Feature classes are fully extracted when no features are selected from that class

[amirrezazare9]

Hello,

I encountered an issue regarding feature selection behavior in PyRadiomics.

When I manually select specific features from different feature classes (e.g., firstorder, glcm, glrlm, etc.), the extraction works as expected for the selected features.

However, if I leave a feature class empty (i.e., I do not select any features from that specific class), instead of skipping that class, PyRadiomics extracts all features from that class.

Expected behavior:
If no features are selected from a specific feature class, the extractor should not compute any features from that class.

Current behavior:
If a feature class is included in the configuration but no individual features are explicitly enabled within it, all features from that class are extracted.

This behavior makes it difficult to implement a custom feature selection interface, since leaving a class empty unintentionally results in full extraction of that class.

Is this the intended behavior? If so, what is the recommended way to explicitly disable a feature class when no features are selected?

Thank you for your help