Dear Guangping Zhang,
I do not know what your "a subspace of a system" meaning, also I
don't know why you want to reconstructing wavefunctions from scratch.
As far as I understand, Siesta's orbital is from subroutine PHIATM
which is in atmfuncs.f. and the return value of dummy parameters of
PHIATM already include spherical harmonics(It already multiply rlylm
in PHIATM, see the code.) If you want to learn how to use it and also
generate a .cube file, you can study from Util/Denchar/Src/wavofr.f.
If you want to do symmetry analysis, i.e, Px, Py, Pz symmetry, maybe
you could study LEV00 & TETR(
http://www.cmmp.ucl.ac.uk/~lev/codes/lev00/ ). It could do some kind
of symmetry operations for geometry. If you just do not want generate
too much files when using DENCHAR, i.e., just output files
corresponding to your selected K point, not all K points. you could do
some 'dirty' modification of wavofr.f.
Best Wishes.
Kuilin, Lu
[email protected]
On 5/21/11, Guangping Zhang <[email protected]> wrote:
> Hi, all
>
> Can anyone tell me whether the basis orbitals used in SIESTA are like
> Px,Py,Pz and so on or the Y_lm like orbitals. I think they are not
> equivalent because Py is a combination of Y_1-1 and Y_11.
>
> I want to reconstruct the basis orbital used in SIESTA. So I wonder when we
> get the matrix element of an operator, whether we use the Py like basis
> orbital or Y_lm like basis orbial.
>
> I see the coefficients in WFSX are for Py like orbitals, and the Mulliken
> population is also for Py like orbitals.
>
> Any advice is welcome.
>
> Guangping
> 2011-05-21
>
>
