Hi JC, Right now, there are two reasons to consider using nek-analyze: 1) You need to do something complicated and would rather do that in python (due to experience, library support, etc) 2) Your data is big enough that you want to avoid loading it all into memory If neither of those applies, then you're better off userchk post-processing within Nek. If you're interested mostly in (2), I can try to help with the python. If you're interested in (1) but not so much (2), you can probably cannibalize some of my tools (the file reader, for example). Max On Tue, Sep 29, 2015 at 1:39 AM <[email protected]> wrote:

Hej,

Thanks a lot! I will take a look at your github. I am not an extremely efficient python coder, but I am willing to learn more about this language and its capabilities and use it as often as possible for my post-processing of DNS data. Maybe we could continue this discussion out of the mailing list. Here is my email : [email protected]

++ JC

On 09/27/2015 02:21 PM, [email protected] wrote:

I post-process in python, using a MapReduce-inspired framework called nek-analyze: https://github.com/maxhutch/nek-analyze I haven't used it to write new *.f* files in a while (instead, I visualize directly in python with matplotlib), so that's probably currently broken, but I'd be willing to fix it. Depending on the mesh geometry, there is some built-in functionality for applying operators, but it is not complete. Again, I'd be willing to put some time into supporting new functionality, and VERY happy to accept contributions. GPL, just like Nek.

Max

_______________________________________________ Nek5000-users mailing [email protected]https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users

_______________________________________________ Nek5000-users mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users