nirs4all.data.loaders.base module

Base file loader interface and registry.

This module defines the abstract FileLoader base class and LoaderRegistry for a pluggable file loading system. It supports multiple file formats with automatic format detection and configurable loading parameters.

Phase 2 Implementation - Dataset Configuration Roadmap

exception nirs4all.data.loaders.base.FormatNotSupportedError[source]

Bases: LoaderError

Raised when a file format is not supported.

exception nirs4all.data.loaders.base.LoaderError[source]

Bases: Exception

Base exception for loader errors.

nirs4all.data.loaders.base.apply_na_policy(data: DataFrame, na_policy: str, na_fill_config: NAFillConfig | dict[str, Any] | None = None) → tuple[DataFrame, dict][source]

Apply NA policy to loaded data.

Parameters:

  • data – The loaded DataFrame, possibly containing NaN.

  • na_policy – One of the canonical NAPolicy values: ‘auto’, ‘abort’, ‘remove_sample’, ‘remove_feature’, ‘replace’, ‘ignore’.

  • na_fill_config – Fill configuration when na_policy=’replace’.

Returns:

Tuple of (processed DataFrame, na_handling report dict).

Raises:

  • NAError – When na_policy=’abort’ and NaN is detected.

  • ValueError – When na_policy is not a recognized value.