Visualizing COVID-19 Data with Shiny
Link to R Shiny project: https://philippe1.shinyapps.io/ShinyProject/
The skills the author demonstrated here can be learned through taking Data Science with Machine Learning bootcamp with NYC Data Science Academy.
Data Science Introduction
Based on data the number COVID-19-related deaths recently surpassed the 100,000+ mark in the United States and is sadly on track to keep growing before an effective vaccine can be researched, manufactured and successfully delivered to the millions of Americans who will depend on it to resume their lives. Until then, disseminating accurate and timely information on the spread of the disease and its associated death toll will be a crucial part of any effective and coordinated response as many world governments have already realized.
To this end, the purpose of this Shiny project was to offer the user different visualizations of US COVID-19 data recording new cases and deaths at the state and county level and which is published on a daily basis by the New York Times (see here for the repository).
Data Science Questions
The app shows bar charts, scatter plots, lollipop charts and state-level maps displaying case and death counts in the US, over time. These visualizations answer the following questions:
1. Which US states have the largest or smallest COVID-19 case counts today?
2. Which US states have the largest or smallest COVID-19 death counts today?
3. Which US states had the largest or smallest COVID-19 case counts at [x] date?
4. Which US states had the largest or smallest COVID-19 case counts at [x] date?
5. How have contagion rates evolved in [x] state since the beginning of the outbreak?
6. How have death rates evolved in [x] state since the beginning of the outbreak?
7. How did absolute COVID-19 case counts compare between [x] and [y] states at [z] date?
8. How did absolute COVID-19 related deaths counts compare between [x] and [y] states at [z] date?
9. How much of the total case count number in the US does [x] state account for?
10. How much of the total death count number in the US does [x] state account for?
Look in to the shiny app
The first page of the App displays a side-by-side histogram comparison of number of cases and COVID-19-related deaths for two states that the user selects using the two left-hand side menu dropdowns. The user can furthermore input the date for which they want to dynamically visualize the data on different dates using the bottom-most dropdown selector.
The second page displays a time-series scatter plot comparison of the growth in number of cases and deaths for two states selected by the user. The app dynamically updates the scatterplots once new state selections are made.
The third page of the app displays a lollipop chart comparison of cases and COVID-19 related deaths for different states. This chart perspective gives the user a more holistic view of state-by-state differences in COVID-19 cases in order to better understand differences at a more macro level.
Lastly, the fourth page presents a color-coded map view of COVID-19 cases and deaths. Similarly to the lollipop chart, this map intends to provide the user with a more macro view of US state differences in COVID-19 cases and deaths. As the user hovers their mouse over the different states shown in the map, pop-ups will reactively appear with the total number of cases and deaths for that state.
Conclusion
Future areas of focus to improve this App will be to join this dataset with a state and county demographic dataset in order to derive per capita death and case rates and to get a more accurate picture of which states and counties have fared the best in containing the spread of the disease. Taking this one step further, this demographic data could also be used to show scatterplot regressions between average age and per capita deaths or cases, literacy rates and per capita deaths or cases, poverty rates and per capita deaths or cases, etc to better inform policy makers as to which communities in which counties are the most vulnerable to the disease.
My contact info is [email protected] if you would like to discuss this project. Thanks for reading!
About Author
