Re: [Wien] Problem with fold2Bloch in real calculations

2022-01-12 Thread Rubel, Oleg 
Dear Veronique,

the option [-r] should work now (you need to fetch the most recent version from 
GitHub and recompile the code). There was indeed a problem with array 
allocation that caused a segfault. Thanks for reporting that!

I tested you FeSe structure. Technically f2b runs, but this structure is _not_ 
a 2x2x1 supercell in a standard way of thinking. If the primitive cell contains 
1 Fe atom, the 2x2x1 should have 4 Fe atoms, while you have 2. Maybe this is a 
sqrt(2) supercell?

Best regards
Oleg


From: Wien  on behalf of Rubel, Oleg 

Sent: Tuesday, January 11, 2022 18:44
To: wien@zeus.theochem.tuwien.ac.at
Subject: Re: [Wien] Problem with fold2Bloch in real calculations

Dear Veronique,

the [-r] option is not much tested (if tested at all). The main reason is that 
typical supercells with disorder have a low symmetry. I do not exclude that 
some supercells can possess the inversion, but I did not have such before in 
practice. Would you mind sharing a simplest struct file of your supercell with 
inversion?

As a temporary solution, one can break the inversion symmetry with labels 
(e.g., Sr1, Sr2, ...). This will force the complex calculation.

Best regards
Oleg


From: Wien  on behalf of Véronique 
BROUET 
Sent: Tuesday, January 11, 2022 10:22
To: wien@zeus.theochem.tuwien.ac.at
Subject: [Wien] Problem with fold2Bloch in real calculations

Dear Oleg,

It's not happening with only one case. I've tried it for the very simple FeSe 
(attached case), trying to unfold with 2:2:1, but also for Sr2IrO4 or SrIrO3 
(the case I'm interested in at the moment, also attached) or your examples. 
Everything works fine in complex mode, but not with another switch.

For example both FeSe and SrIrO3 have inversion symmetry and do not require 
complex calculation, but if I run the case in real mode, fold2Bloch recognizes 
well it's a real case.vector (it asks for the -r switch), but cannot process it 
(segmentation error).

Veronique

___
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] fold2bloch: unfolded band-structure plot have addition bands

2022-01-12 Thread Rubel, Oleg 
Dear Asif,

I am glad it worked for you. I think your unfolded band structure is OK. I am 
not sure what you refer to as "numerical precision of the unfolding procedure", 
but sometime weights go down to 0.0001. So, 0.2 is not so small. The plotting 
script has a weight threshold to display (wth = 0.05; % threshold weight). The 
default is 0.05. Everything less than that will not be displayed. You can set 
it to =0 and see what happens.

> ... bands with Bloch weight < 0.2 in just below the conduction band.

I see a low-weighted conduction band at K point, but it is not below the band 
edge, which is between K-Gamma. Maybe you can highlight the band in question 
and enclose the folded band structure you refer in the text.

> My question is that is this due to the numerical precision of the unfolding 
> procedure which leaves some bands behind or if it is real how to 
> confirm/check such a thing?

It would be beneficial to increase the supercell size to better represent a 
random structure, if this is what you aim at.

Thank you
Oleg


From: Wien  on behalf of Asif Ali 

Sent: Wednesday, January 12, 2022 04:52
To: A Mailing list for WIEN2k users
Subject: Re: [Wien] fold2bloch: unfolded band-structure plot have addition bands

Dear Sir,

I am thankful for your help. I am able to plot the correct unfolded band 
structure with the update plotting script.

I have one more question regarding the bands with small Bloch vector weights. 
In my calculated band structure (M_0.5M'_0.5 X_2), there are bands with Bloch 
weight < 0.2 in just below the conduction band. However these bands are clearly 
seen in the folded band structure.
My question is that is this due to the numerical precision of the unfolding 
procedure which leaves some bands behind or if it is real how to confirm/check 
such a thing?
I am enclosing the figure for your reference. I have increased the marker size 
20 for a clear view of low weight bands.

Looking forward to hearing from you

Thankyou and Regards
Asif

On Wed, Jan 12, 2022 at 3:16 PM Asif Ali 
mailto:as...@iiserb.ac.in>> wrote:
Dear Sir,

I am thankful for your help. I am able to plot the correct unfolded band 
structure with the update plotting script.

I have one more question regarding the bands with small Bloch vector weights. 
In my calculated band structure (M_0.5M'_0.5 X_2), there are bands with bloch 
weight < 0.2 in just below the conduction band. However these bands are clearly 
seen in the folded band structure.
My question is that is this due to the numerical precision of the unfolding 
procedure which leaves some bands behind or if it is real how to confirm/check 
such a thing


I am enclosing the figure for your reference. I have increased the marker size 
20 for a clear view of low weight bands.

Looking forward to hearing from you

Thankyou and Regards
Asif


On Wed, Jan 12, 2022 at 12:41 PM Peter Blaha 
mailto:pbl...@theochem.tuwien.ac.at>> wrote:
Resent from Oleg Rubel:

Dear Asif,

I reproduced your problem. It happens during plotting and related to the
way a distance is calculated between a segment in k space and a point.

The GitHub now contains an updated version of plotting scripts
(https://github.com/rubel75/fold2Bloch-Wien2k/tree/master/Utils).

There is _no_ need to recompile fold2bloch or redo WIEN2k calculation.

Enclosed is the unfolded band structure of 2H-WS2 (2x2x2 supercell) vs
WIEN2k primitive cell to verify that it works.

Thank you for the bug report!

Best regards
Oleg


Am 11.01.2022 um 17:53 schrieb Asif Ali:
> Dear Sir,
>
> Thankyou for your response. Yes, I am working with WS2 and similar
> systems. At the moment,  I was doing LDA calculation and I plan to do
> LDA+SO calculations.
>
> Regards
> Asif
>

--
--
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.at
WIEN2k: http://www.wien2k.at
WWW:   http://www.imc.tuwien.ac.at
-___
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


--
Thanks and Regards
Asif Ali


--
Thanks and Regards
Asif Ali
___
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] fold2bloch: unfolded band-structure plot have addition bands

2022-01-12 Thread Asif Ali 
Dear Sir,

I am thankful for your help. I am able to plot the correct unfolded band
structure with the update plotting script.

I have one more question regarding the bands with small Bloch vector
weights. In my calculated band structure (M_0.5M'_0.5 X_2), there are bands
with Bloch weight < 0.2 in just below the conduction band. However these
bands are clearly seen in the folded band structure.
My question is that is this due to the numerical precision of the unfolding
procedure which leaves some bands behind or if it is real how to
confirm/check such a thing?
I am enclosing the figure for your reference. I have increased the marker
size 20 for a clear view of low weight bands.

Looking forward to hearing from you

Thankyou and Regards
Asif

On Wed, Jan 12, 2022 at 3:16 PM Asif Ali  wrote:

> Dear Sir,
>
> I am thankful for your help. I am able to plot the correct unfolded band
> structure with the update plotting script.
>
> I have one more question regarding the bands with small Bloch vector
> weights. In my calculated band structure (M_0.5M'_0.5 X_2), there are bands
> with bloch weight < 0.2 in just below the conduction band. However these
> bands are clearly seen in the folded band structure.
> My question is that is this due to the numerical precision of the
> unfolding procedure which leaves some bands behind or if it is real how to
> confirm/check such a thing
>
>
> I am enclosing the figure for your reference. I have increased the marker
> size 20 for a clear view of low weight bands.
>
> Looking forward to hearing from you
>
> Thankyou and Regards
> Asif
>
>
> On Wed, Jan 12, 2022 at 12:41 PM Peter Blaha 
> wrote:
>
>> Resent from Oleg Rubel:
>>
>> Dear Asif,
>>
>> I reproduced your problem. It happens during plotting and related to the
>> way a distance is calculated between a segment in k space and a point.
>>
>> The GitHub now contains an updated version of plotting scripts
>> (https://github.com/rubel75/fold2Bloch-Wien2k/tree/master/Utils).
>>
>> There is _no_ need to recompile fold2bloch or redo WIEN2k calculation.
>>
>> Enclosed is the unfolded band structure of 2H-WS2 (2x2x2 supercell) vs
>> WIEN2k primitive cell to verify that it works.
>>
>> Thank you for the bug report!
>>
>> Best regards
>> Oleg
>>
>>
>> Am 11.01.2022 um 17:53 schrieb Asif Ali:
>> > Dear Sir,
>> >
>> > Thankyou for your response. Yes, I am working with WS2 and similar
>> > systems. At the moment,  I was doing LDA calculation and I plan to do
>> > LDA+SO calculations.
>> >
>> > Regards
>> > Asif
>> >
>>
>> --
>> --
>> 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
>>
>> -___
>> 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
>>
>
>
> --
> Thanks and Regards
> Asif Ali
>


-- 
Thanks and Regards
Asif Ali
___
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] Oscillating convergence

2022-01-12 Thread shamik chakrabarti 
Dear Prof. Tran,
 .  .  .  .  .  .  .  .  .  I have started a fresh calculation in a new
directory.

Thanks for your advice.

With regards,

On Wed, Jan 12, 2022, 14:30 Tran, Fabien  wrote:

> I forgot to mention that before restarting the HSE calculation you should
> have executed
> clean_lapw to delete the possibly present case.vectorhf or
> case.vectorhf_from the old HSE
> calculation. They may perturb the convergemce.
>
> 
> From: Wien  on behalf of shamik
> chakrabarti 
> Sent: Wednesday, January 12, 2022 9:52 AM
> To: A Mailing list for WIEN2k users
> Subject: Re: [Wien] Oscillating convergence
>
> Dear Prof. Blaha & Prof. Tran,
>
>  I have started Full HYbrid
> with alpha = 0.05 from scratch ( I have restored the GGA calculation in a
> new folder & started the Hybrid after that). Let see what will happen.
> Thanks for your advice.
>
> with regards,
>
> On Wed, 12 Jan 2022 at 14:02, shamik chakrabarti  > wrote:
> But with GGA+U the simulated voltage is not matching with experiment...
>
> On Wed, 12 Jan 2022 at 14:00, Peter Blaha  > wrote:
> If your present calculation "diverged", I'd restore the GGA+U calculation.
>
> Am 12.01.2022 um 09:19 schrieb shamik chakrabarti:
> > Dear Prof. Blaha,
> > By grid to 0.01, I mean the mixing parameter in
> > case.inm ( I change 0.2 to 0.01). For a trial can I check by reducing
> > the alpha from 0.25 to 0.05 in case.inhf now? Or I need to start with
> > 0.05 as alpha from the scratch?
> >
> > Looking forward to your reply.
> >
> > with regards,
> >
> >
> > On Wed, 12 Jan 2022 at 13:38, Peter Blaha  
> > >>
> wrote:
> >
> > I'm not too surprised. In HSE there is not only a dependency on the
> > density (which is "mixed") but also on the wave functions. This HF
> > potential due to them is "mixed" with 100 %.
> >
> > Maybe it helps when using a small alpha at the beginning (0.05
> instead
> > of 0.25) for the amount of HF (case.inhf), later on increase it when
> > reasonably converged.
> >
> > PS: I would probably not play with PRATT (except you can clearly
> > decrease :DIS significantly without oszillations) and I don't know
> what
> > you mean with:  "grid to 0.01" ???
> >
> > Am 12.01.2022 um 08:51 schrieb shamik chakrabarti:
> >  > Dear Wien2k users,
> >  >
> >  >I have started a simulation of ABCO4 oxide
> >  > material with 64 atomic unit cell using HSE06. However, the
> energy &
> >  > charge convergence is oscillating between 0.08-0.008 and 0.5 -
> > 0.2 for
> >  > around 100 cycles. The same structure was converged using GGA+U.
> > I have
> >  > used 1 k-point & also change the grid to 0.01 & used PRATT for
> > several
> >  > cycles now. What could be the remedy?
> >  >
> >  > (1) Is it that we need to use more than 1 k-point for convergence
> >  > (2) There is something wrong with the structure (However the same
> >  > structure is converged with GGA+U)
> >  >
> >  > Looking forward to hearing from you.
> >  >
> >  > with regards,
> >  >
> >  > --
> >  > Dr. Shamik Chakrabarti
> >  > Research Fellow
> >  > Department of Physics
> >  > Indian Institute of Technology Patna
> >  > Bihta-801103
> >  > Patna
> >  > Bihar, India
> >  >
> >  > ___
> >  > Wien mailing list
> >  > Wien@zeus.theochem.tuwien.ac.at 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
> > <
> 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.at bl...@theochem.tuwien.ac.at>
> > >WIEN2k: http://www.wien2k.at
> > 
> > WWW: http://www.imc.tuwien.ac.at 
> >
>  -
> > ___
> > Wien mailing list
> > Wien@zeus.theochem.tuwien.ac.at Wien@zeus.theochem.tuwien.ac.at>

Re: [Wien] Oscillating convergence

2022-01-12 Thread Tran, Fabien 
I forgot to mention that before restarting the HSE calculation you should have 
executed
clean_lapw to delete the possibly present case.vectorhf or case.vectorhf_from 
the old HSE
calculation. They may perturb the convergemce.


From: Wien  on behalf of shamik 
chakrabarti 
Sent: Wednesday, January 12, 2022 9:52 AM
To: A Mailing list for WIEN2k users
Subject: Re: [Wien] Oscillating convergence

Dear Prof. Blaha & Prof. Tran,

 I have started Full HYbrid with 
alpha = 0.05 from scratch ( I have restored the GGA calculation in a new folder 
& started the Hybrid after that). Let see what will happen.
Thanks for your advice.

with regards,

On Wed, 12 Jan 2022 at 14:02, shamik chakrabarti 
mailto:shamik15041...@gmail.com>> wrote:
But with GGA+U the simulated voltage is not matching with experiment...

On Wed, 12 Jan 2022 at 14:00, Peter Blaha 
mailto:pbl...@theochem.tuwien.ac.at>> wrote:
If your present calculation "diverged", I'd restore the GGA+U calculation.

Am 12.01.2022 um 09:19 schrieb shamik chakrabarti:
> Dear Prof. Blaha,
> By grid to 0.01, I mean the mixing parameter in
> case.inm ( I change 0.2 to 0.01). For a trial can I check by reducing
> the alpha from 0.25 to 0.05 in case.inhf now? Or I need to start with
> 0.05 as alpha from the scratch?
>
> Looking forward to your reply.
>
> with regards,
>
>
> On Wed, 12 Jan 2022 at 13:38, Peter Blaha 
> mailto:pbl...@theochem.tuwien.ac.at>
> >> 
> wrote:
>
> I'm not too surprised. In HSE there is not only a dependency on the
> density (which is "mixed") but also on the wave functions. This HF
> potential due to them is "mixed" with 100 %.
>
> Maybe it helps when using a small alpha at the beginning (0.05 instead
> of 0.25) for the amount of HF (case.inhf), later on increase it when
> reasonably converged.
>
> PS: I would probably not play with PRATT (except you can clearly
> decrease :DIS significantly without oszillations) and I don't know what
> you mean with:  "grid to 0.01" ???
>
> Am 12.01.2022 um 08:51 schrieb shamik chakrabarti:
>  > Dear Wien2k users,
>  >
>  >I have started a simulation of ABCO4 oxide
>  > material with 64 atomic unit cell using HSE06. However, the energy &
>  > charge convergence is oscillating between 0.08-0.008 and 0.5 -
> 0.2 for
>  > around 100 cycles. The same structure was converged using GGA+U.
> I have
>  > used 1 k-point & also change the grid to 0.01 & used PRATT for
> several
>  > cycles now. What could be the remedy?
>  >
>  > (1) Is it that we need to use more than 1 k-point for convergence
>  > (2) There is something wrong with the structure (However the same
>  > structure is converged with GGA+U)
>  >
>  > Looking forward to hearing from you.
>  >
>  > with regards,
>  >
>  > --
>  > Dr. Shamik Chakrabarti
>  > Research Fellow
>  > Department of Physics
>  > Indian Institute of Technology Patna
>  > Bihta-801103
>  > Patna
>  > Bihar, India
>  >
>  > ___
>  > 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.at
> >  
>   WIEN2k: http://www.wien2k.at
> 
> WWW: http://www.imc.tuwien.ac.at 
> -
> ___
> 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] Oscillating convergence

2022-01-12 Thread shamik chakrabarti 
Dear Prof. Blaha & Prof. Tran,

 I have started Full HYbrid
with alpha = 0.05 from scratch ( I have restored the GGA calculation in a
new folder & started the Hybrid after that). Let see what will happen.
Thanks for your advice.

with regards,

On Wed, 12 Jan 2022 at 14:02, shamik chakrabarti 
wrote:

> But with GGA+U the simulated voltage is not matching with experiment...
>
> On Wed, 12 Jan 2022 at 14:00, Peter Blaha 
> wrote:
>
>> If your present calculation "diverged", I'd restore the GGA+U calculation.
>>
>> Am 12.01.2022 um 09:19 schrieb shamik chakrabarti:
>> > Dear Prof. Blaha,
>> > By grid to 0.01, I mean the mixing parameter in
>> > case.inm ( I change 0.2 to 0.01). For a trial can I check by reducing
>> > the alpha from 0.25 to 0.05 in case.inhf now? Or I need to start with
>> > 0.05 as alpha from the scratch?
>> >
>> > Looking forward to your reply.
>> >
>> > with regards,
>> >
>> >
>> > On Wed, 12 Jan 2022 at 13:38, Peter Blaha > > > wrote:
>> >
>> > I'm not too surprised. In HSE there is not only a dependency on the
>> > density (which is "mixed") but also on the wave functions. This HF
>> > potential due to them is "mixed" with 100 %.
>> >
>> > Maybe it helps when using a small alpha at the beginning (0.05
>> instead
>> > of 0.25) for the amount of HF (case.inhf), later on increase it when
>> > reasonably converged.
>> >
>> > PS: I would probably not play with PRATT (except you can clearly
>> > decrease :DIS significantly without oszillations) and I don't know
>> what
>> > you mean with:  "grid to 0.01" ???
>> >
>> > Am 12.01.2022 um 08:51 schrieb shamik chakrabarti:
>> >  > Dear Wien2k users,
>> >  >
>> >  >I have started a simulation of ABCO4 oxide
>> >  > material with 64 atomic unit cell using HSE06. However, the
>> energy &
>> >  > charge convergence is oscillating between 0.08-0.008 and 0.5 -
>> > 0.2 for
>> >  > around 100 cycles. The same structure was converged using GGA+U.
>> > I have
>> >  > used 1 k-point & also change the grid to 0.01 & used PRATT for
>> > several
>> >  > cycles now. What could be the remedy?
>> >  >
>> >  > (1) Is it that we need to use more than 1 k-point for convergence
>> >  > (2) There is something wrong with the structure (However the same
>> >  > structure is converged with GGA+U)
>> >  >
>> >  > Looking forward to hearing from you.
>> >  >
>> >  > with regards,
>> >  >
>> >  > --
>> >  > Dr. Shamik Chakrabarti
>> >  > Research Fellow
>> >  > Department of Physics
>> >  > Indian Institute of Technology Patna
>> >  > Bihta-801103
>> >  > Patna
>> >  > Bihar, India
>> >  >
>> >  > ___
>> >  > 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
>> > <
>> 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.at
>> > WIEN2k:
>> http://www.wien2k.at
>> > 
>> > WWW: http://www.imc.tuwien.ac.at 
>> >
>>  -
>> > ___
>> > Wien mailing list
>> > Wien@zeus.theochem.tuwien.ac.at > 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
>> > <
>> http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html>
>> >
>> >
>> >
>> > --
>> > Dr. Shamik Chakrabarti
>> > Research Fellow
>> > Department of Physics
>> > Indian Institute of Technology Patna
>> > Bihta-801103
>> > Patna
>> > Bihar, India
>> >
>> > ___
>> > 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] Oscillating convergence

2022-01-12 Thread shamik chakrabarti 
But with GGA+U the simulated voltage is not matching with experiment...

On Wed, 12 Jan 2022 at 14:00, Peter Blaha 
wrote:

> If your present calculation "diverged", I'd restore the GGA+U calculation.
>
> Am 12.01.2022 um 09:19 schrieb shamik chakrabarti:
> > Dear Prof. Blaha,
> > By grid to 0.01, I mean the mixing parameter in
> > case.inm ( I change 0.2 to 0.01). For a trial can I check by reducing
> > the alpha from 0.25 to 0.05 in case.inhf now? Or I need to start with
> > 0.05 as alpha from the scratch?
> >
> > Looking forward to your reply.
> >
> > with regards,
> >
> >
> > On Wed, 12 Jan 2022 at 13:38, Peter Blaha  > > wrote:
> >
> > I'm not too surprised. In HSE there is not only a dependency on the
> > density (which is "mixed") but also on the wave functions. This HF
> > potential due to them is "mixed" with 100 %.
> >
> > Maybe it helps when using a small alpha at the beginning (0.05
> instead
> > of 0.25) for the amount of HF (case.inhf), later on increase it when
> > reasonably converged.
> >
> > PS: I would probably not play with PRATT (except you can clearly
> > decrease :DIS significantly without oszillations) and I don't know
> what
> > you mean with:  "grid to 0.01" ???
> >
> > Am 12.01.2022 um 08:51 schrieb shamik chakrabarti:
> >  > Dear Wien2k users,
> >  >
> >  >I have started a simulation of ABCO4 oxide
> >  > material with 64 atomic unit cell using HSE06. However, the
> energy &
> >  > charge convergence is oscillating between 0.08-0.008 and 0.5 -
> > 0.2 for
> >  > around 100 cycles. The same structure was converged using GGA+U.
> > I have
> >  > used 1 k-point & also change the grid to 0.01 & used PRATT for
> > several
> >  > cycles now. What could be the remedy?
> >  >
> >  > (1) Is it that we need to use more than 1 k-point for convergence
> >  > (2) There is something wrong with the structure (However the same
> >  > structure is converged with GGA+U)
> >  >
> >  > Looking forward to hearing from you.
> >  >
> >  > with regards,
> >  >
> >  > --
> >  > Dr. Shamik Chakrabarti
> >  > Research Fellow
> >  > Department of Physics
> >  > Indian Institute of Technology Patna
> >  > Bihta-801103
> >  > Patna
> >  > Bihar, India
> >  >
> >  > ___
> >  > 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
> > <
> 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.at
> > WIEN2k: http://www.wien2k.at
> > 
> > WWW: http://www.imc.tuwien.ac.at 
> >
>  -
> > ___
> > Wien mailing list
> > Wien@zeus.theochem.tuwien.ac.at  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
> > <
> http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html>
> >
> >
> >
> > --
> > Dr. Shamik Chakrabarti
> > Research Fellow
> > Department of Physics
> > Indian Institute of Technology Patna
> > Bihta-801103
> > Patna
> > Bihar, India
> >
> > ___
> > 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
> -
> ___
> Wien mailing list
> Wien@zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>

Re: [Wien] Oscillating convergence

2022-01-12 Thread Tran, Fabien 
Do first a GGA PBE calculation, save it, and then the HSE calculation with 
alpha=0.05 and with the default case.inm (MSR1).


From: Wien  on behalf of shamik 
chakrabarti 
Sent: Wednesday, January 12, 2022 9:19 AM
To: A Mailing list for WIEN2k users
Subject: Re: [Wien] Oscillating convergence

Dear Prof. Blaha,

   By  grid to 0.01, I mean the mixing parameter in 
case.inm ( I change 0.2 to 0.01). For a trial can I check by reducing the alpha 
from 0.25 to 0.05 in case.inhf now? Or I need to start with 0.05 as alpha from 
the scratch?

Looking forward to your reply.

with regards,



On Wed, 12 Jan 2022 at 13:38, Peter Blaha 
mailto:pbl...@theochem.tuwien.ac.at>> wrote:
I'm not too surprised. In HSE there is not only a dependency on the
density (which is "mixed") but also on the wave functions. This HF
potential due to them is "mixed" with 100 %.

Maybe it helps when using a small alpha at the beginning (0.05 instead
of 0.25) for the amount of HF (case.inhf), later on increase it when
reasonably converged.

PS: I would probably not play with PRATT (except you can clearly
decrease :DIS significantly without oszillations) and I don't know what
you mean with:  "grid to 0.01" ???

Am 12.01.2022 um 08:51 schrieb shamik chakrabarti:
> Dear Wien2k users,
>
>I have started a simulation of ABCO4 oxide
> material with 64 atomic unit cell using HSE06. However, the energy &
> charge convergence is oscillating between 0.08-0.008 and 0.5 - 0.2 for
> around 100 cycles. The same structure was converged using GGA+U. I have
> used 1 k-point & also change the grid to 0.01 & used PRATT for several
> cycles now. What could be the remedy?
>
> (1) Is it that we need to use more than 1 k-point for convergence
> (2) There is something wrong with the structure (However the same
> structure is converged with GGA+U)
>
> Looking forward to hearing from you.
>
> with regards,
>
> --
> Dr. Shamik Chakrabarti
> Research Fellow
> Department of Physics
> Indian Institute of Technology Patna
> Bihta-801103
> Patna
> Bihar, India
>
> ___
> 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.at
WIEN2k: http://www.wien2k.at
WWW:   http://www.imc.tuwien.ac.at
-
___
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


--
Dr. Shamik Chakrabarti
Research Fellow
Department of Physics
Indian Institute of Technology Patna
Bihta-801103
Patna
Bihar, India
___
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] Oscillating convergence

2022-01-12 Thread Peter Blaha 

If your present calculation "diverged", I'd restore the GGA+U calculation.

Am 12.01.2022 um 09:19 schrieb shamik chakrabarti:

Dear Prof. Blaha,
                        By grid to 0.01, I mean the mixing parameter in 
case.inm ( I change 0.2 to 0.01). For a trial can I check by reducing 
the alpha from 0.25 to 0.05 in case.inhf now? Or I need to start with 
0.05 as alpha from the scratch?


Looking forward to your reply.

with regards,


On Wed, 12 Jan 2022 at 13:38, Peter Blaha > wrote:


I'm not too surprised. In HSE there is not only a dependency on the
density (which is "mixed") but also on the wave functions. This HF
potential due to them is "mixed" with 100 %.

Maybe it helps when using a small alpha at the beginning (0.05 instead
of 0.25) for the amount of HF (case.inhf), later on increase it when
reasonably converged.

PS: I would probably not play with PRATT (except you can clearly
decrease :DIS significantly without oszillations) and I don't know what
you mean with:  "grid to 0.01" ???

Am 12.01.2022 um 08:51 schrieb shamik chakrabarti:
 > Dear Wien2k users,
 >
 >                        I have started a simulation of ABCO4 oxide
 > material with 64 atomic unit cell using HSE06. However, the energy &
 > charge convergence is oscillating between 0.08-0.008 and 0.5 -
0.2 for
 > around 100 cycles. The same structure was converged using GGA+U.
I have
 > used 1 k-point & also change the grid to 0.01 & used PRATT for
several
 > cycles now. What could be the remedy?
 >
 > (1) Is it that we need to use more than 1 k-point for convergence
 > (2) There is something wrong with the structure (However the same
 > structure is converged with GGA+U)
 >
 > Looking forward to hearing from you.
 >
 > with regards,
 >
 > --
 > Dr. Shamik Chakrabarti
 > Research Fellow
 > Department of Physics
 > Indian Institute of Technology Patna
 > Bihta-801103
 > Patna
 > Bihar, India
 >
 > ___
 > 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.at
    WIEN2k: http://www.wien2k.at

WWW: http://www.imc.tuwien.ac.at 
-
___
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




--
Dr. Shamik Chakrabarti
Research Fellow
Department of Physics
Indian Institute of Technology Patna
Bihta-801103
Patna
Bihar, India

___
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
-
___
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] Oscillating convergence

2022-01-12 Thread shamik chakrabarti 
Dear Prof. Blaha,

   By  grid to 0.01, I mean the mixing parameter in
case.inm ( I change 0.2 to 0.01). For a trial can I check by reducing the
alpha from 0.25 to 0.05 in case.inhf now? Or I need to start with 0.05 as
alpha from the scratch?

Looking forward to your reply.

with regards,



On Wed, 12 Jan 2022 at 13:38, Peter Blaha 
wrote:

> I'm not too surprised. In HSE there is not only a dependency on the
> density (which is "mixed") but also on the wave functions. This HF
> potential due to them is "mixed" with 100 %.
>
> Maybe it helps when using a small alpha at the beginning (0.05 instead
> of 0.25) for the amount of HF (case.inhf), later on increase it when
> reasonably converged.
>
> PS: I would probably not play with PRATT (except you can clearly
> decrease :DIS significantly without oszillations) and I don't know what
> you mean with:  "grid to 0.01" ???
>
> Am 12.01.2022 um 08:51 schrieb shamik chakrabarti:
> > Dear Wien2k users,
> >
> >I have started a simulation of ABCO4 oxide
> > material with 64 atomic unit cell using HSE06. However, the energy &
> > charge convergence is oscillating between 0.08-0.008 and 0.5 - 0.2 for
> > around 100 cycles. The same structure was converged using GGA+U. I have
> > used 1 k-point & also change the grid to 0.01 & used PRATT for several
> > cycles now. What could be the remedy?
> >
> > (1) Is it that we need to use more than 1 k-point for convergence
> > (2) There is something wrong with the structure (However the same
> > structure is converged with GGA+U)
> >
> > Looking forward to hearing from you.
> >
> > with regards,
> >
> > --
> > Dr. Shamik Chakrabarti
> > Research Fellow
> > Department of Physics
> > Indian Institute of Technology Patna
> > Bihta-801103
> > Patna
> > Bihar, India
> >
> > ___
> > 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
> -
> ___
> 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
>


-- 
Dr. Shamik Chakrabarti
Research Fellow
Department of Physics
Indian Institute of Technology Patna
Bihta-801103
Patna
Bihar, India
___
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] Oscillating convergence

2022-01-12 Thread Peter Blaha 
I'm not too surprised. In HSE there is not only a dependency on the 
density (which is "mixed") but also on the wave functions. This HF 
potential due to them is "mixed" with 100 %.


Maybe it helps when using a small alpha at the beginning (0.05 instead 
of 0.25) for the amount of HF (case.inhf), later on increase it when 
reasonably converged.


PS: I would probably not play with PRATT (except you can clearly 
decrease :DIS significantly without oszillations) and I don't know what 
you mean with:  "grid to 0.01" ???


Am 12.01.2022 um 08:51 schrieb shamik chakrabarti:

Dear Wien2k users,

                       I have started a simulation of ABCO4 oxide 
material with 64 atomic unit cell using HSE06. However, the energy & 
charge convergence is oscillating between 0.08-0.008 and 0.5 - 0.2 for 
around 100 cycles. The same structure was converged using GGA+U. I have 
used 1 k-point & also change the grid to 0.01 & used PRATT for several 
cycles now. What could be the remedy?


(1) Is it that we need to use more than 1 k-point for convergence
(2) There is something wrong with the structure (However the same 
structure is converged with GGA+U)


Looking forward to hearing from you.

with regards,

--
Dr. Shamik Chakrabarti
Research Fellow
Department of Physics
Indian Institute of Technology Patna
Bihta-801103
Patna
Bihar, India

___
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
-
___
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