deaths_gdp.csv

Cumulative deaths	Fall in GDP
0	1.73
5	0.36
7	-13.41
9	-4.41
13	-8.45
14	-13.09
20	-9.23
28	-14.18
34	-7.23
37	-5.73
43	-14.55
45	-7.55
51	-6.68
52	-7.68
65	-5.64
69	-14.23
78	-9.23
80	-24.77
93	-11.73
97	-14.27
98	-14.86
105	-11.00
117	-8.82
118	-11.45
128	-9.64
147	-5.82
158	-4.08
204	-17.18
208	-10.46
210	-8.49
219	-9.72
255	-13.38
281	-11.92
304	-16.73
372	-7.95
379	-9.91
488	-18.80
531	-19.68
554	-16.65
573	-8.14
600	-17.85
627	-10.02
641	-21.83
658	-17.97
700	-22.06
701	-10.55
713	-11.86
887	-15.09

deaths_gdp.py

# -*- coding: utf-8 -*-
"""
Deaths vs GDP fall in 2020

Response to the following tweet:

https://twitter.com/ianbremmer/status/1317924980659785728

@author: Nicolas Guarin-Zapata
@date: October 2020
"""
import numpy as np
import matplotlib.pyplot as plt
from scipy.stats import linregress


plt.style.use("fivethirtyeight")

deaths = np.loadtxt("deaths_gdp.csv", usecols=(0,), skiprows=1)
gdp = np.loadtxt("deaths_gdp.csv", usecols=(1,), skiprows=1)

# Linear regression
slope, intercept, r_value, p_value, std_err = linregress(deaths,
                                                         gdp)

# Plots
plt.plot(deaths, gdp, "o")
plt.plot(deaths, slope*deaths + intercept)
plt.text(600, 0, "$R^2 = %.3f$" % (r_value**2))
plt.xlabel("Cumulative deaths per million, Oct 14 2020")
plt.ylabel("Fall in GDP 2020 H1 (%)")
plt.savefig("deaths_gdp_fall.png", dpi=300)
plt.show()

deaths_gdp_fall.png