view +scheme/errorSbp.m @ 1120:7963a9cab637 feature/poroelastic

In RungeKuttaDiscreteData, make K and U non-sparse to improve performance by roughly a factor of 3.
author Martin Almquist <malmquist@stanford.edu>
date Tue, 07 May 2019 11:41:15 -0700
parents ed6a704b028d
children
line wrap: on
line source

function e = errorSbp(discr, v1, v2)
    % If v1 and v2 are more complex types, something like grid functions... Then we may use .getVectorFrom here!
    H = discr.H;
    err = v2 - v1;
    e = sqrt(err'*H*err);
end