rworldmap :
world maps from
country data
Andy South
[email protected]
andysouth.co.uk
CambridgeR, November 27th 2014
1
2
What do you want to say with your
global data ?
1. Geographical pattern
2. Identify key countries quickly
3. Ranking
4. Relative values
5. Extremes
3
rworldmap
• Released 2010
• Aims
improve visualisation of global data in R
easy to start, flexible to extend
publication quality outputs
aid interdisciplinary work
• ~50 downloads per day via Rstudio mirror
• Contains up-to-date country boundaries
• RJournal article
4
The R Journal Vol. 3/1, June 2011
http://journal.r-project.org/archive/2011-1/RJournal_2011-1_South.pdf
5
rworldmap shiny UI
https://andysouth.shinyapps.io/rworldmapUI/
6
rworldmap uses package sp
• spatial classes
• Spatial Polygons Dataframe (sPDF)
@data
7
maps in rworldmap are
Spatial Polygons Dataframes
sPDF <- getMap()
class(sPDF)
plot(sPDF)
#to get at the dataframe
dF <- sPDF@data
str(dF)
8
vector
raster
9
Joining 1
Your data
Map data
country
temperature
UK
Spain
cold
hot
Columns in common
10
Joining 2
dF <- data.frame( country=c("spain", "uk"),
weather=c("hot", "cold") )
sPDF <- joinCountryData2Map( dF,
joinCode = "NAME",
nameJoinColumn = "country")
mapCountryData( sPDF, nameColumnToPlot= "weather")
11
Note that these data are categorical & rworldmap copes.
12
HOWEVER …
Joining using country codes instead of country
names avoids potential pitfalls.
13
14
• North Korea
Korea, Democratic People's Republic of
Democratic People's Republic of Korea
Korea, Dem. Rep.
• South Korea
Korea, Republic of
Republic of Korea
Korea
15
Côte d'Ivoire
was my
bête noire
16
17
Joining data using country codes
#example data in rworldmap
data(countryExData)
sPDF <- joinCountryData2Map( countryExData,
joinCode = "ISO3",
nameJoinColumn = "ISO3V10")
mapParams <- mapCountryData(sPDF,
nameColumnToPlot="EPI")
18
Examples with world bank indicators
library(WDI)
# http://data.worldbank.org/indicator/
# Improved sanitation facilities (% of popn)
indicator <- "SH.STA.ACSN"
dFsanitation <- WDI( indicator=indicator,
start=2005,
end=2005 )
19
i. choropleth map
library(rworldmap)
sPDFsanitation <joinCountryData2Map( dFsanitation,
nameJoinColumn= "iso2c",
joinCode='ISO2')
library(RColorBrewer)
numCats <- 5
colourPalette <- rev(brewer.pal(numCats, "PuBuGn"))
mapCountryData( sPDFsanitation,
nameColumnToPlot=indicator,
colourPalette=colourPalette,
numCats=numCats )
20
ii. bubble map
library(rworldmap)
mapBubbles( sPDFsanitation ,
nameZSize = indicator ,
nameZColour = 'blue' ,
fill = FALSE )
#optional interactive labelling of countries
identifyCountries( sPDFsanitation ,
nameColumnToPlot = indicator )
22
iii. Ranked bar plot
library(rworldmap)
barplotCountryData( sPDFsanitation ,
nameColumnToPlot=indicator ,
nameCountryColumn = "NAME" ,
numCats = numCats ,
colourPalette = colourPalette ,
na.last = NA ,
decreasing = FALSE ,
scaleSameInPanels = TRUE ,
numPanels = 7 ,
cex = 1.1 )
24
Access to sanitation from barplotCountryData()
non world maps
## US states map downloaded from :
## http://www2.census.gov/cgibin/shapefiles2009/national-files
inFile <- 'tl_2009_us_state.shp'
sPDF <- readShapePoly(inFile)
str(sPDF@data)
## use mapPolys to map the sPDF
mapPolys(sPDF,nameColumnToPlot = "AWATER“
,mapRegion='North America')
26
non world maps
27
• These are the basics of rworldmap
• Add bits together to make more
interesting plots …
28
http://stackoverflow.com/a/19695755/1718356
29
https://alienplantation.wordpress.com/2013/06/18
/nuts-and-bolts-modern-maps-of-eurasia-in-r/30
Things to remember
• spatial polygons dataframes
• joining country data to maps
• country codes
• rworldmap::barplotCountryData()
31
rworldmap positives
• modern maps (including South Sudan)
• matches multiple names for same country &
reports on failures in country matching
• country codes (ISO 2 letter, 3 letter & numeric)
• maps at 3 resolutions
• copes with map projections and holes
• flexibility
32
https://andysouth.shinyapps.io/shinytse7/
33
Meta-mapping
(using rworldmap to map the downloads of …)
[email protected]
andysouth.co.uk
34
Xtra bits after here …
35
modifying map legends
data(countryExData)
sPDF <- joinCountryData2Map(countryExData ,
joinCode = "ISO3" ,
nameJoinColumn = "ISO3V10")
mapParams <- mapCountryData(sPDF ,
nameColumnToPlot="EPI" ,
addLegend=FALSE )
do.call( addMapLegend ,
c(mapParams ,
legendLabels="all" ,
legendWidth=0.5 ))
36
adding multiple elements to a plot
mapCountryData()
mapBubbles( getMap()
, nameZSize="GDP_MD_EST"
, nameZColour="black"
, add=TRUE )
37
interactive identification & labelling
mapCountryData(nameColumnToPlot="POP_EST")
identifyCountries(nameColumnToPlot="POP_EST")
Then click close to the centre of countries to bring up
country labels.
38
Resources
http://andysouth.co.uk/softwarerworldmap/
https://groups.google.com/group/rworldmap
39
40