Sarah Yurick

Correlated subqueries

2

Previously, we would get a `ValueError: Not all divisions are known, can't align partitions. Please use set_index to set the index.` for something like: ``` from dask_sql import Context import...

Address `@xfail_if_external_scheduler` TODOs in `test_model`

2

Several of our pytests in `tests/integration/test_model.py` have this note: ``` # TODO - many ML tests fail on clusters without sklearn - can we avoid this? @skip_if_external_scheduler ``` We should...

- Extract microseconds/millennium/millisecond - Last_day - Timestampadd - Ceil - Floor

Check for `np.timedelta64` in `as_timelike`

1

When I try: ``` from dask_sql import Context import pandas as pd import dask.dataframe as dd from datetime import datetime c = Context() date = datetime(2021, 10, 3, 15, 53,...

Replace `dask_ml.wrappers.ParallelPostFit` with custom `ParallelPostFit` class

2

As part of our initiative to move away from Dask-ML, I've migrated some code from Dask-ML into Dask-SQL to support ParallelPostFit.

This is currently being tracked by #832 and #839 to replace the `ParallelPostFit` and `Incremental` classes, respectively. In addition to these, we should look into removing all Dask-ML references in...

Closes #839. In addition to #832, we want to create a custom implementation for Dask-ML's `Incremental` class as well. So as not to create any merge conflicts, I've only added...

**Is your feature request related to a problem? Please describe.** After #813 is resolved, we should also add support for the `TIMESTAMPDIFF` function, as described [here](https://www.w3resource.com/mysql/date-and-time-functions/mysql-timestampdiff-function.php). **Describe the solution you'd...

Support `to_timestamp`

1

Closes #831.

As part of our initiative to move away from Dask-ML, we should replace the `dask_ml.wrappers.Incremental` class with our own custom class. This implementation should be very similar to the `ParallelPostFit`...