date:20160131

[Wien] A basic question about full hybrid functional calculation

2016-01-31 Thread Hu, Wenhao

Hi, All:


According to the link:


http://www.mail-archive.com/wien%40zeus.theochem.tuwien.ac.at/msg13559.html



It seems that it's impossible to realize atomic position relaxation with full 
relaxation functional. Then, does it mean that the forces obtained with MSR1a 
are all wrong? I'm also wondering whether the calculation without atomic 
relaxation is accurate enough to describe the system.


Best,

Wenhao

[http://www.mail-archive.com/logo.png]

Re: [Wien] relaxation of atomic positions with the full 
...
www.mail-archive.com
Re: [Wien] relaxation of atomic positions with the full hybrid functionals. 
tran Fri, 25 Dec 2015 11:11:41 -0800

___
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] A basic question about full hybrid functional calculation

2016-01-31 Thread tran


Forces for full hybrid functionals are wrong. The same when SOC is
included whatever is the functional used (LDA, GGA, LDA+U, etc.).
However, for on-site eece hybrids (without SOC), forces are ok.

Onn, 31 Jan 2016, Hu, Wenhao wrote:



Hi, All:


According to the link:


http://www.mail-archive.com/wien%40zeus.theochem.tuwien.ac.at/msg13559.html



It seems that it's impossible to realize atomic position relaxation with full 
relaxation functional. Then, does it mean that the forces obtained with MSR1a 
are
all wrong? I'm also wondering whether the calculation without atomic relaxation 
is accurate enough to describe the system.


Best,

Wenhao

[logo.png]
Re: [Wien] relaxation of atomic positions with the full ...
www.mail-archive.com
Re: [Wien] relaxation of atomic positions with the full hybrid functionals. 
tran Fri, 25 Dec 2015 11:11:41 -0800


___
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] A basic question about full hybrid functional calculation

2016-01-31 Thread Laurence Marks

As a clarification, this means that neither MSR1a nor PORT will work
correctly with full hybrids or SOC.

What could be done is to incorporate a code such as PRAXIS (e.g.
https://people.sc.fsu.edu/~jburkardt/f_src/praxis/praxis.html) into PORT do
do a derivative-free optimization. If anyone is interested in doing some
coding let me know.

On Sun, Jan 31, 2016 at 2:02 PM, t...@theochem.tuwien.ac.at <
t...@theochem.tuwien.ac.at> wrote:

> Forces for full hybrid functionals are wrong. The same when SOC is
> included whatever is the functional used (LDA, GGA, LDA+U, etc.).
> However, for on-site eece hybrids (without SOC), forces are ok.
>
> Onn, 31 Jan 2016, Hu, Wenhao wrote:
>
> >
> > Hi, All:
> >
> >
> > According to the link:
> >
> >
> >
> http://www.mail-archive.com/wien%40zeus.theochem.tuwien.ac.at/msg13559.html
> >
> >
> >
> > It seems that it's impossible to realize atomic position relaxation with
> full relaxation functional. Then, does it mean that the forces
> obtained with MSR1a are
> > all wrong? I'm also wondering whether the calculation without atomic
> relaxation is accurate enough to describe the system.
> >
> >
> > Best,
> >
> > Wenhao
> >
> > [logo.png]
> > Re: [Wien] relaxation of atomic positions with the full ...
> > www.mail-archive.com
> > Re: [Wien] relaxation of atomic positions with the full hybrid
> functionals. tran Fri, 25 Dec 2015 11:11:41 -0800
> >
> >
> >
>



-- 
Professor Laurence Marks
Department of Materials Science and Engineering
Northwestern University
www.numis.northwestern.edu
Corrosion in 4D: MURI4D.numis.northwestern.edu
Co-Editor, Acta Cryst A
"Research is to see what everybody else has seen, and to think what nobody
else has thought"
Albert Szent-Gyorgi
___
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] A basic question about full hybrid functional calculation

2016-01-31 Thread delamora

Wenhao,

I do not know about the MSR1a, but sometimes the unrelaxed structure 
can give you very wrong results, a curious example is Bi, which has a structure 
very close to cubic P, but if you put it as cubic P the results are completely 
different.

Now, relaxing a structure is quite demanding, but my feeling is that 
the functional is not that important since all the electrons contribute to the 
forces.

Pablo

De: wien-boun...@zeus.theochem.tuwien.ac.at 
 en nombre de Hu, Wenhao 

Enviado: domingo, 31 de enero de 2016 01:30 p. m.
Para: wien@zeus.theochem.tuwien.ac.at
Asunto: [Wien] A basic question about full hybrid functional calculation


Hi, All:


According to the link:


http://www.mail-archive.com/wien%40zeus.theochem.tuwien.ac.at/msg13559.html



It seems that it's impossible to realize atomic position relaxation with full 
relaxation functional. Then, does it mean that the forces obtained with MSR1a 
are all wrong? I'm also wondering whether the calculation without atomic 
relaxation is accurate enough to describe the system.


Best,

Wenhao

[http://www.mail-archive.com/logo.png]

Re: [Wien] relaxation of atomic positions with the full 
...
www.mail-archive.com
Re: [Wien] relaxation of atomic positions with the full hybrid functionals. 
tran Fri, 25 Dec 2015 11:11:41 -0800

___
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] case.clmsum_so and case.clmsum files

2016-01-31 Thread Majid Yazdani

Dear Wien2k authors and users

I‘m interested to study the topology of electron charge density of these
cases using the critic2 code.  To calculate the electronic structures, I
used the wien2k13.1 using the B3PW91 with alpha=0.2 and PBE-GGA
exchange-correlation functionals.   The spin polarization and spin-orbit
coupling (SOC) have been considered in these calculations. After
convergence the charge and stopping scf calculations, I compared the
case.clmsum  file generated within  PBE-GGA with that of generated within
B3PW91. They are very different.

However, the case.clmsum_so file within PBE-GGA is the same as that of
B3PW91:


[yazdani@cm3 diff]$ diff case.clmsum_so-B3PW91 case.clmsum_so-GGA

[yazdani@cm3 diff]$


Could you help me to know that why the case.clmsum_so files of GGA and
B3PW91 are similar while this is not the case for case.clmsum?

What is the difference of case.clmsum and case.clmsum_so?

Does Wien2k update the case.clmsum_so during the scf calculations?

If yes, why is the case.clmsum_so generated by GGA is the same as that of
B2PW91?

If No, what is the role of case.clmsum_so in the claculations?

Does Wien2k consider the SOC effects in generating the case.clmsum file?

Which file must be used for critic2 code if SOC is considered? case.clmsum
or case.clmsum_so?

Thanks for your helps

Sincerely Yours

Majid

This
email has been sent from a virus-free computer protected by Avast.
www.avast.com

<#DDB4FAA8-2DD7-40BB-A1B8-4E2AA1F9FDF2>
___
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] A basic question about full hybrid functional calculation

2016-01-31 Thread delamora

One thing is the relaxation and another is a calculation of electronic 
properties; the better the functional the better the results.
Pablo

Thank you all for the suggestions. I was considering doing a full hybrid 
functional calculation. Since the corresponding structure optimization is not 
implemented so far, I may postpone it until I have extra node to try it.

BTW, as Pablo mentioned in his post, a unrelaxed structure can give a really 
wrong results. I’m wondering whether the full hybrid function calculation in 
WIEN2K is valuable or not. Or maybe they can still be used in some certain 
systems?

Best,
Wenhao
___
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] A basic question about full hybrid functional calculation

2016-01-31 Thread Hu, Wenhao

Thank you all for the suggestions. I was considering doing a full hybrid 
functional calculation. Since the corresponding structure optimization is not 
implemented so far, I may postpone it until I have extra node to try it. 

BTW, as Pablo mentioned in his post, a unrelaxed structure can give a really 
wrong results. I’m wondering whether the full hybrid function calculation in 
WIEN2K is valuable or not. Or maybe they can still be used in some certain 
systems? 

Best,
Wenhao
___
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] BerryPi

2016-01-31 Thread lokanath patra

Dear Oleg,

Thanks a lot. I can find my mistake now. I will try for an insulator.

On Mon, Feb 1, 2016 at 6:17 AM, Oleg Rubel  wrote:

> Well, the fractional occupancies indicate that this is a metal. I afraid
> you cannot proceed with he effective charge calculation. BerryPI implies
> the presence of a band gap.
>
> > So if i have understood you, I have to rerun init_lapw after the
> displacement.
>
> Not quite, you need to change the order and start with the distorted
> structure (similar to
> https://github.com/spichardo/BerryPI/wiki/Tutorial-3:-Non-orthogonal-lattice-vectors).
> Again, it makes sense only if it is not a metal. Even a semimetal works
> with a shifted k-mesh.
>
>
> Oleg
>
> > On Jan 31, 2016, at 2:03 AM, lokanath patra 
> wrote:
> >
> > Dear Oleg,
> >
> > Here I am listing the result
> >
> > :BAN00034:  340.2210690.307010  1.
> > :BAN00035:  350.2463190.307010  1.
> > :BAN00036:  360.2688320.335387  1.
> > :BAN00037:  370.2843330.339720  1.
> > :BAN00038:  380.3035260.345742  1.
> > :BAN00039:  390.3184700.383233  1.
> > :BAN00040:  400.3243830.388931  1.
> > :BAN00041:  410.3616690.412751  1.
> > :BAN00042:  420.3868300.422636  1.
> > :BAN00043:  430.5354300.570228  0.88992023
> > :BAN00044:  440.5365260.576437  0.13433882
> > :BAN00045:  450.5535270.582444  0.
> > :BAN00046:  460.5682560.585172  0.
> > :BAN00047:  470.5723390.596519  0.
> > :BAN00048:  480.5723390.604896  0.
> > :BAN00049:  490.6236560.699132  0.
> > :BAN00042:  420.3122230.353933  1.
> > :BAN00043:  430.3248150.359415  1.
> > :BAN00044:  440.3331620.375854  1.
> > :BAN00045:  450.3521200.381727  1.
> > :BAN00046:  460.3521200.391737  1.
> > :BAN00047:  470.3911940.435997  1.
> > :BAN00048:  480.4003320.435997  1.
> > :BAN00049:  490.4138160.516497  1.
> > :BAN00050:  500.4138160.516497  1.
> > :BAN00051:  510.4964440.603847  0.88561301
> > :BAN00052:  520.5164970.603974  0.09045336
> > :BAN00053:  530.6728400.832567  0.
> > :BAN00054:  540.7121050.832567  0.
> > :BAN00055:  550.7638370.869655  0.
> > :BAN00056:  560.8068160.894698  0.
> > :BAN00057:  570.8278470.894698  0.
> > :BAN00034:  340.2210800.307021  1.
> > :BAN00035:  350.2463300.307021  1.
> > :BAN00036:  360.2688430.335401  1.
> > :BAN00037:  370.2843440.339732  1.
> > :BAN00038:  380.3035360.345754  1.
> > :BAN00039:  390.3184820.383243  1.
> > :BAN00040:  400.3243940.388944  1.
> > :BAN00041:  410.3616800.412761  1.
> > :BAN00042:  420.3868400.422646  1.
> > :BAN00043:  430.5354230.570222  0.88968694
> > :BAN00044:  440.5365200.576432  0.13424669
> > :BAN00045:  450.5535220.582439  0.
> > :BAN00046:  460.5682500.585166  0.
> > :BAN00047:  470.5723350.596516  0.
> > :BAN00048:  480.5723350.604893  0.
> > :BAN00049:  490.6236500.699132  0.
> >
> > So if i have understood you, I have to rerun init_lapw after the
> displacement.
> >
> > Thanks
> >
> >
> > On Sat, Jan 30, 2016 at 4:48 AM, Oleg Rubel  wrote:
> > Just a small addition: It is a “must" to start with the least symmetric
> structure first. This implies that you need to introduce displacements,
> init_lapw, run(sp)_lapw, berrypy and then bring atoms back to the high
> symmetry position, dstart (-up/-dn), run(sp)_lapw, berrypi.
> >
> > In your structure alpha and beta are 90 deg. If you are interested in
> Z*(3,3) (i.e., along Z axis) then there is no need to worry about
> nonorthogonality of lattice vectors, similar to
> https://github.com/spichardo/BerryPI/wiki/Tutorial-4:-Polarization-in-GaN
> >
> > Please be cautious with the band occupancy. Can you list the results of
> “grep :BAN *scf”? If the structure is metallic, berrypi cannot be continued.
> >
> > Oleg
> >
> > > On Jan 29, 2016, at 05:59, Dr. K. C. Bhamu 
> wrote:
> > >
> > > What you need it to shift both position i.e. 0.5 to 0.501 and
> 0. to 0.0010.
> > >
> > > Problem will be solved.
> > >
> > > regards
> > >
> > >
> > >
> > > 
> > > Dr. K. C. Bhamu
> > > (UGC-Dr. D. S. Kothari Postdoc Fellow)
> > > Department of Physics
> > > Goa University, Goa-403 206
> > > India
> > > Mob. No.  +91-9782911977
> > >
> > > On Fri, Jan 29, 2016 at 3:59 PM, lokanath patra <
>

Re: [Wien] BerryPi

2016-01-31 Thread Oleg Rubel

Well, the fractional occupancies indicate that this is a metal. I afraid you 
cannot proceed with he effective charge calculation. BerryPI implies the 
presence of a band gap.

> So if i have understood you, I have to rerun init_lapw after the displacement.

Not quite, you need to change the order and start with the distorted structure 
(similar to 
https://github.com/spichardo/BerryPI/wiki/Tutorial-3:-Non-orthogonal-lattice-vectors).
 Again, it makes sense only if it is not a metal. Even a semimetal works with a 
shifted k-mesh.


Oleg

> On Jan 31, 2016, at 2:03 AM, lokanath patra  
> wrote:
> 
> Dear Oleg,
> 
> Here I am listing the result
> 
> :BAN00034:  340.2210690.307010  1.
> :BAN00035:  350.2463190.307010  1.
> :BAN00036:  360.2688320.335387  1.
> :BAN00037:  370.2843330.339720  1.
> :BAN00038:  380.3035260.345742  1.
> :BAN00039:  390.3184700.383233  1.
> :BAN00040:  400.3243830.388931  1.
> :BAN00041:  410.3616690.412751  1.
> :BAN00042:  420.3868300.422636  1.
> :BAN00043:  430.5354300.570228  0.88992023
> :BAN00044:  440.5365260.576437  0.13433882
> :BAN00045:  450.5535270.582444  0.
> :BAN00046:  460.5682560.585172  0.
> :BAN00047:  470.5723390.596519  0.
> :BAN00048:  480.5723390.604896  0.
> :BAN00049:  490.6236560.699132  0.
> :BAN00042:  420.3122230.353933  1.
> :BAN00043:  430.3248150.359415  1.
> :BAN00044:  440.3331620.375854  1.
> :BAN00045:  450.3521200.381727  1.
> :BAN00046:  460.3521200.391737  1.
> :BAN00047:  470.3911940.435997  1.
> :BAN00048:  480.4003320.435997  1.
> :BAN00049:  490.4138160.516497  1.
> :BAN00050:  500.4138160.516497  1.
> :BAN00051:  510.4964440.603847  0.88561301
> :BAN00052:  520.5164970.603974  0.09045336
> :BAN00053:  530.6728400.832567  0.
> :BAN00054:  540.7121050.832567  0.
> :BAN00055:  550.7638370.869655  0.
> :BAN00056:  560.8068160.894698  0.
> :BAN00057:  570.8278470.894698  0.
> :BAN00034:  340.2210800.307021  1.
> :BAN00035:  350.2463300.307021  1.
> :BAN00036:  360.2688430.335401  1.
> :BAN00037:  370.2843440.339732  1.
> :BAN00038:  380.3035360.345754  1.
> :BAN00039:  390.3184820.383243  1.
> :BAN00040:  400.3243940.388944  1.
> :BAN00041:  410.3616800.412761  1.
> :BAN00042:  420.3868400.422646  1.
> :BAN00043:  430.5354230.570222  0.88968694
> :BAN00044:  440.5365200.576432  0.13424669
> :BAN00045:  450.5535220.582439  0.
> :BAN00046:  460.5682500.585166  0.
> :BAN00047:  470.5723350.596516  0.
> :BAN00048:  480.5723350.604893  0.
> :BAN00049:  490.6236500.699132  0.
> 
> So if i have understood you, I have to rerun init_lapw after the displacement.
> 
> Thanks
> 
> 
> On Sat, Jan 30, 2016 at 4:48 AM, Oleg Rubel  wrote:
> Just a small addition: It is a “must" to start with the least symmetric 
> structure first. This implies that you need to introduce displacements, 
> init_lapw, run(sp)_lapw, berrypy and then bring atoms back to the high 
> symmetry position, dstart (-up/-dn), run(sp)_lapw, berrypi.
> 
> In your structure alpha and beta are 90 deg. If you are interested in Z*(3,3) 
> (i.e., along Z axis) then there is no need to worry about nonorthogonality of 
> lattice vectors, similar to 
> https://github.com/spichardo/BerryPI/wiki/Tutorial-4:-Polarization-in-GaN
> 
> Please be cautious with the band occupancy. Can you list the results of “grep 
> :BAN *scf”? If the structure is metallic, berrypi cannot be continued.
> 
> Oleg
> 
> > On Jan 29, 2016, at 05:59, Dr. K. C. Bhamu  wrote:
> >
> > What you need it to shift both position i.e. 0.5 to 0.501 and 
> > 0. to 0.0010.
> >
> > Problem will be solved.
> >
> > regards
> >
> >
> >
> > 
> > Dr. K. C. Bhamu
> > (UGC-Dr. D. S. Kothari Postdoc Fellow)
> > Department of Physics
> > Goa University, Goa-403 206
> > India
> > Mob. No.  +91-9782911977
> >
> > On Fri, Jan 29, 2016 at 3:59 PM, lokanath patra 
> >  wrote:
> > Dear Fecher and Bhamu,
> > I also think that the error is due to the symmetry after changing the 
> > struct file. But the tutorial says ''Do not rerun the init_lapw as it may 
> > realize a higher symmetry. The intention is to keep symmetry unchanged 
> > between subsequent runs''. Just check the 3rd and

[Wien] A basic question about full hybrid functional calculation

Re: [Wien] A basic question about full hybrid functional calculation

Re: [Wien] A basic question about full hybrid functional calculation

Re: [Wien] A basic question about full hybrid functional calculation

[Wien] case.clmsum_so and case.clmsum files

Re: [Wien] A basic question about full hybrid functional calculation

Re: [Wien] A basic question about full hybrid functional calculation

Re: [Wien] BerryPi

Re: [Wien] BerryPi

9 matches

Site Navigation

Mail list logo

Footer information