num::mixed_precision

Defined in header <linear/factorization/mixed_lu.hpp>

constexpr mixed_precision_structure mixed_precision{};

Select LU in float with iterative refinement to double accuracy.