Experts -
I am using the multi-row formula to partition data by 2 dimensions and then do a row count within each partition, i.e.:
IF ISNULL([Row-1:filter_partition])
THEN 1
ELSE [Row-1:filter_partition] + 1
ENDIF
I am grouping by the 2 dimensions in the multi-row formula. Initially I put a sort tool before the multi-row to sort by those dimensions so "filter_partition" would calculate correctly, but I just noticed if I remove the sort tool it still calculates correctly.
I can only assume this is because the multi-row formula sorts the data as per the group by fields. Can someone confirm? It is not mentioned in the online documentation.
Thank you!
Solved! Go to Solution.
Yes grouping data in the Multi Row sorts data during the caculation of the tool. Users typically use a sort tool beforehand as they can see what order their data is in before it goes into the multi-row tool.
Thanks Joseph - any idea why this isn't included in the documentation? It seems like a pretty fundamental feature of the tool, and eliminating the unnecessary pre-sort can have a big effect on performance when chugging large data sets.