convolution() [2/2]

template<typename T >

void cntr::convolution	(	herm_matrix< T > &	C,
		herm_matrix< T > &	A,
		herm_matrix< T > &	Acc,
		function< T > &	ft,
		herm_matrix< T > &	B,
		herm_matrix< T > &	Bcc,
		T	beta,
		T	h,
		int	SolveOrder
	)

Returns the result of the contour convolution of two matrices and a function object

Purpose

Calls convolution routines to compute contour convolution C=A*fxB. The objects A,B and C are of the class 'herm_matrix'. If 'n=-1', one performs Matsubara convolution with a function 'F(-1)', otherwise with 'F(t)'. Works for a scalar and square matrices.

Parameters

C	[herm_matrix] Matrix to which the result of the convolution is given
A	[herm_matrix] contour Green's function
Acc	[herm_matrix] complex conjugate to A
ft	[function] contour function
B	[herm_matrix] contour Green's function
Bcc	[herm_matrix] complex conjugate to B
beta	inversed temperature
h	time interval
SolveOrder	[int] integrator order

Definition at line 3170 of file cntr_convolution_impl.hpp.

References cntr::herm_matrix< T >::nt(), and cntr::function< T >::ptr().

                                              {
    int tstp;
    convolution_matsubara(C, A, ft.ptr(-1), B, integration::I<T>(SolveOrder), beta);
    for (tstp = 0; tstp <= C.nt(); tstp++)
        convolution_timestep<T>(tstp, C, A, Acc, ft, B, Bcc, beta, h, SolveOrder);
}

Here is the call graph for this function: