On Thu, Jan 1, 2015 at 10:00 PM, Kyle Lawlor <[email protected]> wrote:
> To start the error crops up running the following lines of code:
>
> from fipy import *
> file_path="C:\path\square.msh"
> mesh=Gmsh2D(file_path)
>
> The error I get is:
>
> line 3, in <module>
> mesh=Gmsh2D(file_path)
> File "C:\Python27\lib\site-packages\fipy\meshes\gmshMesh.py", line 1578,
> in __init__
> background=background)
> File "C:\Python27\lib\site-packages\fipy\meshes\gmshMesh.py", line 151, in
> openMSHFile
> raise EnvironmentError("Gmsh version must be >= 2.0.")
> EnvironmentError: Gmsh version must be >= 2.0.
This often pops up in the context of parallel simulations when the
environment isn't the same on all PIDs, but also for other reasons
that I can't recall.
> As far as the trouble shooting I've done thus far, I opened up the
> gmshMesh.py file to check out where the errors were coming from. I copy and
> paste up to the openMSHFile() definition in that code, and run it to see if
> I could get some info about what the code was getting in terms of
> recognizing gmsh and what version it was getting. As a note I have gmsh 2.8
> installed.
>
> Oddly enough, when I run the segment of code from the gmshMesh.py I find
> that that:
> _checkForGmsh() -- True
> _gmshVersion() -- 2.8
>
> I'm puzzled as to why this works when I run the segment of code but not when
> I call on the Gmsh2D() function. I think that likely, it is how I have gmsh
> installed on my computer. But the code seems to recognize it. Is there a
> specific location that I should have the gmsh program? Hopefully, this is an
> easy fix and I have complicated things. Looking forward to hearing back.
No idea why there is a discrepancy and it's difficult to diagnose
remotely. I think you'll just have to debug the broken version. Start
here
https://github.com/usnistgov/fipy/blob/develop/fipy/meshes/gmshMesh.py#L149
and find out what "gmshVersion" is. I suspect it's "None" or something
like that and Gmsh is not even being found. I often use the
Interactive Python Debugger for this sort of thing and it's a great
help.
Cheers,
Daniel
--
Daniel Wheeler
_______________________________________________
fipy mailing list
[email protected]
http://www.ctcms.nist.gov/fipy
[ NIST internal ONLY: https://email.nist.gov/mailman/listinfo/fipy ]
