This message was deleted.
# ask-questionsw
worried-planet-2191
04/12/2023, 12:15 PMThis message was deleted.
f
fresh-football-47124
04/12/2023, 3:26 PM
Interesting
e
eager-wall-93028
04/13/2023, 9:10 AMHey @fresh-football-47124 is there a way to fix it? This query was generated by Growthbook and I have no power over it. I get the same error when running it in IDE.
Just to give more context, I’ve declared session_marketing_source dimension in assignment query and it throws error when applying this dimension to the experiment results.
There’s a same topic in Github: https://github.com/metabase/metabase/issues/12544
f
fresh-football-47124
04/13/2023, 9:31 AM
We’re working on a fix
e
eager-wall-93028
04/13/2023, 9:52 AMGreat news, thank you very much! If you have an ETA for it, I’d be glad to know
eager-wall-93028
04/13/2023, 4:06 PMAlso @fresh-football-47124 I’m not sure why, but this is only applicable to dimensions that were specified in assignment query. Dimensions from ‘Dimensions’ tab seem to work
h
helpful-application-7107
04/13/2023, 8:22 PMHi Katerina. This is a problem on redshift with some special handling we added to use the first dimension value per user as part of our SQL refactor. This "first" notion only applies to assignment query dimensions (which have a timestamp) and not user dimensions (which do not have a timestamp; for those we just use the MAX dimension and instead ask users to only have a 1:1 user to dimension mapping).
So we need to fix this logic for the assignment query dimensions and we know how to do so. Just give us a second to test it.
helpful-application-7107
04/13/2023, 9:46 PMAs for an ETA, we should land a fix tomorrow pacific time.
e
eager-wall-93028
04/14/2023, 2:03 PMHey Luke, thanks a lot for explaining 🙌
h
helpful-application-7107
04/18/2023, 4:55 PMHey Katerina! Can you let us know if the fix we landed on Friday worked for you?
e
eager-wall-93028
04/19/2023, 4:22 PMHey Luke, I’ve checked and it works, thank you!
🙌 1
149 Views