Corona Virus And Social Distancing
Is this you?

Time To Go Out?

Lindsay Moir


It can be difficult to get accurate and time sensitive information on how countries are doing at present during the SARS-2 pandemic (COVID-19 is the illness, the pandemic is the virus). Exploring factors potentially related to outcomes is important as a means to consider further outcome prediction. Key factors are likely related to government and public health leadership and the potential for actual public adoption of mitigation strategies for the virus. In this post, I answer 4 questions that should provide some insight into this very important issue.

John Hopkins Hospital

What I Did

I took the John Hopkins Covid-19 data, which is the defacto centralized source for Covid-19 data globally. I then combined that with data on Median Per Capita Income and Population from Finally I added a column of data from the Transparency International dataset that provides an honesty score for each country. This allows me to answer the following 4 questions.

Q1: How are countries doing thus far dealing with Covid-19?

Ranking Of Countries To Covid-19 Response

The above graph shows who is performing the best and the worst on the covid-19 pandemic. The only countries included in this bar graph have a Transparency International score of >= 60 and a Median Per Capita Income >= 50% of the worlds countries. Countries with the shortest bars are doing better. Remember this IS adjusted for population. So, the fact that New Zealand is right at the top is interesting. We should be looking at Australia and New Zealand to understand how they achieved this.

The USA is in the bottom third and the UK is dead last. Perhaps a wake up call for their populations? The universe of countries that I chose was significantly restricted by having the above cutoffs.

Q2: Which countries are flattening the curve?

Germany Is Flattening The Covid-19 Curve

The term”Flattening The Curve” has come into popular use by the mainstream media. It describes a goal of seeing the number of covid cases being reduced day by day, for typically 14 days. If the “curve” is going down, that is a good thing. Going up, is NOT good.

For me in this notebook, this is defined as the number of New_Cases per day over the last 14 days. We use a scatter plot and a regression line that is fitted to the data. As you can see from the above example graph, Germany has flattened the curve. Here are the lists.

Flattening The Curve

[‘Australia’, ‘Austria’, ‘Belgium’, ‘Botswana’, ‘Denmark’, ‘Estonia’, ‘Finland’, ‘France’, ‘Germany’, ‘Israel’, ‘Japan’, ‘Lithuania’, ‘Luxembourg’, ‘Netherlands’, ‘New Zealand’, ‘Norway’, ‘Portugal’, ‘Singapore’, ‘Slovenia’, ‘Spain’, ‘Sweden’, ‘Taiwan’, ‘United Kingdom’, ‘United States’]

NOT Flattening The Curve

[‘Canada’, ‘Chile’, ‘Qatar’, ‘Uruguay’]

Chile Is Not Flattening The Covid-19 Curve

Chile is one of the countries that is NOT flattening the curve. We do not limit the countries based on Median Per Capita Income on these lists. The only cutoff is the Transparency International score needs to be >= 60.

Q3: Can I see a global geographic representation of infections?

Covid-19 Deaths Visualized By Country

The above image quickly tells you where the challenges are. Any country that shows color on the above map is in trouble. The grey countries are not reporting data. If you look at the ‘Transparency International Score By Country (below), you can immediately see the issue about reporting. For example, China is reporting but likely not being 100% transparent. Countries that ARE transparent are reporting and as a result often report higher deaths. The vast majority of countries are either not reporting or have challenges with being transparent.

Transparency International Score Visualized On A Global Map

Q4: What is the projected global mortality by December 31, 2020?

Test vs Prediction for Covid-19 Deaths Over Model Test Data Period

What would you do if you had a model that predicted this accurately (above) and then did a forecast like the next graph (below).

I created a machine learning model based on the Autoregressive Integrated Moving Average (ARIMA) technology. ARIMA is pretty much the gold standard for univariate time series prediction.

ARIMA predicts ~ 1.1 million people will not see 2021.

So, … Is It Time To Go Out?

There are some encouraging signs over the last 2 weeks that we are making progress. For example, 24 of the 28 countries looked at were flattening the curve. However, this analysis clearly points to some fundamental differences between countries in terms of their response to this crisis. When you control for population, transparency, and median income there are some countries that are doing much better (and obviously worse) than their peers.

In addition, we are probably only looking at the tip of the iceberg here. The vast majority of countries are not reporting or are under reporting their cases and deaths. It is extremely likely that, the 1.1 million number above (updated June 6, 2020) is grossly under stated.

The answer to this question really depends on where you live and what your financial situation is. Hopefully this analysis will give you some additional information upon which to make an informed decision.

Do you want to see how your country, province/state, city, is doing?

Do you want to predict the mortality right down to your individual city (if the data is there, which is VERY likely)?

Then go to my GitHub repo that has all of the Python/Pandas code for this notebook.

Please clap if you thought this post was insightful!



Lindsay Moir

I am a Data Scientist with deep industry experience in business, software, life sciences, bioinformatics, financial markets, and genetics.