[ase-users] Test Error for fresh installation
Ask Hjorth Larsen
asklarsen at gmail.com
Tue Mar 27 17:56:04 CEST 2018
Dear Yunzhe Wang,
2018-03-26 22:14 GMT+02:00 Yunzhe Wang via ase-users
<ase-users at listserv.fysik.dtu.dk>:
>
> Dear ASE-developer,
>
>
> I just subscribed the mailing list since last email was blocked.
>
>
> The last email is about installing ASE with an old version of numpy(1.11.0) and scipy (0.15.0). I managed to use virtualenv to upgrade the version of numpy(1.14.2) and scipy(1.0.0) to the lastest ones. However, when I tried the demo python code on the tutorial for H2, the nwchem failed. The error message and the input is attached. The test result after the installation is below:
As far as I can see the segfault happened in the 37th test, which is
eam_pot_test.py. This test uses the EAM potential which in turn uses
splines from scipy. It may be this bug in scipy:
https://github.com/scipy/scipy/issues/3691
If you can stlil use the older scipy you may wish to confirm this.
Also maybe run "ase test -v" to see the names of the tests, so we will
know for sure if it was the EAM test.
>
>
> Is the failure of H2 calculation due to installation or the tutorial codes? I tried other demo code like the N2Cu.py and it works fine.
The H2 example requires NWChem, and appears to fail only because the
nwchem binary is missing. So this does not indicate a problem.
It appears that netCDF4 is not correctly installed: It cannot find the
shared library for netCDF. This will make netCDF I/O unavailable,
probably, but should not have any other consequence. I think this is
not strictly an issue with ASE, but maybe the environment
(LD_LIBRARY_PATH or whatever).
Best regards
Ask
>
>
> Best Regards,
>
> Yunzhe Wang (Phil)
>
> ________________________________
> From: Yunzhe Wang
> Sent: Monday, March 26, 2018 2:33:41 PM
> To: ase-users at listserv.fysik.dtu.dk
> Subject: Test Error for fresh installation
>
>
> Dear ASE-developer,
>
>
> I just installed the ase package. There is an error shown below:
>
>
>
> I know the versions of python and numpy are a little old. Is that the reason that the package doesn't work? The installation is done on the supercomputer of our school, which I don't have the root permission. Can you recommend some previous version of ase that might work?
>
>
> Thanks,
>
> Yunzhe Wang
>
>
> _______________________________________________
> ase-users mailing list
> ase-users at listserv.fysik.dtu.dk
> https://listserv.fysik.dtu.dk/mailman/listinfo/ase-users
More information about the ase-users
mailing list