Df Model: 1 \n",
"Covariance Type: nonrobust \n",
"=======================================================================================\n",
" coef std err t P>|t| [0.025 0.975]\n",
"---------------------------------------------------------------------------------------\n",
"Intercept 0.4715 0.037 12.917 0.000 0.399 0.544\n",
"elevation_geometric 9.362e-06 2.27e-05 0.413 0.681 -3.57e-05 5.45e-05\n",
"==============================================================================\n",
"Omnibus: 51.982 Durbin-Watson: 1.871\n",
"Prob(Omnibus): 0.000 Jarque-Bera (JB): 187.208\n",
"Skew: 1.975 Prob(JB): 2.23e-41\n",
"Kurtosis: 9.004 Cond. No. 2.19e+03\n",
"==============================================================================\n",
"\n",
"Notes:\n",
"[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
"[2] The condition number is large, 2.19e+03. This might indicate that there are\n",
"strong multicollinearity or other numerical problems.\n",
"\"\"\""
]
},
"execution_count": 35,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"smf.ols('prate ~ elevation_geometric', data=paises_andinos_df).fit().summary()"
]
},
{
"cell_type": "code",
"execution_count": 43,
"id": "classical-coast",
"metadata": {},
"outputs": [],
"source": [
"paises_andinos_df_ = paises_andinos_df.copy()\n",
"\n",
"paises_andinos_df_ = pd.merge(\n",
" paises_andinos_df_.groupby('country_name').median(), paises_andinos_df,\n",
" left_on='country_name',\n",
" right_on='country_name',\n",
" suffixes=['_country','']\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 44,
"id": "ranging-ghost",
"metadata": {},
"outputs": [],
"source": [
"paises_andinos_df_['prate'] = paises_andinos_df_['prate'] - paises_andinos_df_['prate_country']"
]
},
{
"cell_type": "code",
"execution_count": 45,
"id": "monetary-superior",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"<AxesSubplot:xlabel='elevation_geometric', ylabel='prate'>"
]
},
"execution_count": 45,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 1417.32x708.661 with 2 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"fig, axs = pyplot.subplots(figsize=(50/2.54, 25/2.54), ncols=2)\n",
"\n",
"sns.scatterplot(data=paises_andinos_df_, x='elevation_arithmetic', y='prate', size='population', ax=axs[0])\n",
"sns.scatterplot(data=paises_andinos_df_, x='elevation_geometric', y='prate', size='population', ax=axs[1])"
]
},
{
"cell_type": "code",
"execution_count": 41,
"id": "advance-mauritius",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<table class=\"simpletable\">\n",
"<caption>OLS Regression Results</caption>\n",
"<tr>\n",
" <th>Dep. Variable:</th> <td>prate</td> <th> R-squared: </th> <td> 0.007</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Model:</th> <td>OLS</td> <th> Adj. R-squared: </th> <td> -0.004</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Method:</th> <td>Least Squares</td> <th> F-statistic: </th> <td> 0.6244</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Date:</th> <td>Thu, 30 Dec 2021</td> <th> Prob (F-statistic):</th> <td> 0.432</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Time:</th> <td>11:40:01</td> <th> Log-Likelihood: </th> <td> 10.474</td>\n",
"</tr>\n",
"<tr>\n",
" <th>No. Observations:</th> <td> 87</td> <th> AIC: </th> <td> -16.95</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Df Residuals:</th> <td> 85</td> <th> BIC: </th> <td> -12.02</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Df Model:</th> <td> 1</td> <th> </th> <td> </td> \n",
"</tr>\n",
"<tr>\n",
" <th>Covariance Type:</th> <td>nonrobust</td> <th> </th> <td> </td> \n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <td></td> <th>coef</th> <th>std err</th> <th>t</th> <th>P>|t|</th> <th>[0.025</th> <th>0.975]</th> \n",
"</tr>\n",
"<tr>\n",
" <th>Intercept</th> <td> -0.9565</td> <td> 0.033</td> <td> -28.779</td> <td> 0.000</td> <td> -1.023</td> <td> -0.890</td>\n",
"</tr>\n",
"<tr>\n",
" <th>elevation_arithmetic</th> <td>-1.549e-05</td> <td> 1.96e-05</td> <td> -0.790</td> <td> 0.432</td> <td>-5.44e-05</td> <td> 2.35e-05</td>\n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <th>Omnibus:</th> <td>43.846</td> <th> Durbin-Watson: </th> <td> 1.651</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Prob(Omnibus):</th> <td> 0.000</td> <th> Jarque-Bera (JB): </th> <td> 139.764</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Skew:</th> <td> 1.661</td> <th> Prob(JB): </th> <td>4.47e-31</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Kurtosis:</th> <td> 8.246</td> <th> Cond. No. </th> <td>2.42e+03</td>\n",
"</tr>\n",
"</table><br/><br/>Notes:<br/>[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.<br/>[2] The condition number is large, 2.42e+03. This might indicate that there are<br/>strong multicollinearity or other numerical problems."
],
"text/plain": [
"<class 'statsmodels.iolib.summary.Summary'>\n",
"\"\"\"\n",
" OLS Regression Results \n",
"==============================================================================\n",
"Dep. Variable: prate R-squared: 0.007\n",
"Model: OLS Adj. R-squared: -0.004\n",
"Method: Least Squares F-statistic: 0.6244\n",
"Date: Thu, 30 Dec 2021 Prob (F-statistic): 0.432\n",
"Time: 11:40:01 Log-Likelihood: 10.474\n",
"No. Observations: 87 AIC: -16.95\n",
"Df Residuals: 85 BIC: -12.02\n",
"Df Model: 1 \n",
"Covariance Type: nonrobust \n",
"========================================================================================\n",
" coef std err t P>|t| [0.025 0.975]\n",
"----------------------------------------------------------------------------------------\n",
"Intercept -0.9565 0.033 -28.779 0.000 -1.023 -0.890\n",
"elevation_arithmetic -1.549e-05 1.96e-05 -0.790 0.432 -5.44e-05 2.35e-05\n",
"==============================================================================\n",
"Omnibus: 43.846 Durbin-Watson: 1.651\n",
"Prob(Omnibus): 0.000 Jarque-Bera (JB): 139.764\n",
"Skew: 1.661 Prob(JB): 4.47e-31\n",
"Kurtosis: 8.246 Cond. No. 2.42e+03\n",
"==============================================================================\n",
"\n",
"Notes:\n",
"[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
"[2] The condition number is large, 2.42e+03. This might indicate that there are\n",
"strong multicollinearity or other numerical problems.\n",
"\"\"\""
]
},
"execution_count": 41,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"smf.ols('prate ~ elevation_arithmetic', data=paises_andinos_df_).fit().summary()"
]
},
{
"cell_type": "code",
"execution_count": 42,
"id": "revolutionary-marketing",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<table class=\"simpletable\">\n",
"<caption>OLS Regression Results</caption>\n",
"<tr>\n",
" <th>Dep. Variable:</th> <td>prate</td> <th> R-squared: </th> <td> 0.005</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Model:</th> <td>OLS</td> <th> Adj. R-squared: </th> <td> -0.007</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Method:</th> <td>Least Squares</td> <th> F-statistic: </th> <td> 0.4338</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Date:</th> <td>Thu, 30 Dec 2021</td> <th> Prob (F-statistic):</th> <td> 0.512</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Time:</th> <td>11:40:05</td> <th> Log-Likelihood: </th> <td> 10.377</td>\n",
"</tr>\n",
"<tr>\n",
" <th>No. Observations:</th> <td> 87</td> <th> AIC: </th> <td> -16.75</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Df Residuals:</th> <td> 85</td> <th> BIC: </th> <td> -11.82</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Df Model:</th> <td> 1</td> <th> </th> <td> </td> \n",
"</tr>\n",
"<tr>\n",
" <th>Covariance Type:</th> <td>nonrobust</td> <th> </th> <td> </td> \n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <td></td> <th>coef</th> <th>std err</th> <th>t</th> <th>P>|t|</th> <th>[0.025</th> <th>0.975]</th> \n",
"</tr>\n",
"<tr>\n",
" <th>Intercept</th> <td> -0.9611</td> <td> 0.032</td> <td> -30.302</td> <td> 0.000</td> <td> -1.024</td> <td> -0.898</td>\n",
"</tr>\n",
"<tr>\n",
" <th>elevation_geometric</th> <td>-1.298e-05</td> <td> 1.97e-05</td> <td> -0.659</td> <td> 0.512</td> <td>-5.22e-05</td> <td> 2.62e-05</td>\n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <th>Omnibus:</th> <td>43.918</td> <th> Durbin-Watson: </th> <td> 1.653</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Prob(Omnibus):</th> <td> 0.000</td> <th> Jarque-Bera (JB): </th> <td> 139.283</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Skew:</th> <td> 1.667</td> <th> Prob(JB): </th> <td>5.69e-31</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Kurtosis:</th> <td> 8.226</td> <th> Cond. No. </th> <td>2.19e+03</td>\n",
"</tr>\n",
"</table><br/><br/>Notes:<br/>[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.<br/>[2] The condition number is large, 2.19e+03. This might indicate that there are<br/>strong multicollinearity or other numerical problems."
],
"text/plain": [
"<class 'statsmodels.iolib.summary.Summary'>\n",
"\"\"\"\n",
" OLS Regression Results \n",
"==============================================================================\n",
"Dep. Variable: prate R-squared: 0.005\n",
"Model: OLS Adj. R-squared: -0.007\n",
"Method: Least Squares F-statistic: 0.4338\n",
"Date: Thu, 30 Dec 2021 Prob (F-statistic): 0.512\n",
"Time: 11:40:05 Log-Likelihood: 10.377\n",
"No. Observations: 87 AIC: -16.75\n",
"Df Residuals: 85 BIC: -11.82\n",
"Df Model: 1 \n",
"Covariance Type: nonrobust \n",
"=======================================================================================\n",
" coef std err t P>|t| [0.025 0.975]\n",
"---------------------------------------------------------------------------------------\n",
"Intercept -0.9611 0.032 -30.302 0.000 -1.024 -0.898\n",
"elevation_geometric -1.298e-05 1.97e-05 -0.659 0.512 -5.22e-05 2.62e-05\n",
"==============================================================================\n",
"Omnibus: 43.918 Durbin-Watson: 1.653\n",
"Prob(Omnibus): 0.000 Jarque-Bera (JB): 139.283\n",
"Skew: 1.667 Prob(JB): 5.69e-31\n",
"Kurtosis: 8.226 Cond. No. 2.19e+03\n",
"==============================================================================\n",
"\n",
"Notes:\n",
"[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
"[2] The condition number is large, 2.19e+03. This might indicate that there are\n",
"strong multicollinearity or other numerical problems.\n",
"\"\"\""
]
},
"execution_count": 42,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"smf.ols('prate ~ elevation_geometric', data=paises_andinos_df_).fit().summary()"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.8.10"
}
},
"nbformat": 4,
"nbformat_minor": 5
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment