Re: [Wien] Initialization with spin-orbit

2016-02-17 Thread Peter Blaha

Sorry, it took some time to find time for debugging.


symmetso.f had a couple of accuracy issues whach made it working for the 
unrelaxed structure (because there some positions are exactly 0.0 or 
0.5), while in the relaxed structure they are very slightly modified.


Anyway, the fix is simple and consisted of changing

0.1d0 to 0.001d0   in a couple of places.

I attach the correct symmetso.f, which should be placed into 
SRC_symmetso, compiled (make) and   cp symmetso ..


Regards

On 01/28/2016 05:14 PM, Islam, Md F wrote:

Hi Prof. Blaha,

 Thanks for your reply. I have sent the structure to your email as you have 
mentioned.

Regards,
Fhokrul


From: wien-boun...@zeus.theochem.tuwien.ac.at 
[wien-boun...@zeus.theochem.tuwien.ac.at] On Behalf Of Peter Blaha 
[pbl...@theochem.tuwien.ac.at]
Sent: Tuesday, January 26, 2016 5:11 PM
To: A Mailing list for WIEN2k users
Subject: Re: [Wien] Initialization with spin-orbit

Not exactly sure what can cause the problem.

Please search the mailing list, there was some time ago a patch for
symmetso.

  From your description of the problem (optimized/non-optimized
structure) it could be a symmetry problem, i.e. during optimization
(hexagonal case ???) the symmetry of the positions in the struct file
was slightly broken (rounding errors, something like 0.34 or
similar ...)

Otherwise you have to send me your struct file.



1. Is it ok to initialize the job using initso with unrelaxed structure but 
after initialization, replace it by relaxed
  structure file and run the job?


No, at least not in general as symmetso may/may not reduce symmetry
depending on the direction of magnetism.



2. For spin polarized calculations, spin-orbit may reduce the symmetry 
depending on the choice of magnetization
  direction. According to user guide, initso picks up correct symmetry for 
such cases. But if the system does not
  have any symmetry to begin with, then is it ok to initialize the 
structure treating as non-magnetic case?


Yes, if you are already in P1 you are save and a further reduction of
symmetry is obviously not possible. Symmetso will not do anything.

However,x kgen -so  (or for P1: x kgen -fbz) is still necessary, as
for a non-so case, kgen will by default add inversion symmetry to P1,
which it should not do for a spin-polarized case.

--
--
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300 FAX: +43-1-58801-165982
Email: bl...@theochem.tuwien.ac.atWIEN2k: http://www.wien2k.at
WWW:   http://www.imc.tuwien.ac.at/staff/tc_group_e.php
--
___
Wien mailing list
Wien@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
___
Wien mailing list
Wien@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: bl...@theochem.tuwien.ac.atWIEN2k: http://www.wien2k.at
WWW:   http://www.imc.tuwien.ac.at/staff/tc_group_e.php
--
   program symmetso
!  Modified program SYMMETRY of WIEN97 - analysis of symmetry in the
!  presence of spin-orbit coupling and magnetization. Program creates
!  modified structure file case.struct_so.
!  Analysis is made in two steps:
!  1/ File case.struct is read, check is made which operations are
! retained with M in (tm1,tm2,tm3) direction. Distribution of atoms
! into groups of magnetically equivalent atoms is performed and
! intermediate case.struct file created (SUBROUTINE SYMSO).
!  2/ For each nonequivalent group of atoms local rotation matrix
! local symmetry and nonzero (l,m) components are found - output
! is then analogous as when using SYMMETRY package.
!
! --
!   gives the symmetry operations of all atoms of the unit cell  
!  
  IMPLICIT REAL*8 (A-H,O-Z)
  INCLUDE 'param.inc'
  CHARACTER*4   LATTIC,type 
  CHARACTER*11  STATUS,FORM 
  CHARACTER*79  TITLE
  character*69  help2 
  CHARACTER*80  DEFFN, ERRFN
  CHARACTER*180  fname

Re: [Wien] Initialization with spin-orbit

2016-01-29 Thread Islam, Md F
Thanks, Prof. Blaha. 

Fhokrul  



From: wien-boun...@zeus.theochem.tuwien.ac.at 
[wien-boun...@zeus.theochem.tuwien.ac.at] On Behalf Of Peter Blaha 
[pbl...@theochem.tuwien.ac.at]
Sent: Thursday, January 28, 2016 1:16 PM
To: A Mailing list for WIEN2k users
Subject: Re: [Wien] Initialization with spin-orbit

I can verify the problem.

However, I could not find a fix quickly. Since I'm very busy in the next
2 weeks, it may take some time to debug this completely.

As mentioned before, when SO does not change symmetry (you can see this
in outsymso file if all operations are either type A or B), you can
neglect symmetso.

Am 28.01.2016 um 17:14 schrieb Islam, Md F:
> Hi Prof. Blaha,
>
>  Thanks for your reply. I have sent the structure to your email as you 
> have mentioned.
>
> Regards,
> Fhokrul
>
> 
> From: wien-boun...@zeus.theochem.tuwien.ac.at 
> [wien-boun...@zeus.theochem.tuwien.ac.at] On Behalf Of Peter Blaha 
> [pbl...@theochem.tuwien.ac.at]
> Sent: Tuesday, January 26, 2016 5:11 PM
> To: A Mailing list for WIEN2k users
> Subject: Re: [Wien] Initialization with spin-orbit
>
> Not exactly sure what can cause the problem.
>
> Please search the mailing list, there was some time ago a patch for
> symmetso.
>
>   From your description of the problem (optimized/non-optimized
> structure) it could be a symmetry problem, i.e. during optimization
> (hexagonal case ???) the symmetry of the positions in the struct file
> was slightly broken (rounding errors, something like 0.34 or
> similar ...)
>
> Otherwise you have to send me your struct file.
>
>
>> 1. Is it ok to initialize the job using initso with unrelaxed structure but 
>> after initialization, replace it by relaxed
>>   structure file and run the job?
>
> No, at least not in general as symmetso may/may not reduce symmetry
> depending on the direction of magnetism.
>
>
>> 2. For spin polarized calculations, spin-orbit may reduce the symmetry 
>> depending on the choice of magnetization
>>   direction. According to user guide, initso picks up correct symmetry 
>> for such cases. But if the system does not
>>   have any symmetry to begin with, then is it ok to initialize the 
>> structure treating as non-magnetic case?
>
> Yes, if you are already in P1 you are save and a further reduction of
> symmetry is obviously not possible. Symmetso will not do anything.
>
> However,x kgen -so  (or for P1: x kgen -fbz) is still necessary, as
> for a non-so case, kgen will by default add inversion symmetry to P1,
> which it should not do for a spin-polarized case.
>
> --
> --
> Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
> Phone: +43-1-58801-165300 FAX: +43-1-58801-165982
> Email: bl...@theochem.tuwien.ac.atWIEN2k: http://www.wien2k.at
> WWW:   http://www.imc.tuwien.ac.at/staff/tc_group_e.php
> --
> ___
> Wien mailing list
> Wien@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
> ___
> Wien mailing list
> Wien@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
>

--
--
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300 FAX: +43-1-58801-165982
Email: bl...@theochem.tuwien.ac.atWIEN2k: http://www.wien2k.at
WWW:   http://www.imc.tuwien.ac.at/staff/tc_group_e.php
--
___
Wien mailing list
Wien@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
___
Wien mailing list
Wien@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


Re: [Wien] Initialization with spin-orbit

2016-01-28 Thread Peter Blaha

I can verify the problem.

However, I could not find a fix quickly. Since I'm very busy in the next 
2 weeks, it may take some time to debug this completely.


As mentioned before, when SO does not change symmetry (you can see this 
in outsymso file if all operations are either type A or B), you can 
neglect symmetso.


Am 28.01.2016 um 17:14 schrieb Islam, Md F:

Hi Prof. Blaha,

 Thanks for your reply. I have sent the structure to your email as you have 
mentioned.

Regards,
Fhokrul


From: wien-boun...@zeus.theochem.tuwien.ac.at 
[wien-boun...@zeus.theochem.tuwien.ac.at] On Behalf Of Peter Blaha 
[pbl...@theochem.tuwien.ac.at]
Sent: Tuesday, January 26, 2016 5:11 PM
To: A Mailing list for WIEN2k users
Subject: Re: [Wien] Initialization with spin-orbit

Not exactly sure what can cause the problem.

Please search the mailing list, there was some time ago a patch for
symmetso.

  From your description of the problem (optimized/non-optimized
structure) it could be a symmetry problem, i.e. during optimization
(hexagonal case ???) the symmetry of the positions in the struct file
was slightly broken (rounding errors, something like 0.34 or
similar ...)

Otherwise you have to send me your struct file.



1. Is it ok to initialize the job using initso with unrelaxed structure but 
after initialization, replace it by relaxed
  structure file and run the job?


No, at least not in general as symmetso may/may not reduce symmetry
depending on the direction of magnetism.



2. For spin polarized calculations, spin-orbit may reduce the symmetry 
depending on the choice of magnetization
  direction. According to user guide, initso picks up correct symmetry for 
such cases. But if the system does not
  have any symmetry to begin with, then is it ok to initialize the 
structure treating as non-magnetic case?


Yes, if you are already in P1 you are save and a further reduction of
symmetry is obviously not possible. Symmetso will not do anything.

However,x kgen -so  (or for P1: x kgen -fbz) is still necessary, as
for a non-so case, kgen will by default add inversion symmetry to P1,
which it should not do for a spin-polarized case.

--
--
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300 FAX: +43-1-58801-165982
Email: bl...@theochem.tuwien.ac.atWIEN2k: http://www.wien2k.at
WWW:   http://www.imc.tuwien.ac.at/staff/tc_group_e.php
--
___
Wien mailing list
Wien@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
___
Wien mailing list
Wien@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



--
--
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300 FAX: +43-1-58801-165982
Email: bl...@theochem.tuwien.ac.atWIEN2k: http://www.wien2k.at
WWW:   http://www.imc.tuwien.ac.at/staff/tc_group_e.php
--
___
Wien mailing list
Wien@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


Re: [Wien] Initialization with spin-orbit

2016-01-28 Thread Islam, Md F
Hi Prof. Blaha,

Thanks for your reply. I have sent the structure to your email as you have 
mentioned.

Regards,
Fhokrul


From: wien-boun...@zeus.theochem.tuwien.ac.at 
[wien-boun...@zeus.theochem.tuwien.ac.at] On Behalf Of Peter Blaha 
[pbl...@theochem.tuwien.ac.at]
Sent: Tuesday, January 26, 2016 5:11 PM
To: A Mailing list for WIEN2k users
Subject: Re: [Wien] Initialization with spin-orbit

Not exactly sure what can cause the problem.

Please search the mailing list, there was some time ago a patch for
symmetso.

 From your description of the problem (optimized/non-optimized
structure) it could be a symmetry problem, i.e. during optimization
(hexagonal case ???) the symmetry of the positions in the struct file
was slightly broken (rounding errors, something like 0.34 or
similar ...)

Otherwise you have to send me your struct file.


> 1. Is it ok to initialize the job using initso with unrelaxed structure but 
> after initialization, replace it by relaxed
>  structure file and run the job?

No, at least not in general as symmetso may/may not reduce symmetry
depending on the direction of magnetism.


> 2. For spin polarized calculations, spin-orbit may reduce the symmetry 
> depending on the choice of magnetization
>  direction. According to user guide, initso picks up correct symmetry for 
> such cases. But if the system does not
>  have any symmetry to begin with, then is it ok to initialize the 
> structure treating as non-magnetic case?

Yes, if you are already in P1 you are save and a further reduction of
symmetry is obviously not possible. Symmetso will not do anything.

However,x kgen -so  (or for P1: x kgen -fbz) is still necessary, as
for a non-so case, kgen will by default add inversion symmetry to P1,
which it should not do for a spin-polarized case.

--
--
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300 FAX: +43-1-58801-165982
Email: bl...@theochem.tuwien.ac.atWIEN2k: http://www.wien2k.at
WWW:   http://www.imc.tuwien.ac.at/staff/tc_group_e.php
--
___
Wien mailing list
Wien@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
___
Wien mailing list
Wien@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


Re: [Wien] Initialization with spin-orbit

2016-01-26 Thread Peter Blaha

Not exactly sure what can cause the problem.

Please search the mailing list, there was some time ago a patch for 
symmetso.


From your description of the problem (optimized/non-optimized 
structure) it could be a symmetry problem, i.e. during optimization 
(hexagonal case ???) the symmetry of the positions in the struct file 
was slightly broken (rounding errors, something like 0.34 or 
similar ...)


Otherwise you have to send me your struct file.



1. Is it ok to initialize the job using initso with unrelaxed structure but 
after initialization, replace it by relaxed
 structure file and run the job?


No, at least not in general as symmetso may/may not reduce symmetry 
depending on the direction of magnetism.




2. For spin polarized calculations, spin-orbit may reduce the symmetry 
depending on the choice of magnetization
 direction. According to user guide, initso picks up correct symmetry for 
such cases. But if the system does not
 have any symmetry to begin with, then is it ok to initialize the structure 
treating as non-magnetic case?


Yes, if you are already in P1 you are save and a further reduction of 
symmetry is obviously not possible. Symmetso will not do anything.


However,x kgen -so  (or for P1: x kgen -fbz) is still necessary, as 
for a non-so case, kgen will by default add inversion symmetry to P1, 
which it should not do for a spin-polarized case.


--
--
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300 FAX: +43-1-58801-165982
Email: bl...@theochem.tuwien.ac.atWIEN2k: http://www.wien2k.at
WWW:   http://www.imc.tuwien.ac.at/staff/tc_group_e.php
--
___
Wien mailing list
Wien@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