Merge polygons and plot using spplot()

Question

I would like to merge some regions in gadm data and then plot the map. So far I have the following:

#install.packages("sp",dependencies=TRUE)
#install.packages("RColorBrewer",dependencies=TRUE)
#install.packages("maptools",dependencies=TRUE)
library(sp)
library(maptools)
#library(RColorBrewer)

# get spatial data
con <- url("http://gadm.org/data/rda/CZE_adm2.RData")
print(load(con))
close(con)

IDs <- gadm$ID_2
IDs[IDs %in% c(11500:11521)] <- "11500"
gadm_new <- unionSpatialPolygons(gadm, IDs)

# plot map
spplot(gadm_new, "NAME_2", col.regions=col, main="Test",colorkey = FALSE, lwd=.4, col="white")

However this results in error:

Error in function (classes, fdef, mtable)  : 
  unable to find an inherited method for function "spplot", for signature "SpatialPolygons"

Now I have no idea what can possibly fix this error.

Answer 1

I'm not sure about what you're trying to do here.

The error is due to the fact that spplot is used to draw spatial objects with attributes, ie with associated data. Your gadm object is of class SpatialPolygonsDataFrame, so it defines polygons and associated data that can be accessed via the slot gadm@data. When you use UnionSpatialPolygons, you only get a SpatialPolygons class object, which can be plotted with plot, but not with spplot :

IDs <- gadm$ID_2
IDs[IDs %in% c(11500:11521)] <- "11500"
gadm_new <- unionSpatialPolygons(gadm, IDs)
plot(gadm_new)

If you want to use spplot, you have to merge your associated data manually, the same way you merged your polygons, and then build back a SpatialPolygonsDataFrame. One way to do it is the following :

gadm_new <- gadm
## Change IDs
gadm_new$ID_2[gadm_new$ID_2 %in% c(11500:11521)] <- "11500"
## Merge Polygons
gadm_new.sp <- unionSpatialPolygons(gadm_new, gadm_new$ID_2)
## Merge data
gadm_new.data <- unique(gadm_new@data[,c("ID_2", "ENGTYPE_2")])
## Rownames of the associated data frame must be the same as polygons IDs
rownames(gadm_new.data) <- gadm_new.data$ID_2
## Build the new SpatialPolygonsDataFrame
gadm_new <- SpatialPolygonsDataFrame(gadm_new.sp, gadm_new.data) 

Then you can use spplot to plot a map with an associated attribute :

spplot(gadm_new, "ENGTYPE_2", main="Test", lwd=.4, col="white")

Note that here I only used the ENGTYPE_2 variable of your data, not the NAME_2 variable, as I don't see the point to represent a variable where each value seems unique for each polygon.