using a character string of row indices to subset a data frame in R

Question

I have a large data frame that I need to subset based on a character string of row indices provided in a single column within another table (from outside of R). These strings include several styles of index (e.g. c(1,3), c(1:10,15), 3:10, but all are formats that should be acceptable to R. I've provided a small example here, but in practice the indices typically refer to large row chunks within a very large data frame, so I'm keen to stick with this approach if possible.

Example data:

my.df<-data.frame(group=c("a","a","b","b","c","c"),value=c(1,5,5,15,20,3))

set1<-"c(1,3)"

set2<-"c(1:3,5)"

what I'd like:

my.df[set1,]

group value
1     a     1
3     b     5


my.df[set2,]

group value
1     a     1
2     a     5
3     b     5

I can read the index strings into R successfully and convert them to a form that looks right as per the code above, but haven't been successful with the subsetting. I'm sure there's an easy trick so any help is much appreciated.

Answer 1

Something like this maybe

 my.df[eval(parse(text=set1)),]