MBrouns

@ghusse I'm also in need of a single slider and would gladly help in testing.

It probably makes the most sense to return all possible states with a 0 probability, right?

I think it's the other way around @koaning. The train set should not contain the outliers, so if A is the training set in step 2, we remove observation with...

> With regard to a possible implementation, I'm not super familiar with the types of arguments available. I know that CV iterators return indices, so if `y` is available it...

I think it makes sense in the way you've currently written it, although it would be nice to see a few test cases to show the impact and behaviour

The thing is that if there are only zero's in your y_train, there is no data left for the regressor to train on. We filter out all zero entries before...

I'm thinking something along the lines of `handle_unknown` in sklearns `OneHotEncoder` ``` ZeroInflatedRegressor(handle_zero='error') ZeroInflatedRegressor(handle_zero='ignore') ```

Hm we probably need to implement a classes_ attribute that patches through to the base estimator, right? On Fri, 18 Jun 2021 at 19:35, Victor Oliveira ***@***.***> wrote: > I...

hi @nipnipj! Could you provide a small example that reproduces this behaviour?

I'd argue the feature in this case would be consistency with the main sklearn API. When we built this transformer either the `ColumnTransformer` wasn't out yet or was still experimental,...