revert autoformatter changes

scikit-learn-contrib · Nov 7, 2024 · bd3a02f · bd3a02f
1 parent cdc7768
commit bd3a02f
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 10 deletions.
diff --git a/skglm/datafits/single_task.py b/skglm/datafits/single_task.py
@@ -211,8 +211,7 @@ def gradient_scalar(self, X, y, w, Xw, j):
     def gradient_scalar_sparse(self, X_data, X_indptr, X_indices, y, Xw, j):
         XjTXw = 0.
         for i in range(X_indptr[j], X_indptr[j + 1]):
-            XjTXw += X_data[i] * \
-                self.sample_weights[X_indices[i]] * Xw[X_indices[i]]
+            XjTXw += X_data[i] * self.sample_weights[X_indices[i]] * Xw[X_indices[i]]
         return (XjTXw - self.Xty[j]) / self.sample_weights.sum()
 
     def gradient(self, X, y, Xw):

diff --git a/skglm/estimators.py b/skglm/estimators.py
@@ -601,8 +601,7 @@ def fit(self, X, y):
             Fitted estimator.
         """
         if self.weights is None:
-            warnings.warn(
-                'Weights are not provided, fitting with Lasso penalty')
+            warnings.warn('Weights are not provided, fitting with Lasso penalty')
             penalty = L1(self.alpha, self.positive)
         else:
             penalty = WeightedL1(self.alpha, self.weights, self.positive)
@@ -735,8 +734,7 @@ def path(self, X, y, alphas, coef_init=None, return_n_iter=True, **params):
             The number of iterations along the path. If return_n_iter is set to
             ``True``.
         """
-        penalty = compiled_clone(L1_plus_L2(
-            self.alpha, self.l1_ratio, self.positive))
+        penalty = compiled_clone(L1_plus_L2(self.alpha, self.l1_ratio, self.positive))
         datafit = compiled_clone(Quadratic(), to_float32=X.dtype == np.float32)
         solver = AndersonCD(
             self.max_iter, self.max_epochs, self.p0, tol=self.tol,
@@ -1324,8 +1322,7 @@ def fit(self, X, y):
                 f"two columns. Got one column.\nAssuming that `y` "
                 "is the vector of times and there is no censoring."
             )
-            y = np.column_stack((y, np.ones_like(y))).astype(
-                X.dtype, order="F")
+            y = np.column_stack((y, np.ones_like(y))).astype(X.dtype, order="F")
         elif y.shape[1] > 2:
             raise ValueError(
                 f"{repr(self)} requires the vector of response `y` to have "
@@ -1543,8 +1540,7 @@ def path(self, X, Y, alphas, coef_init=None, return_n_iter=False, **params):
             The number of iterations along the path. If return_n_iter is set to
             ``True``.
         """
-        datafit = compiled_clone(QuadraticMultiTask(),
-                                 to_float32=X.dtype == np.float32)
+        datafit = compiled_clone(QuadraticMultiTask(), to_float32=X.dtype == np.float32)
         penalty = compiled_clone(L2_1(self.alpha))
         solver = MultiTaskBCD(
             self.max_iter, self.max_epochs, self.p0, tol=self.tol,