Posts

Introduction The long term objective for this modelling project is to investigate the relationship between demographic variables such as race, ethnicity, gender, and others and the mortality rates due to COVID-19. The primary dataset that I will be using for this project is the number of deaths per county in the US, tabulated daily. This means that each county is a timeseries of counts of deaths due to COVID-19. Time series models and especially regressions with timeseries is a large subject area, and in this blog post we are going to dive into the basics of timeseries modelling with the added twist of working with count data.

Introduction This blog post is dedicated to downloading and cleaning data for an exploratory analysis of COVID-19 in the United States. I will be using COVID-19 cases and deaths data from the New York Times, demographic and spatial data from the US Census Bureau, and social distancing data from the Opportunity Project. The end goal is to have four separate tidy datasets: two that contain a unique identifier and spatial geometries for the counties and states of the US, and two that contain combined time series and demographic data for US counties and states. Along the way, we will deal with idiosyncrasies between how different data is collected, and with a small number of missing variables.