Hi Joel,

Thanks for this idea - but can you confirm that processing large data sets in batches like this is faster overall for the entire dataset? My (admittedly hazy) memory of this approach is that the overall time taken is around the same?

THanks for the input!

Adam