[yt-users] Analyzing SPH data
Nathan Goldbaum
nathan12343 at gmail.com
Sat Mar 4 09:16:43 PST 2017
To test my pull request you're going to want to build yt from a clone of
the mercurial repository after pulling in my changes.
$ conda uninstall yt
$ hg clone https://bitbucket.org/yt_analysis/yt
$ cd yt
$ hg pull -r f4c5c13 https://bitbucket.org/yt_analysis/yt
$ hg update f4c5c13
$ python setup.py develop
You'll need a compilation environment set up for that to succeed. If you
don't have compilers installed it will be less straightforward to test
this. I'd urge you *not* to hand-edit your installed version of yt from
conda, as that will create more headaches for you if you make a mistake and
forget to correct it.
On Sat, Mar 4, 2017 at 10:34 AM Alankar Dutta <dutta.alankar at gmail.com>
wrote:
> Hello,
>
> I have modified the yt files as given in the pull request. I get the
> following error code:
>
> Traceback (most recent call last):
> File "dev.py", line 9, in <module>
> import yt
> File
> "/home/alankar/anaconda3/lib/python3.6/site-packages/yt/__init__.py", line
> 133, in <module>
> frontends = _frontend_container()
> File
> "/home/alankar/anaconda3/lib/python3.6/site-packages/yt/frontends/api.py",
> line 52, in __init__
> setattr(self, frontend, importlib.import_module(_mod))
> File "/home/alankar/anaconda3/lib/python3.6/importlib/__init__.py", line
> 126, in import_module
> return _bootstrap._gcd_import(name[level:], package, level)
> File
> "/home/alankar/anaconda3/lib/python3.6/site-packages/yt/frontends/eagle/api.py",
> line 17, in <module>
> from .data_structures import \
> File
> "/home/alankar/anaconda3/lib/python3.6/site-packages/yt/frontends/eagle/data_structures.py",
> line 21, in <module>
> from yt.frontends.gadget.data_structures import \
> File
> "/home/alankar/anaconda3/lib/python3.6/site-packages/yt/frontends/gadget/data_structures.py",
> line 28, in <module>
> from yt.frontends.sph.data_structures import \
> ImportError: cannot import name 'SPHDataset'
>
> The following is my code:
> import yt
> ds = yt.load("snapshot_068.0")
>
> Alankar Dutta
>
> On Sat, Mar 4, 2017 at 3:13 PM, Alankar Dutta <dutta.alankar at gmail.com>
> wrote:
>
> Oh I missed telling that the full dataset is around 950 GB
>
> On Sat, Mar 4, 2017 at 1:02 PM, Alankar Dutta <dutta.alankar at gmail.com>
> wrote:
>
> Hello,
>
> This is only one part of the multipart snapshot of the simulation. I am
> sending you the header information from the snapshot file:
>
> {'O0': 0.27500000000000002,
> 'Ol': 0.72499999999999998,
> 'boxsize': 100000.0,
> 'flag_age': 1,
> 'flag_cooling': 1,
> 'flag_delaytime': 0,
> 'flag_fb': 1,
> 'flag_fh2': 0,
> 'flag_metals': 1,
> 'flag_potential': 3,
> 'flag_sfr': 1,
> 'flag_tmax': 0,
> 'h': 0.70199999999999996,
> 'massTable': array([ 0. , 0.00110449, 0. , 0. , 0. , 0. ]),
> 'nbndry': 76247,
> 'nbulge': 0,
> 'ndisk': 0,
> 'ndm': 1459617792,
> 'nfiles': 1024,
> 'ngas': 1343721867,
> 'npartThisFile': array([10045237, 10239661, 0, 0, 451148, 67], dtype=uint32),
> 'npartTotal': array([1343721867, 1459617792, 0, 0, 503500456,
> 76247], dtype=uint32),
> 'npartTotalHW': array([1, 1, 0, 0, 0, 0], dtype=uint32),
> 'nstar': 503500456,
> 'redshift': 1.0000000010567627,
> 'rhocrit': 2.707660428120944e-29,
> 'time': 0.49999999973580939}
>
>
>
> I extracted this information using a different python program.
>
> Alankar Dutta
>
> On Sat, Mar 4, 2017 at 12:45 PM, Nathan Goldbaum <nathan12343 at gmail.com>
> wrote:
>
>
>
> On Sat, Mar 4, 2017 at 12:25 AM, Nathan Goldbaum <nathan12343 at gmail.com>
> wrote:
>
>
>
> On Sat, Mar 4, 2017 at 12:13 AM, Alankar Dutta <dutta.alankar at gmail.com>
> wrote:
>
> Hello,
>
> Modifying the GADGET simulation isn't possible the moment because it has
> already been developed by someone else who has a paper published on this
> simulation and I want to use that simulation snapshots to make mock xray
> map from it.
>
>
> I'm talking about modifying yt, not Gadget.
>
> Is the file you attached to your other e-mail just a single file in a
> multi-file dataset? How large is the full dataset? Do you not have a way to
> produce a full dataset in this output format that's not prohibitively large?
>
>
> I've opened a pull request that allows me to do IO on the data you
> attached: https://bitbucket.org/yt_analysis/yt/pull-requests/2537
>
> This allows me to read your data in, getting sensible values for e.g.
> position. I suspect we're not using the correct field specification because
> I see this warning:
>
> yt : [WARNING ] 2017-03-04 01:06:09,109 Your Gadget-2 file may have extra
> columns or different precision! (1814947576 file vs 1486279952 computed)
>
> yt supports a number of field specifications out of the box, see:
>
>
> https://bitbucket.org/yt_analysis/yt/src/3eca2ae80ab14a48b643d3055d7d3c0933fa77ae/yt/frontends/gadget/definitions.py?at=yt&fileviewer=file-view-default#definitions.py-50
>
> Do you happen to know which fields are in your output file?
>
>
>
>
>
>
> Alankar Dutta
>
> On Mar 4, 2017 11:38 AM, "Nathan Goldbaum" <nathan12343 at gmail.com> wrote:
>
> I think the most straightforward thing to do here is to fix the Gadget
> frontend so it properly reads in gadget binary data with positions written
> in double precision.
>
> Is there any chance you can generate a smallish test dataset in your
> Gadget output format that we can use for debugging purposes? With that
> available it should be straightforward to add support. You can share the
> dataset using the yt curldrop (https://docs.hub.yt/services.html#curldrop)
> or a cloud filesharing service like dropbox or google drive.
>
> Unfortunately there isn't a way to load SPH data without a full-fledged
> frontend right now.
>
> We do have a load_particles function which allows creating a dataset from
> particle data loaded as numpy arrays, but it's currently not possible to
> use it to load SPH data.
>
> I'm currently actively working on improving support for SPH data in yt and
> adding the ability to load SPH data with load_particles is one of the
> things I've added in that branch of the code. Hopefully this work will be
> stabilized sometime in the next few months.
>
>
>
> On Fri, Mar 3, 2017 at 11:45 PM, Alankar Dutta <dutta.alankar at gmail.com>
> wrote:
>
> Hello yt-community,
>
> I am a beginner on yt. I have an array of data which I have read from a
> Gadget simulation snapshot. It is not directly supported by yt at present
> (I have ensured this by already discussing this issue in the community
> before). The array has position,velocity, density, mass, internal energy
> and smoothing length information on the gas particles. Now how can I use
> this to make the slice plots or other useful visualizations?
>
> Alankar Dutta,
> Third year Undergraduate,
> Physics Department,
> Presidency University, India
>
> _______________________________________________
> yt-users mailing list
> yt-users at lists.spacepope.org
> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
>
>
>
> _______________________________________________
> yt-users mailing list
> yt-users at lists.spacepope.org
> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
>
>
> _______________________________________________
> yt-users mailing list
> yt-users at lists.spacepope.org
> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
>
>
>
>
> _______________________________________________
> yt-users mailing list
> yt-users at lists.spacepope.org
> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
>
>
>
>
> _______________________________________________
> yt-users mailing list
> yt-users at lists.spacepope.org
> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.spacepope.org/pipermail/yt-users-spacepope.org/attachments/20170304/b40de1f8/attachment.html>
More information about the yt-users
mailing list