DCL:MATH2:FFTLIB: Fast Fourier Transformation
This is a package of subroutines that perform fast Fourier transformation on data of arbitrary length, transported from the NCAR math library.
In the section "Definitions" in the Explanation of Subroutines,
the following notations will be used in the mathematical descriptions of IO
parameters. For the i th element (i =1,...,N) of array X
(length N) to be processed, the input value will be written as xi (lowercase),
and the output as Xi (capitalized).
When using a conversion routine that belongs to one of the 7 subroutine groups,
the subroutine for initialization (with name ending with I) for that
group must be called once at the beginning. After that, the initialization
routine only needs to be called when specifying data with different number of
elements. The
working area used by the initialization routine must not be changed while using
a conversion routine of the same subroutine group. (Tables of factors and trigonometric
functions are stored in the working area.)