PlateProof v0.1.0

PlateProof v0.1.0

PlateProof's first public release audits an existing microplate-map CSV for
transparent spatial design risks without uploading experimental data.

Highlights

• Audits arbitrary categorical factors against plate, row, column, and edge
  position with deterministic normalized mutual information metrics.
• Reports factor levels confined to one position, orthogonally adjacent
  replicate groups, and required roles missing from a plate.
• Distinguishes a clean audit (0), completed audit with review findings (1),
  and invalid input or operational failure (2).
• Produces versioned JSON, portable Markdown, and escaped self-contained
  HTML/SVG plate maps.
• Runs fully offline with a standard-library-only Python 3.11+ runtime.

Install

python -m pip install "https://github.com/KanadeK/plateproof/releases/download/v0.1.0/plateproof-0.1.0-py3-none-any.whl"

Quick example

After unpacking plateproof-0.1.0-demo.zip:

plateproof audit examples/risky.csv \
  --factor treatment \
  --factor cohort \
  --replicate sample_id \
  --role-column role \
  --require-role positive_control \
  --out plateproof-report

The intentionally risky example exits 1, writes all three reports, and finds
four factor/plate confinement cases, one adjacent replicate group, and one
missing per-plate positive control.

Assets

• plateproof-0.1.0-py3-none-any.whl: directly installable universal wheel.
• plateproof-0.1.0.tar.gz: source distribution.
• plateproof-0.1.0-demo.zip: examples, documentation, and a pre-generated
  risky-layout report.

Verification

The release gate passes 40 tests at 94.68% branch-aware coverage, formatting,
lint, strict type checking, all committed success/boundary/review/failure
examples, locked-dependency vulnerability audit, archive contract checks, and a
clean-environment wheel install followed by a real installed-entry-point run.

Limits

PlateProof reports descriptive preflight evidence, not statistical significance
or proof of experimental validity. It does not generate layouts, model liquid
handling, or analyze assay readings. See the README for the full capability
boundary and troubleshooting guidance.