Get subplot of density histogram

[arno.kimeswenger]

Hi, maybe this is a fussy question, but I am interested in the following:
I have 2 data frames. The first one is just to have simple example. The second one is the well known diamonds data frame.

library(ggplot2)
my.data1<-data.frame(x=c(1,2,3,4,20))
my.data2<-diamonds

I want to build a density-histogram:

ggplot(my.data1)+
  geom_histogram(aes(x,y=after_stat(density)),breaks=c(0,10,20))
ggplot(my.data2)+
  geom_histogram(aes(x=carat,y=after_stat(density)),breaks=seq(0,6,length.out=400))

But now I am interested in a subplot:

ggplot(my.data1)+
  geom_histogram(aes(x=x,y=after_stat(density)),breaks=c(0,10,20))+
  xlim(0,10)
ggplot(my.data2)+
  geom_histogram(aes(x=carat,y=after_stat(density)),breaks=seq(0,6,length.out=400))+
  xlim(0,1)

But these are no real subplots of the above ones. These are new histograms which use a subset of the data.

For the simple data frame we get the width=10 and height=0.1 (so 10*0.1=1). But I am interested in width=10 and height=0.8 (because the remaining part is in the range (10,20)).

With diamonds data we can see a similar behaviour.

Do you know how to solve this issue?
Thanks for help!
Arno

[arno.kimeswenger]

I think the following does the trick:

ggplot(my.data1)+
  geom_histogram(aes(x=x,y=after_stat(density)),breaks=c(0,10,20))+
  coord_cartesian(xlim=c(0,3))

I got the idea from
http://rstudio-pubs-static.s3.amazonaws.com/209392_437ec4da7fa2432d831320f3591e7491.html