[CIG-MC] tracer.dat file cannot work

tan2 tan2tan2 at gmail.com
Sun Oct 18 08:28:12 PDT 2015


On Sun, Oct 18, 2015 at 11:01 PM, 姜辉 <jianghui at mail.iggcas.ac.cn> wrote:

> If I run with tracers generated pseudo-randomly, it could run. But if I
> run with the tracer.dat file to locate the certain tracer area, it always
> abnormally ended without working any steps. Has anyone encountered this
> problem before? How to solve it?



I only have time to look at your first problem.

The error message in tracer_log file is:

WARNING(check_initial_composition)-number of tracers is low, 4799 elements
contain no tracer initially
Using neighboring elements for initial composition...
Error(fill_composition_from_neighbors)-all neighboring elements are empty

The warning message shows that 4799 elements contain no tracers, which
indicates that the total number of tracer is too small to fill the whole
domain. When an element contains no tracers, its composition is undefined.
In the code, we will take the average of the compositions of neighboring
elements as the ad hoc measure for the composition. But all neighboring
elements are empty in tracers, hence the code quits. The fix is simple: put
more tracers in your tracer.dat such that every elements will contain more
than 1 tracers.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.geodynamics.org/pipermail/cig-mc/attachments/20151018/513e41de/attachment.html>


More information about the CIG-MC mailing list