Fix prop_above_and_below typing and empty concat handling

[neuralsorcerer]

No description provided.

[Copilot]

Pull request overview

This PR fixes the prop_above_and_below() function in balance/stats_and_plots/weights_stats.py by:

1. Replacing the untyped dict[Any, Any] return type with a new PropAboveBelowResult TypedDict.
2. Adding @overload signatures to let static type checkers narrow the return type based on the return_as_series literal argument.
3. Fixing a latent bug where pd.concat could be called with an all-None list (when one of below/above is None) by building the pieces list with only non-None Series.

Changes:

• Introduced PropAboveBelowResult TypedDict and @overload signatures for prop_above_and_below, plus a safe concat fix when only one threshold group is provided
• Added edge-case tests covering empty threshold iterables, mixed-None threshold groups in dict mode, and all-None both-groups behavior
• Updated CHANGELOG.md with entries for both the fix and the new tests

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 2 comments.

File	Description
balance/stats_and_plots/weights_stats.py	Introduces PropAboveBelowResult TypedDict, adds @overload signatures, and replaces unsafe pd.concat with filtered-pieces guard
tests/test_stats_and_plots.py	Adds test_prop_above_and_below_edge_cases covering empty iterables, mixed-None dict mode, and both-None case
CHANGELOG.md	Records the return-path stabilization and expanded test coverage

---

Two issues were identified:

1. Second @overload has an unnecessary and misleading default value (line 222): return_as_series: Literal[False] = False declares False as the default, creating ambiguity for Pyre/mypy when the argument is omitted — both overloads now claim to apply. The standard pattern is to omit the default on the non-default overload: return_as_series: Literal[False] (no = False).

2. Docstring Returns section and below/above parameter descriptions are stale (lines 227–232 area): The Returns: type hint still says pd.Series | dict instead of pd.Series | PropAboveBelowResult | None, and the descriptions for below and above still say "Using None returns None," which is only true when both are None.

[Copilot]

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated 3 comments.

[Copilot]

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated no new comments.

[meta-codesync]

@talgalili has imported this pull request. If you are a Meta employee, you can view this in D95019739.

[meta-codesync]

@talgalili merged this pull request in ddae91c.