Apply Black code formatting

Author: github-actions[bot]

setup.py

@@ -23,7 +23,9 @@ def build_extension(self, ext):
         self.spawn(["cmake", "--build", build_temp, "--target", "QR_cpp"])
 
         # Dynamically find the compiled shared library
-        matches = glob.glob(os.path.join(ext.sourcedir, "src", "pyclassify", "QR_cpp*.so"))
+        matches = glob.glob(
+            os.path.join(ext.sourcedir, "src", "pyclassify", "QR_cpp*.so")
+        )
         if not matches:
             raise RuntimeError(
                 "Could not find compiled QR_cpp shared library in expected location."

src/pyclassify/eigenvalues.py

@@ -14,7 +14,8 @@
     max_iteration_warning,
 )
 
-#from parallel_tridiag_eigen import parallel_eigen
+# from parallel_tridiag_eigen import parallel_eigen
+
 
 def eigenvalues_np(A, symmetric=True):
     """
@@ -221,7 +222,7 @@ def __init__(self, A: np.ndarray, max_iter=5000, tol=1e-8, tol_deflation=1e-12):
         self.diag = None
         self.off_diag = None
         self.Q = None
-        self.tol_deflation=tol_deflation
+        self.tol_deflation = tol_deflation
 
     # @jit(nopython=True, parallel=True) # removed because is it not compatible with C++ functions!
     def Lanczos_PRO(self, A=None, q=None, m=None, tol=np.sqrt(np.finfo(float).eps)):
@@ -383,7 +384,6 @@ def eig(self, diag=None, off_diag=None):
         Q_triangular = np.array(Q_triangular)
         return np.array(eig), Q_triangular @ self.Q.T
 
-
     # def parallel_tridiagMatrix_eig_solver(self, diag=None, off_diag=None):
     #     if diag is None and off_diag is None:
     #         if self.diag is None:
@@ -394,9 +394,10 @@ def eig(self, diag=None, off_diag=None):
     #         off_diag = self.off_diag
     #     if len(diag) != (len(off_diag) + 1):
     #         raise ValueError("Mismatch  between diagonal and off diagonal size")
-        
+
     #     return(parallel_eigen(self.diag, self.off_diag, self.tol_QR, self.max_iterQR, self.tol_deflation))
 
+
 # def power_method_cp(A, max_iter=500, tol=1e-4, x=None):
 #    """
 #    Compute the dominant eigenvalue of a square matrix using the power method.

src/pyclassify/parallel_tridiag_eigen.py

@@ -139,7 +139,17 @@ def deflate_eigenpairs(D, v, beta, tol_factor=1e-12):
 
 
 @profile
-def parallel_tridiag_eigen(diag, off, comm=None, tol_factor=1e-16, min_size=1,  depth=0, profiler=None, tol_QR=1e-8, max_iterQR=5000):
+def parallel_tridiag_eigen(
+    diag,
+    off,
+    comm=None,
+    tol_factor=1e-16,
+    min_size=1,
+    depth=0,
+    profiler=None,
+    tol_QR=1e-8,
+    max_iterQR=5000,
+):
     """
     Computes eigenvalues and eigenvectors of a symmetric tridiagonal matrix.
         Input:
@@ -183,10 +193,8 @@ def parallel_tridiag_eigen(diag, off, comm=None, tol_factor=1e-16, min_size=1,
 
     if n <= min_size or size == 1:
         eigvals, eigvecs = QR_algorithm(diag, off, tol_QR, max_iterQR)
-        eigvecs=np.array(eigvecs)
-        eigvals=np.array(eigvals)
-
-        
+        eigvecs = np.array(eigvecs)
+        eigvals = np.array(eigvals)
 
         # profiler[-1].disable_by_count()
         # with open(prof_filename, "w") as f:
@@ -384,9 +392,16 @@ def parallel_eigen(main_diag, off_diag, tol_QR, max_iterQR, tol_deflation):
     comm = MPI.COMM_WORLD
     main_diag = comm.bcast(main_diag, root=0)
     off_diag = comm.bcast(off_diag, root=0)
-    eigvals, eigvecs = parallel_tridiag_eigen(main_diag, off_diag, comm=comm, min_size=1, tol_factor=tol_deflation,  tol_QR=tol_QR, max_iterQR=max_iterQR)
+    eigvals, eigvecs = parallel_tridiag_eigen(
+        main_diag,
+        off_diag,
+        comm=comm,
+        min_size=1,
+        tol_factor=tol_deflation,
+        tol_QR=tol_QR,
+        max_iterQR=max_iterQR,
+    )
     return eigvals, eigvecs
-    
 
 
 if __name__ == "__main__":

test/test_eigensolvers.py

@@ -17,6 +17,7 @@
 )
 from pyclassify.utils import make_symmetric
 
+
 @pytest.fixture(autouse=True)
 def set_random_seed():
     seed = 1422
@@ -115,8 +116,6 @@ def test_EigenSolver(size):
         _ = eigensolver.compute_eigenval(diag=np.arange(2), off_diag=np.arange(49))
 
 
-
-
 # @pytest.mark.parametrize("size", sizes)
 # @pytest.mark.parametrize("density", densities)
 # def test_cupy(size, density):

test/test_zero_finder.py

@@ -38,7 +38,6 @@ def test_psi_s(rho, d, v, i):
     assert psi2(lambda_guess) * rho >= 0, "Error. Inconsistent with the theory."
 
 
-
 @pytest.mark.parametrize("d", d_s)
 @pytest.mark.parametrize("rho", rho_s)
 @pytest.mark.parametrize("v", v_s)
@@ -61,6 +60,7 @@ def test_compute_eigenvalues(rho, d, v):
             np.abs(computed_eigs[i] - exact_eigs[i]) < 1e-8
         ), "Error. The eigenvalues were not computed correctly."
 
+
 @pytest.mark.parametrize("d", d_s)
 @pytest.mark.parametrize("rho", rho_s)
 @pytest.mark.parametrize("v", v_s)
@@ -82,4 +82,3 @@ def test_compute_eigenvalues(rho, d, v):
         assert (
             np.abs(computed_eigs[i] - exact_eigs[i]) < 1e-8
         ), "Error. The eigenvalues were not computed correctly."
-