standard example not work in Centos7

Asked by Fan

Hi Sir,

I am so sorry to bother your. Because I am a freshman in use you esys-escript program, so I have running a lot questions.

At present, I use conda installed the esys-escript in Centos7 (python2.7.13). my problems is follows:

(1) use run-escript to running, the error is "This example requires scipy's netcdf support which does not appear to be installed", However, I has installed the netcdf4, it still show the same error.

(2) use run-escript to running, the error is "line 481: 20161 Segmentation fault (core dumped) ${EXEC_CMD}".

I search a lot of measure to solve above problem on internet, but cann't solve it. So , I want to get your help.

Thank you very much.

Question information

English Edit question
esys-escript Edit question
No assignee Edit question
Solved by:
Lutz Gross
Last query:
Last reply:
Revision history for this message
Lutz Gross (l-gross) said :

(1) is still using the *scipy* netcdf3 reader (sorry!). you need to install scipy.
(2) not so sure about this one. The first thing that comes to mind is a memory overflow. The problem in the documentation is small so this should not be a problem?

Revision history for this message
Fan (fan188) said :

Hi Dr Gross,

Thank you for your reply. I will follow you advice to test and build my environment.

Recently, I read your publish title "Australia's lithospheric density field, and its isostatic equilibration"in the Geophysical Journal International. It makes me have a new understanding of lithospheric research. However, I have some questions about this paper:

(1) How to build reference model, I see you use the density and depth information from the SEEBASE, Crust1.0, AusCrust model, but I don't know use what measure to build reference model and to running inversion?

(2) I am confused about the inversion input data you use the residual Bouguer gravity data, this data weather represent the upward Bouguer data or remove the crust gravity anomaly to get the mantle Bouguer data?

Looking forward you reply.

Thanks again.


Revision history for this message
Best Lutz Gross (l-gross) said :

Try <email address hidden> - he put these data together and run all the inversions.
There are also more details on

Revision history for this message
Fan (fan188) said :

Thanks Lutz Gross, that solved my question.