Run ruff fix again

src/pylemur/tl/_design_matrix_utils.py

@@ -1,4 +1,5 @@
 from collections.abc import Mapping
+
 import numpy as np
 import pandas as pd
 

src/pylemur/tl/_grassmann_lm.py

@@ -1,4 +1,5 @@
 import numpy as np
+
 from pylemur.tl._design_matrix_utils import row_groups
 from pylemur.tl._grassmann import grassmann_log, grassmann_map
 from pylemur.tl._lin_alg_wrappers import fit_pca, ridge_regression
@@ -18,7 +19,7 @@ def grassmann_geodesic_regression(coord_systems, design, base_point, weights=Non
     weights : array-like, shape (n_groups,)
 
     Returns
-    ----------
+    -------
     beta: array-like, shape (n_emb, n_features, n_coef)
     """
     n_obs = design.shape[0]
@@ -33,7 +34,7 @@ def grassmann_geodesic_regression(coord_systems, design, base_point, weights=Non
         weights = np.ones(n_obs)
 
     tangent_vecs = [
-        grassmann_log(base_point.T, coord_systems[i].T).T.reshape((n_emb * n_features)) for i in range(n_obs)
+        grassmann_log(base_point.T, coord_systems[i].T).T.reshape(n_emb * n_features) for i in range(n_obs)
     ]
     tangent_vecs = np.vstack(tangent_vecs)
     if tangent_vecs.shape[0] == 0:
@@ -48,6 +49,7 @@ def grassmann_geodesic_regression(coord_systems, design, base_point, weights=Non
 def grassmann_lm(Y, design_matrix, base_point):
     """
     Solve Sum_i||Y_i: - Y_i: Proj(Exp_p(Sum_k V_k:: * X_ik))||^2 for V.
+
     Parameters
     ----------
     Y : array-like, shape (n_samples, n_features)
@@ -56,6 +58,7 @@ def grassmann_lm(Y, design_matrix, base_point):
         The design matrix.
     base_point : array-like, shape (n_emb, n_features)
         The base point.
+
     Returns
     -------
     beta: array-like, shape (n_emb, n_features, n_coef)

src/pylemur/tl/_lin_alg_wrappers.py

@@ -1,6 +1,7 @@
-import sklearn.decomposition as skd
 from typing import NamedTuple
+
 import numpy as np
+import sklearn.decomposition as skd
 
 
 class PCA(NamedTuple):
@@ -12,6 +13,7 @@ class PCA(NamedTuple):
 def fit_pca(Y, n, center=True):
     """
     Calculate the PCA of a given data matrix Y.
+
     Parameters
     ----------
     Y : array-like, shape (n_samples, n_features)
@@ -20,6 +22,7 @@ def fit_pca(Y, n, center=True):
         The number of principal components to return.
     center : bool, default=True
         If True, the data will be centered before computing the covariance matrix.
+
     Returns
     -------
     pca : sklearn.decomposition.PCA
@@ -41,6 +44,7 @@ def fit_pca(Y, n, center=True):
 def ridge_regression(Y, X, ridge_penalty=0, weights=None):
     """
     Calculate the ridge regression of a given data matrix Y.
+
     Parameters
     ----------
     Y : array-like, shape (n_samples, n_features)
@@ -51,6 +55,7 @@ def ridge_regression(Y, X, ridge_penalty=0, weights=None):
         The ridge penalty.
     weights : array-like, shape (n_features,)
         The weights to apply to each feature.
+
     Returns
     -------
     ridge: array-like, shape (n_coef, n_features)

src/pylemur/tl/alignment.py

@@ -1,13 +1,8 @@
-from typing import Union
 import harmonypy
 import numpy as np
-import pandas as pd
 
 from pylemur.tl._design_matrix_utils import row_groups
-from pylemur.tl._lin_alg_wrappers import multiply_along_axis, ridge_regression
-
-
-
+from pylemur.tl._lin_alg_wrappers import ridge_regression
 
 
 def _align_impl(

src/pylemur/tl/lemur.py

@@ -1,18 +1,23 @@
 import re
-from typing import Any, Literal, Union
-from collections.abc import Iterable, Mapping
 import warnings
+from collections.abc import Iterable, Mapping
+from typing import Any, Literal, Union
+
+import anndata as ad
 import formulaic
 import numpy as np
-import anndata as ad
 from sklearn.exceptions import NotFittedError
 
 from pylemur.tl._design_matrix_utils import *
-from pylemur.tl._grassmann_lm import grassmann_lm, project_data_on_diffemb
 from pylemur.tl._grassmann import grassmann_map
+from pylemur.tl._grassmann_lm import grassmann_lm, project_data_on_diffemb
 from pylemur.tl._lin_alg_wrappers import *
-from pylemur.tl.alignment import _align_impl, _apply_linear_transformation, _init_harmony, _reverse_linear_transformation
-
+from pylemur.tl.alignment import (
+    _align_impl,
+    _apply_linear_transformation,
+    _init_harmony,
+    _reverse_linear_transformation,
+)
 
 
 class LEMUR:
@@ -122,7 +127,6 @@ def fit(self, verbose: bool = True):
         `self`
             The fitted LEMUR model.
         """
-
         Y = self.data_matrix
         design_matrix = self.design_matrix
         n_embedding = self.n_embedding
@@ -154,7 +158,7 @@ def fit(self, verbose: bool = True):
         self.linear_coefficients = lin_coef
 
         return self
-    
+
     def align_with_harmony(
         self, ridge_penalty: Union[float, list[float], np.ndarray] = 0.01, max_iter: int = 10, verbose: bool = True
     ):
@@ -287,7 +291,7 @@ def transform(self, adata: ad.AnnData, layer: Union[str, None] = None,
         design_matrix, _ = handle_design_parameter(self.formula, adata.obs)
         dm = design_matrix.to_numpy()
         Y_clean = Y - dm @ self.linear_coefficients
-        embedding = project_data_on_diffemb(Y_clean, design_matrix = dm, coefficients = self.coefficients, 
+        embedding = project_data_on_diffemb(Y_clean, design_matrix = dm, coefficients = self.coefficients,
                                             base_point = self.base_point)
         embedding = _apply_linear_transformation(embedding, self.alignment_coefficients, dm)
         if return_type == "embedding":
@@ -298,7 +302,7 @@ def transform(self, adata: ad.AnnData, layer: Union[str, None] = None,
             fit.design_matrix = design_matrix
             fit.embedding = embedding
             return fit
-    
+
 
     def predict(self,
         embedding: Union[np.ndarray, None] = None,
@@ -334,7 +338,6 @@ def predict(self,
         array-like, shape (n_cells, n_genes)
             The predicted expression of the cells in the new condition.
         """
-
         if embedding is None:
             if self.embedding is None:
                 raise NotFittedError("The model has not been fitted yet.")
@@ -443,13 +446,13 @@ def _get_var_from_colname(colname):
         df = pd.DataFrame([kwargs])
 
         return design_matrix.model_spec.get_model_matrix(df)
-    
+
     def __str__(self):
         if self.embedding is None:
             return f"LEMUR model (not fitted yet) with {self.n_embedding} dimensions"
         else:
             return f"LEMUR model with {self.n_embedding} dimensions"
-        
+
 
 
 def order_axis_by_variance(embedding, coefficients, base_point):

tests/test_grassmann.py

@@ -1,4 +1,5 @@
 import numpy as np
+
 from pylemur.tl._grassmann import *
 
 

tests/test_lin_alg_utils.py

@@ -1,5 +1,6 @@
 import numpy as np
 from sklearn.linear_model import LinearRegression, Ridge
+
 from pylemur.tl._lin_alg_wrappers import *