piaso-tools 1.0.2

PIASO v1.0.2 - SCALAR, StitchSpace & Parallel Stability

on Python PyPI

New Features

• Ligand-Receptor Analysis: Added piaso.tl.runSCALAR for single-cell L-R inference.
• Dataset Integration: Added piaso.tl.stitchSpace for integrating datasets.
• Preprocessing: Added utility for 2D coordinate rotation in piaso.pp.

Visualization

• SCALAR Plots: Added piaso.pl.plotLigandReceptorInteraction and piaso.pl.plotLigandReceptorLollipop.
• Metrics: Added piaso.pl.plotConfusionMatrix.

Critical Stability

• Fixed "Unsafe Fork" Crash: Switched ProcessPoolExecutor to use the 'spawn' start method. This resolves Terminating: fork() called from a process already using GNU OpenMP errors on Linux/macOS.
• Strict Seeding: random_seed is now enforced across internal steps (Neighbors, Leiden).
• Race Condition Resolved: Fixed a logic error in parallel functions where results were processed by completion order rather than batch order. Results are now explicitly sorted.

Compatibility & Maintenance

• Updated for Pandas 2.1+ (applymap -> map) and fixed Scanpy/AnnData warnings.

Thank you for your feedback and support!