5 Ways to Effectively Visualize Elapsed Time Data (COVID-19 Edition)
Published 26 March 2020/Blog
It’s crucial that visualizations are clear on the data they represent, especially around crises like a virus outbreak. The question is how do we present the data in the most effective way to help us analyze and understand what is going on in our country and around the world?
It’s been about two weeks since the lockdown started in the US due to COVID-19. We are still adjusting to new norms in our daily lives and everyone is checking the latest news in hopes of a decrease in new cases and deaths. We see a sea of data being presented and there are a lot of data visualizations around this topic. The majority of them are about confirmed cases, deaths and recoveries. But we also know that there are significant impacts on the economy due to this pandemic.
Stock Market Reacts to COVID-19
We have identified a few data sources to show the relationship between COVID-19 and stock prices. How fast did the market react? Is the price for crude oil down because of less demand for oil? In order to compare stock prices in a comparable manner, we have calculated indexes for each stock by taking the price on January 22, 2020, to be 100.
The crude oil price was slightly going upwards at the end of 2019. In fact, it was one of the highest in 5 years. However, the price started declining in mid-January, and by the end of March, the price dropped to less than a half. Nikkei 225 (^N225) was hit negatively when Japan started seeing an increase in new cases in mid-January. The US noticed a rise in confirmed cases the second week of March; however, the US market had started a decline a few weeks prior. Since then, it has been decreasing consistently. How do I know the correlation? It’s because the increase of COVID-19 cases is also captured in the background within the same chart. This is a “seamless” way to capture two different elements (stock prices vs cases) in one chart without distracting each other.
Time-Series Snapshot Bar Charts
The best way to see a ranking is by using a bar chart. It’s the simplest visualization format and ANYONE can easily understand the information.
By placing multiple bar charts next to each other as a snapshot of a time series is an effective way to tell a story. You can see how the ranking changes rapidly every few weeks. Having the reference line to point to the stock price line chart below is also effective to give a specific timeline to associate two different charts.
Next, let’s look at confirmed cases from a different angle. In this distribution plot, each dot represents a reported date. January is green, February is yellow, and March is orange. Only a few countries display green dots, and China is the only country with the highest concentration of visible green dots since COVID-19 originated from there. Yellow dots are visible for a few more countries including China, Italy, Iran, South Korea, and Japan. The majority of countries mainly have orange dots. These clusters of colored dots indicate how and when COVID-19 cases have increased. This is a unique and effective way to show the timing of data flow over time.
Spread and Density
Instead of using color for each month, I have colored squares based on the number of cases reported each day. Clearly, China showed an epidemic earlier than any other country and numbers there have since decreased when other countries started showing spread by a higher volume (darker-color squares). Also, while Italy has dark squares spread out, squares for the US are more concentrated towards the end of the graph (right) and have much darker intense colors. Unlike the above distribution plot visualization, a heatmap gives equal weight in a time series so that you can see a pattern of spread as well as its density.
Another way to show the same information is to sort the chart by measure, in this scenario, the number of cases. Regardless of when the cases started, you see the ranking by states for total cases.
Geography Spread Stages
Placing several maps side by side gives a time dimension to the overall picture. This shows each stage of the virus spread in the US. A map is a very effective way of displaying data without necessary needing additional explanations.
From the same dataset, you can tell several different stories. Data visualization is all about what to communicate and how to best communicate it. You’ve seen fancy wow-factor visualizations before. But if it doesn’t tell a story, it’s just a pretty picture to look at. You want readers to understand the information within 5 seconds, or at least that has always been my objective when creating a data visualization.
COVID-19 is affecting all of us and it’s very difficult to comprehend that only a couple of weeks ago, we had a “normal” life. We’re trying to adapt to this new norm and not knowing when we can go back to our old norm is daunting, but let’s start with what we can do today. That is washing our hands and self-quarantine and we will get through this together. Help flatten the curve and stay safe!
Infographics: Apeksha Pathak
BI (Qlik/Tableau): Duc Ngo
Sources: www.finance.yahoo.com, www.kaggle.com
An updated infographic with data through May is available here.
Published 2 June 2023
TX Tuesdays provide short technical, how-to videos about TimeXtender and the data platforms they support. Here you'll find informa...
Infozone Intelligence Celebrates 10th Anniversary
Published 28 April 2023
The word-cloud image above shows feedback from our recent customer 360 survey and for the past 10 years Infozone Intelligence has ...