The ABINIT Users Mailing List ( CLOSED )

[Takeshi Nishimatsu <takeshi@physics.rutgers.edu>]

Dear Marc Torrent and all,

> So, the solution is to "convert" your dos_fractions_m in
> partial_dos_fractions_paw.F90 from REAL to COMPLEX spherical harmonics
> (apply some "rotation" matrix)...

It seems that what I only have to do is just an averaging of +m and -m
parts of dos_fractions_m in the end of partial_dos_fractions_paw.F90, 
because LDOS (and any other absolute-squared things) of +m and -m are
same even if there is no symmetry around the atom.

This is additional patch to my previous 14-27.diff:
==============================================================================
--- src/13paw/partial_dos_fractions_paw.F90     (revision 27)
+++ src/13paw/partial_dos_fractions_paw.F90     (working copy)
@@ -172,6 +172,18 @@
  deallocate(int1,int2)
  end do !iatom
 
+ ! Averaging: A quick hack for m-decomposed LDOS
+ do iat=1,dtset%natsph
+    do il = 0, mbesslang-1
+       do im = 1, il
+          dos_fractions_m(:,:,:,mbesslang**2*(iat-1)+il**2+il+1+im) = &
+         (dos_fractions_m(:,:,:,mbesslang**2*(iat-1)+il**2+il+1+im) + &
+          dos_fractions_m(:,:,:,mbesslang**2*(iat-1)+il**2+il+1-im))/2
+          dos_fractions_m(:,:,:,mbesslang**2*(iat-1)+il**2+il+1-im) = &
+          dos_fractions_m(:,:,:,mbesslang**2*(iat-1)+il**2+il+1+im)
+       end do
+    end do
+ end do !iatom
 
 end subroutine partial_dos_fractions_paw
 !!***
==============================================================================

So far, I have to give up the cubic (or hexagonal) harmonics decomposition 
(p_x,
p_y, p_z decomposition and that for d (l=2)), because m-decomposition may be
enough for my current subject. And I also gave up to use symmetry. Therefore,
spgroup=1 is necessary for this m-decomosition. (I guess kptopt=3 is *not*
necessary.) Sorry.

cf. kptopt=3: Do not take into account any symmetry : k points will be 
generated
in the full Brillouin zone. Not in the half of that.


Thank you very much!
-- 
 Takeshi Nishimatsu
 http://loto.sourceforge.net/feram/   Fast MD program for perovskite-type 
ferroelectrics