Re: [QE-users] nscf calculation of 2x2x2 Si supercell

2020-05-14 Thread Stefano Baroni 
PS. Start practicing with a small unit cell that you can draw with your hands. 
Get familiar with the numerical parameters (thresholds, cutoffs, etc.) in that 
case and develop your feeling on what happens when you change each of them. 
Starting with examples provided with the distribution may be a good idea. Then, 
but only then, try more complex (and possibly meaningful) cases. Above all, do 
not do anything that you do not understand, no matter who recommends that. Good 
luck! SB

> On 14 May 2020, at 17:32, Stefano Baroni  wrote:
> 
> Fariba: luck is no substitute for practice & understanding.
> 
>> On 14 May 2020, at 16:34, Fariba Islam > > wrote:
>> 
>> Initially it was 1e-8 but nscf calculation was crashing. So I searched for a 
>> solution and in ResearchGate I found that in order to solve this I need to 
>> decrease conv threshold or increase ecutwfc. That is why I increased it to 
>> such high values gradually. But didn’t have any luck.
>> 
>> On Thursday, May 14, 2020, Stefano de Gironcoli > > wrote:
>> do you really need conv_thr=1e-20 ?
>> 
>> stefano
>> 
>> On 14/05/20 09:18, Fariba Islam wrote:
>>> Dear all,
>>> My  conv_thr=1e-20 and ecutwfc=200 ecutrho=800
>>> My nscf calculation is crashing with the error 
>>> 
>>> task #10
>>>  from c_bands : error # 1
>>>  too many bands are not converged
>>> 
>>> How can I solve? From suggestions found online, I have reduced the 
>>> threshold and increased ecutwfc and ecutrho. 
>>> Regards
>>> Fariba Islam
>>> Bangladesh University of Engineering and Technology
>>> 
>>> 
>>> ___
>>> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso 
>>> )
>>> users mailing list users@lists.quantum-espresso.org 
>>> 
>>> https://lists.quantum-espresso.org/mailman/listinfo/users 
>>> ___
>> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso 
>> )
>> users mailing list users@lists.quantum-espresso.org 
>> 
>> https://lists.quantum-espresso.org/mailman/listinfo/users 
>> 
> — 
> Stefano Baroni - Trieste —  http://stefano.baroni.me 
>  
> 
> 
> 
> 
> ___
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
> users mailing list users@lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users

— 
Stefano Baroni - Trieste —  http://stefano.baroni.me 
 




___
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
users mailing list users@lists.quantum-espresso.org
https://lists.quantum-espresso.org/mailman/listinfo/users

Re: [QE-users] nscf calculation of 2x2x2 Si supercell

2020-05-14 Thread Stefano Baroni 
Fariba: luck is no substitute for practice & understanding.

> On 14 May 2020, at 16:34, Fariba Islam  wrote:
> 
> Initially it was 1e-8 but nscf calculation was crashing. So I searched for a 
> solution and in ResearchGate I found that in order to solve this I need to 
> decrease conv threshold or increase ecutwfc. That is why I increased it to 
> such high values gradually. But didn’t have any luck.
> 
> On Thursday, May 14, 2020, Stefano de Gironcoli  > wrote:
> do you really need conv_thr=1e-20 ?
> 
> stefano
> 
> On 14/05/20 09:18, Fariba Islam wrote:
>> Dear all,
>> My  conv_thr=1e-20 and ecutwfc=200 ecutrho=800
>> My nscf calculation is crashing with the error 
>> 
>> task #10
>>  from c_bands : error # 1
>>  too many bands are not converged
>> 
>> How can I solve? From suggestions found online, I have reduced the threshold 
>> and increased ecutwfc and ecutrho. 
>> Regards
>> Fariba Islam
>> Bangladesh University of Engineering and Technology
>> 
>> 
>> ___
>> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso 
>> )
>> users mailing list users@lists.quantum-espresso.org 
>> 
>> https://lists.quantum-espresso.org/mailman/listinfo/users 
>> ___
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
> users mailing list users@lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users

— 
Stefano Baroni - Trieste —  http://stefano.baroni.me 
 




___
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
users mailing list users@lists.quantum-espresso.org
https://lists.quantum-espresso.org/mailman/listinfo/users

Re: [QE-users] nscf calculation of 2x2x2 Si supercell

2020-05-14 Thread Fariba Islam 
Initially it was 1e-8 but nscf calculation was crashing. So I searched for
a solution and in ResearchGate I found that in order to solve this I need
to decrease conv threshold or increase ecutwfc. That is why I increased it
to such high values gradually. But didn’t have any luck.

On Thursday, May 14, 2020, Stefano de Gironcoli  wrote:

> do you really need conv_thr=1e-20 ?
>
> stefano
> On 14/05/20 09:18, Fariba Islam wrote:
>
> Dear all,
> My  conv_thr=1e-20 and ecutwfc=200 ecutrho=800
> My nscf calculation is crashing with the error
>
> task #10
>  from c_bands : error # 1
>  too many bands are not converged
>
> How can I solve? From suggestions found online, I have reduced the
> threshold and increased ecutwfc and ecutrho.
> Regards
> Fariba Islam
> Bangladesh University of Engineering and Technology
>
> ___
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
> users mailing list 
> users@lists.quantum-espresso.orghttps://lists.quantum-espresso.org/mailman/listinfo/users
>
>
___
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
users mailing list users@lists.quantum-espresso.org
https://lists.quantum-espresso.org/mailman/listinfo/users

Re: [QE-users] nscf calculation of 2x2x2 Si supercell

2020-05-14 Thread Stefano de Gironcoli 

do you really need conv_thr=1e-20 ?

stefano

On 14/05/20 09:18, Fariba Islam wrote:

Dear all,
My  conv_thr=1e-20 and ecutwfc=200 ecutrho=800
My nscf calculation is crashing with the error

task #        10
     from c_bands : error #         1
     too many bands are not converged

How can I solve? From suggestions found online, I have reduced the 
threshold and increased ecutwfc and ecutrho.

Regards
Fariba Islam
Bangladesh University of Engineering and Technology

___
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
users mailing list users@lists.quantum-espresso.org
https://lists.quantum-espresso.org/mailman/listinfo/users
___
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
users mailing list users@lists.quantum-espresso.org
https://lists.quantum-espresso.org/mailman/listinfo/users

Re: [QE-users] nscf calculation of 2x2x2 Si supercell

2020-05-14 Thread Stefano Baroni 
The convergence threshold looks unreasonable, and so does the kinetic-energy 
cutoff (is it Si?)
It’s always a good practice to play around with the parameter you use to 
develop your own feeling about what is is reasonable and what is not. In this 
matter, nobody can teach you better than yourself.
SB

PS: the code did not crash —  it just informed you that you probably did 
something unreasonable ...


> On 14 May 2020, at 09:18, Fariba Islam  wrote:
> 
> Dear all,
> My  conv_thr=1e-20 and ecutwfc=200 ecutrho=800
> My nscf calculation is crashing with the error 
> 
> task #10
>  from c_bands : error # 1
>  too many bands are not converged
> 
> How can I solve? From suggestions found online, I have reduced the threshold 
> and increased ecutwfc and ecutrho. 
> Regards
> Fariba Islam
> Bangladesh University of Engineering and Technology
> ___
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
> users mailing list users@lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users

— 
Stefano Baroni - Trieste —  http://stefano.baroni.me 
 




___
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
users mailing list users@lists.quantum-espresso.org
https://lists.quantum-espresso.org/mailman/listinfo/users

Re: [QE-users] nscf calculation of 2x2x2 Si supercell

2020-05-14 Thread pdelugas 
Hiconv_thr is exceedingly low. For nscf calculations anything lower than 1e-12 is likely going to cause convergence problems.Il 14 mag 2020 9:18 AM, Fariba Islam  ha scritto:Dear all,My  conv_thr=1e-20 and ecutwfc=200 ecutrho=800My nscf calculation is crashing with the error task #        10     from c_bands : error #         1     too many bands are not convergedHow can I solve? From suggestions found online, I have reduced the threshold and increased ecutwfc and ecutrho. RegardsFariba IslamBangladesh University of Engineering and Technology
___
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
users mailing list users@lists.quantum-espresso.org
https://lists.quantum-espresso.org/mailman/listinfo/users