Small Multiples Map

Despite using R and its various graphing facilities I have somehow managed not to have stumbled across Bob Rudis. I recently noticed his ggalt package which has some similarities overlap to stuff I have tried to do in phylogeo. This inspired me to give a quick whirl to the small-multiples style of charts, of which the following is a small, reproducible example. When put into combination with the updated projections and higher-res mapdata of the [maps package] ( (natural earth 1:50m.)

After googling around looking for some links I’m seeing some really nice maps made with the current R-geography stack. Check out Austin Wehrwein and Bob Rudis. Very nice. I love d3.js for maps but its tough to compete with R for making static maps within your analytic workflow.

So heres a simple small multiple map with ggplot, and ggalt.

Loading the data#

#get data
dflats     <- read.csv(file="", sep=",")
twoletters <- read.csv(file="", sep=",")

crimedf <- USArrests %>%
    add_rownames(var = "state") %>%
    reshape2::melt() %>%
    left_join(twoletters, by=c("state" = "name_long")) %>%
    left_join(dflats,     by= c("name_short" = "state")) %>%
    filter(!name_short %in% c('AK','HI'))

usa <- map_data("usa")
ggplot() +  
    geom_map(data = usa,
             aes(x=long, y=lat, map_id=region), fill="gray90") +
    theme_bw() +
               aes(x= longitude, y=latitude, size=value)) +

plot of chunk small-multiples-map-1-4