nirs4all.pipeline.runner module

Pipeline runner - Main entry point for pipeline execution.

This module provides the PipelineRunner class, which serves as the main interface for executing ML pipelines on spectroscopic datasets. It delegates execution to PipelineOrchestrator and provides prediction/explanation capabilities via Predictor and Explainer classes.

nirs4all.pipeline.runner.init_global_random_state(seed: int | None = None)

Initialize global random state for reproducibility.

Sets random seeds for numpy, Python’s random module, TensorFlow, PyTorch, and sklearn to ensure reproducible results across runs.

Parameters:

seed – Random seed value. If None, uses default seed of 42 for TensorFlow and PyTorch.