mazamascience / covid-19 Goto Github PK

Some proof-of-concept code for generating movies exists in local_jon/state_map.R. These use the relatively new MazamaSpatialPlots package stateMap() function.

I would like to exercise MazamaSpatialPlots::stateMap() by making a series of COVID movies using state-level data. The COVID19* package make the data side of things very straightforward.

On my Mac I was unable to get the system(cmd) line to run but I could run the resulting cmd from the command line after I installed the latest version of ffmpeg. Once you have installed ffmpeg I would expect no problems running the R code on a linux box.

Subtasks include:

• Start with a single map for just confirmed_per_100.
• Figure out appropriate breaks so that we end up with a common legend scale for all images in a movie.
• Figure out appropriate sizing information for both tmap::tmap_save() and cmd_ffmpeg.
• Adjust other ffmpeg parameters as you see fit but do so using variables that we can define elsewhere.
• Create a 1-month movie as a test
• After everything has been sorted out, create a movie for all dates
• All of this should be encapsulated in a function that accepts a dataframe and a suite of parameters so that we can easily generate movies from other COVID parameters, or other state-level datasets.

Bonus task:

• Create a left-right side-by-side map of two parameters that are interesting together.
• Adjust sizing appropriately. (left-right vs. over-under may depend on the plotting space allotted)
• Create a full-year movie (all available data) for this left-right visualization.