issue in geometry optimization with constraints

Agostino Migliore

unread,

Oct 15, 2020, 3:51:33 PM10/15/20

to NWChem Forum

Hello,

I have unsuccessfully tried to run the job with input reported below using NWChem 7.0. The job was correctly working with NWChem 6.6, while using NWChem version 7.0 it stops at the beginning producing the error message

------------------------------------------------------------------------
cons_input: geom_rtdb? 71
------------------------------------------------------------------------
------------------------------------------------------------------------
current input line :
46: constraints
------------------------------------------------------------------------
------------------------------------------------------------------------
There is an error related to the specified geometry
------------------------------------------------------------------------

The compilation did not appear to produce any error message and I do not see any mistake in the input file (and in fact it works using version 6.6 of the program). Any suggestion on the possible causes of this issue / things to check / ways to fix the problem would be highly appreciated!

Many thanks,

Best regards,

Agostino Migliore

University of Padua

scratch_dir /scratch/agostino.migliore/ae_12_NF_pbe0_6311pdf_relax/

start ae_12_NF_pbe0_6311pdf_relax

memory 500 mb

geometry system nocenter noautoz

C 2.80472967 0.00002519 -0.40891184

C 3.92934326 -0.00000122 0.62676940

C 5.32531965 0.00002716 -0.00007225

C 6.45492015 0.00000222 1.03223225

C 7.82889916 0.00001704 0.36373180

C 9.05146950 -0.00001794 1.26165661

C 10.36173670 -0.00008524 0.51131136

C 11.71575887 0.00000747 1.22581279

C 12.86154846 -0.00000795 0.21002918

C 14.24800048 0.00002074 0.85618628

C 15.38935342 0.00001235 -0.16325331

C 16.77805092 0.00002293 0.47480580

N 8.78518382 -0.00001490 2.50400482

F 9.97215933 -0.00006352 3.31240001

H 1.82097479 0.00000396 0.06591136

H 2.86108258 0.88108601 -1.05266943

H 2.86109328 -0.88099409 -1.05272531

H 3.82680372 0.87459205 1.27616816

H 3.82681446 -0.87463832 1.27611083

H 5.42588290 0.87437986 -0.65181933

H 5.42589366 -0.87428544 -0.65187153

H 6.36390270 0.87242211 1.68194246

H 6.36390303 -0.87245026 1.68189878

H 7.91954279 0.87078099 -0.29570503

H 7.91953398 -0.87071233 -0.29575432

H 10.31273521 0.86563954 -0.16078068

H 10.31275226 -0.86596342 -0.16058857

H 11.79806936 0.87309139 1.87266619

H 11.79813322 -0.87299667 1.87276416

H 12.76873121 0.87455752 -0.44304102

H 12.76874861 -0.87461026 -0.44299405

H 14.34532927 0.87514202 1.50700416

H 14.34534690 -0.87507408 1.50703702

H 15.28890801 0.87364855 -0.81450374

H 15.28891279 -0.87363962 -0.81448336

H 16.92528382 0.88213845 1.10272588

H 16.92528772 -0.88207862 1.10274460

H 17.56516029 0.00001636 -0.28244022

symmetry c1

end

constraints

spring bond 13 14 3.0 2.57

end

basis noprint

C library "6-311++g(3df,3pd)"

N library "6-311++g(3df,3pd)"

F library "6-311++g(3df,3pd)"

H library "6-311++g(3df,3pd)"

end

set geometry system

charge -1

dft

XC pbe0

iterations 3000

convergence energy 4e-11

convergence density 4e-10

convergence gradient 3e-9

mult 2

vectors input atomic output ae_12_NF_pbe0_6311pdf_relax.movecs

sym off

adapt off

direct

noio

end

driver

maxiter 3000

end

task dft optimize

Edoardo Aprà

unread,

Oct 15, 2020, 4:54:14 PM10/15/20

to NWChem Forum

If you move the line

set geometry system

before the constraint section, the error goes away

...

end

set geometry system

Agostino Migliore

unread,

Oct 15, 2020, 6:57:50 PM10/15/20

to nwchem...@googlegroups.com

Many thanks! So, I see that this changed compared to previous versions (where I was successfully using the kind of file I sent).

Best regards,

Agostino

--
You received this message because you are subscribed to a topic in the Google Groups "NWChem Forum" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/nwchem-forum/DVRiEY3A5zw/unsubscribe.
To unsubscribe from this group and all its topics, send an email to nwchem-forum...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/nwchem-forum/e5dd2707-612b-491a-8081-469a85b79d33o%40googlegroups.com.

Edoardo Aprà

unread,

Oct 15, 2020, 7:11:11 PM10/15/20

to NWChem Forum

You can blame me for this.

It's a code change to use dynamical allocation of objects in the constraints module (see more details below).

Constraint needs to be applied to a specific geometry object ... what's wrong with this?

https://github.com/nwchemgit/nwchem/commit/2e230bf7a21f3ca210876655ea8f5e21f5af234a#diff-e2afeb58a952dfef9d97240e288f6e1ad19b02bf4746ffc94111e8ba7148ea8d

The old code was breaking when the number of fixed atoms is greater than 100

https://github.com/nwchemgit/nwchem/issues/39

https://nwchemgit.github.io/Special_AWCforum/st/id2650

Agostino Migliore

unread,

Oct 15, 2020, 7:46:31 PM10/15/20

to nwchem...@googlegroups.com

I understand the issue. Fortunately it did not occur in my calculations using previous code versions, since I never fixed so many atoms in my simulations.

Many thanks for the clarification.

Best,

Agostino

--

You received this message because you are subscribed to a topic in the Google Groups "NWChem Forum" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/nwchem-forum/DVRiEY3A5zw/unsubscribe.
To unsubscribe from this group and all its topics, send an email to nwchem-forum...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/nwchem-forum/edf8fd58-2c64-428a-9438-b26a31c12e1fo%40googlegroups.com.

Reply all

Reply to author

Forward