Dimensionality Reduction

We have been experimenting with input data for classification, notably adding interesting features (various indices and such) and trying to reduce the dimensionality of our input data (mostly through PCA analysis). However I have just discovered the is already a script to do so within iota (https://framagit.org/iota2-project/iota2/blob/develop/iota2/Sampling/DimensionalityReduction.py), but have not found any documentation related to this feature.

Could you clarify how this feature works? Is it active by default or do we have to parametrize it? What options are available? Can we input additional features alongside original spectral bands?

I apologize for the numerous questions but I am very curious to see what can be done with this feature.

Thanks