Visualizing the Vaccine

March 2021 · 3 minute read

Hello March, it’s been a few months since the first vaccinations took place and I find myself Googling about COVID-19 more than ever. Some questions I had:

“What types of vaccines are used?”

“What country is vaccinating the most?”

“How fast is the US vaccinating?”

“What state is vaccinating the most?”

“When will I be eligible for a vaccine?”

I wanted all the information I could get my hands on. Fortunately, there are datasets that answer my questions. The data comes from Our World in Data’s GitHub repository, Kaggle’s COVID-19 World Vaccination Progress dataset, Kaggle’s Countries of the World dataset, and a state postal codes dataset. Let’s get started!

What countries are using mRNA, non-mRNA, or both?

To see the different types of vaccines countries use, a stacked bar plot from Matplotlib or Seaborn work, but I wanted something a little more “worldly”. Instead I used Plotly Express to generate a Choropleth Map. Making choropleth maps requires two main types of input:

  1. Geometry information

    • I used plotly built-in geometry for world countries.
  2. A list of values indexed by feature identifier.

There’s some interesting geopolitics happening here. Most of the Global South use non-mRNA vaccines, while most of the Global North use mRNA vaccines. Of course, we see some Global North countries in Europe use both types of vaccines.

What countries are vaccinating the fastest?

Next, I wanted to look at the top 10 countries with highest vaccinations per hundred. To compare the United States with other larger countries, I filtered by countries with populations >= 10 million. A Plotly Express Bar Chart works great here.

In this order, Great Britain, United States, Chile, Turkey, Morocco, Poland, Greece, Portugal, Spain, and Romania. Great Britain leads with 30% of vaccinations, the United States at 20%, Chile at 18% and all other countries with <= 10% vaccinations.

How do these countries look over time?

A standard Plotly Express Line Plot shows that Great Britain and Chile started with an upward trend, but recently saw dramatic dips. The United States experienced a slight dip in daily vaccinations in mid-February, but now trends upwards.

How well are the states doing in vaccinating?

For this plot, I used another Plotly Express Choropleth map, but this time with Plotly’s built-in geometry for states.

Alaska, New Mexico, North Dakota, and South Dakota are the leading in total vaccinations per hundred. I initially thought states with higher density of people would vaccinate faster, but the data proved me wrong.

What are the United States total vaccinations over time?

Good news, everyone! Over time, the United States total vaccinations has increased exponentially.

Will we reach 150 million vaccinations in 100 days?

On January 25, 2021, President Joe Biden pledged 150 million vaccinations (shots in arms) in 100 days. Let’s see what the data can tell us!

For predicting future vaccination numbers, I used open source software, called Prophet. According to Prophet, this forecasting procedure handles seasonality and outliers well without too much tuning. Prophet also lets you make forecasts using a logistic growth trend model with a specified carrying capacity. Our carrying capacity is the population of the United States.

Prophet’s model says by April 30, 2021 around 220 million vaccinations will have happened. It looks like President Joe Biden’s prediction is a little conservative!

I know that it feels like a lifetime, but we are on our way to getting those shots to folks. I hope you will have a more positive outlook on what the future has in store! If you like what you’ve read, be sure to check out my talk on “Visualizing the Vaccine” and its corresponding Jupyter Notebook.