Incorrect spffte documentation

[DavidHuber-NOAA]

It looks like the spffte documentation for the G and W arrays is incorrect. According to the documentation

NCEPLIBS-ip/src/spffte.f

Lines 39 to 44 in 2344525

	C> @param[out] W complex amplitudes if IDIR>0
	C> @param[out] G real values if IDIR<0
	C> @param IDIR direction flag
	C> - IDIR=0 to initialize trigonometric data
	C> - IDIR>0 to transform from Fourier to physical space
	C> - IDIR<0 to transform from physical to Fourier space

W should return complex amplitudes in Fourier space if IDIR > 0 or G should return real values if IDIR < 0. However, if IDIR > 0, then the transform should be going from Fourier to real space, and if IDIR < 0, then the transform should be going from real space to Fourier. This implies that either the algorithm is misdocumented (i.e. IDIR > 0 means we are going to Fourier space) or the variables are misdocumented.

[AlexanderRichert-NOAA]

Yeah, I think what's going on is that the "if IDIR>0" "if IDIR<0" descriptions are referring to when those variables serve as input. So I think the fix will be to remove the '[out]' from the W and G descriptions (those should be automatically inferred by Doxygen), then clarify when W and G are input vs. output. I'll update this when I get a chance. Thanks for catching this!