feat: add stats/incr/nanwmean

[Arjan-P]

Description

What is the purpose of this pull request?

This pull request:

• add incremental accumulators for handling NaN values in stats/incr/nanwmean

Related Issues

Does this pull request have any related issues?

This pull request has the following related issues:

• [RFC]: add stats/incr/nanwmean #5628

Questions

Any questions for reviewers of this pull request?

No.

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.".

Generated package.json, repl.txt from gemini

---

@stdlib-js/reviewers

[stdlib-bot]

Coverage Report

Package	Statements	Branches	Functions	Lines
stats/incr/nanwmean	$\\color{green}174/174$ $\\color{green}+0.00\\%$	$\\color{green}9/9$ $\\color{green}+0.00\\%$	$\\color{green}2/2$ $\\color{green}+0.00\\%$	$\\color{green}174/174$ $\\color{green}+0.00\\%$

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

[Neerajpathak07]

/stdlib update-copyright-years

[Neerajpathak07]

@Arjan-P Thanks for the PR!! Left out a few comments. Commonly we want to prevent a straight away copy-paste between subsequent packages and think a bit more on what is the true intention and nature of nanwmean being added here.

[Neerajpathak07]

Copied from incr/wmean, it's the opposite of what nanwmean does (NaN is ignored). Sibling incr/nanewmean omits this note.

[Neerajpathak07]

Should read "... weighted arithmetic mean, ignoring NaN values." to match nanewmean and the module description.

[Neerajpathak07]

incorrect package name. Copy paste error from incrwmean

[Neerajpathak07]

Should be updated below as well!!!

[Neerajpathak07]

package.json, repl.txt, and sibling packages say "ignoring NaN values". Let's Align here on "values".

[Neerajpathak07]

The full TeX derivation is duplicated from incr/wmean, which this package merely wraps it will drift out of sync. incr/nanewmean keeps a one-line description; recommend trimming to that plus a NaN args are ignored note.

[Neerajpathak07]

Suggested change

	if ( isnan( x ) || isnan( w ) ) {
	return acc();
	if ( arguments.length < 2 || isnan( x ) || isnan( w ) ) {
	return acc();

Calling acc( 2.0 ) with no weight leaves w === undefined, isnan( undefined ) is false, so this guard is bypassed and acc( 2.0, undefined ) corrupts wsum/mu to NaN for every subsequent call unrecoverable, and exactly what a nan-accumulator should prevent.

[Neerajpathak07]

A bit high value for ULP Diff here. Consider a relative-error tolerance (as other incr/* tests use) or fixtures bounded away from zero.

[Arjan-P]

My bad, will be more mindful going forth. I think the package is also missing a README.md. Thank you for the feedback, will address these as soon as possible.

[Arjan-P]

@Neerajpathak07 I have addressed most of the changes. Regarding the ULP difference in test, even with relative error it requires a high tolerance:

tape( 'the accumulator function incrementally computes a weighted arithmetic mean', function test( t ) {
	var expected;
	var actual;
	var delta;
	var tol;
	var acc;
	var N;
	var x;
	var w;
	var i;

	N = data.length;

	acc = incrnanwmean();

	for ( i = 0; i < N; i++ ) {
		x = data[ i ].x;
		w = data[ i ].w;
		expected = data[ i ].mean;
		actual = acc( x, w );

		if ( actual === expected ) {
			t.strictEqual( actual, expected, 'returns expected value: ' +expected );
		} else {
			delta = abs( actual - expected );
			tol = 150 * EPS * abs( expected );
			t.strictEqual( delta <= tol, true, 'within tolerance. x: ' + x + '. w: ' + w + '. Value: ' + actual + '. Expected: ' + expected + '.' );
		}
	}
	t.end();
});

Is the difference due to the fixtures being generated by different implementation than the one begin tested?