Wait, so we'd have both an 'all' ftype and an 'n-body' ftype and the 'n-body' ftype would just include non-gas particles (ie ones without the 'smoothing_length' field)?  I'm assuming this won't add more computational load when reading in the dataset?  If that's the case, then I'm +0.5 on it.  I haven't had a need for it up to this point, but maybe other people really need it?

On Wed, Mar 29, 2017 at 2:21 PM, John ZuHone <jzuhone@gmail.com> wrote:
+1.

"n_body"?

On Mar 29, 2017, at 5:19 PM, Matthew Turk <matthewturk@gmail.com> wrote:

+1, and I think updating YTEP-0031 is sufficient.  Not sure that "n-body" specifically is my preference, since it's not tokenizable, but maybe it's fine.

On Wed, Mar 29, 2017 at 4:18 PM, Nathan Goldbaum <nathan12343@gmail.com> wrote:
Hi all,

I'd like to propose adding a new particle union that should be defined for all datasets that include particles. This came up in the context of the demeshening work (see https://bitbucket.org/yt_analysis/ytep/pull-requests/67 for more details).

Right now many of the derived quantities make a distinction between calculating results using just the gas or just the particles or both. Up until now they have calculated the results for particles using particle fields from the 'all' particle union. This makes perfect sense for AMR data but doesn't really make sense for SPH data, since it will double-count SPH particles. In fact, I think this is an issue even without the demeshening, but the demeshening makes it more starkly apparent.

I'd like to propose defining a new "n-body" particle union (suggestions for alternate names are very welcome) that will be defined for all yt datasets. This union will be identical to the 'all' particle union for AMR data and N-body particle data, but for SPH data will only include the particle types that aren't SPH particles (if any). That means the "n-body" particle type represents infinitesimal particles but not particles that have finite extents (e.g. an SPH particle's smoothing region).

I think this new particle type would probably be generically useful beyond just the derived quantities, maybe even more useful than "all". I also kind of prefer the name "n-body" to "all" since it more prominently indicates that it's associated with particle data.

Please let me know if you have thoughts or suggestions about this proposal. I'm happy to draft a YTEP or update YTEP-0031 with more details if people want to see that.

-Nathan

_______________________________________________
yt-dev mailing list
yt-dev@lists.spacepope.org
http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org


_______________________________________________
yt-dev mailing list
yt-dev@lists.spacepope.org
http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org


_______________________________________________
yt-dev mailing list
yt-dev@lists.spacepope.org
http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org




--
Cameron Hummels
NSF Postdoctoral Fellow
Department of Astronomy
California Institute of Technology
http://chummels.org