rename grammian to gramian

CI/unit_tests/utils/test_matrix_utils.py

@@ -25,7 +25,7 @@
 from numpy.testing import assert_array_almost_equal, assert_array_equal, assert_raises
 
 from papyrus.utils import (
-    compute_grammian_diagonal_distribution,
+    compute_gramian_diagonal_distribution,
     compute_hermitian_eigensystem,
     compute_l_pq_norm,
     flatten_rank_4_tensor,
@@ -140,9 +140,9 @@ def test_normalizing_gram_matrix(self):
         assert_array_almost_equal(normalized_matrix[0], np.zeros(4))
         assert_array_almost_equal(normalized_matrix[:, 0], np.zeros(4))
 
-    def test_compute_grammian_diagonal_distribution(self):
+    def test_compute_gramian_diagonal_distribution(self):
         """
-        Test the computation of the grammian diagonal distribution.
+        Test the computation of the gramian diagonal distribution.
 
         The test is done in the following steps:
             - Compute a gram matrix
@@ -158,7 +158,7 @@ def test_compute_grammian_diagonal_distribution(self):
         array_norm_density = array_norm / array_norm.sum()
 
         # Evaluate the magnitude density with the function that is to be tested
-        mag_density = compute_grammian_diagonal_distribution(matrix)
+        mag_density = compute_gramian_diagonal_distribution(matrix)
         assert_array_almost_equal(array_norm_density, mag_density)
 
     def test_compute_l_pq_norm(self):

papyrus/measurements/measurements.py

@@ -33,7 +33,7 @@
     compute_von_neumann_entropy,
 )
 from papyrus.utils.matrix_utils import (
-    compute_grammian_diagonal_distribution,
+    compute_gramian_diagonal_distribution,
     compute_hermitian_eigensystem,
 )
 
@@ -415,7 +415,7 @@ def apply(self, ntk: np.ndarray) -> np.ndarray:
                 "To compute the self-entropy of the NTK, the NTK matrix must"
                 f" be a tensor of rank 2, but got a tensor of rank {len(ntk.shape)}."
             )
-        distribution = compute_grammian_diagonal_distribution(gram_matrix=ntk)
+        distribution = compute_gramian_diagonal_distribution(gram_matrix=ntk)
         return compute_shannon_entropy(distribution, effective=self.effective)
 
 
@@ -486,7 +486,7 @@ def apply(self, ntk: np.ndarray) -> np.ndarray:
                 "To compute the magnitude distribution of the NTK, the NTK matrix must"
                 f" be a tensor of rank 2, but got a tensor of rank {len(ntk.shape)}."
             )
-        return compute_grammian_diagonal_distribution(gram_matrix=ntk)
+        return compute_gramian_diagonal_distribution(gram_matrix=ntk)
 
 
 class NTKEigenvalues(BaseMeasurement):

papyrus/utils/__init__.py

@@ -27,7 +27,7 @@
     compute_von_neumann_entropy,
 )
 from papyrus.utils.matrix_utils import (
-    compute_grammian_diagonal_distribution,
+    compute_gramian_diagonal_distribution,
     compute_hermitian_eigensystem,
     compute_l_pq_norm,
     flatten_rank_4_tensor,
@@ -38,7 +38,7 @@
 __all__ = [
     compute_hermitian_eigensystem.__name__,
     normalize_gram_matrix.__name__,
-    compute_grammian_diagonal_distribution.__name__,
+    compute_gramian_diagonal_distribution.__name__,
     compute_l_pq_norm.__name__,
     flatten_rank_4_tensor.__name__,
     unflatten_rank_4_tensor.__name__,

papyrus/utils/matrix_utils.py

@@ -103,7 +103,7 @@ def normalize_gram_matrix(gram_matrix: np.ndarray):
     return gram_matrix * normalizing_matrix
 
 
-def compute_grammian_diagonal_distribution(gram_matrix: np.ndarray) -> np.ndarray:
+def compute_gramian_diagonal_distribution(gram_matrix: np.ndarray) -> np.ndarray:
     """
     Compute the normalized distribution of the diagonals of a gram matrix.