Tutorial request

[j-bac]

Thanks for the great library !

It would be nice to have an example of how to run things within an already defined pipeline, using the lower-level API. This could be showing a simple run of a knockoff sampler on training data within each fold, fitting one's model manually, then running one of the package statistics ?

For example, it's not fully clear to me the high-level API is flexible enough in my case, or what would be the best practice to use it:

• already running a custom (nested) cross-validation scheme, including hyperparameter selection
• fitting models that don't follow sklearn API or have extra steps