[Wien] Bandstructure with SOC

[prasad jayasena]

Dear Prof. Blaha

Thank you very much for clearing me the correct procedure. This time, when I run "x spaghetti -up -so" (without -p as you asked) it finished without any problem.



In the previous process, I submitted following command in a single job script:
x lapw1 -band -up -p 
x lapw1 -band -dn -p 
x lapwso -up -p
x lapw2 -band -qtl -so -up -p 

But the x spaghetti -so -up -p was in a second job. And it crashed.
But without -p , it works.


Thank you all.

Prasad



     On Friday, November 8, 2019, 3:23:06 p.m. CST, Peter Blaha <pblaha at theochem.tuwien.ac.at> wrote:  
 
 I think  x spaghetti -p -so -up works fine.

My guess: You probably destroyed your .processes file before the 
spaghetti step ? Most likely because of doing this in an extra batch job ?

You must describe your procedure much more exactly, otherwise one can 
only "guess" what could cause the problem.
Please cut and paste your actual job/commands, not type them in (with 
missing some important details).

Apparently you did this in batch jobs (one or 2 or 3 different ones) ???

> After I tried doing bandstructure. Here is my procedure;
>  1. created klist_band
> 2.
> x lapw1 -band -up -p
> x lapw1 -band -dn -p > x lapwso -up -p

so far ok.
Is the lapw2 step done in the same batch job ??

> x lapw2 -band -qtl -up -p

This misses a  -so  switch !!!
and what about  x lapw2 -band -so -qtl -dn -p  ???
You will get only the qtlup file, thus cannot do the spin-dn projections 
in a fat band plot.

> All these runs finished correct.
> 
> 3. I added the E_fermi in case.insp
> 4. xspaghetti -so -up -p

I would never use a batch job for the spaghetti step. This is so short 
and should be done on the frontend!!!

> In the slurm job file I see the following message .
> 
> case.outputso created from 0 parallel files
>    ERROR IN OPENING UNIT:          7
>          FILENAME:
>  case.outputso
>      STATUS: old          FORM:formatted
> OPEN FAILED
> 0.001u 0.009s 0:00.10 0.0%    0+0k 8784+0io 28pf+0w

It says:  ... created from 0 parallel files.

The x script needs to know how many parallel jobs you had used for 
lapw1/so. It does it using a  .processes file

what gives:
ls -als .processes
cat .processes

Does it still exist and contains the 32 processes you talked in some 
other mail ??  I expect you destroyed it in the second batch job.

-------------------
Anyway, the fix is easy as indicated by Oleg in a previous mail. 
Concatenate all outputso_* files into one single outputso file.
    cat case.outputso_1 > case.outputso
    cat case.outputso_2 >> case.outputso
    cat case.outputso_3 >> case.outputso
    etc...
Then run  x spaghetti -so -up  (but WITHOUT  -p)
and probably  also  x spaghetti -so -dn (at least when you do fat bands).

-- 
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300            FAX: +43-1-58801-165982
Email: blaha at theochem.tuwien.ac.at    WIEN2k: http://www.wien2k.at
WWW: 
http://www.imc.tuwien.ac.at/tc_blaha------------------------------------------------------------------------- 

_______________________________________________
Wien mailing list
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@zeus.theochem.tuwien.ac.at/index.html
  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20191108/5b4ea4eb/attachment.html>