• Points, Regions, Routes
• Variables can be continuous and/or discrete data
• Static or Interactive

library(choroplethr)
library(choroplethrMaps)
data(df_state_demographics)
names(df_state_demographics)

## [1] "region"            "total_population"  "percent_white"    
## [4] "percent_black"     "percent_asian"     "percent_hispanic" 
## [7] "per_capita_income" "median_rent"       "median_age"

kable(head(df_state_demographics))

dfstatemedrent=df_state_demographics[,c(1,8)] # Median Rent
colnames(dfstatemedrent)=c("region","value")
state_choropleth(dfstatemedrent, title="Median Rent by State")

data("df_county_demographics")
#names(df_county_demographics)
# Federal Information Processing Standard (FIPS) https://en.wikipedia.org/wiki/FIPS_county_code
kable(head(df_county_demographics))

dfcountymedrent=df_county_demographics[,c(1,8)] 
colnames(dfcountymedrent)=c("region","value")
county_choropleth(dfcountymedrent, title="Median Rent by County")

library(choroplethrZip)#devtools::install_github("arilamstein/choroplethrZip")
#Zip Code Tabulation Areas are generalized area representations of the United States Postal Service (USPS) ZIP code service areas
library(ggplot2)
data("df_zip_demographics")
kable(head(df_zip_demographics))

Use population density (population per area) rather than raw count for a sound spatial analysis. A good explanation of the trouble with choropleth maps of raw count data is available in Monmonier, Mark, Mapping It Out: Expository Cartography for the Humanities and Social Sciences (1993: University of Chicago Press, Chicago). Pages 159-167.

library(dplyr)
data(zip.map)  #Pull ZCTA area from `choroplethrZip` package data sets to compute density
zarea <- zip.map %>% group_by(region) %>% 
  summarize(zland = first(ALAND10), zwater = first(AWATER10))
df_pop_zip <- df_pop_zip %>% merge(zarea, all.x = TRUE)
df_pop_zip$popdensity <- with(df_pop_zip, count / (zland + zwater) * 1000 ^ 2)
kable(head(df_pop_zip))

Use an equal area projection since the measure is normalized by area.

df_pop_zip$value <- df_pop_zip$popdensity
zip_choropleth(df_pop_zip, state_zoom = 'washington', legend = 'Count per km^2') + 
  coord_map(projection = 'albers', lat0 = 41, lat1 = 47)  # equal area projection for pacific NW  http://spatialreference.org/ref/sr-org/nad83-us-pacific-northwest-albers/html/

zip_choropleth(df_pop_zip, county_zoom = 53063,  legend = 'Count per km^2') + 
  coord_map(projection = 'albers', lat0 = 41, lat1 = 47)

thingstodo=read.table(text="
                      Attraction  lat lon  Population
                      Jepson  47.667268 -117.405114 45
                      Starbucks  47.669088 -117.396847  50
                      WSU 47.660960 -117.405697 250
                      EWU 47.661061 -117.404044 400",header=TRUE)

library(ggmap)

whatislatlon=function(mydata,addressindata){
locs=geocode(as.character(unique(mydata[,addressindata])))
locs$address=unique(mydata[,addressindata])
mydata$latitude=locs$lat[ match(mydata[,addressindata],locs$address)]
mydata$longitude=locs$lon[ match(mydata[,addressindata],locs$address)]
return(mydata)
}

Address=c("502 E Boone Ave, Spokane, WA, 99258","502 E Boone Ave, Spokane, WA, 99258")
mydummydata=data.frame(Address=Address)
mysmartdata=whatislatlon(mydummydata,"Address")
kable(mysmartdata)

location=c(-117.402209,47.665330)
map=get_map(location=location,maptype="roadmap",source="google",zoom=16)
spokanemap=ggmap(map)
print(spokanemap)

spokanemap=spokanemap+geom_point(data=thingstodo,
            aes(lon,lat,color=Attraction),size=5)
print(spokanemap)

spokanemap+theme(panel.grid.major = element_blank(),
                 panel.grid.minor = element_blank(),
                axis.text = element_blank(),axis.title = element_blank(),
                axis.ticks = element_blank())

routes=data.frame(x=thingstodo$lon,y=thingstodo$lat)
newmap=get_googlemap(center=location,zoom=16,
                     markers=routes,
                     path = routes,scale=2,maptype = "satellite")
ggmap(newmap,darken=.3)+geom_text(data=thingstodo,aes(lon,lat,label=Attraction),
                                                  color="white",size=3)

• Interface to leaflet JS

library(leaflet) #rstudio package
leaflet() %>% addTiles()

leaflet() %>% addTiles()%>%addCircleMarkers(data=thingstodo)

leaflet() %>% addTiles()%>%addCircleMarkers(data=thingstodo,popup=~Attraction,radius=~Population*.05)

leaflet() %>% addTiles()%>%addMarkers(data=thingstodo,popup=~Attraction)

leaflet() %>% addTiles()%>%addMarkers(data=thingstodo,popup=~Attraction)%>% 
  addPolylines(thingstodo$lon,thingstodo$lat)

Some cool stuff possible:http://bit.ly/INRUGMapShiny

http://patilv.com/airpollution/

http://patilv.com/MortalityUnder5/

http://bit.ly/patilanimatedchoropleths

http://patilv.com/USOpenCountries/

• R-Studio: https://rstudio.github.io/leaflet/
• Ari Lamstein: http://www.arilamstein.com/
• Kyle Walker: http://walkerke.github.io/
• Erik Erhardt: http://statacumen.com/

  Thanks are due to the many package developers and the wonderful R community.