Just to clarify, you want to roll up the example rows above into a single row? Since everything is the same, except Measurement, that would be the big question: what do you want to happen to Measurement? Do the values get added together, averaged, keep only one of them (and which one?)
Or maybe there's another solution altogether? For example, is the join splitting out what used to be a single row into the rows above? If so, why are there different values for Measurement?
I guess it might just take a bit more context or at least clarification to get to an answer.
Looking at it, I think your note about the join splitting it out might be on to something. Each of my data sources have start_time, end_time, and a measurement. The times from one don't necessarily match the times from the other though, so I can't use them as a join condition. I'm joining on one of the dimensions but I'm still ending up with multiple measurements per start_time. Maybe that's the problem? I'm joining on a dimension, so the granularity difference is too great?
Sorry if this seems like rambling. Here's what I'm looking at. actual_hours is the measurement I want for each start_time.
So I found a temporary workaround, but my initial confusion about the Aggregate step still stands. What I did was use Prep to create my custom SQL extract, then use that as a primary data source in Tableau Desktop. I then brought in my other data as a secondary data source, and established a relationship based on the dimension AND the month/year of my start_time fields. From here I can compare at a month level, and it gets me most of the way there.
As far as the Agg step goes, should you just put the fields you want to keep the same on the left side (grouped fields) and the fields you want aggregated on the right side? Any you don't place don't make it through to the other side of the step?
This is really helpful, Deepak. Those obtuse triangles in Prep always throw me for a loop, but this is a good way to think about it!