From 1f66907d2b5bc430629aa66d97ab26a05bee777f Mon Sep 17 00:00:00 2001 From: "Kevin J. Sung" Date: Wed, 27 Mar 2024 08:44:42 -0400 Subject: [PATCH] fix diag coulomb indices doc --- python/ffsim/linalg/double_factorized_decomposition.py | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/python/ffsim/linalg/double_factorized_decomposition.py b/python/ffsim/linalg/double_factorized_decomposition.py index dca2ad4b3..be30033a0 100644 --- a/python/ffsim/linalg/double_factorized_decomposition.py +++ b/python/ffsim/linalg/double_factorized_decomposition.py @@ -132,11 +132,11 @@ def double_factorized( It uses `scipy.optimize.minimize`, passing both the objective function and its gradient. The diagonal Coulomb matrices returned by the optimization can be optionally constrained to have only certain elements allowed to be nonzero. - This is achieved by passing the `diag_coulomb_mask` parameter, which is an - :math:`N \times N` matrix of boolean values where :math:`N` is the number of - orbitals. The nonzero elements of this matrix indicate where the diagonal Coulomb - matrices are allowed to be nonzero. Only the upper triangular part of the matrix is - used because the diagonal Coulomb matrices are symmetric. + This is achieved by passing the `diag_coulomb_indices` parameter, which is a + list of matrix entry indices (integer pairs) specifying where the diagonal Coulomb + matrices are allowed to be nonzero. Since the diagonal Coulomb matrices are + symmetric, only upper triangular indices should be given, i.e., + pairs :math:`(i, j)` where :math:`i \leq j`. References: - `arXiv:1808.02625`_