OpenStreetMapR an R interface to Leaflet

Created by Royalty Analytics

GitHub Project page link:

https://github.com/greentheo/OpenStreetMapR

A Cool Example

Here’s a neat example to start with:

## Loading required package: RJSONIO
## Loading required package: tools

df = data.frame(state=sample(unique(map_data("state")$region), 49,replace=F),
                metric=runif(49)
)

#choropleth by state with custom colors assigned to the state
df[["hexColor"]]=sample(rgb(runif(10), runif(10), runif(10)), nrow(df), replace = T)
choroMap = choroplethMap(df, state='state', plotVariable='metric', color="hexColor", layerName = 'States')

#over lay some points and lines
df = data.frame(lat=runif(100, 30,50), long=-runif(100, 70,120), 
                size=runif(100)*20, color=sample(letters[1:3], 100, replace=TRUE),
                group=sample(1:3, 100,replace=TRUE),
                greys=gray(seq(0,1,length.out=100)))

pointPlot = OSMMap(df, color='greys', size='size', layer = 'Points',colorByFactor = F)

print(addLayers(choroMap, pointPlot))

Plots with RMarkdown and OpenStreetMapR

Open StreetMapR is designes to work either interactively with plots or interactively within an RMarkdown document. Reporting and reproducible research at it’s best!

This page is (mostly) generated from the RMarkdown @ inst/examples/RMrkdownExample.Rmd inside the package. Start there for a feel for how OpenStreetMapR works in Rmd.

There’s also a shiny app in the inst/examples/shinyApp folder which makes use of the package. Check it out.

Package loading and dataset creation

First let’s take a look at a dataset.

library(OpenStreetMapR)
library(htmltools)
library(knitr)
library(ggplot2)
df = data.frame(lat=runif(100, 38,40), long=-runif(100, 104,106), 
                size=runif(100)*20, color=sample(rainbow(10), 100, replace=T),
                popup=rep(sample(letters, 10),100))

head(df)

Plot in seperate browser window

You can plot this in it’s own window easily with traditional R plot interface

plot(OSMMap(df))
plot(OSMMap(df, size='size', color='color'))

Include in RMarkdown

But you can also include this data as part of your RMarkdown just as easily, just print the OSMap object and make sure to turn off echo and results=“asis”.

There are also extended print options to control Titles and subtitles.

### This code block is just to show what the next inline plot is generated from.
print(OSMMap(df, size='size', color='color', popup='popup'))
print(OSMMap(df, size='size', color='color'), 
      title='OpenStreetMapR',
      subtitle='<a href="https://github.com/greentheo/OpenStreetMapR/">Github Link</a>',
      desc='',
      returnText=F)

Shiny Apps

Including this in a shiny app is just as easy, but now you’ll have to inject the results into a renderHTML ui object and set returnText=T.

A load of HTML text will be returned.

print(OSMMap(df, size='size', color='color'), 
      title='OpenStreetMapR',
      subtitle='<a href="https://github.com/greentheo/OpenStreetMapR/">Github Link</a>',
      desc='',
      returnText=T)

Choropleth Maps

Choropleth maps are straight forward.

counties = map_data("county")
df = data.frame(state="colorado", county=unique(counties[counties$region=='colorado',"subregion"]),
                metric=runif(64)
)
choroMapRed = choroplethMap(df, state='state', county='county',plotVariable='metric',palette = 'red', layerName='Red Layers')

print(choroMapRed, 
     title = 'A Random coloring of Counties',
     subtitle = 'With Red, Blue and Green Color scales',
     desc='The colors range from white to dark, darker = higher level of metrics.')

Even at the township level (Kansas and Oklahoma data included…roll your own for other states).

A selection of 100 twps across KS and OK.

#and one at the township level
twp = readRDS('~/githubrepo//OpenStreetMapR/inst/OK.rds')
df = data.frame(township=sample(unique(twp$region), 100),
                metric=sample(1:4, 5, replace = T)
)

choroMapRed = choroplethMap(df, state=NULL,township='township',plotVariable='metric',palette = 'red', layerName='Red Township Layers',townshipData = twp)
print(choroMapRed)

And Beyond

So are polygon maps:

#generate rand lats and longs in Colorado
df = data.frame(lat=runif(100, 38,40), long=-runif(100, 104,106), 
                size=runif(100)*20, color=sample(letters[1:3], 100, replace=TRUE),
                group=sample(1:3, 100,replace=TRUE))

polyPlot = OSMMap(df, group='group', color='color', groupType="Polygon")
print(polyPlot)

Adding seperate maps into a multi layered map is also straightforward. Let’s do a choropleth of Denver and Nebraska with some points laid on top and lines connecting them.

#generate rand lats and longs in Colorado and New Mexico
df = data.frame(lat=runif(100, 33,41), long=-runif(100, 102,108), 
                size=runif(100)*20, color=sample(letters[1:3], 100, replace=TRUE),
                group=sample(1:3, 100,replace=TRUE))
dfState = data.frame(state=c('colorado','new mexico'), metric=c(1,2))

linePlot = OSMMap(df, group='group', color='color', groupType="LineString", layer='Lines')
pointPlot = OSMMap(df, color='color', size='size', layer = 'Points')
choroPlot = choroplethMap(dfState, state='state',plotVariable = 'metric', layerName = 'States', static=T)
comboPlot = addLayers(choroPlot, 
                addLayers(linePlot, pointPlot ))
print(comboPlot)

The OSMMap function can also have a single color applied to points, lines and polygons, or it can have a vector of points specified. It will (sort of) figure it out.

#generate rand lats and longs in Colorado and New Mexico

pointPlot = OSMMap(df, color='#FF0000', size='size', layer = 'Points',colorByFactor = F)
print(pointPlot)

df[["greys"]] = gray(seq(0,1,length.out=100))
pointPlot = OSMMap(df, color='greys', size='size', layer = 'Points',colorByFactor = F)
print(pointPlot)