1FFTs: module{ 2 PATH: con "/dis/math/ffts.dis"; 3 4 ffts: fn(a,b:array of real, ntot,n,nspan,isn:int); 5}; 6 7# multivariate complex fourier transform, computed in place 8# using mixed-radix fast fourier transform algorithm. 9# arrays a and b originally hold the real and imaginary 10# components of the data, and return the real and 11# imaginary components of the resulting fourier coefficients. 12# multivariate data is indexed according to the fortran 13# array element successor function, without limit 14# on the number of implied multiple subscripts. 15# the subroutine is called once for each variate. 16# the calls for a multivariate transform may be in any order. 17# ntot is the total number of complex data values. 18# n is the dimension of the current variable. 19# nspan/n is the spacing of consecutive data values 20# while indexing the current variable. 21# the sign of isn determines the sign of the complex 22# exponential, and the magnitude of isn is normally one. 23# univariate transform: 24# ffts(a,b,n,n,n,1) 25# trivariate transform with a(n1,n2,n3), b(n1,n2,n3): 26# ffts(a,b,n1*n2*n3,n1,n1,1) 27# ffts(a,b,n1*n2*n3,n2,n1*n2,1) 28# ffts(a,b,n1*n2*n3,n3,n1*n2*n3,1) 29# the data can alternatively be stored in a single vector c 30# alternating real and imaginary parts. the magnitude of isn changed 31# to two to give correct indexing increment, and a[0:] and a[1:] used 32# for a and b 33