Incorporating splitstackshape into loop

Question

I have the following code that selects (4 rows of iris x 1000) *100 and calculates the bias of each column.

library(SimDesign)
library(data.table)

do.call(rbind,lapply(1:100, function(x) {
  bias(
    setDT(copy(iris))[as.vector(sapply(1:1000, function(X) sample(1:nrow(iris),4)))][
      , lapply(.SD, mean), by=rep(c(1:1000),4), .SDcols=c(1:4)][,c(2:5)],
    parameter=c(5,3,2,1), #parameter is the true population value used to calculate bias
    type='relative' #denotes the type of bias being calculated 
  )
}))

This takes 1000 samples of 4 rows, calculates the mean by sample #, giving me 1000 means. The bias for the 1000 means is found for each column, and then is done 99 more times giving me a distribution of bias estimates for each column. This is mimicking a random sampling design. However, I also want to do this for a stratified design. So I use splitstackshape's stratified function.

do.call(rbind,lapply(1:100, function(x) {
  bias(
    setDT(copy(iris))[as.vector(sapply(1:1000, function(X) stratified(iris,group="Species", size=1)))][
      , lapply(.SD, mean), by=rep(c(1:1000),4), .SDcols=c(1:4)][,c(2:5)],
    parameter=c(5,3,2,1), 
    type='relative'
  )
}))

I would've thought that it is just a matter of swapping out the functions, but I keep on getting errors (i is invalid type (matrix)). Perhaps in future a 2 column matrix could return a list of elements of DT . I think it might be something related to setDT, but I'm not really sure how to fix it. Anybody know where I'm going wrong?