U.S. Unemployment Data: Animated Choropleth Maps

August 4, 2010
By

(This article was first published on R-Chart, and kindly contributed to R-bloggers)


I was curious about creating maps with ggplot2 - and was happy to find that great minds have long since tackled this topic.  Folks who are interested in how to learn about R (and ggplot2) might be interested in the process.  (There should be a "greatest hits" listing related to R topics like this - the best I can think of to do is revive them in a forum like this).

After looking through the ggplot2 book (which had some simple examples) and the online documentation, I did a few searches and came across the topic on stackoverflow.  At the end of this post there is a reference to previous discussions in the R community.  It seems that an article on Flowing Data (using Python) inspired a bit of discussion of implementations of choropleth maps with R.  And so a challenge was issued to do this in R.  Several solutions were created and reviewed - and I decided to replicate the first response.  Fortunately Hadley Wickham included his solution in his github repository.  A slightly modified version of his original chart is shown below.



My revised version included a title, legend name and call to coord_map.  I followed the same process as the original challenge - downloaded the data from the Bureau of Labor Statistics and saved the relevant sheet as a csv.




m=ggplot(choropleth, aes(long, lat, group = group)) +
   geom_polygon(aes(fill = rate_d), colour = alpha("white", 1/2), size = 0.2) + 
   geom_polygon(data = state_df, colour = "white", fill = NA) +
   scale_fill_brewer(pal = "PuRd", name="Rate")


m + coord_map() + opts(title="US Unemployment Data (2009)")


Worked like a charm.  Note that there are some known problems with this solution.  Some additional mapping/data cleanup is required to get a complete view of the data set (which explains why Louisiana is missing).  I thought it would be neat to animate the results.  So I subcontracted out animation using Gimp to a local expert.  He created the following - a quick attempt to illustrate the changes over time.  It is not ideal, you kind of have to pick a point stare at it and watch for changes over time.

Incidentally, blogger does not support animated gifs directly, so I added the image in Google docs and linked to it from there.  This does not seem to work due to authentication issues.  Uploaded it to Picasa and you can view it there instead.

I know that there is an R animation package, but was not sure of its applicability in this scenario.  How would you create an animation that illustrates changes on a map over time?

To leave a comment for the author, please follow the link and comment on his blog: R-Chart.

R-bloggers.com offers daily e-mail updates about R news and tutorials on topics such as: visualization (ggplot2, Boxplots, maps, animation), programming (RStudio, Sweave, LaTeX, SQL, Eclipse, git, hadoop, Web Scraping) statistics (regression, PCA, time series, trading) and more...



If you got this far, why not subscribe for updates from the site? Choose your flavor: e-mail, twitter, RSS, or facebook...

Tags: ,

Comments are closed.