[Pw_forum] BFGS relaxation ignoring forc_conv_thr

Leonid Kahle leonid.kahle at epfl.ch
Wed Sep 20 16:02:55 CEST 2017


Thanks a lot for the fast answer,

as you say, after this warning is printed another step is performed, and 
the message:

bfgs converged in   4 scf cycles and   3 bfgs steps
      (criteria: energy <  1.0E-06 Ry, force <  1.0E-15 Ry/Bohr)

is printed after the last step. The keyword 'converged' is of course 
misleading in such a case.

Should others encounter the same problem, I switched to

&IONS
     ion_dynamics = 'damp'
/

from the default 'bfgs'. This makes it possible to reduce the force 
components to 1.0D-7.

Leonid


On 09/20/2017 02:03 PM, Paolo Giannozzi wrote:
> If you look carefully in the output, you will notice this:
> WARNING: bfgs curvature condition failed, Theta= 0.815
> Message from routine bfgs:
> history already reset at previous step: stopping
> This kind of problem arises when one is very close to convergence and 
> forces contain too much numerical noise for the minimization to be 
> effective. Not sure why it performs a further step.
> Paolo
>
> On Wed, Sep 20, 2017 at 9:56 AM, Leonid Kahle <leonid.kahle at epfl.ch 
> <mailto:leonid.kahle at epfl.ch>> wrote:
>
>     Dear all,
>
>     A relaxation of a rather large system (Li21 Si3 P3 S24) is shown
>     as converged even though the force convergence threshold is not
>     reached. I will paste the final output below, and the full input
>     and output are attached. I put a very low threshold 1D-15 to prove
>     my point here, but also more sensible thresholds are ignored.
>
>     As explained in the documentation, but also in this forum (
>     http://qe-forge.org/pipermail/pw_forum/2004-November/076156.html
>     <http://qe-forge.org/pipermail/pw_forum/2004-November/076156.html>
>     ) convergence of the relaxation is achieved when all the
>     components of the force vector are smaller than forc_conv_thr (and
>     energy criterion is met as well). In the example below this is
>     obviously not the case, but bfgs is shown as converged without any
>     warning.
>
>     Can anyone help me find an explanation for this? I am using the
>     parallel version of QE 6.1 compiled with intel compilers on an
>     x86_64 architecture.
>
>     Many thanks in advance,
>
>         Leonid Kahle
>
>     -- 
>
>     Doctoral Assistant
>     EPFL STI IMX THEOS
>     ME-D2 1019 (Bâtiment ME)
>     Station 9
>     CH-1015 Lausanne (Switzerland)
>
>
>          Forces acting on atoms (cartesian axes, Ry/au):
>
>          atom    1 type  1   force =     0.00000940 0.00000005   
>     0.00000759
>          atom    2 type  1   force =     0.00003922 -0.00005066  
>     -0.00003891
>          atom    3 type  1   force =    -0.00000527 -0.00001786  
>     -0.00000047
>          atom    4 type  1   force =     0.00000695 -0.00000525  
>     -0.00000490
>          atom    5 type  1   force =     0.00000837 0.00000136   
>     0.00000695
>          atom    6 type  1   force =    -0.00000139 0.00000800  
>     -0.00000768
>          atom    7 type  1   force =    -0.00000042 -0.00000074  
>     -0.00000888
>          atom    8 type  1   force =    -0.00002354 -0.00001155  
>     -0.00003861
>          atom    9 type  1   force =     0.00001305 -0.00001546   
>     0.00000555
>          atom   10 type  1   force =    -0.00001756 0.00002269  
>     -0.00001184
>          atom   11 type  1   force =    -0.00000255 -0.00001734   
>     0.00001741
>          atom   12 type  1   force =     0.00001174 0.00001759  
>     -0.00001735
>          atom   13 type  1   force =    -0.00000633 -0.00000979   
>     0.00000484
>          atom   14 type  1   force =    -0.00001325 0.00001294  
>     -0.00000493
>          atom   15 type  1   force =     0.00000782 -0.00002956  
>     -0.00000586
>          atom   16 type  1   force =    -0.00000317 0.00001363   
>     0.00000878
>          atom   17 type  1   force =    -0.00002203 0.00000270  
>     -0.00000480
>          atom   18 type  1   force =    -0.00000998 0.00000531   
>     0.00001788
>          atom   19 type  1   force =     0.00000060 -0.00000321  
>     -0.00001756
>          atom   20 type  1   force =    -0.00000688 0.00002017   
>     0.00000404
>          atom   21 type  1   force =    -0.00000129 -0.00000416  
>     -0.00001532
>          atom   22 type  2   force =    -0.00000319 -0.00000167   
>     0.00000395
>          atom   23 type  2   force =     0.00001103 0.00001040   
>     0.00000626
>          atom   24 type  2   force =     0.00000396 -0.00000186   
>     0.00000201
>          atom   25 type  3   force =    -0.00001044 0.00000478  
>     -0.00000086
>          atom   26 type  3   force =    -0.00001291 0.00000415  
>     -0.00000411
>          atom   27 type  3   force =     0.00001756 -0.00000296   
>     0.00000913
>          atom   28 type  4   force =    -0.00001578 0.00000879   
>     0.00000423
>          atom   29 type  4   force =    -0.00001227 0.00000946  
>     -0.00000839
>          atom   30 type  4   force =     0.00001416 0.00001301   
>     0.00001436
>          atom   31 type  4   force =     0.00001660 -0.00000202   
>     0.00000659
>          atom   32 type  4   force =     0.00003452 -0.00001067   
>     0.00001793
>          atom   33 type  4   force =     0.00000938 -0.00001056   
>     0.00000237
>          atom   34 type  4   force =    -0.00001496 -0.00000472   
>     0.00000305
>          atom   35 type  4   force =    -0.00001625 -0.00000739  
>     -0.00000918
>          atom   36 type  4   force =     0.00001063 0.00001606  
>     -0.00000310
>          atom   37 type  4   force =    -0.00000110 0.00000394  
>     -0.00000268
>          atom   38 type  4   force =    -0.00000163 0.00000039   
>     0.00000700
>          atom   39 type  4   force =    -0.00000066 -0.00000016   
>     0.00000132
>          atom   40 type  4   force =    -0.00000813 0.00000603   
>     0.00000016
>          atom   41 type  4   force =    -0.00001613 0.00000196   
>     0.00000515
>          atom   42 type  4   force =    -0.00000014 0.00000820  
>     -0.00000460
>          atom   43 type  4   force =     0.00000091 0.00000952   
>     0.00001063
>          atom   44 type  4   force =     0.00000012 -0.00000246   
>     0.00000554
>          atom   45 type  4   force =     0.00000154 0.00000256   
>     0.00001045
>          atom   46 type  4   force =     0.00000846 0.00000608  
>     -0.00000078
>          atom   47 type  4   force =     0.00000508 -0.00000261   
>     0.00001345
>          atom   48 type  4   force =     0.00000340 0.00000050   
>     0.00000619
>          atom   49 type  4   force =     0.00000354 -0.00000332   
>     0.00000588
>          atom   50 type  4   force =    -0.00000428 0.00000637   
>     0.00000854
>          atom   51 type  4   force =    -0.00000650 -0.00000066  
>     -0.00000641
>
>          Total force =     0.000149     Total SCF correction =    
>     0.000000
>
>          bfgs converged in   4 scf cycles and   3 bfgs steps
>          (criteria: energy <  1.0E-06 Ry, force < 1.0E-15 Ry/Bohr)
>
>          End of BFGS Geometry Optimization
>
>
>
>
>     _______________________________________________
>     Pw_forum mailing list
>     Pw_forum at pwscf.org <mailto:Pw_forum at pwscf.org>
>     http://pwscf.org/mailman/listinfo/pw_forum
>     <http://pwscf.org/mailman/listinfo/pw_forum>
>
>
>
>
> -- 
> Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche,
> Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
> Phone +39-0432-558216, fax +39-0432-558222
>
>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20170920/fc6756df/attachment.html>


More information about the users mailing list