Overlapping a portion of an image with a Google map in R

Question

I am trying to add this graph of a function defined on Veneto (Italian region)

obtained using image and contour :

 image(X,Y,evalmati,col=heat.colors(100), xlab="", ylab="", asp=1,zlim=zlimits,main=title) contour(X,Y,evalmati,add=T) 

(here you can find objects: https://dl.dropboxusercontent.com/u/47720440/bounty.RData )

on the background of a Google map.

I tried two ways:

RGoogleMaps PACK

I uploaded a mbackground map

 MapVeneto<-GetMap.bbox(lonR=c(10.53,13.18),latR=c(44.7,46.76),size = c(640,640),MINIMUMSIZE=TRUE) PlotOnStaticMap(MapVeneto) 

but I don’t know the commands useful for adding a graph defined by image and contour to the map

PACKAGE loa

I tried this way:

 lat.loa<-NULL lon.loa<-NULL z.loa<-NULL nx=dim(evalmati)[1] ny=dim(evalmati)[2] for (i in 1:nx) { for (j in 1:ny) { if(!is.na(evalmati[i,j])) { lon.loa<-c(lon.loa,X[i]) lat.loa<-c(lat.loa,Y[j]) z.loa<-c(z.loa,evalmati[i,j]) } } } GoogleMap(z.loa ~ lat.loa*lon.loa,col.regions=c("red","yellow"),labels=TRUE,contour=TRUE,alpha.regions=list(alpha=.5, alpha=.5),panel=panel.contourplot) 

but the plot was not like the first:

in the legend of this plot, I have 7 colors, and the plot uses only these values. image graph is more accurate.

How to add image graph to GoogleMaps background?

Answer 1

If using a GoogleMap map is optional (for example, if you only need to visualize the coastline + depth / height information on the map), you can use the marmap package to accomplish what you want. Please note that you will need to install the latest version of marmap available on github to use readGEBCO.bathy() , as the file format generated when downloading GEBCO files has been changed recently. Data from NOAA servers is beautiful, but not very accurate in the region you are interested in (only a one-minute resolution versus half a minute for GEBCO). Here is the data from GEBCO that I used to create the map: the GEBCO file

 library(marmap) # Get hypsometric and bathymetric data from either NOAA or GEBCO servers # bath <- getNOAA.bathy(lon1=10, lon2=14, lat1=44, lat2=47, res=1, keep=TRUE) bath <- readGEBCO.bathy("GEBCO_2014_2D_10.0_44.0_14.0_47.0.nc") # Create color palettes for sea and land blues <- c("lightsteelblue4", "lightsteelblue3", "lightsteelblue2", "lightsteelblue1") greys <- c(grey(0.6), grey(0.93), grey(0.99)) # Plot the hypsometric/bathymetric map plot(bath, land=T, im=T, lwd=.03, bpal = list(c(0, max(bath), greys), c(min(bath), 0, blues))) plot(bath, n=1, add=T, lwd=.5) # Add coastline # Transform your data into a bathy object rownames(evalmati) <- X colnames(evalmati) <- Y class(evalmati) <- "bathy" # Overlay evalmati on the map plot(evalmati, land=T, im=T, lwd=.1, bpal=col2alpha(heat.colors(100),.7), add=T, drawlabels=TRUE) # use deep= shallow= step= to adjust contour lines plot(outline.buffer(evalmati),add=TRUE, n=1) # Outline of the data # Add cities locations and names library(maps) map.cities(country="Italy", label=T, minpop=50000) 

Since your evalmati data evalmati now evalmati objects, you can adjust its appearance on the map, as well as for the background of the map (adjust the number and width of contour lines, adjust the color gradient, etc.). plot.bath() uses both image() and contour() so you can get the same results as when building with image() . Please see plot.bathy() help and package vignettes for more examples.

Answer 2

I'm not quite at that, but Lovelace, R. "An Introduction to Spatial Data Visualization in R" can help you https://github.com/Robinlovelace/Creating-maps-in-R/raw/master/intro-spatial- rl.pdf From the section “Adding base maps to ggplot2 with ggmap” with minor changes and data from https://github.com/Robinlovelace/Creating-maps-in-R/archive/master.zip

 library(dplyr) library(ggmap) library(rgdal) lnd_sport_wgs84 <- readOGR(dsn = "./Creating-maps-in-R-master/data", layer = "london_sport") %>% spTransform(CRS("+init=epsg:4326")) lnd_wgs84_f <- lnd_sport_wgs84 %>% fortify(region = "ons_label") %>% left_join( lnd_sport_wgs84@data , by = c("id" = "ons_label")) ggmap(get_map(location = bbox(lnd_sport_wgs84) )) + geom_polygon(data = lnd_wgs84_f, aes(x = long, y = lat, group = group, fill = Partic_Per), alpha = 0.5)