num::sparse_convertible

Defined in header <operator/concepts.hpp>

template <class Op>
concept sparse_convertible = linear_operator<Op> && requires(const Op &A) {
  { A.to_sparse() };
};

Linear operator that can materialize itself as explicit sparse CSR storage.

A statement about representation rather than mathematics: it says an operator can hand over its entries, which a factorization needs and a matrix-free operator cannot do.

Template parameters

• Op: Operator type.