[USPEX 10.3] Error in Variable Composition (301)

[Carlo Nervi]

Dear all,

I'm facing a problem running Variable composition (calculationType 301).

I'm calculating alloy with 3 metals, but with the following composition:

% numSpecies

1 0 1

0 1 1

% EndNumSpecies

3 : minAt

9 : maxAt

The calculation starts normally and perform completely the first generation. The following error appears after the first generation has been calculated:

status = Local optimisation finished

Calculating fitness in 0.044507 sec

Opps: NaN appears in coordinates/lattice!

Coordinates:

NaN   NaN   NaN

NaN   NaN   NaN

NaN   NaN   NaN

NaN   NaN   NaN

Lattice:

NaN 0.000 0.000

NaN   NaN 0.000

NaN   NaN   NaN

Opps: NaN appears in coordinates/lattice!

Coordinates:

NaN   NaN   NaN

...

... huge list like this...

...

Lattice:

NaN 0.000 0.000

NaN   NaN 0.000

NaN   NaN   NaN

Structure 3 generated by heredity

Structure 4 built with the symmetry group 12 (A2/m) , composition 1  3  4

Structure 5 built with the topology 1490 , composition 4  0  4

/bin/bash: synclient: comando non trovato

*********************************************************

*                                                       *

  _|    _|     _|_|_|   _|_|_|     _|_|_|_|   _|      _|

  _|    _|   _|         _|    _|   _|           _|  _|

  _|    _|     _|_|     _|_|_|     _|_|_|         _|

  _|    _|         _|   _|         _|           _|  _|

    _|_|     _|_|_|     _|         _|_|_|_|   _|      _|

*                                                       *

** USPEX v10.3                           Oganov's Lab! **

*********************************************************

still_runing is present, USPEX has to exit, please see NOT_YET file for details

I repeated two times and I got the same error.

Any ideas?

Thank you,

Carlo

[Nursultan Sagatov]

Dear Carlo,

I have updated USPEX from 9.4.4 to 10.3 and now I am facing with the same problem. Did you solve it? If so, could you tell me how?

Regards,

Nursultan

[Carlo Nervi]

One of the problem was subtle:

you have to set (bash shell):

export LC_ALL=C.UTF-8

export LANG=C.UTF-8

With international settings (i.e.  , instead of . for example) it seems to me that many problems are introduced.

In this way it seems ok...

[Nursultan Sagatov]

Thank you for your reply. I will try it.

[Adam Carlsson]

Dear USPEX users, 

I have the same error in USPEX 10.4 and I have tried to follow the advice of Carlo and set 

export LC_ALL=C.UTF-8

export LANG=C.UTF-8

However, I am still generating mostly NAN structures when using the heredity variational operator. 
I would appreciate any comments or guidance in this regard. 

Thank you, 
Adam 

[Carlo Nervi]

First I would suggest to update to USPEX 10.4.1

Then please check:

python version: python -V

installed general python programs: pip list

user-installed: pip list --user

You should have all the required programs, like ase, numpy, scipy, matplotlib, spglib, pysqlite3,  Cython

(I hope I listed all, but please check the USPEX installation manual/README).

The settings may vary depending on the machine and version you are working on. Different versions of Unix and libraries may require slightly different settings/program versions.

then:

 export LC_ALL=C.UTF-8

 export LANG=C.UTF-8

I usually include these setting also in the script that will run VASP code.

Concerning VASP, 

 whether there are many environmental settings that I found could influence the way in which USPEX is running.

Sometimes the libstdc++.so.6 contained in the v91/sys/os/glnxa64 directory are creating problems.

You may try to move the library in a non-accessible directory, so that the corresponding library of the current machine will be used.

Sometimes also the libgcc_s.so.1 libgfortran.so.3 and libquadmath.so.0 may interfere as well.

You will have to try and test your local machine.

However, the NaN error should be mostly related to the impossibility for python to correctly read numbers from the output, and this is strictly related to the local settings of the unix machine.

At least this is my personal experience.

HTH,

Carlo

--
You received this message because you are subscribed to a topic in the Google Groups "USPEX" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/uspex/SNFO_LiDdmU/unsubscribe.
To unsubscribe from this group and all its topics, send an email to uspex+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/uspex/64733d96-35cb-4876-a5d3-9f02b745e20dn%40googlegroups.com.

--

------------------------------------------------------------
Prof. Carlo Nervi carlo...@unito.it  Tel:+39 0116707507/8
Fax: +39 0116707855      -      Dipartimento di Chimica, via
P. Giuria 7, 10125 Torino, Italy.    http://lem.ch.unito.it/

ICCC2020 has been postponed at 2022
ICCC 2022 28 August - 2 September 2022, Rimini, Italy: http://www.iccc2020.com
International Conference on Coordination Chemistry (ICCC 2022)