[QE-users] Problem with tetrahedra method: from tetra_init : error # 26, cannot remap grid on k-point list

Thu Feb 14 09:52:49 CET 2019

Hi Oleksandr

Yes you are right, I just checked in the code, initializinfg the 
tetrahedra 2D grids are just  considered as 3D grids  because of the 
periodicity.

As for results I used tetrahedra for dos and pdos ( in 3D systems) and 
results are fine. Never used it for scf calculations though.
Pietro

On 02/14/2019 09:11 AM, Oleksandr Motornyi wrote:
>
> Hello Pacome and Pietro,
>
> I can confirm that tetrahedra_opt works for the 2D k-point meshes, 
> both for scf/nscf and dos.x/projwfc.x calculations, so it should be 
> implemented at least in some way. The other thing is that I am not 
> convinced with the credibility of results I obtain, but it is a 
> different story.
>
> Regards,
>
> Oleksandr
>
> Oleksandr Motornyi
> PhD
>
> Laboratoire des Solides Irradies
> Ecole Polytechnique (Palaiseau, France)
> On 02/14/2019 08:58 AM, Pietro Davide Delugas wrote:
>> Hello Pacome
>>
>> using qe-6.3  you can actually run the nscf  calculation without 
>> specifying occupation=tetrahedra leave smearing or whatever else and  
>> just take care to check that you are computing as many bands you 
>> need, which as I read you are already doing specifying nbnd.
>> In the dos.x input  you have just to add the option bz_sum = 
>> 'tetrahedra_opt'.
>>
>> if you use qe-6.3 remember that the format of xml and charge density 
>> files are different from those of 6.2 and older versions, so you have 
>> to rerun the scf computation as well.
>>
>> One thing that I notice now is that you are using a 9X9X1 mesh, I am 
>> not sure whether 2D version of the method is actually implemented in 
>> Q.E. you could try using a 9X9X2 mesh and see if that works.
>>
>> greetings - Pietro
>>
>>
>>
>> On 02/14/2019 01:03 AM, Pacome NGUIMEYA wrote:
>>> Hi all,
>>>
>>> I have been trying to run the calculation of the Density of State 
>>> (DOS) for TaAs using *QE-6.1* but It always crashes and returns the 
>>> following error message:
>>>
>>>      task #         6
>>>      from tetra_init : error #        26
>>>      cannot remap grid on k-point list
>>>
>>> There are several forums on this subject where many workaround have 
>>> been proposed.
>>>
>>> I went through all of them and I tried each of the proposed 
>>> workaround that I found, but none of them could solve the problem.
>>>
>>> Neither adding bands *(ndnd=144)* nor *the increase of the 
>>> convergence threshold to 1.0e-11* worked 
>>> (http://blog.sina.com.cn/s/blog_5f15ead20100cw4d.html), not even the 
>>> workaround propose by Prof Paolo Giannozzi here 
>>> https://www.mail-archive.com/users@lists.quantum-espresso.org/msg32997.html
>>>
>>> I even *changed the version of quantum espresso*; I used *QE 5.4*, 
>>> *6.2* and *6.3* but I keep getting errors
>>>
>>> with *QE-5.4* and *QE-6.2* I am getting the following error
>>>
>>>      task #        21
>>>      from tetrahedra : error #        26
>>>      cannot remap grid on k-point list
>>>
>>>  and this one with *QE-6.3*
>>>
>>>      task #         0
>>>      from pw_readfile : error #         1
>>>      error opening xml data file
>>>
>>> In the *particular case* of *QE-6.3* the *problem seems to be 
>>> solved* but it looks like the program *can not find the xml file*; 
>>> *it looks like it does not exist*.
>>>
>>> I am using the CHPC(centre for high performance computing)'s cluster 
>>> to run my calculations
>>>
>>> I need your help to sort this out
>>>
>>> Is it possible to calculate the DOS with *occupations = 'smearing', 
>>> smearing = 'M-P', degauss = 0.011* and with the *automatically 
>>> chosen k-points 9 9 1* 1 1 1 instead of *occupations = 'tetrahedra'* 
>>> with *k-points 9 9 1* 1 1 1.
>>>
>>>  I look forward to hearing from you,
>>>
>>>  Below is my input file (also attached to this email)
>>>
>>>
>>> &CONTROL
>>>  calculation = 'nscf',
>>>     prefix = 'TaAs',
>>> restart_mode = 'from_scratch',
>>> pseudo_dir = '/...,
>>> outdir = '/...,
>>>      etot_conv_thr = 1.0e-5,
>>> wf_collect = .true.,
>>>  /
>>> &SYSTEM
>>>  ibrav = 7,
>>>  A = 3.4348002616,
>>>  B = 3.4348002616,
>>>  C = 11.641,
>>>      cosAB = 0,
>>>      cosAC = 0,
>>>      cosBC = 0,
>>>  nat = 8,
>>> ntyp = 2,
>>> nbnd = 72,
>>>  ecutwfc = 70,
>>>  ecutrho = 400,
>>> occupations = 'tetrahedra',
>>>  /
>>> &ELECTRONS
>>>           conv_thr = 1.0e-11,
>>>        mixing_mode = 'plain',
>>>  mixing_beta = 0.7,
>>>  diagonalization = 'cg',
>>>  /
>>>  ATOMIC_SPECIES
>>>    Ta   180.94788 Ta.pbe-spn-kjpaw_psl.0.2.UPF
>>>    As   74.9216 As.pbe-n-kjpaw_psl.0.2.UPF
>>>  ATOMIC_POSITIONS {crystal}
>>> Ta       0.256888939 -0.145590417  -0.031926398
>>> Ta       0.575434474  0.440904275   0.567465416
>>> Ta      -0.284237537  0.808259399   0.316294705
>>> Ta       0.027951774  0.396761377   0.919029854
>>> As       0.041545914 -0.136018486   0.451188370
>>> As       0.704536879  0.274274640   0.880297072
>>> As       0.172406031  0.385662500   0.492672865
>>> As       0.505473527 -0.024253289   0.072978100
>>>  K_POINTS {automatic}
>>>   9 9 1  1 1 1
>>>
>>>  Regards,
>>>  Pacome
>>> *___**_____________
>>> *Pacome NGUIMEYA
>>> Ph.D. Candidate
>>> Computational Condensed Matter Physics
>>> University of Cape Town (UCT), South Africa
>>> */“Be Yourself; everyone else is already taken.”///*
>>> */Oscar Wilde/*
>>>
>>>
>>> _______________________________________________
>>> users mailing list
>>> users at lists.quantum-espresso.org
>>> https://lists.quantum-espresso.org/mailman/listinfo/users
>>
>>
>>
>>
>> _______________________________________________
>> users mailing list
>> users at lists.quantum-espresso.org
>> https://lists.quantum-espresso.org/mailman/listinfo/users
>
>
>
> _______________________________________________
> users mailing list
> users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20190214/f5ed45d6/attachment.html>