For each level of factor I need to extract values aggregated over all subsets of data.frame except the current one. For example, there is a several subjects doing a reaction time task during several days, and I need to compute mean reaction time for all subjects and all days, but not including the subject for whom the mean is computed. Currently, I do it like this:

 library(lme4)
 ddply(sleepstudy, .(Subject, Days), summarise , avg_rt=mean(sleepstudy[sleepstudy$Subject!=Subject&sleepstudy$Days==Days,"Reaction"]), .progress="text")

It works fine for small data sets, but for large ones it can be very slow. Is there a way to do it faster?