This would be good for consistency across the fi_ family of functions. Also, it would allow the implementation of a fi_pool_random() much more easily and consistently with fi_pool_cluster_random().
Downside is that it is slightly more clunky to vectorise fi_pool wrt pool_size and pool_number, but I don't think is major. Would need to check that fi_pool isn't being used in vectorised form already.
This would be good for consistency across the
fi_family of functions. Also, it would allow the implementation of afi_pool_random()much more easily and consistently withfi_pool_cluster_random().Downside is that it is slightly more clunky to vectorise
fi_poolwrtpool_sizeandpool_number, but I don't think is major. Would need to check thatfi_poolisn't being used in vectorised form already.