Hi!

Me and @blzserdos have a notebook describing the replicability procedure for choosing the number of components for PARAFAC2. We already made a PR for this check for CP models to tlviz, but the respective PARAFAC2 test is somehow different. We could also extend the notebook to include other component selection techniques if desired. I wanted to kindly ask if that would be a better fit here or in tlviz as an example notebook.

Thanks in advance!