Skip to content

About

Python 历史依赖兼容性与约束修复研究 | A research artifact for historical ML dependency compatibility and budgeted constraint repair.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Repository files navigation

ReproShift

简体中文 · v1.1.0

A research artifact for studying historical dependency compatibility in Python machine-learning repositories and evaluating budgeted constraint repair.

The evaluation distinguishes direct package availability, wheel availability, and binary dependency resolution. Resolver success is a prerequisite for further reproduction work; it does not establish successful imports, tests, training, or numerical reproduction.

Artifact contents

  • Versioned protocols and analysis amendments under config/.
  • Public compatibility and repair projections under data/public/.
  • Parsers, compatibility checks, and repair analysis under src/.
  • Analysis and figure-generation entry points under scripts/.
  • Regression tests under tests/.
  • Manuscript figures and tables under manuscript/.

The v6 extension evaluates the frozen declaration set across CPython 3.10/3.12 and Windows AMD64/Linux x86-64 targets. Outcomes distinguish observed failures, successes, and operational unknowns.

Environment and tests

From the repository root in PowerShell:

py -3.10 -m venv .venv
.\.venv\Scripts\python.exe -m pip install -r requirements.txt
$env:PYTHONPATH=(Resolve-Path .\src)
.\.venv\Scripts\python.exe -m unittest discover -s tests -v

Rebuild the public v6 analysis

.\.venv\Scripts\python.exe scripts\run_public_analysis_v6.py --out data\processed\analysis_v6_public_rebuild.json
.\.venv\Scripts\python.exe scripts\make_tables_v6.py --analysis data\processed\analysis_v6_public_rebuild.json
.\.venv\Scripts\python.exe scripts\make_figures_v6.py --analysis data\processed\analysis_v6_public_rebuild.json
.\.venv\Scripts\python.exe scripts\make_alt_text_v6.py --analysis data\processed\analysis_v6_public_rebuild.json --out docs\FIGURE_ALT_TEXT_V6.md

These commands use the committed public projections and sealed-index receipt. Recollecting live package metadata produces a different dataset. Some upstream classifications require excluded frozen inputs, so exact public reanalysis and full source-data reconstruction have different requirements.

Evaluation boundary

The dependency probes use metadata and wheel-only pip --dry-run. They do not execute target repository setup scripts, import its modules, run its tests, or install its code. Non-index and executable dependency sources are rejected according to the protocol.

Documentation and citation

The release manifest describes the frozen v1.1.0 package. Later documentation changes are recorded in presentation_manifest.json.

First-party code uses MIT. Public derived data, configurations, figures, tables, and documentation use CC BY 4.0. Upstream materials retain their original terms.

About

Python 历史依赖兼容性与约束修复研究 | A research artifact for historical ML dependency compatibility and budgeted constraint repair.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages