Folks who’ve been tracking this blog on R-bloggers probably remember [this post](https://rud.is/b/2014/11/16/moving-the-earth-well-alaska-hawaii-with-r/) where I showed how to create a composite U.S. map with an Albers projection (which is commonly referred to as AlbersUSA these days thanks to D3).
I’m not sure why I didn’t think of this earlier, but you don’t _need_ to do those geographical machinations every time you want a prettier & more inclusive map (Alaska & Hawaii have been states for a while, so perhaps we should make more of an effort to include them in both data sets and maps). After doing the map transformations, the composite shape can be saved out to a shapefile, preferably GeoJSON since (a) you can use `geojsonio::geojson_write()` to save it and (b) it’s a single file vs a ZIP/directory.
I did just that and saved both state and country maps out with FIPS codes and other useful data slot bits and created a small data package : [`albersusa`](https://github.com/hrbrmstr/albersusa) : with some helper functions. It’s not in CRAN yet so you need to `devtools::install_github(“hrbrmstr/albersusa”)` to use it. The github repo has some basic examples, heres a slightly more complex one.
### Mapping Obesity
I grabbed an [obesity data set](http://www.cdc.gov/diabetes/data/county.html) from the CDC and put together a compact example for how to make a composite U.S. county choropleth to show obesity rates per county (for 2012, which is the most recent data). I read in the Excel file, pull out the county FIPS code and 2012 obesity rate, then build the choropleth. It’s not a whole lot of code, but that’s one main reason for the package!
library(readxl) library(rgeos) library(maptools) library(ggplot2) # devtools::install_github("hadley/ggplot2") only if you want subtitles/captions library(ggalt) library(ggthemes) library(albersusa) # devtools::install_github("hrbrmstr/albersusa") library(viridis) library(scales) # get the data and be nice to the server and keep a copy of the data for offline use URL <- "http://www.cdc.gov/diabetes/atlas/countydata/OBPREV/OB_PREV_ALL_STATES.xlsx" fil <- basename(URL) if (!file.exists(fil)) download.file(URL, fil) # it's not a horrible Excel file, but we do need to hunt for the data # and clean it up a bit. we just need FIPS & 2012 percent info wrkbk <- read_excel(fil) obesity_2012 <- setNames(wrkbk[-1, c(2, 61)], c("fips", "pct")) obesity_2012$pct <- as.numeric(obesity_2012$pct) / 100 # I may make a version of this that returns a fortified data.frame but # for now, we just need to read the built-in saved shapefile and turn it # into something ggplot2 can handle cmap <- fortify(counties_composite(), region="fips") # and this is all it takes to make the map below gg <- ggplot() gg <- gg + geom_map(data=cmap, map=cmap, aes(x=long, y=lat, map_id=id), color="#2b2b2b", size=0.05, fill=NA) gg <- gg + geom_map(data=obesity_2012, map=cmap, aes(fill=pct, map_id=fips), color="#2b2b2b", size=0.05) gg <- gg + scale_fill_viridis(name="Obesity", labels=percent) gg <- gg + coord_proj(us_laea_proj) gg <- gg + labs(title="U.S. Obesity Rate by County (2012)", subtitle="Content source: Centers for Disease Control and Prevention", caption="Data from http://www.cdc.gov/diabetes/atlas/countydata/County_ListofIndicators.html") gg <- gg + theme_map(base_family="Arial Narrow") gg <- gg + theme(legend.position=c(0.8, 0.25)) gg <- gg + theme(plot.title=element_text(face="bold", size=14, margin=margin(b=6))) gg <- gg + theme(plot.subtitle=element_text(size=10, margin=margin(b=-14))) gg
Note that some cartographers think of this particular map view the way I look at a pie chart, but it’s a compact & convenient way to keep the states/counties together and will make it easier to include Alaska & Hawaii in your cartographic visualizations.
The composite GeoJSON files are in:
– `system.file(“extdata/composite_us_states.geojson.gz”, package=”albersusa”)`
– `system.file(“extdata/composite_us_counties.geojson.gz”, package=”albersusa”)`
if you want to use them in another program/context.
Drop an issue [on github](https://github.com/hrbrmstr/albersusa) if you want any more default fields in the data slot and if you “need” territories (I’d rather have a PR for the latter tho :-).
Pingback: Easier Composite U.S. Choropleths with albersusa – Mubashir Qasim
I’m trying to transpose your example above to my much more simple (binary) data. However, I’m using your new function counties_sf and having difficulty transposing the code. I assume, perhaps wrongly, that the code would be simpler with this new function. Would you kindly transpose the above code based on that new function?
Pingback: Accessing and Examining Covid-19 Data On Your Own – Data Science Austria