Module utils 

algorithmic primitives for sparse matrices

Functions

adjoint

computes the adjoint of the matrix $A$ and returns a view over it.

permute_dedup_self_adjoint

computes the self-adjoint permutation $P A P^\top$ of the matrix $A$ and deduplicate the elements of the output matrix

permute_dedup_self_adjoint_scratch

computes the workspace size and alignment required to apply a two sided permutation to a self-adjoint sparse matrix and deduplicate its elements

permute_self_adjoint

computes the self-adjoint permutation $P A P^\top$ of the matrix $A$

permute_self_adjoint_scratch

computes the workspace size and alignment required to apply a two sided permutation to a self-adjoint sparse matrix

permute_self_adjoint_to_unsorted

computes the self-adjoint permutation $P A P^\top$ of the matrix $A$ without sorting the row indices, and returns a view over it

sort_dedup_indices

sorts and deduplicates row_indices and values simultaneously so that row_indices is nonincreasing and contains no duplicate indices.

sort_indices

sorts row_indices and values simultaneously so that row_indices is nonincreasing.

transpose

computes the transpose of the matrix $A$ and returns a view over it.

transpose_dedup

computes the transpose of the matrix $A$ and returns a view over it.

transpose_dedup_scratch

computes the workspace size and alignment required to transpose a matrix and deduplicate the output elements

transpose_scratch

computes the workspace size and alignment required to transpose a matrix