so
m + geom_line(data=dfmed, color="red", aes(group=1,y=V1))
or
dfmed <- ddply(df, .(bl, jahr), function(x) c(v1=median(x$v1)))
m + geom_line(data=dfmed, color="red", aes(group=1))
will do that.
> --
> You received this message because you are subscribed to the ggplot2 mailing list.
> Please provide a reproducible example: http://gist.github.com/270442
>
> To post: email ggp...@googlegroups.com
> To unsubscribe: email ggplot2+u...@googlegroups.com
> More options: http://groups.google.com/group/ggplot2
>