I have a similar use case. Since I don't have access to our corporate license, I am testing the free version. I think partitioning by a discrete column value would solve my problem, but partitioning is not allowed on the free license.
2nd potential option: I'm familiar with Pandas' funcitonality of referencing a groupby as all of its records, not just as an aggregation. As I've tried Grouping in a flow in DSS without code, I don't see a way to access all the records. I only see aggregations. I would hope that Dataiku Group recipe could accomplish this:
3rd potential option: a short piece of Python code as above.
Can anyone provide a successful example of splitting one dataset into several with the 2nd or 3rd option? Other ideas?