Hello,
I'm having trouble understanding what finally happened to the idea of introducing a handle_missing parameter for the OneHotEncoder. My current project could still benefit from such an implementation.
There are many existing issues regarding this topic, however, I cannot deduct what was finally decided/implemented and what wasn't.

• Handle missing values in OneHotEncoder #11996
• ENH: handle missing values in OneHotEncoder #12025
• ENH Adds missing value support to OneHotEncoder #17317
• Include drop='last' to OneHotEncoder #23436

Considering the following features:

import pandas as pd

test_df = pd.DataFrame(
    {"col1": ["red", "blue", "blue"], "col2": ["car", None, "plane"]}
)

when using the encoder:

from sklearn.preprocessing import OneHotEncoder

ohe = OneHotEncoder(
    handle_unknown="ignore",
    sparse_output=False,
    #handle_missing="ignore"
)
ohe.fit_transform(test_df)

I get the output:

array([[0., 1., 1., 0., 0.],
       [1., 0., 0., 0., 1.],
       [1., 0., 0., 1., 0.]])

but what I'm actually looking for is to remove the None, i.e. not create a new feature but set all the others to zero:

array([[0., 1., 1., 0.],
       [1., 0., 0., 0.],
       [1., 0., 0., 1.]])

Is there a way to achieve this without using another transformer object?