Fix flake8 errors (#62)

scikit-adaptation · Jan 26, 2024 · 3e84668 · 3e84668
1 parent d4eb894
commit 3e84668
Show file tree

Hide file tree

Showing 2 changed files with 14 additions and 9 deletions.
diff --git a/skada/_subspace.py b/skada/_subspace.py
@@ -254,7 +254,10 @@ def fit(self, X, y=None, sample_domain=None, **kwargs):
             allow_multi_source=True,
             allow_multi_target=True,
         )
-        self.X_source_, self.X_target_ = source_target_split(X, sample_domain=sample_domain)
+        self.X_source_, self.X_target_ = source_target_split(
+            X,
+            sample_domain=sample_domain
+        )
 
         Kss = pairwise_kernels(self.X_source_, metric=self.kernel)
         Ktt = pairwise_kernels(self.X_target_, metric=self.kernel)

diff --git a/skada/metrics.py b/skada/metrics.py
@@ -170,8 +170,10 @@ def _score(self, estimator, X, y, sample_domain=None, **params):
 
         X, y, sample_domain = check_X_y_domain(X, y, sample_domain)
         X_source, X_target, y_source, _ = source_target_split(
-            X, y, sample_domain=sample_domain
-        ) 
+            X,
+            y,
+            sample_domain=sample_domain
+        )
         self._fit(X_source, X_target)
         ws = self.weight_estimator_source_.score_samples(X_source)
         wt = self.weight_estimator_target_.score_samples(X_source)
@@ -333,10 +335,10 @@ def __init__(
         self._sign = 1 if greater_is_better else -1
 
     def _no_reduc_log_loss(self, y, y_pred):
-         return np.array(
-             [
-                 log_loss(y[i : i + 1], y_pred[i : i + 1], labels=np.unique(y))
-                 for i in range(len(y))
+        return np.array(
+            [
+                log_loss(y[i : i + 1], y_pred[i : i + 1], labels=np.unique(y))
+                for i in range(len(y))
              ]
          )
 
@@ -362,12 +364,12 @@ def _score(self, estimator, X, y, sample_domain=None, **kwargs):
         features_train = estimator.get_features(X_train)
         features_val = estimator.get_features(X_val)
         features_target = estimator.get_features(X[~source_idx])
-        
+
         self._fit_adapt(features_train, features_target)
         N, N_target = len(features_train), len(features_target)
         predictions = self.domain_classifier_.predict_proba(features_val)
         weights = N / N_target * predictions[:, :1] / predictions[:, 1:]
-        
+
         y_pred = estimator.predict_proba(X_val, sample_domain=sample_domain_val)
         error = self._loss_func(y_val, y_pred)
         assert weights.shape[0] == error.shape[0]