[kgryte]

Description

What is the purpose of this pull request?

This pull request:

• migrates the tests for math/base/special/gammaln from relative-tolerance (EPS-scaled) assertions to ULP-based assertions using @stdlib/assert/is-almost-same-value, per [RFC]: Migrate math/base/special packages from relative tolerance testing to ULP difference testing (tracking issue) #11352.
• applies the same change to both test/test.js and test/test.native.js.
• removes the now-unused @stdlib/math/base/special/abs and @stdlib/constants/float64/eps requires from both test files.

ULP bounds

Each fixture group was measured against its expected values and the bound tightened to the minimum integer that still passes:

Test	Fixtures	Previous tolerance	Final ULP bound	Measured max ULP
positive integers	data1 / expected1 (171 values)	21.2 * EPS * abs(expected)	2	2
decimal values	data2 / expected2 (1000 values)	1.1e4 * EPS * abs(expected)	15	15

The bounds are minimal: lowering them to 1 and 14 respectively produces 9 failing assertions (8 in the positive-integer set, 1 in the decimal set). At the final bounds the suite passes 1184/1184, and was run twice with identical results to rule out non-determinism.

The same bounds are used in test.native.js. To measure rather than assume this, the C implementation (src/main.c plus its transitive C dependencies) was compiled standalone and evaluated over both fixture sets: it is bit-for-bit identical to the JavaScript implementation across all 1171 fixture points (max ULP difference between the two implementations is 0), so the native bounds are also exactly 2 and 15.

Related Issues

Does this pull request have any related issues?

This pull request has the following related issues:

• [RFC]: Migrate math/base/special packages from relative tolerance testing to ULP difference testing (tracking issue) #11352

Questions

Any questions for reviewers of this pull request?

The native test suite could not be executed in the authoring environment because the native add-on was not built there (test.native.js skips when the add-on is absent). The native ULP bounds were instead derived by compiling and evaluating the C implementation directly, as described above. Confirmation from CI, where the add-on is built, would be welcome.

Other

Any other information relevant to this pull request? This may include screenshots, references, and/or implementation notes.

Only the two test files are changed; the implementation and fixtures are untouched.

Note on verification environment: make install-node-modules could not complete in the authoring environment (the configured npm registry has no es-object-atoms@^1.1.2, which a transitive dependency requires), so the standard make lint/test targets were unavailable. As substitutes, tape was installed standalone to run the suite, ESLint was run against the repository's own rule set from etc/eslint/rules with the etc/eslint/.eslintrc.tests.js overrides applied (clean, and cross-checked against an already-merged converted package as a control), and the EditorConfig rules were validated directly. Full CI verification is still needed.

Checklist

Please ensure the following tasks are completed before submitting this pull request.

• Read, understood, and followed the contributing guidelines.

AI Assistance

When authoring the changes proposed in this PR, did you use any kind of 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

If you answered "yes" to using AI assistance, please provide a short disclosure indicating how you used AI assistance.

This PR was written by Claude Code running as an unattended scheduled task. It selected the package, studied already-merged conversions to match the established idiom, applied the conversion, and measured and tightened the ULP bounds empirically.

---

@stdlib-js/reviewers

---

Generated by Claude Code