Michael Levy

We currently have model info in a lot of different places: model names, hyperparameters, whether they are good for variable importance/interpretation, etc. Putting all that in one table would make...

See https://topepo.github.io/recipes/reference/step_ordinalscore.html

- [ ] Define a custom recipe and attach it to a model_list - [ ] Train a caret model or two and make a model_list of them with `as.model_list`...

E.g. look for "xNBR" columns of character type and convert. If this is a custom recipe step, need to make sure the conversion is optional in case, e.g. a NBR...

Give align_models_hyperparameters some love

1

Consider messages and getting defaults if only some models' hyperparameter values were provided