Ayush Dattagupta
Ayush Dattagupta
> In my opinion we should handle merges b/w columns of similar dtype , say like int32/int64 and float32/float64 etc and FAIL loudly in other cases (like int32/float32, int32/uint32 )...
> However in the last day, I upgraded my dependencies and found that this was now not being respect. To followup does this mean that in a newer environment the...
Thanks for looking into this @rjzamora. We can update the implementation here once the dask dataframe pr makes it in.
Thanks for reporting @randerzander! I can't reproduce the error with the latest dask-sql and dask releases button reproduce with dask from main since the merge of dask/dask#8341. cc: @charlesbluca
Thanks for raising the issue @DaceT! I believe there's a couple of things going on here: 1. Dask-sql does not support window rank functions today. Here's a list of initial...
Not sure it's the same error but should revisit after the fix for #532 lands
I have been looking into this a bit more and updating #293 locally to work with the latest branch. On further inspection it looks like the main issue with `month`...
Thanks a lot for opening this issue. Would be happy to go over the pr once you have something up 😄
@PeterLappo Do you have more information or details on what kind of schema evolution you're looking for. In general dask (and by extension dask-sql) expects all partitions in a dataframe...
> @ayushdg this is marked as a draft but seems ready. Are you ready for a review? Not yet. The core issues is fixed but I'm also trying to update...