Skip to content

benchmark: establish systematic linear and GLM inference coverage #105

Description

@TheHiddenObserver

Goal

Add methodical benchmark and validation coverage for maintained linear-model and GLM inference outputs.

Scope

  • align coefficient, covariance, standard-error, test-statistic, p-value, confidence-interval, and likelihood criteria with appropriate external references;
  • cover NumPy/CuPy/Torch paths and strict failure behavior;
  • separate inference correctness from fit timing;
  • retain model, family, covariance mode, scale, tolerance, and seed identity;
  • produce structured sources, parser tests, registration, generated assets, and coverage-matrix updates.

Parent: #100

Metadata

Metadata

Assignees

No one assigned

    Labels

    enhancementNew feature or request

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions