feat: add `stats/base/dists/log-logistic/pdf`#11203

Open

parthodas23 wants to merge 1 commit intostdlib-js:developfrom

parthodas23:feat/stat-base-dists-log-logistic-pdf

Contributor

parthodas23 commented Mar 29, 2026

none

Description

What is the purpose of this pull request?

This pull request:

feat: add stats/base/dists/log-logistic/pdf

Related Issues

Does this pull request have any related issues?

This pull request has the following related issues:

RFC: add the log-logistic distribution #175

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.".

{{TODO: add disclosure if applicable}}

@stdlib-js/reviewers

stdlib-bot added Statistics Needs Review labels

Contributor

stdlib-bot commented Mar 29, 2026 •

edited

Loading

Coverage Report

Package	Statements	Branches	Functions	Lines
stats/base/dists/log-logistic/pdf	$\color{green}307/307$ $\color{green}+0.00%$	$\color{green}25/25$ $\color{green}+0.00%$	$\color{green}4/4$ $\color{green}+0.00%$	$\color{green}307/307$ $\color{green}+0.00%$

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

parthodas23 force-pushed the feat/stat-base-dists-log-logistic-pdf branch 3 times, most recently from 5cf562d to 13f4fad Compare

March 29, 2026 17:58


          feat: add stats/base/dists/log-logistic/pdf

cbec15c

parthodas23 force-pushed the feat/stat-base-dists-log-logistic-pdf branch from 13f4fad to cbec15c Compare

March 29, 2026 18:08

github-actions bot mentioned this pull request

[aw] No-Op Runs #11033

Open

manit2004 reviewed

View reviewed changes

lib/node_modules/@stdlib/stats/base/dists/log-logistic/pdf/test/fixtures/julia/runner.jl

Comment on lines +19 to +20

		import Distributions: pdf, Log-logistic
		import JSON

Contributor

manit2004 Apr 9, 2026

Julia doesn't have log-logistic or fisk distribution. For that reason upon running this script, module import error takes place. Scipy has fisk distribution, but the parameterization is different. Refer to scipy docs for more info. This pdf function import for log logistic isn't even used in later part of the script.

manit2004 reviewed

View reviewed changes

lib/node_modules/@stdlib/stats/base/dists/log-logistic/pdf/test/fixtures/julia/runner.jl

Comment on lines +43 to +52

+              function gen( x, alpha, beta, name )
+              	z = Array{Float64}( undef, length(x) );
+              	for i in eachindex(x)
+              		if x[i] < 0.0
+              			z[i] = 0.0;
+              		else
+              			r = x[i] / alpha[i];
+              			z[i] = (beta[i] / alpha[i]) * r^(beta[i] - 1.0) / (1.0 + r^beta[i])^2;
+              		end
+              	end

Contributor

manit2004 Apr 9, 2026

To generate fixtures, generally already implemented library functions are used. For this case use scipy fisk distribution pdf.

manit2004 reviewed

View reviewed changes

lib/node_modules/@stdlib/stats/base/dists/log-logistic/pdf/test/test.pdf.js

Comment on lines +47 to +57

+              	y = pdf( NaN, 1.0, 1.0 );
+              	t.strictEqual( isnan( y ), true, 'returns expected value' );
+              	y = pdf( 0.5, NaN, 1.0 );
+              	t.strictEqual( isnan( y ), true, 'returns expected value' );
+              	y = pdf( 0.5, 1.0, NaN );
+              	t.strictEqual( isnan( y ), true, 'returns expected value' );
+              	y = pdf( NaN, NaN, NaN );
+              	t.strictEqual( isnan( y ), true, 'returns expected value' );

Contributor

manit2004 Apr 9, 2026

Suggested change

      
            	y = pdf( NaN, 1.0, 1.0 );
          
            	t.strictEqual( isnan( y ), true, 'returns expected value' );
          
            	y = pdf( 0.5, NaN, 1.0 );
          
            	t.strictEqual( isnan( y ), true, 'returns expected value' );
          
            	y = pdf( 0.5, 1.0, NaN );
          
            	t.strictEqual( isnan( y ), true, 'returns expected value' );
          
            	y = pdf( NaN, NaN, NaN );
          
            	t.strictEqual( isnan( y ), true, 'returns expected value' );
          
            	y = pdf( NaN, 1.0, 1.0 );
          
            	t.strictEqual( isnan( y ), true, 'returns expected value' );
          
            	y = pdf( 0.5, NaN, 1.0 );
          
            	t.strictEqual( isnan( y ), true, 'returns expected value' );
          
            	y = pdf( 0.5, 1.0, NaN );
          
            	t.strictEqual( isnan( y ), true, 'returns expected value' );
          
            	y = pdf( NaN, 1.0, NaN );
          
            	t.strictEqual( isnan( y ), true, 'returns expected value' );
          
            	y = pdf( NaN, NaN, 1.0 );
          
            	t.strictEqual( isnan( y ), true, 'returns expected value' );
          
            	y = pdf( 0.5, NaN, NaN );
          
            	t.strictEqual( isnan( y ), true, 'returns expected value' );
          
            	y = pdf( NaN, NaN, NaN );
          
            	t.strictEqual( isnan( y ), true, 'returns expected value' );

Contributor

manit2004 Apr 9, 2026

Refer to normal dist pdf tests. All 3 combinations of two NaN occurences should also be tested.

manit2004 reviewed

View reviewed changes

lib/node_modules/@stdlib/stats/base/dists/log-logistic/pdf/test/test.pdf.js

Comment on lines +98 to +102

+              	y = pdf( -1.0, 1.0, 1.0 );
+              	t.strictEqual( y, 0.0, 'returns expected value' );
+              	y = pdf( -10.0, 2.0, 3.0 );
+              	t.strictEqual( y, 0.0, 'returns expected value' );

Contributor

manit2004 Apr 9, 2026

Both of these test cases test the same thing. My suggestion would be to remove one of them, since both of them check for x<0

manit2004 reviewed

View reviewed changes

Contributor

manit2004 left a comment

comments made for test.pdf.js is also applicable for test.native.js and test.factory.js

lib/node_modules/@stdlib/stats/base/dists/log-logistic/pdf/test/test.pdf.js


		y = pdf( 1.0, 1.0, NINF );
		t.strictEqual( isnan( y ), true, 'returns expected value' );

Contributor

manit2004 Apr 9, 2026

Suggested change

      
            	y = pdf( 1.0, PINF, NINF );
          
            	t.strictEqual( isnan( y ), true, 'returns expected value' );
          
            	y = pdf( 1.0, NaN, NINF );
          
            	t.strictEqual( isnan( y ), true, 'returns expected value' );
          
            	y = pdf( 1.0, NINF, NINF );
          
            	t.strictEqual( isnan( y ), true, 'returns expected value' );

Refer to normal pdf tests, combinations of PINF and NaN should also be added with NINF. Import PINF at the top of the file. Same for non positive alpha test cases.

lib/node_modules/@stdlib/stats/base/dists/log-logistic/pdf/test/test.pdf.js

+              	beta = data.beta;
+              	for ( i = 0; i < x.length; i++ ) {
+              		y = pdf( x[ i ], alpha[ i ], beta[ i ] );
+              		if ( expected[ i ] !== null && !( expected[ i ] === 0.0 && y < EPS ) ) {

Contributor

manit2004 Apr 9, 2026

Suggested change

      
            		if ( expected[ i ] !== null && !( expected[ i ] === 0.0 && y < EPS ) ) {
          
            		if ( expected[ i ] !== null && !( expected[ i ] === 0.0 && y < EPS ) )

!( expected[ i ] === 0.0 && y < EPS this expression is not needed. In your Julia Script x can be equal to 0 thus it's a valid possibilty that expected[ i ] === 0.0 and y can be very small. In normal pdf tests or any other test files we don't use this. Also expected[ i ] !== null is unnecessary, fixtures are already reliably generated.

lib/node_modules/@stdlib/stats/base/dists/log-logistic/pdf/test/test.pdf.js

Comment on lines +132 to +133

		t.ok( delta <= tol, 'within tolerance. x: '+x[ i ]+'. alpha: '+alpha[ i ]+'. beta: '+beta[ i ]+'. y: '+y+'. E: '+expected[ i ]+'. Δ: '+delta+'. tol: '+tol+'.' );
		}

Contributor

manit2004 Apr 9, 2026

use @stdlib/assert/is-almost-same-value' instead of delta tolerance tests

...ode_modules/@stdlib/stats/base/dists/log-logistic/pdf/docs/img/equation_log_logistic_pdf.svg

Contributor

manit2004 Apr 9, 2026

These images are auto generated, we can remove this file

lib/node_modules/@stdlib/stats/base/dists/log-logistic/pdf/benchmark/c/benchmark.c

+              	int i;
+              	for ( i = 0; i < 100; i++ ) {
+              		x[ i ] = random_uniform( STDLIB_CONSTANT_FLOAT64_EPS, 20.0 );

Contributor

manit2004 Apr 9, 2026

Suggested change

      
            		x[ i ] = random_uniform( STDLIB_CONSTANT_FLOAT64_EPS, 20.0 );
          
            		x[ i ] = random_uniform( 0.0, 20.0 );

the domain for x is [0,inf) so here the lower bound should be zero.

lib/node_modules/@stdlib/stats/base/dists/log-logistic/pdf/benchmark/benchmark.js

+              	opts = {
+              		'dtype': 'float64'
+              	};
+              	x = uniform( 100, EPS, 20.0, opts );

Contributor

manit2004 Apr 9, 2026

Suggested change

      
            	x = uniform( 100, EPS, 20.0, opts );
          
            	x = uniform( 100, 0.0, 20.0, opts );

same comment

lib/node_modules/@stdlib/stats/base/dists/log-logistic/pdf/benchmark/benchmark.js

+              	opts = {
+              		'dtype': 'float64'
+              	};
+              	x = uniform( 100, EPS, 20.0, opts );

Contributor

manit2004 Apr 9, 2026

Suggested change

      
            	x = uniform( 100, EPS, 20.0, opts );
          
            	x = uniform( 100, 0.0, 20.0, opts );

same comment.

lib/node_modules/@stdlib/stats/base/dists/log-logistic/pdf/benchmark/benchmark.native.js

+              	bopts = {
+              		'dtype': 'float64'
+              	};
+              	x = uniform( 100, EPS, 20.0, bopts );

Contributor

manit2004 Apr 9, 2026

Suggested change

      
            	x = uniform( 100, EPS, 20.0, bopts );
          
            	x = uniform( 100, 0.0, 20.0, bopts );

same comment

lib/node_modules/@stdlib/stats/base/dists/log-logistic/pdf/lib/factory.js

+              		beta <= 0.0
+              	) {
+              		return constantFunction( NaN );
+              	}

Contributor

manit2004 Apr 9, 2026

before returning pdf function we can do some micro-optimization by precomputing beta/alpha and (beta -1) for the exponent.

Contributor

manit2004 commented Apr 9, 2026

comments made for test.pdf.js is also applicable for test.factory.js and test.native.js

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Needs Review Statistics