dfftz2 - initialize the trigonometric weight and factor tables or compute the two-dimensional forward Fast Fourier Transform of a two-dimensional double precision array. =head1 SYNOPSIS
SUBROUTINE DFFTZ2( IOPT, N1, N2, SCALE, X, LDX, Y, LDY, TRIGS, IFAC, * WORK, LWORK, IERR) DOUBLE COMPLEX Y(LDY,*) INTEGER IOPT, N1, N2, LDX, LDY, LWORK, IERR INTEGER IFAC(*) DOUBLE PRECISION SCALE DOUBLE PRECISION X(LDX,*), TRIGS(*), WORK(*)
SUBROUTINE DFFTZ2_64( IOPT, N1, N2, SCALE, X, LDX, Y, LDY, TRIGS, * IFAC, WORK, LWORK, IERR) DOUBLE COMPLEX Y(LDY,*) INTEGER*8 IOPT, N1, N2, LDX, LDY, LWORK, IERR INTEGER*8 IFAC(*) DOUBLE PRECISION SCALE DOUBLE PRECISION X(LDX,*), TRIGS(*), WORK(*)
SUBROUTINE FFT2( IOPT, [N1], [N2], [SCALE], X, [LDX], Y, [LDY], * TRIGS, IFAC, WORK, [LWORK], IERR) COMPLEX(8), DIMENSION(:,:) :: Y INTEGER :: IOPT, N1, N2, LDX, LDY, LWORK, IERR INTEGER, DIMENSION(:) :: IFAC REAL(8) :: SCALE REAL(8), DIMENSION(:) :: TRIGS, WORK REAL(8), DIMENSION(:,:) :: X
SUBROUTINE FFT2_64( IOPT, [N1], [N2], [SCALE], X, [LDX], Y, [LDY], * TRIGS, IFAC, WORK, [LWORK], IERR) COMPLEX(8), DIMENSION(:,:) :: Y INTEGER(8) :: IOPT, N1, N2, LDX, LDY, LWORK, IERR INTEGER(8), DIMENSION(:) :: IFAC REAL(8) :: SCALE REAL(8), DIMENSION(:) :: TRIGS, WORK REAL(8), DIMENSION(:,:) :: X
#include <sunperf.h>
void dfftz2(int iopt, int n1, int n2, double scale, double *x, int ldx, doublecomplex *y, int ldy, double *trigs, int *ifac, double *work, int lwork, int *ierr);
void dfftz2_64(long iopt, long n1, long n2, double scale, double *x, long ldx, doublecomplex *y, long ldy, double *trigs, long *ifac, double *work, long lwork, long *ierr);
dfftz2 initializes the trigonometric weight and factor tables or computes the two-dimensional forward Fast Fourier Transform of a two-dimensional double precision array. In computing the two-dimensional FFT, one-dimensional FFTs are computed along the columns of the input array. One-dimensional FFTs are then computed along the rows of the intermediate results. .Ve
N2-1 N1-1
Y(k1,k2)
= scale * SUM SUM W2*W1*X(j1,j2)
j2=0 j1=0 .Ve
where
k1 ranges from 0 to N1-1 and k2 ranges from 0 to N2-1
i = sqrt(-1)
isign = -1 for forward transform
W1 = exp(isign*i*j1*k1*2*pi/N1)
W2 = exp(isign*i*j2*k2*2*pi/N2)
In real-to-complex transform of length N1, the (N1/2+1) complex output data points stored are the positive-frequency half of the spectrum of the Discrete Fourier Transform. The other half can be obtained through complex conjugation and therefore is not stored.
IOPT = 0 computes the trigonometric weight table and factor table
IOPT = -1 computes forward FFT
0 = normal return
-1 = IOPT is not 0 or -1
-2 = N1 < 0
-3 = N2 < 0
-4 = (LDX < N1) or (LDX not equal 2*LDY when X and Y are same array)
-5 = (LDY < N1/2+1)
-6 = (LWORK not equal 0) and (LWORK < MAX(N1,2*N2))
-7 = memory allocation failed
fft
On exit, output array Y(1:LDY, 1:N2) is overwritten.