Skip to content

Replace performance claims with reproducible baselines #62

Description

@stefan-jansen

Findings: BR-026.

Establish controlled retained baselines for deterministic single-asset, 250-asset daily,
quote-aware, rebalance, and partial-fill workloads. Remove or correct every unsupported headline
runtime, throughput, memory, and relative-framework claim.

Primary surfaces: benchmark harness, performance CI, README and performance documentation,
retained result metadata.

Acceptance:

  • Every public performance number reproduces within 10 percent under its stated environment.
  • Runtime excludes or includes setup explicitly, and memory uses defined whole-process boundaries.
  • Framework comparisons use identical inputs and equivalent measurement boundaries.
  • Output checksums prove benchmark behavior did not change during measurement.

Dependencies: Issues 11-12, 16, and 21.


Plan reference: /home/stefan/ml4t/libraries/ml4t-backtest-dev/.workspace/work/backtest-stable-release/milestone-4/plan.md - Issue 22 in milestone 0.1.0-m4 - Maintainability, performance, and adoption.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions