Question

My objective is to plot the path of a river with points indicating important sites near the river.

I have two data frames, giving the river and site coordinates respectively:

river<-data.frame(
    long=c(-2.816452494909265,-2.845487331898639,-2.883036393822358),
    lat=c(56.38229290416972,56.36346886284386,56.36577994637793))

samploc<-data.frame(
    site=c("Site1","Site2","Site3"),
    long=c(-2.826213585663894,-2.816519300644918,-2.868437228090127),
    lat=c(56.3649482229089,56.38166100310631,56.36716019476281))

Using an old school R plot, with par(new=T) and conserving xlim and ylim, I would get something like this:

old school plot http://users.utu.fi/susjoh/Riverplot.png

But I would like to do it using ggplot2. The river and points can be easily called individually:

ggplot(river,aes(x=long,y=lat)) + geom_path()
ggplot(samploc,aes(x=long,y=lat,lab=site)) + geom_point() + geom_text(vjust=2)

I have tried to cheat, by creating the following data frame from the previous two:

> rivsamp
  river.long river.lat samp.site samp.long samp.lat
1  -2.816452  56.38229      NA        NA       NA
2  -2.845487  56.36347      NA        NA       NA
3  -2.883036  56.36578      NA        NA       NA
4         NA        NA     Site1 -2.826214 56.36495
5         NA        NA     Site2 -2.816519 56.38166
6         NA        NA     Site3 -2.868437 56.36716

ggplot(rivsamp) +
  geom_path(aes(x=river.long,y=river.lat)) +
  geom_point(aes(x=samp.long,y=samp.lat)) +
  geom_text(aes(x=samp.long,y=samp.lat,lab=samp.site),vjust=2)

ggplot2 plot http://users.utu.fi/susjoh/riverggplot.png

It works, but creating this new data frame is not as straightforward as the old par(new=T) method.

Is there a simpler way to overplot from the individual data frames using ggplot2?

Thanks!

Answer 1

Here is one way to do it

ggplot(samploc, aes(x = long, y = lat)) + 
  geom_point() + 
  geom_text(aes(label = site), vjust = 2) + 
  geom_line(data = river, aes(y = lat))