Replacing csd in sample selection with california because the larger …

…dataset means a few redundant samples, which is not good for stable testing

docs/source/conf.py

@@ -209,7 +209,7 @@
 # (source start file, target name, title,
 #  author, documentclass [howto, manual, or own class]).
 latex_documents = [
-    (master_doc, "skcosmo.tex", u"scikit-COSMO Documentation", author, "manual"),
+    (master_doc, "skcosmo.tex", "scikit-COSMO Documentation", author, "manual"),
 ]
 
 # The name of an image file (relative to this directory) to place at the top of

examples/PlotLFRE.ipynb

@@ -140,7 +140,7 @@
     "ax34.set_title(r\"$X^-$ LFRE(3-body, 4-body)\")\n",
     "ax43.set_title(r\"$X^-$ LFRE(4-body, 3-body)\")\n",
     "\n",
-    "cbar = fig.colorbar(pcm, ax=(ax34, ax43), label=\"LFRE\", location=\"bottom\")\n",
+    "cbar = fig.colorbar(pcm, ax=[ax34, ax43], label=\"LFRE\", location=\"bottom\")\n",
     "\n",
     "plt.show()"
    ]

skcosmo/_selection.py

@@ -932,7 +932,7 @@ def _init_greedy_search(self, X, y, n_to_select):
 
         super()._init_greedy_search(X, y, n_to_select)
 
-        self.norms_ = (X ** 2).sum(axis=abs(self._axis - 1))
+        self.norms_ = (X**2).sum(axis=abs(self._axis - 1))
         self.haussdorf_ = np.full(X.shape[self._axis], np.inf)
         self.haussdorf_at_select_ = np.full(X.shape[self._axis], np.inf)
 

skcosmo/decomposition/_kernel_pcovr.py

@@ -624,7 +624,7 @@ def _decompose_full(self, mat):
         U, Vt = svd_flip(U, Vt)
 
         # Get variance explained by singular values
-        explained_variance_ = (S ** 2) / (self.n_samples_ - 1)
+        explained_variance_ = (S**2) / (self.n_samples_ - 1)
         total_var = explained_variance_.sum()
         explained_variance_ratio_ = explained_variance_ / total_var
 

skcosmo/metrics/_reconstruction_measures.py

@@ -441,8 +441,8 @@ def pointwise_local_reconstruction_error(
     Y_test = scaler.transform(Y_test)
 
     squared_dist = (
-        np.sum(X_train ** 2, axis=1)
-        + np.sum(X_test ** 2, axis=1)[:, np.newaxis]
+        np.sum(X_train**2, axis=1)
+        + np.sum(X_test**2, axis=1)[:, np.newaxis]
         - 2 * X_test @ X_train.T
     )
 

skcosmo/sample_selection/_voronoi_fps.py

@@ -195,7 +195,7 @@ def _init_greedy_search(self, X, y, n_to_select):
 
         super()._init_greedy_search(X, y, n_to_select)
 
-        self.norms_ = (X ** 2).sum(axis=abs(self._axis - 1))
+        self.norms_ = (X**2).sum(axis=abs(self._axis - 1))
 
         if self.initialize == "random":
             random_state = check_random_state(self.random_state)

skcosmo/utils/_pcovr_utils.py

@@ -177,8 +177,8 @@ def pcovr_covariance(
                 random_state=random_state,
             )
 
-            UC = UC.T[:, (vC ** 2) > rcond]
-            vC = vC[(vC ** 2) > rcond]
+            UC = UC.T[:, (vC**2) > rcond]
+            vC = vC[(vC**2) > rcond]
 
         C_isqrt = UC @ np.diagflat(1.0 / vC) @ UC.T
 

tests/test_sample_simple_cur.py

@@ -3,14 +3,15 @@
 import numpy as np
 from sklearn import exceptions
 
-from skcosmo.datasets import load_csd_1000r as load
+from sklearn.datasets import fetch_california_housing as load
 from skcosmo.sample_selection import CUR
 
 
 class TestCUR(unittest.TestCase):
     def setUp(self):
-        self.X, self.y = load(return_X_y=True)
-        self.n_select = 20
+        self.X, _ = load(return_X_y=True)
+        self.X = self.X[:1000]
+        self.n_select = min(20, min(self.X.shape) // 2)
 
     def test_bad_transform(self):
         selector = CUR(n_to_select=2)