Adding additional regressors to Nilearn.decoding.SearchLight

Hi all,

I’m running a wholebrain searchlight analysis and want to add a behavioral regressor to my analysis, so that each searchlight has the X number of voxels + a few other data points. Is that possible to do using this function, if so how?

I’ve tried building a manual searchlight but that has generally been significantly slower and would like to avoid rediving into that if possible.