omicverse.metabol.plsda

Contents

omicverse.metabol.plsda#

omicverse.metabol.plsda(adata, *, group_col='group', group_a=None, group_b=None, n_components=2, scale=False)[source]#

Partial Least Squares Discriminant Analysis (wraps sklearn PLS).

Parameters:

  • n_components (int (default: 2)) – Number of latent components. 2 is standard for visualization; use leave-one-out Q² to pick the optimal count for classification.

  • scale (bool (default: False)) – Scale features inside sklearn PLS (z-score). Usually False here because you’ve already Pareto-scaled via transform().

Return type:

PLSDAResult

Contents