Use different estimator for different features in IterativeImputer

[jnothman]

An idealised IterativeImputer would allow some features to be predicted as categorical variables (with a classifier) and others to be predicted as continuous variables (with a regressor) or even other specific distributions.

We could consider an interface to IterativeImputer that would allow users to specify the estimator used for a particular column selector, much like ColumnTransformer does with transformers.

On the other hand we may decide that this adds rarely needed power with excessive complexity.

[amueller]

A mixture of categorical and continuous features seems not that rare but handling categorical missing values with adding a new category might be enough for most cases?

[yacth]

Hi @jnothman can I work on this issue ?

[jnothman]

It's fairly complicated, @yacth, and not at all ensured to be merged.

[vkhodygo]

Any progress in this direction? As more and more people start working with data nowadays, imputing missing values properly is a must, and this part of scikit is a bit lacking.