[Wannier] Problems restarting from *.chk file

[Robert Peyton Cline]

 wannier90.amn
<https://drive.google.com/a/colorado.edu/file/d/1eACtnrXIXaGgZ2WzujfEUQOEm4olEBih/view?usp=drive_web>
 wannier90.chk
<https://drive.google.com/a/colorado.edu/file/d/10bJhKlHzcIUygUH8nJzqGeIoGPvhyquK/view?usp=drive_web>
 wannier90.eig
<https://drive.google.com/a/colorado.edu/file/d/1K9wsaFlDU4gFwBZpSQCj2MjRkhZJ-GfZ/view?usp=drive_web>
 wannier90.mmn
<https://drive.google.com/a/colorado.edu/file/d/1NI58ekNlhGB3dY8hSxjtF5KSPG7YDvV7/view?usp=drive_web>
 wannier90.win
<https://drive.google.com/a/colorado.edu/file/d/19U1k5ZFoaJlx7eiNd22XL_5CL6C_Mctk/view?usp=drive_web>
Dear Wannier developers and users,

I cannot restart one of my wannier90 calculations from a checkpoint file,
and I honestly have no idea why.   Actually to be more specific - the
wannier90.wout file does say that the *.chk file is read, but the following
wannierisation phase starts from scratch instead of starting from the
contents of the checkpoint.  This is true for both 'restart = default' and
for 'restart = wannierise'.  The latter option is what I usually use to
restart calculations, and I've never had any issues with this in the past.

The only main difference between my current calculation and previous
calculations is I am currently performing a full wannierisation (num_wann =
num_bands) where no disentanglement step is required, and I use the
'exclude_bands' tag to get rid of high-energy conduction bands from the DFT
calculation that I do not need.  Typically I don't exclude any bands, and I
use disentanglement to pick out the appropriate bands to make a subset of
orbitals.  Perhaps this difference is part of the problem?  I'm not sure
why that would matter, though.

I've attached the relevant files to this message in case those would be
useful.  The calculation is rather large and memory-intensive, but it ran
perfectly fine on my university supercomputer before writing the final
*.chk file.  It just won't restart from that file.

Many thanks,
Peyton Cline
4th year PhD Student
CU-Boulder
Eaves Group
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/wannier/attachments/20190606/fbe7c6be/attachment.html>