The ABINIT Users Mailing List ( CLOSED )

[Adam Sorini <asorini@gmail.com>]

On Tue, Aug 4, 2009 at 2:48 PM, Markus Wohlgenannt <markus-wohlgenannt@uiowa.edu> wrote:

Dear Matthieu,

 

Thank you for your answers, please see my responses below

On Tue, Aug 4, 2009 at 1:50 PM, matthieu verstraete <matthieu.jean.verstraete@gmail.com> wrote:

Hello Markus,

1) have you checked that you are only looking at the real part of the
wf? As indicated, by default you get both real and imaginary parts,
and abinit has no particular preference for real valued wf for finite
systems. You should probably plot the norm squared, using both values,
or plot both values side by side.

 

Yes, I have checked that, the imaginary part numbers are all zero (or 1E-16).

2) are you sure you have the correct band? This looks like a high
lying box state with all the weight at the edges

 

Yes, I am using the tutorial example 1.1 unchanged, in this example we calculate 2 bands, and I am plotting band 1 of two.

3) the plot looks like it is not centered properly: are you certain
about the position of the molecule (around 0,0,0 I presume) and the
coordinates of the box you plot? If the molecule is actually in a
corner of the box this makes more sense, and suggests the box is not
big enough as the density in the middle is not very small...

 

Yes, I centered it around 0 0 0. Also I am pretty sure it is not merely a problem of centering, because I also calculated other molecules which have non-spherical, very characteristic shapes, and the wavefunction I plot never has any resemblence to the density plot (the latter is always as expected). Regarding the box-size, I am using tutorial example 1.1 unchanged, so the box may be a little small according to the convergence studies done in later tutorials, but it shouldn't be too much off.

4) try the xcrysden format. That gives you the norm squared
automatically, and you can visualize it more easily, without messing
around with scripts etc... (option 9 in the main cut3d menu for wf).
In this case, if your molecule is not centered, you can shift the data
by a certain number of grid points to center it.

 

I have two options (option 9 in my cut3d is 3D Data Explorer, only imaginary part). I have option 11 "Xcrysden formatted data and position files" and option 13 "XCrysden/Venus wavefunction real data". When I choose option 11 I obtain a file which, when I open it, seems to contain the norm^2 (since all numbers are positive), but there are no indexed coordinates. When I open it in XCrysDen, I only obtain 4 dots (which presumably signify some lattice, but no plot of the data itself.

You can see isosurfaces of the data if you click on "Tools" in the upper right. Then click on "data grid". Then click "ok". Then choose an iso-value in the window that pops up. Then click "submit".

 

When I chose option 13, cut3d terminates with an error "At line 1454 of file wffile.F90 (unit=12,file='myfilename') Fortran runtime error: Cannot change STATUS parameter in OPEN statement".

 

I would appreciate further help from you,

 

Many thanks,

 

Markus

Cheers

Matthieu

On Tue, Aug 4, 2009 at 7:56 PM, Markus
Wohlgenannt<markus-wohlgenannt@uiowa.edu> wrote:
> Dear forum,
>
> I am a new user to abinit and have a question. I hope someone can help me.
>
> I want to plot the real space wavefunction. I was trying to use cut3d. Cut3d
> works well for plotting the density. I attached the plot of the density in
> the molecular plane for tutorial one, H2. But when I try to plot the
> real-space wavefunction, I seem to get nonsensical results, the
> corresponding wavefunction plot is shown as the second graph in my
> attachement. I did the following to obtain this plot. I start cut3d, and
> tell it to open the WFK file. Then I choose option "0" to analyze the
> wavefunction, choose band 1, I do not want atomic analysis (option 0), then
> I choose option 4 "3D indexed real and imaginary data".
>
> Then I manually select the molecular plane by deleting all data outside the
> plane I want to view, i.e. all data for which z neq Z, where I choose Z
> half-way between the minum and maximum z to obtain the molecular plane.
>
> What am I doing wrong? Or else, is there an easier way to plot the
> real-space wavefunction?
>
> Someone please help,
>
> Thanks,
>
> Markus
>
> --
> Markus Wohlgenannt
> Associate Professor
> Department of Physics and Astronomy
> University of Iowa
> markus-wohlgenannt@uiowa.edu
> 319-353-1974
>
>

--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Dr. Matthieu Verstraete

European Theoretical Spectroscopy Facility (ETSF)
Dpto. Fisica de Materiales,
U. del Pais Vasco,
Centro Joxe Mari Korta, Av. de Tolosa, 72,   Phone: +34-943018393
E-20018 Donostia-San Sebastian, Spain        Fax  : +34-943018390

Mail : matthieu.jean.verstraete@gmail.com
http://www-users.york.ac.uk/~mjv500

--

Markus Wohlgenannt
Associate Professor
Department of Physics and Astronomy
University of Iowa
markus-wohlgenannt@uiowa.edu
319-353-1974