test: migrate stats/base/dists/t/variance to ULP-based assertions

[kgryte]

Resolves a part of #11352.

Description

What is the purpose of this pull request?

This pull request:

• migrates the tests for stats/base/dists/t/variance from relative tolerance assertions to ULP-based assertions using @stdlib/assert/is-almost-same-value.

Only test/test.native.js contained the computed-tolerance idiom (delta = abs( y - expected[ i ] ) / tol = 2.0 * EPS * abs( expected[ i ] )), so that is the only file changed. The fixture loop in test/test.js already used exact assertions, and, per the guidance in #11352 to avoid extraneous test changes, it has been left as is.

Final ULP constant: 0.

The bound was measured rather than guessed. For every one of the 50 fixture cases, the smallest maxULP satisfying isAlmostSameValue( y, expected[ i ], maxULP ) was computed for both the JavaScript and the C implementation; the maximum over the fixture set is 0 in both cases. In other words, both implementations reproduce the Julia fixtures bit-for-bit, so 0 is the tightest admissible bound and no padding was added. This is consistent with test/test.js, which already asserted exact equality, and it means the old else branch was in practice unreachable on this fixture set.

A 0 ULP bound is well-established in the codebase for this situation; for example, stats/base/dists/beta/mean and stats/base/dists/beta/skewness use isAlmostSameValue( y, expected[ i ], 0 ) in their fixture loops.

Verification

• The native add-on was built locally (make install-node-addons NODE_ADDONS_PATTERN="stats/base/dists/t/variance"), so the test.native.js assertions actually execute rather than being skipped: test/test.js reports 58 passing and test/test.native.js reports 59 passing.
• make test TESTS_FILTER=".*/stats/base/dists/t/variance/.*" was run twice at the final bound, with identical results both times (no arch/FMA-dependent variation).
• eslint is clean against etc/eslint/.eslintrc.tests.js, and lint-license-headers-files passes.
• The only file changed is test/test.native.js.

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?

One, on a judgment call: since the measured minimum is 0 and test/test.js was already exact, an alternative reading is that this package has no tolerance to migrate and should be skipped entirely. This PR instead removes the dead tolerance branch and states the 0 ULP budget explicitly, matching the beta/mean precedent. Happy to close this if you would rather leave such packages untouched.

Other

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

No.

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 helps reviewers determine how much scrutiny to apply when reviewing your contribution. Example disclosures: "This PR was written primarily by Claude Code." or "I consulted ChatGPT to understand the codebase, but the proposed changes were fully authored manually by myself.".

This PR was written by Claude Code running as an unattended scheduled task. It selected the package, mirrored the conversion idiom from already-merged PRs in the same family, measured the ULP bound programmatically over the full fixture set, and ran the tests and linters locally. The changes have not been reviewed by a human prior to submission, which is why it is opened as a draft.

---

@stdlib-js/reviewers

🤖 Generated with Claude Code

https://claude.ai/code/session_01JJ2JUhP1NnZRW3mrjB7UpH

---

Generated by Claude Code

[stdlib-bot]

Coverage Report

Package	Statements	Branches	Functions	Lines
stats/base/dists/t/variance	$\\color{green}169/169$ $\\color{green}+100.00\\%$	$\\color{green}10/10$ $\\color{green}+100.00\\%$	$\\color{green}2/2$ $\\color{green}+100.00\\%$	$\\color{green}169/169$ $\\color{green}+100.00\\%$

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