Update to 1.6

[kostrykin]

After merging, also the requirements in https://github.com/bioconda/bioconda-recipes/blob/master/recipes/segmetrics/meta.yaml need to be updated.

[github-actions]

Test coverage of 94a3317: 88%

[Copilot]

Pull request overview

This PR updates the segmetrics package to version 1.6, adds the Aggregated Jaccard Coefficient (AJC) measure, and modernizes packaging/CLI/docs while adjusting the test suite (including removing network download of ISBI evaluation software).

Changes:

• Add AggregatedJaccardCoefficient measure with new postprocessing support in the measure/study pipeline.
• Replace the old CLI entrypoint (segmetrics.cli) with a python -m segmetrics module CLI and update docs/tests accordingly.
• Modernize packaging/versioning (setuptools, __version__, python_requires, install_requires) and update CI/docs configs.

Reviewed changes

Copilot reviewed 20 out of 22 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
tests/testsuite.py	Adds AJC coverage, updates CLI invocation, and adjusts dataframe comparison logic + ISBI SEG test gating.
tests/isbi_seg.py	Switches ISBI SEG “official” helper to use bundled .7z and a provided password (no network download).
tests/full-study-test.csv	Updates golden CSV output to include AJC and rounded numeric formatting.
setup.py	Migrates to setuptools, loads version from segmetrics/version.py, adds runtime requirements, sets python_requires.
segmetrics/version.py	Introduces canonical __version__ (1.6.0).
segmetrics/study.py	Extends result handling to allow intermediate representations and applies postprocess() before aggregation/output.
segmetrics/regional.py	Fixes minor naming/docs, and adds AggregatedJaccardCoefficient implementation.
segmetrics/measures.py	Adds a centralized measures module with explicit __all__ exports.
segmetrics/measure.py	Extends measure protocol to support intermediate outputs via postprocess().
segmetrics/contour.py	Switches ndimage import style and uses scipy.ndimage directly for dilation/distance transform.
segmetrics/__main__.py	Implements new module CLI (python -m segmetrics) and dynamic measure construction.
segmetrics/__init__.py	Refactors exports, adds __version__, preserves VERSION alias for backward compatibility.
README.rst	Updates Galaxy link wording and copyright year.
LICENSE	Updates copyright year.
docs/source/usage.rst	Updates CLI usage examples and help reference to python -m segmetrics.
docs/source/index.rst	Removes changelog from the toctree.
docs/source/conf.py	Updates project name and copyright year.
docs/source/changelog.rst	Removes the changelog page content (file deleted).
.readthedocs.yaml	Updates RTD build OS/python and points Sphinx to docs/source/conf.py.
.gitignore	Adds .venv and macOS metadata files.
.github/workflows/testsuite.yml	Updates action versions, attempts to pass ISBI password into tests, adds setup-install smoke test job.

Comments suppressed due to low confidence (1)

segmetrics/main.py:90

• The CLI still uses eval(measure_spec, ..., measures_dict) to instantiate measures. Even with restricted globals/locals, this is exploitable because callers can access __globals__ via any provided class (e.g., Dice.__init__.__globals__[...]) and regain __import__/builtins. Please replace eval with a safe parser (e.g., parse the string with ast and only allow Call(Name(...)) with whitelisted class names/kwargs), or implement an explicit mapping from allowed measure names to constructors.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

compare_dataframe now relies on DataFrame.equals() for float comparisons, which is an exact (bitwise) equality check. This is likely to be flaky across platforms/Python/pandas versions due to float parsing/rounding differences; additionally only study_df is rounded while expected_df is not. Consider using pandas.testing.assert_frame_equal with a tolerance (or round both dataframes before comparing) so the test is robust while still enforcing the intended precision.