Been doing a lot of research and experimentation with interpolation of spatial data with R – for example trying to derive block level estimates of election results from election districts – assuming that blocks in Democratic leaning EDs near more Republican EDs will be more Republican. Practical, not sure but interesting. R has some really interesting interpolation libraries and it usually only requires a few lines of code thanks to the ease of the tidyverse.

Also been learning about doing essentially the reverse of that – taking point data or smaller polygons and aggregating it upwards to larger shapes. Tomorrow’s map of the day is an example of that – I calculated based on tax rolls the percentage of buildings each block in New York State where the buildings likely contain asbestos materials – buildings built from 1920 to 1979. Turns out tidyverse is good for that too, mutate a new field that shows if building in that range, do a st_join, then pipe into a group by and then summarize by what you want such as creating a percentage. Neat stuff.