refactor: add explicit NaN guard to stats/base/dists/arcsine/median native source

[Planeshifter]

Description

This pull request:

• Aligns the native stats/base/dists/arcsine/median implementation with the validation prologue used by 6 of 8 sibling 2-argument distribution functions in the same namespace and by 23 of 25 (92%) */median/src/main.c files across stats/base/dists.

Namespace summary

• Namespace: @stdlib/stats/base/dists/arcsine (14 members)
• Features analyzed: file tree, package.json / manifest.json keys, README structure, JS validation prologue and JSDoc shape, presence of native/Julia/factory artifacts
• Features with a clear majority (≥75% conformance): native implementation suite (12/14), Julia fixture suite (12/14), README skeleton [Usage, Examples, C APIs, Usage, Examples] (11/14), stdlib_base_is_nan guard in 2-arg C sources (6/8)
• Features excluded for lack of majority: JS validation prologue isnan guard (4/8 for 2-arg, 9/13 across non-ctor), lib/factory.js presence (5/14)

stats/base/dists/arcsine/median

median/src/main.c previously guarded with if ( a >= b ) only, where the sibling C sources mean, stdev, entropy, kurtosis, mode, and skewness all open with if ( stdlib_base_is_nan( a ) || stdlib_base_is_nan( b ) || a >= b ). Behavior is unchanged — NaN inputs already propagate to a NaN return via the arithmetic path; this only makes the early-return explicit and matches the namespace convention. Adds the corresponding #include "stdlib/math/base/assert/is_nan.h" and registers @stdlib/math/base/assert/is-nan under build, benchmark, and examples in manifest.json so the native build resolves the header.

Related Issues

None.

Questions

No.

Other

Validation

Structural features were extracted by reading the file tree and package.json / manifest.json / README.md of each of the 14 namespace members. Semantic features (JS public signature, validation prologue, JSDoc shape) were extracted directly from lib/main.js of every member; native-source content was inspected manually as a follow-on once structural analysis showed near-uniform native-file presence (12/14). Each surviving outlier was put through three independent validation agents (semantic review, cross-reference, structural review) before any patch was applied — the cross-reference agent caught that the original patch sketch was incomplete (manifest.json also needed the is-nan dependency or the build would break), and the patch was extended before commit.

Deliberately excluded from this PR:

• Missing C implementation for stats/base/dists/arcsine/logcdf — open-PR collision with feat: add C implementation for stats/base/dists/arcsine/logcdf #3389 and feat: add C implementation for stats/base/dists/arcsine/logcdf #10739
• Missing C implementation for stats/base/dists/arcsine/ctor — ctor is a constructor class, not a scalar function; the native skeleton is not applicable
• Missing Julia fixtures for kurtosis and skewness — both return mathematical constants (-1.5 and 0.0 respectively), so reference values would be uninformative
• Missing JS isnan guard in mean, median, stdev, variance — 4-of-8 split in the 2-arg JS sources, below the 75% threshold for declaring a majority pattern
• stats/base/dists/arcsine/variance/src/main.c NaN guard — ecosystem-wide presence for */variance/src/main.c is 78%, below the 90% threshold for treating absence as drift

Checklist

• Read, understood, and followed the contributing guidelines.

AI Assistance

• Yes
• No

If you answered "yes" above, how did you use AI assistance?

• Code generation (e.g., when writing an implementation or fixing a bug)
• Test/benchmark generation
• Documentation (including examples)
• Research and understanding

Disclosure

This PR was authored by Claude Code as part of the automated cross-package drift detection routine. Candidate-outlier enumeration and within-namespace majority computation ran locally; each surviving candidate was validated by three independent agents (semantic review, cross-reference against tests / docs / consumers, structural review including manifest implications) before the patch was applied. A human maintainer should verify the patch before promoting this PR out of draft.

---

@stdlib-js/reviewers

---

Generated by Claude Code

[stdlib-bot]

Coverage Report

Package	Statements	Branches	Functions	Lines
stats/base/dists/arcsine/median	$\\color{green}177/177$ $\\color{green}+100.00\\%$	$\\color{green}7/7$ $\\color{green}+100.00\\%$	$\\color{green}2/2$ $\\color{green}+100.00\\%$	$\\color{green}177/177$ $\\color{green}+100.00\\%$

The above coverage report was generated for the changes in this PR.