[Wien] Problem running fold2bloch interface with wien2k_17

[Rubel, Oleg]

I made the benchmarking on the same system (one of supercomputers) with pre-installed Matlab (2021a.5) and Octave (v5.2.0). I did not do the install and cannot comment on libraries. Both run on a single core of a compute node (I can see it in top). The Matlab runs through the task in 5-10 sec, while Octave needs 20 mins (!) for the same task. On my home desktop I do not have Octave, but Matlab timing is comparable with a standard installation.

I can provide files if anyone is interested in running the benchmark.

Oleg

> -----Original Message-----
> From: Wien <wien-bounces at zeus.theochem.tuwien.ac.at> On Behalf Of
> Laurence Marks
> Sent: Sunday, April 23, 2023 10:27 PM
> To: A Mailing list for WIEN2k users <wien at zeus.theochem.tuwien.ac.at>
> Subject: Re: [Wien] Problem running fold2bloch interface with wien2k_17
> 
> A question out of curiosity. When you are comparing the Matlab and Octave
> versions, are you using mkl + OMP_NUM_THREADS for both?
> 
> 
> --
> Professor Laurence Marks (Laurie)
> Department of Materials Science and Engineering, Northwestern University
> www.numis.northwestern.edu <http://www.numis.northwestern.edu>
> "Research is to see what everybody else has seen, and to think what nobody
> else has thought" Albert Szent-Györgyi
> 
> On Sun, Apr 23, 2023, 16:08 Rubel, Oleg <rubelo at mcmaster.ca
> <mailto:rubelo at mcmaster.ca> > wrote:
> 
> 
> 	Dear Gavin,
> 
> 	Many thanks for letting me know about all of those inconsistencies. I
> fixed the links. Also the ubs_dots_w2k_octave.m script was revisited to fix
> issues that you mentioned. I had to make some kind of a  progress bar
> because it takes relatively long to run for a realistic scenario. For some reason
> the MATLAB version of the same (ubs_dots.m) runs an order of magnitude
> faster. All updates were pushed to GitHub.
> 
> 	Thank you once again
> 	Oleg
> 
> 	> -----Original Message-----
> 	> From: Wien <wien-bounces at zeus.theochem.tuwien.ac.at
> <mailto:wien-bounces at zeus.theochem.tuwien.ac.at> > On Behalf Of
> 	> Gavin Abo
> 	> Sent: Wednesday, April 19, 2023 1:04 AM
> 	> To: wien at zeus.theochem.tuwien.ac.at
> <mailto:wien at zeus.theochem.tuwien.ac.at>
> 	> Subject: Re: [Wien] Problem running fold2bloch interface with
> wien2k_17
> 	>
> 	> Dear Oleg,
> 	>
> 	> The tutorial 1 at:
> 	>
> 	> https://github.com/rubel75/fold2Bloch-Wien2k/wiki/Tutorial-1:-
> Lattice-of-
> 	> Hydrogen-atoms
> 	>
> 	> It has:
> 	>
> 	>
> 	> ~/fold2Bloch/fold2Bloch -c 6-atom2D.vector "''1 0 0:0 2 0:0 0 3''"
> 	>
> 	> Cannot remember for sure but that might not have worked starting
> with
> 	> (double quote)(single quote)(single quote).  I think it needed to be
> (double
> 	> quote)(single quote).  Similarly ending with (single quote)(double
> quote).
> 	>
> 	> The ubs_dot.m link seems to be broken pointing to:
> 	>
> 	>
> 	>
> https://github.com/rubel75/fold2Bloch/blob/master/Utils/ubs_dots.m
> 	>
> 	>
> 	> The working link now appears to be:
> 	>
> 	> https://github.com/rubel75/fold2Bloch-
> 	> Wien2k/blob/master/utils/ubs_dots.m
> 	>
> 	> It might just be me, but instead of the matlab script, I like using the
> octave
> 	> script at:
> 	>
> 	>
> 	> https://github.com/rubel75/fold2Bloch-
> 	> Wien2k/blob/master/utils/ubs_dots_w2k_octave.m
> 	>
> 	>
> 	> However, if I recall correctly, the script failed to run right.
> 	>
> 	> The ubs_dots.m has line 43:
> 	>
> 	> roundOffErrK = 0.000001; % this is the round off error 1/3 =
> 0.333333 + err
> 	>
> 	> and it has lines 70-72:
> 	>
> 	>
> 	> epsk = [roundOffErrK roundOffErrK roundOffErrK]; % k rounding
> error epsk =
> 	> coordTransform(epsk,G); % transform to Cart. coords epsk =
> 	> sqrt(dot(epsk,epsk)); % get magnitude of the vector
> 	>
> 	>
> 	> I believe I had to add line 43 after line 54 and lines 70-72 between
> lines 72 and
> 	> 73 in the current ubs_dots_w2k_octave.m for it to work.
> 	>
> 	>
> 	> Regarding Tutorial 2 at:
> 	>
> 	>
> 	> https://github.com/rubel75/fold2Bloch-Wien2k/wiki/Tutorial-2:-
> Bismuth-in-
> 	> GaAs
> 	>
> 	>
> 	> It has:
> 	>
> 	> ~/fold2Bloch/fold2Bloch -c Ga8As7Bi1.vector 2:2:2
> 	>
> 	> According to your post below, that needs to be:
> 	>
> 	>
> 	> ~/fold2Bloch/fold2Bloch -c 6-atom2D.vector "'2 0 0:0 2 0:0 0 2'"
> 	>
> 	> The link to ubs_dot.m for Ga8As7Bi1.unfolded for reproducing the
> tutorial 2
> 	> plot seems to be broken (for me it leads me to a github page with a
> 404  error).
> 	>
> 	>
> 	> Thanks,
> 	>
> 	> Gavin
> 	> WIEN2k user
> 
> 	_______________________________________________
> 	Wien mailing list
> 	Wien at zeus.theochem.tuwien.ac.at
> <mailto:Wien at zeus.theochem.tuwien.ac.at>
> 	http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> 	SEARCH the MAILING-LIST at:  http://www.mail-
> archive.com/wien at zeus.theochem.tuwien.ac.at/index.html
>