4

I have created a function DevCstat().

It takes the arguments: indat, mod, Covar,txtMat, PatCovar.

indat is a list, I would like to apply the function to each element of the list.

mod, Covar, txtMat, PatCovar, are objects that I would like to use for each call to the function, these do not change over the list elements.

This seems to work:

lapply(test, DevCstat, mod='A', Covar=Covar,txtMat=txtMat, PatCovar=PatCovar)

However, the parallel version does not work:

mclapply(test,DevCstat,mod = 'A', Covar=Covar, txtMat=txtMat, PatCovar=PatCovar, SIMPLIFY = F, mc.cores = getOption("mc.cores", numcore))

I get the error 

all scheduled cores encountered errors in the user code

I think the problem is that mclapply is not passing the additional arguments.

Does anyone know how to correctly do this?

Thanks

user1375871
  • 1,199
  • 1
  • 12
  • 23

1 Answers1

7

You can see from the following example that mclapply does allow extra arguments in this way:

mclapply(2:4, function(i,j,k) c(i,j,k), i=1, k=5)

I think your problem is that mclapply doesn't take a SIMPLIFY argument, so it passes it on to your function causing the error. You might be getting confused with mcmapply, which does take a SIMPLIFY argument.

Steve Weston
  • 19,197
  • 4
  • 59
  • 75