The ABINIT Users Mailing List ( CLOSED )

["Erik Ylvisaker" <eylvisaker@physics.ucdavis.edu>]

I think this may have to do with the istwfk variable.  Abinit will take into 
account time-reversal symmetry for points like (1/2 0 0) to cut the number 
of planewave coefficients in half and use real to complex FFTs.

You can turn off doing this by figuring out how many k-points you have (ie, 
give a quick short run with nstep=1 or something) and then adding
istwfk = nkpt*1
where nkpt has to be the number of kpoints in the calculation.

----- Original Message ----- 
From: <joydeep@jncasr.ac.in>
To: <forum@abinit.org>
Sent: Friday, December 30, 2005 12:00 AM
Subject: [abinit-forum] reverse FFT by fourwf.F90


> Dear Abinitians,
>
> I am trying to use the fourwf.F90 subroutine
> to do an reverse FFT using option 3. I am having
> problem at the (0.0,0.0,0.0) and BZ edge
> points ,ie, (0.5,*,*) type points in the k-mesh.
> It seems to be doing fine at other k-points in the BZ.
> May be these k-points need some extra packing of the
> real space wave function which I am missing out.
> I see that that normalization of wave function within
> the G sphere st these k-points is 0.5 instead of 1.0.
>
> Thanks in advance for comments or pointing to relevant
> documentation which may already exist in the distribution.
>
> Joydeep