Skip to content

FIX GaussianProcessRegression(normalize_y=True).predict(X, return_cov=True) #19706

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

Closed
wants to merge 10 commits into from
Closed

FIX GaussianProcessRegression(normalize_y=True).predict(X, return_cov=True) #19706

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
10 commits
Select commit Hold shift + click to select a range
43126ec
[FIX] of the issue scikit-learn#17394 and scikit-learn#18065
AidarShakerimoff Mar 17, 2021
c0247c0
Update _gpr.py
AidarShakerimoff Mar 18, 2021
d119b07
linting
AidarShakerimoff Mar 29, 2021
865240a
linting2
AidarShakerimoff Mar 29, 2021
db500ff
linting3
AidarShakerimoff Mar 29, 2021
86df0ee
doc
AidarShakerimoff Mar 29, 2021
f4b702b
Update _gpr.py
AidarShakerimoff Apr 9, 2021
b45372c
Update _gpr.py
AidarShakerimoff Apr 10, 2021
5f85260
multioutput test is improved
AidarShakerimoff Apr 10, 2021
4d2087d
Update test_gpr.py
AidarShakerimoff Apr 10, 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
41 changes: 35 additions & 6 deletions sklearn/gaussian_process/_gpr.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -203,8 +203,13 @@ def fit(self, X, y):
y = (y - self._y_train_mean) / self._y_train_std

else:
self._y_train_mean = np.zeros(1)
self._y_train_std = 1
if hasattr(y[0], "__len__"):
self._y_train_mean = np.zeros(y.shape[1])
self._y_train_std = np.ones(y.shape[1])
else:
self._y_train_mean = np.zeros(1)
self._y_train_std = 1


if np.iterable(self.alpha) \
and self.alpha.shape[0] != y.shape[0]:
Expand Down Expand Up @@ -305,11 +310,12 @@ def predict(self, X, return_std=False, return_cov=False):
y_mean : ndarray of shape (n_samples, [n_output_dims])
Mean of predictive distribution a query points.

y_std : ndarray of shape (n_samples,), optional
y_std : ndarray of shape (n_samples, [n_output_dims]), optional
Standard deviation of predictive distribution at query points.
Only returned when `return_std` is True.

y_cov : ndarray of shape (n_samples, n_samples), optional
y_cov : ndarray of shape (n_samples, n_samples, [n_output_dims]),
optional
Covariance of joint predictive distribution a query points.
Only returned when `return_cov` is True.
"""
Expand Down Expand Up @@ -352,7 +358,23 @@ def predict(self, X, return_std=False, return_cov=False):
y_cov = self.kernel_(X) - K_trans.dot(v) # Line 6

# undo normalisation
y_cov = y_cov * self._y_train_std**2

# for multitarget data:
if hasattr(self._y_train_std, "__len__"):
y_cov_x = y_cov.shape[0]
y_cov_y = y_cov.shape[1]
y_cov_copy = y_cov.reshape((y_cov_x, y_cov_y, 1))
y_cov = np.zeros((y_cov_x, y_cov_x,
self._y_train_std.shape[0]))
idx = 0
for line in y_cov_copy:
line = line * self._y_train_std**2
y_cov[idx] = line
idx += 1

# for single target data
else:
y_cov = y_cov * self._y_train_std**2

return y_mean, y_cov
elif return_std:
Expand All @@ -378,7 +400,14 @@ def predict(self, X, return_std=False, return_cov=False):
y_var[y_var_negative] = 0.0

# undo normalisation
y_var = y_var * self._y_train_std**2

# for multitarget data
if hasattr(self._y_train_std, "__len__"):
y_var = y_var.reshape((-1, 1))
y_var = y_var * self._y_train_std**2
# for single target data
else:
y_var = y_var * self._y_train_std**2

return y_mean, np.sqrt(y_var)
else:
Expand Down
7 changes: 5 additions & 2 deletions sklearn/gaussian_process/tests/test_gpr.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -358,8 +358,11 @@ def test_y_multioutput():
assert_almost_equal(y_pred_1d, y_pred_2d[:, 1] / 2)

# Standard deviation and covariance do not depend on output
assert_almost_equal(y_std_1d, y_std_2d)
assert_almost_equal(y_cov_1d, y_cov_2d)
assert_almost_equal(y_std_1d, y_std_2d[:, 0])
assert_almost_equal(y_std_1d, y_std_2d[:, 1])

assert_almost_equal(y_cov_1d, y_cov_2d[:, :, 0])
assert_almost_equal(y_cov_1d, y_cov_2d[:, :, 1])

y_sample_1d = gpr.sample_y(X2, n_samples=10)
y_sample_2d = gpr_2d.sample_y(X2, n_samples=10)
Expand Down