SDV
SDV copied to clipboard
HMA1 with non parametric tabular models: improve the error message
Error Description
When HMA1 is used with a non-parametrical model the message that informs the user about the model not being parametric is being wrongly caught on this try/except block and it ends up producing a cryptic error message. See #340.
Steps to reproduce
See #340.
How to Fix
We should modify the except block to let the NonParametricError go up.
Another possibility to consider is to add a control during instantiation to certify that the model is actually parametric before even starting to fit on the data.