Phillip Cloud
Phillip Cloud
We don't have any guarantees about what the repr looks like, so this doesn't need to happen in a major release.
I think starting with `to_dask` makes sense. Supporting a general batching API doesn't (yet) seem worth the effort.
Not seeing exactly what `to_batches` is getting us here. Is this motivated by an ibis-ml use case?
I'll run the clouds.
Snowflake is passing, but BigQuery is raising this exception: ``` E google.api_core.exceptions.BadRequest: 400 ORDER BY clause expression references t0.id which is not visible after SELECT DISTINCT at [11:7]; reason: invalidQuery,...
👋🏻 Thanks for the PR! Would you mind adding a test for this that fails without your change and then passes with it?
All good points, definitely changing `grouping` -> `grouping_id`. I'm half tempted to call it `group_id`, those extra three characters are killing me.
For a full featured and correct implementation, grouping sets will depend on the earliest version of sqlglot contains https://github.com/tobymao/sqlglot/pull/3985.
Ok, now that I think about this some more, I think having separate `ibis.rollup/grouping_sets/cube` objects adds unnecessary complication. What do people think about making these keyword arguments to `group_by`? E.g.,...
This would break users who having columns named `rollup`, `cube`, or `grouping_sets` but that seems okay?