I'm writing a mapreduce script for my mongodb database. The computation requires two mapreduce stages. Currently I write to an output collection and then run the second stage on that collection.
Is it possible to chain mapreduce jobs together without having to manually specify the output collection?