On Jan 22, 2013, at 2:24 PM, Dominic Roye wrote:
> Hello,
>
> How i can use a costum function in tapply which has more than one variable?
>
> I mean sum(x) only needs one object but what when i have a function
> function(x,y) with more, how i indicate where are the other variables
> to use?7
You can use:
lapply(split( multi_col_object, category_vec) , function(x,y){sum(x,y)} )
aggregate(dat, category, FUN=sum)
Or:
do.call(rbind, by( multi_col_object, category_vec, function(x,y){ ....} )
Sometimes `Reduce` is more compact. Other times `mapply` is needed.
--
David Winsemius
Alameda, CA, USA