Stephen,
This is impressive work. What is your roadmap for making it more
generally available? Furthermore, I think when you do make it
available, we will put out a release.
-Matt
On Sat, Mar 13, 2010 at 10:08 PM, Stephen Skory
Hi all,
I have just added to the hg-devel branch a parallel (in memory and workload) structure function generator. It's very alpha in the sense that I haven't even really done any comparisons to other generators, so it may be completely wrong. It is not be trusted right now. The results I have got in testing are reasonable, so if it is wrong, it's not in a big way.
I've written some documentation for it, and there is a simple example in here:
http://hg.enzotools.org/yt-doc/file/c1f1442dc71e/source/extensions/struct_fc...
There can be multiple structure functions operating at once, and the structure functions are completely flexible. You can use defined fields in yt, and output multiple outputs, meaning you can make multidimensional structure functions. You can define density cutoffs for the structure function points as well by putting that into your function. The analysis region can be the entire volume, or a subregion you define. It uses very little blocking global MPI, so it should be fairly scalable, but I haven't done any big tests, so I really don't know.
Even though I just wrote software to do this, I still find the whole structure function stuff confusing. So if the documentation isn't clear enough, can you help me out and say what could improve it? And if you get the chance, or you need it, please try running it, and let me know how it goes. _______________________________________________________ sskory@physics.ucsd.edu o__ Stephen Skory http://physics.ucsd.edu/~sskory/ _.>/ _Graduate Student ________________________________(_)_\(_)_______________
_______________________________________________ Yt-dev mailing list Yt-dev@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org