[Wien] 'LAPW2: semicore band-ranges too large error' for spin-orbit calculation

[Peter Blaha]

> I am not sure if the problem is with lapwso. All the output files from 
> lapw1 and lapwso (case.scf1, case.scfso) look ok to me. 

The scf files look "ok", but you need to check the output1* and 
outputso* files.

lapw2 detects that the first eigenvalue on all the k-points has a large 
spread, i.e. they vary from -9 to -7 Ry.
This is unphysical (such a deep semicore state MUST NOT vary in energy 
by more than a few mRy) for the different k-points and therefore lapw2 
stops with an error messge, but the error must happened already before.

Check if such a variation already occurs in output1 or only in outputso.

Of course, these problems could also come from non-optimal sphere radii 
(with approximate linear dependency), bad energy parameters, a bad 
potential, ....


But case.scf2
> files have only these lines:
> 
> ----------
> case.scf2up
> 
>         TEMP.-SMEARING WITH    0.00500 Ry
>            -S / Kb           =  -5.64060904
>            -(T*S)/2          =  -0.00705076
>            Chem Pot          =   0.25857200
>           Bandranges (emin - emax) and occupancy:
> :WARN :BAN00001:   1   -9.849452   -7.837911  1.00000000
> ------------
> 
> 
> It is bit confusing for me that I am encountering this problem only for 
> this system. As I mentioned in my previous message, I have worked with 
> different systems with spin-orbit coupling (some supercells 
> containingmore than 250 atoms) but I didn't have problem with this 
> version of intel.
> 
> 
> Thanks,
> Fhokrul
> 
> 
> ------------------------------------------------------------------------
> *From:* Wien <wien-bounces at zeus.theochem.tuwien.ac.at> on behalf of 
> Gavin Abo <gsabo at crimson.ua.edu>
> *Sent:* Thursday, April 19, 2018 2:55 AM
> *To:* wien at zeus.theochem.tuwien.ac.at
> *Subject:* Re: [Wien] 'LAPW2: semicore band-ranges too large error' for 
> spin-orbit calculation
> 
> 
>> Thanks, for the reply. I am using intel 2016. I did apply the patch 
>> get_nloat.patch in SRC_lapwso but I still have the same problem.
> 
> Which 2016 ifort?  I haven't used Update 3 (16.0.3.210) but that version 
> seemed particularly bad in mailing list posts probably because of the 
> unformatted file I/O bug [ 
> https://software.intel.com/en-us/articles/read-failure-unformatted-file-io-psxe-16-update-3 
> ].
> 
> Every so often it happens to me that I think I recompiled a change to 
> the code, but I make as mistake and the change doesn't get compiled into 
> the executable.  Though, maybe this doesn't happen to you.  When I get 
> paranoid about that, I remove the existing executable and .o files using 
> 'make clean'.
> 
> In this case for example:
> 
> username at computername:~/Desktop$ cd $WIENROOT
> username at computername:~/WIEN2k$ ls -l lapwso
> -rwxrwxr-x 1 username username 1533555 Apr 18 20:06 lapwso
> username at computername:~/WIEN2k$ rm lapwso
> username at computername:~/WIEN2k$ cd SRC_lapwso
> username at computername:~/WIEN2k/SRC_lapwso$ make clean
> rm -f *.o _tmp_.* *.P .real .complex .sequential .parallel *.mod
> username at computername:~/WIEN2k/SRC_lapwso$ ls -l get_nloat.f
> -rw-rw-r-- 1 username username 682 Apr  2  2014 get_nloat.f
> username at computername:~/WIEN2k/SRC_lapwso$ wget 
> https://raw.githubusercontent.com/gsabo/WIEN2k-Patches/master/17.1/get_nloat.patch
> ...
> username at computername:~/WIEN2k/SRC_lapwso$ patch -b get_nloat.f 
> get_nloat.patch
> patching file get_nloat.f
> username at computername:~/WIEN2k/SRC_lapwso$ ls -l get_nloat.f
> -rw-rw-r-- 1 username username 1782 Apr 18 20:12 get_nloat.f <- Notice 
> how the get_nloat.f file changes from 682 to1782 after the patch
> username at computername:~/WIEN2k/SRC_lapwso$ cd ..
> username at computername:~/WIEN2k$ ./siteconfig
> ...
>    Selection: R
> ...
>     ***********************************************************************
>     *                      Compile/Recompile programs                     *
>     ***********************************************************************
> 
>       A   Compile all programs
>       S   Select program
> 
>       Q   Quit
> 
>       Selection: S
>     Which program to recompile? lapwso
> ...
> Compile time errors (if any) were:
> 
> 
> Check file   compile.msg   in the corresponding SRC_* directory for the
> compilation log and more info on any compilation problem.
> 
>       Press RETURN to continue
> ...
> username at computername:~/WIEN2k$ ls -l lapwso
> -rwxrwxr-x 1 username username 1520752 Apr 18 20:16 lapwso <- Notice how 
> the lapwso file size decreases from 1533555 to 1520752.  However, file 
> sizes for the executable generated by the compiler specifically for your 
> system might be of different sizes.
> 
> 
> _______________________________________________
> 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
> 

-- 

                                       P.Blaha
--------------------------------------------------------------------------
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
--------------------------------------------------------------------------