Hey guys, I have some questions regarding scalding steps:
1. Does scalding writes the intermediate data (that is persisted to hdfs between steps) steps compressed? Can I change it?
2. I want to control the number of input splits (and thus control the number of mappers) when reading the output of the reducers in a previous step. How should I go about that?
Thanks,
Hagai.