MULM provide basic features similar to “statmodels” (like OLS) where Y is a matrix of many responses where many independant fit are requested. Usefull to perform univariate statistics and residualization to remove site, or age effect. MULM provides a Residualizer module that can be used, as scalers, in scikit-learn pipelines.

Univariate statistics

Univariate T and F tests. Usefull in neuroimaging to perform Voxel Based Morphometry (VBM) in python.

Residualizer

Residualizer basic usage.

Residualization in scikit-learn pipelines.

Example in neuroimaging

Univariate statistics (Voxel Based Morphometry) & Residualization (remove sex) before age prediction with scikit-learn