[Wannier] wannier_plot = true crashes the calculation after wannierisation for gamma_only = true

[Robert Peyton Cline]

Dear Wannier90 developers,

I have a strange problem I cannot figure out concerning the execution of
wannier90 with gamma_only = true.  I am forming localized molecular
orbitals for perylene dimer pairs, with one orbital (either the monomeric
HOMO or LUMO) on each monomer.  I am using only the gamma point for these
calculations.  For the electronic structure, I use VASP.6.3.0, and this is
linked with wannier90 v3.1.0.

If I use vasp_std followed by a normal wannier calculation without
gamma_only = true, the calculation completes in full: disentanglement and
wannierisation both complete without issue, and I can plot the resultant
molecular orbitals in the form of *.xsf files.  However, since I only
sample with the gamma point, I figured I could compare this calculation
with vasp_gam and wannier90 computed with gamma_only = true.  If I do this,
the disentanglement and wannierisation phases still complete just fine, and
a similar final spread is reached for the wannierisation phase; however, I
cannot plot the orbital that comes out of the calculation within an *.xsf
file.  The code crashes and leaves an error message.

I have attached some salient example files in ZIP folders to this email.
Folder 'std' corresponds to the initial start using vasp_std and with
gamma_only = false in wannier90.  Folder 'gam' corresponds to the initial
start using vasp_gam with gamma_only = true in wannier90.  Within these
folders are e.* files; for the folder 'gam' you will find the error message
from above in this e.* file.  Note that I had to remove the *.mmn files in
order to meet the file size requirement.

I could not include the UNK* files for these calculations due to their
immense sizes, but I will say that the UNK* file for the gamma_only = true
calculation is half that of the gamma_only = false calculation, which makes
sense if only the former is purely real.  For the std calculation, since I
could plot the resultant *.xsf file, I have provided an image of the target
orbital in lieu of the actual *.xsf file.  The target orbital is correct.
You might notice that the spreads of these calculations are quite large,
but that's because I'm finding localized molecular orbitals, not
atomic-like orbitals.  So this is not an error, if that was something you
might think.

Any help would be appreciated.  It seems strange to me that I cannot plot
the orbitals if I use the gamma_only version of the code, so I'm wondering
if I'm missing an input flag, or if there is some compilation error on my
end.  Or maybe I'm mistaken that plotting is possible for gamma_only = true.

Many thanks,
Peyton Cline

--------------------------------------------------------------------
R. Peyton Cline
Pronouns: he/him
Postdoctoral Associate
Prof. Joel Eaves' Group
University of Colorado Boulder
Department of Chemistry
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/wannier/attachments/20220420/9fea1d66/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: gam.zip
Type: application/zip
Size: 28712 bytes
Desc: not available
URL: <http://lists.quantum-espresso.org/pipermail/wannier/attachments/20220420/9fea1d66/attachment-0002.zip>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: std.zip
Type: application/zip
Size: 226916 bytes
Desc: not available
URL: <http://lists.quantum-espresso.org/pipermail/wannier/attachments/20220420/9fea1d66/attachment-0003.zip>