[Oscar-man-shrestha]

What happened to labels 6 and 7

The external grouping Series is named a and labels the rows [6, 6, 7]. The DataFrame also has a data column named a with values [1, 1, 2], so the sums are 2 for group 6 and 2 for group 7. With as_index=False, pandas skipped inserting the external group labels because the result already had a column named a holding [2, 2]. The old output showed a = [2, 2] and omitted the group labels.

Why ValueError matches reset_index() here

With as_index=True, the labels sit in an index named a next to a data column named a. Calling reset_index() raises ValueError: cannot insert a, already exists. Both sets of values can be kept if the grouping key is renamed. With the name left as a, though, pandas would have to invent a new name or produce duplicate a columns. Raising the same error from as_index=False lets the caller choose the name instead of having pandas guess.

Changes

• _insert_inaxis_grouper now raises ValueError when a named external grouper collides with a result column, instead of skipping the group labels.

• DataFrameGroupBy.aggregate with a callable returns the result of _python_agg_general directly, since that already inserts the grouping columns; previously they would be inserted twice.

• Behavior change: test_groupby_as_index_agg grouped by an external Series named jim that collides with a column. mean, max, count, idxmax and all now raise ValueError there, matching reset_index(). cumsum is a transform, inserts no grouping column, and is unchanged.

• Tests cover the collision, the renamed key (external.rename("group") keeps [6, 7] next to the sums), and a callable aggregation inserting the key once.

• closes BUG: groupby with as_index=False drops groupings when names collide #58337

• Tests added and passed if fixing a bug or adding a new feature

• All code checks passed.

• Added an entry in the latest doc/source/whatsnew/vX.X.X.rst file if fixing a bug or adding a new feature.

• I have reviewed and followed all the contribution guidelines

Check exactly one of the following, per the automated contributions policy:

• I did not use AI to develop this pull request.
• I used AI to develop this pull request. I prompted it to follow AGENTS.md, I have reviewed and understood every change, and I have described above how I used it and exactly which tool, model version, and effort setting — e.g. claude opus 4.8 (xhigh), not just claude.