[yt-dev] Gadget status

Matthew Turk matthewturk at gmail.com
Wed Jul 16 08:27:49 PDT 2014 

I think it should be a quick fix, it just requires a bit of time.  In
theory we don't need to do any splitting of the octree at all -- it
just needs to construct an ordering of the octree leafs.  Sam, what do
you think?

On Wed, Jul 16, 2014 at 10:56 AM, Chris Malone <chris.m.malone at gmail.com> wrote:
> Indeed, we were seeing several of the 'Failed to split.' messages.  I don't
> know enough about the kD-tree stuff to know whether or not that was a
> game-breaker.  It didn't throw an error during split_grid, so I assumed it
> was still going about its business.
>
>
> On Wed, Jul 16, 2014 at 5:41 AM, Matthew Turk <matthewturk at gmail.com> wrote:
>>
>> On Tue, Jul 15, 2014 at 6:18 PM, Sam Skillman <samskillman at gmail.com>
>> wrote:
>> > Hi all,
>> >
>> > VR for particle-based simulations doesn't currently work, but I'm not
>> > entirely sure why. At some point I thought we could render the deposited
>> > fields but I haven't checked/tried in a while.
>>
>> My recollection is that it's unrelated to the camera interface, but is
>> instead something with the kD-tree trying to split octs
>> (unsuccessfully), which affects RAMSES and ART data as well.
>>
>> >
>> > As Cameron mentioned, we're moving towards a new (and hopefully better!)
>> > camera system (albeit slowly due to me being away from this for a
>> > while),
>> > but the timescale on that is not set in stone. I expect the new camera
>> > interface will first be implemented for grid-based datasets. Then we can
>> > hopefully have others that are more familiar with rendering options for
>> > particle datasets help implement new methods that are beyond just
>> > particle
>> > splatting.
>> >
>> > Cheers,
>> > Sam
>> >
>> >
>> >
>> >
>> > On Tue, Jul 15, 2014 at 3:16 PM, Chris Malone <chris.m.malone at gmail.com>
>> > wrote:
>> >>
>> >> Ok, thanks guys.  I figured that was the case with VR, but wanted to
>> >> double check.
>> >>
>> >> The field docs will be a great boon; as it is, things are quite
>> >> confusing.
>> >>
>> >> Nice talk Nathan!
>> >>
>> >> Chris
>> >>
>> >>
>> >> On Tue, Jul 15, 2014 at 4:06 PM, Cameron Hummels <chummels at gmail.com>
>> >> wrote:
>> >>>
>> >>> I don't think VR is working for particle-based datasets as of yet.
>> >>> The
>> >>> volume rendering interface is being redesigned right now, but its
>> >>> release
>> >>> has been postponed until yt 3.1 unfortunately.  When it is released it
>> >>> will
>> >>> handle all particle-based datasets as well as grid-based datasets.
>> >>>
>> >>> I know Matt is working on the field docs right now, so they should be
>> >>> ready somewhat soon.
>> >>>
>> >>>
>> >>> On Tue, Jul 15, 2014 at 3:00 PM, Nathan Goldbaum
>> >>> <nathan12343 at gmail.com>
>> >>> wrote:
>> >>>>
>> >>>>
>> >>>>
>> >>>>
>> >>>> On Tue, Jul 15, 2014 at 2:56 PM, Chris Malone
>> >>>> <chris.m.malone at gmail.com>
>> >>>> wrote:
>> >>>>>
>> >>>>> Thanks for the feedback, Nathan.
>> >>>>>
>> >>>>> In slices, I was plotting 'all_density', which is a 'deposit' field.
>> >>>>> I'll try looking at 'Gas_smoothed_density' ...
>> >>>>>
>> >>>>
>> >>>> ('deposit', 'all_density') uses nearest neighbor deposition.  There's
>> >>>> also ('deposit', 'all_cic') which does cloud-in-cell.
>> >>>>
>> >>>> ('deposit', 'Gas_smoothed_density') is aliased to ('gas', 'density')
>> >>>> or
>> >>>> just 'density'.  This way we can access gas fields in an SPH dataset
>> >>>> just
>> >>>> like with a grid dataset.
>> >>>>
>> >>>>>
>> >>>>> I'll see about getting permission to pass along the dataset.  For
>> >>>>> the
>> >>>>> time being, do you know of any notebook or cookbook recipe that
>> >>>>> shows how to
>> >>>>> do this for particle data?  I couldn't seem to find anything.
>> >>>>
>> >>>>
>> >>>> Yes, there's an example in the docs using tipsy data:
>> >>>>
>> >>>> http://yt-project.org/docs/dev-3.0/examining/loading_data.html#tipsy-data
>> >>>>
>> >>>> I also went into a bit of depth about working with SPH data in my
>> >>>> scipy
>> >>>> talk: www.youtube.com/watch?v=sNkN7nyj4nE
>> >>>>
>> >>>> I don't think there's an example of volume rendering SPH data
>> >>>> floating
>> >>>> around, thus my uncertainty about it working or not.
>> >>>>
>> >>>>>
>> >>>>>
>> >>>>>
>> >>>>> Chris
>> >>>>>
>> >>>>>
>> >>>>> On Tue, Jul 15, 2014 at 3:43 PM, Nathan Goldbaum
>> >>>>> <nathan12343 at gmail.com> wrote:
>> >>>>>>
>> >>>>>>
>> >>>>>>
>> >>>>>>
>> >>>>>> On Tue, Jul 15, 2014 at 2:35 PM, Chris Malone
>> >>>>>> <chris.m.malone at gmail.com> wrote:
>> >>>>>>>
>> >>>>>>> Hi all,
>> >>>>>>>
>> >>>>>>> Caveat: I'm not that familiar with particle-based datasets/codes.
>> >>>>>>>
>> >>>>>>> I'm unsuccessfully trying to help a summer student do some
>> >>>>>>> visualization of someone else's Gadget binary data dump.  What is
>> >>>>>>> the status
>> >>>>>>> of volume rendering SPH/Gadget data in 3.0?
>> >>>>>>
>> >>>>>>
>> >>>>>> Not sure about volume rendering.  I believe it should work for
>> >>>>>> particle/octree data but haven't tried it recently.
>> >>>>>>
>> >>>>>>>
>> >>>>>>>
>> >>>>>>> When I've tried what I know works for grid-based data, the
>> >>>>>>> returned
>> >>>>>>> image is simply empty.  Projections look okay (for some
>> >>>>>>> variables), and
>> >>>>>>> Slices look worse as there are blank splotches, likely due to
>> >>>>>>> emptiness in
>> >>>>>>> the tree.
>> >>>>>>>
>> >>>>>>
>> >>>>>> What field are you slicing?  There shouldn't be empty spots for SPH
>> >>>>>> smoothed fields.
>> >>>>>>
>> >>>>>>>
>> >>>>>>> Also, is there a list somewhere that describes/defines what the
>> >>>>>>> various fluid types mean (e.g. 'all', 'gas', 'deposit', etc)?  If
>> >>>>>>> not, it
>> >>>>>>> might be nice to have somewhere in the docs.
>> >>>>>>>
>> >>>>>>
>> >>>>>> There are several open issues about this. We won't be releasing
>> >>>>>> yt-3.0
>> >>>>>> until these docs are written.
>> >>>>>>
>> >>>>>> 'all' is the union that contains all particle data.
>> >>>>>>
>> >>>>>> 'gas' (lowercase g) corresponds to SPH smoothed quantities for
>> >>>>>> Gadget.
>> >>>>>> In binary datasets, the Gas particle type corresponds to the gas
>> >>>>>> particles.
>> >>>>>>
>> >>>>>> 'deposit' corresponds to fields that are deposited onto the octree
>> >>>>>> mesh using a particle type. The SPH smoothed fields are actually
>> >>>>>> defined
>> >>>>>> here, 'gas' is just an alias to the smoothed deposition fields.
>> >>>>>>
>> >>>>>> If you have more specific questions (even better if you have an
>> >>>>>> example script and a dataset) one of us should be able to help you
>> >>>>>> out more.
>> >>>>>>
>> >>>>>> -Nathan
>> >>>>>>
>> >>>>>>>
>> >>>>>>> Chris
>> >>>>>>>
>> >>>>>>> _______________________________________________
>> >>>>>>> yt-dev mailing list
>> >>>>>>> yt-dev at lists.spacepope.org
>> >>>>>>> http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
>> >>>>>>>
>> >>>>>>
>> >>>>>>
>> >>>>>> _______________________________________________
>> >>>>>> yt-dev mailing list
>> >>>>>> yt-dev at lists.spacepope.org
>> >>>>>> http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
>> >>>>>>
>> >>>>>
>> >>>>>
>> >>>>> _______________________________________________
>> >>>>> yt-dev mailing list
>> >>>>> yt-dev at lists.spacepope.org
>> >>>>> http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
>> >>>>>
>> >>>>
>> >>>>
>> >>>> _______________________________________________
>> >>>> yt-dev mailing list
>> >>>> yt-dev at lists.spacepope.org
>> >>>> http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
>> >>>>
>> >>>
>> >>>
>> >>>
>> >>> --
>> >>> Cameron Hummels
>> >>> Postdoctoral Researcher
>> >>> Steward Observatory
>> >>> University of Arizona
>> >>> http://chummels.org
>> >>>
>> >>> _______________________________________________
>> >>> yt-dev mailing list
>> >>> yt-dev at lists.spacepope.org
>> >>> http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
>> >>>
>> >>
>> >>
>> >> _______________________________________________
>> >> yt-dev mailing list
>> >> yt-dev at lists.spacepope.org
>> >> http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
>> >>
>> >
>> >
>> > _______________________________________________
>> > yt-dev mailing list
>> > yt-dev at lists.spacepope.org
>> > http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
>> >
>> _______________________________________________
>> yt-dev mailing list
>> yt-dev at lists.spacepope.org
>> http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
>
>
>
> _______________________________________________
> yt-dev mailing list
> yt-dev at lists.spacepope.org
> http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
>

More information about the yt-dev mailing list