[WIP] Handle NaNs in OneHotEncoder #16749

Closed

nilichen wants to merge 14 commits intoscikit-learn:masterfrom

nilichen:ohe_missing

Commits on Sep 6, 2018

Docstring Update, Modules docs update and initial logic implementation

olamilekan
authored and
olamilekan
committed
Merge branch 'master' of https://github.com/scikit-learn/scikit-learn

olamilekan
authored and
olamilekan
committed

Commits on Sep 8, 2018

Fixed line length and updated docstring

olamilekan
authored and
olamilekan
committed
Fixed line length and updated docstring

olamilekan
authored and
olamilekan
committed

Commits on Sep 12, 2018

Merge branch 'master' of https://github.com/scikit-learn/scikit-learn into handlemissing-onehotencoder

olamilekan
authored and
olamilekan
committed
Updated logic

olamilekan
authored and
olamilekan
committed

Commits on Mar 22, 2020

resove merge conflict
Katrina Ni
committed
first attempt to allow na in _encode_python and deal with handle_missing of 'indicator', 'all-zero'
Katrina Ni
committed

Commits on Mar 23, 2020

all-zero seems to work
Katrina Ni
committed
inverse_tranform for 'all-zero'; force_all_finite as class variable
Katrina Ni
committed
handle_missing = 'indicator'
Katrina Ni
committed
fixed test_encode_check_unknown
Katrina Ni
committed

Commits on Mar 24, 2020

handle when NaN only present during transform, default to handle_missing='ignore'
Katrina Ni
committed
bug in _encode_python w. NaN; self.categores_ during fit
Katrina Ni
committed