Skip to content

D01 19352 imputer #1

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
ISofiko merged 2 commits into from
Mar 7, 2021
Merged

D01 19352 imputer #1

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
1c70c8e
Added checker for imputed values shape to avoid shape mismatch, added…
ISofiko Mar 1, 2021
4be3bf2
added a test case and comments for shape mismatch issue
ISofiko Mar 2, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 6 additions & 0 deletions sklearn/impute/_iterative.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -344,6 +344,12 @@ def _impute_one_feature(self,
self._min_value[feat_idx],
self._max_value[feat_idx])

# If estimator prediction returns a multi-dimensional numpy array we
# need to re-cast it to the wanted shape
shape_imputed_values = imputed_values.shape
if len(shape_imputed_values) > 1:
imputed_values = imputed_values.reshape(shape_imputed_values[0])

# update the feature
X_filled[missing_row_mask, feat_idx] = imputed_values
return X_filled, estimator
Expand Down
32 changes: 30 additions & 2 deletions sklearn/impute/tests/test_impute.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,7 @@

import io

from sklearn.cross_decomposition import PLSRegression
from sklearn.utils._testing import assert_allclose
from sklearn.utils._testing import assert_allclose_dense_sparse
from sklearn.utils._testing import assert_array_equal
Expand All @@ -16,7 +17,7 @@
# make IterativeImputer available
from sklearn.experimental import enable_iterative_imputer # noqa

from sklearn.datasets import load_diabetes
from sklearn.datasets import load_diabetes, fetch_california_housing
from sklearn.impute import MissingIndicator
from sklearn.impute import SimpleImputer, IterativeImputer
from sklearn.dummy import DummyRegressor
Expand Down Expand Up @@ -633,7 +634,7 @@ def test_iterative_imputer_imputation_order(imputation_order):

@pytest.mark.parametrize(
"estimator",
[None, DummyRegressor(), BayesianRidge(), ARDRegression(), RidgeCV()]
[None, DummyRegressor(), BayesianRidge(), ARDRegression(), RidgeCV(), PLSRegression()]
)
def test_iterative_imputer_estimators(estimator):
rng = np.random.RandomState(0)
Expand All @@ -660,6 +661,33 @@ def test_iterative_imputer_estimators(estimator):
assert len(set(hashes)) == len(hashes)


def test_iterative_imputer_multiple_components():

rng = np.random.RandomState(42)

# get sample data from california housing dataset
X_california, y_california = fetch_california_housing(return_X_y=True)
n_samples, n_features = X_california.shape

# Add missing values in 75% of the lines
missing_rate = 0.75
n_missing_samples = int(n_samples * missing_rate)
missing_samples = np.zeros(n_samples, dtype=bool)
missing_samples[: n_missing_samples] = True

rng.shuffle(missing_samples)
missing_features = rng.randint(0, n_features, n_missing_samples)
X_missing = X_california.copy()
X_missing[missing_samples, missing_features] = np.nan

# PLSRegression returns multi-dimensional numpy array as
# opposed to other estimators which return 1-D array,
# but this should not cause issues in the imputer
imputer = IterativeImputer(estimator=PLSRegression(n_components=2))
X_imputed = imputer.fit_transform(X_missing)
assert X_imputed.shape == X_california.shape


def test_iterative_imputer_clip():
rng = np.random.RandomState(0)
n = 100
Expand Down