distance_norm2() [3/12]

template<typename T >

T cntr::distance_norm2	(	int	tstp,
		herm_matrix< T > &	g1,
		herm_matrix_timestep< T > &	g2
	)

Evaluate the Euclidean norm between a herm_matrix and herm_matrix_timestep at a given time step.

Purpose

Evaluate the Euclidean norm between of a herm_matrix \(g_1\) and a herm_matrix_timestep ( \(g_2\)) at a given time step (tstp). To evaluate the norm, the elements of retarded, lesser and left-mixing components at the time step is used. The norm is not normalized per elements, but it is the summention of all the elements.

Parameters

tstp	time step
g1	herm_matrix
g2	herm_matrix_timestep

Definition at line 1685 of file cntr_utilities_impl.hpp.

References cntr::herm_matrix< T >::nt(), cntr::herm_matrix_timestep< T >::ntau(), cntr::herm_matrix< T >::ntau(), cntr::herm_matrix_timestep< T >::size1(), cntr::herm_matrix< T >::size1(), and cntr::herm_matrix_timestep< T >::tstp().

                                                                            {
    assert(g1.size1() == g2.size1());
    assert(g1.ntau() == g2.ntau());
    assert(g1.nt() >= tstp);
    assert(g2.tstp() == tstp);
    herm_matrix_timestep_view<T> g1_view(tstp, g1);
    herm_matrix_timestep_view<T> g2_view(tstp, g2);
    if (g1.size1() == 1)
        return distance_norm2_dispatch<T, 1>(tstp, g1_view, g2_view);
    else
        return distance_norm2_dispatch<T, LARGESIZE>(tstp, g1_view, g2_view);
}

Here is the call graph for this function: