tree 1b2df6deada84cf6aa8c6e3b72f9087f54eec482
parent d5b62ad99f0ef316a106be7df2766d78d45a7e1f
author Brian Hulette <bhulette@google.com> 1620764046 -0700
committer GitHub <noreply@github.com> 1620764046 -0700
gpgsig -----BEGIN PGP SIGNATURE-----
 
 wsBcBAABCAAQBQJgmuWOCRBK7hj4Ov3rIwAAMS4IAGZ/yClvJ+VrHv9pDDWI76RT
 EY16UAp7HOu1SQJKgEIDZjct4ljQ1K/Eej/0rYA7ASMwbNYwXyNAC0nvmKslmiUl
 Dy8Am7yIvszAm8JlDynCflOtMqxJW582OI4zCxfuaYD1ffpcMVKI2PF/zDeDqkqD
 DyTxm7p7Suz8+VEvMQwCBT3/iU7OEtRtuHoUYdagndJFRhBRbc9T494jvzPbQn8b
 qEVDJ0NsTkQz4ii8q8z2gVQ9JV2bBBvVeyhh2OjSjwmlHREDCZTkgi0LwPmajKd6
 RElNkqbt6csECJ+9jhW8SdsGDkxPS4mVYVVoLTXcU8qwV2Y3/HP3YzwYJ+NFKec=
 =3OVs
 -----END PGP SIGNATURE-----
 

[BEAM-11777][BEAM-11978] Add support for all kwargs in DataFrame, Series aggregation methods (#14438)

* agg(level=) implementation that defers to GroupBy(level=).agg

* min_count

* lint

* Disable numeric_only test for pandas <1.2

* fix skip

* yapf

* remove min_count raises test

* Reuire numeric_only= to be specified for numeric aggregations on DataFrames with non-numeric columns

* lint

* cleanups

* yapf

* _run_test correctly verifies nan outputs

* Use new _run_test infra

* Add additional testing for invalid kwargs

* yapf

* Track different singleton reasons in agg

* Remove TODOs

* yapf

* Address PR comments

* Switch to getattr in _(un)liftable_agg expressions