[yt-users] Profile plot of a 2d projection FRB object

David Collins dcollins4096 at gmail.com
Wed Jun 3 06:20:38 PDT 2015


>
> > Probably doable as a 2D dataset, loading the quadtree in-memory. Would
> take
> > some work.
>
> Yes, this is totally doable, but would as you note take some work.  I
> can see a few possible ways to it, all of which we should consider,
>

That would be awesome if it were possible.

For now I think John's FRB at max res. will be the way I go for me.  I'm
already looking at FRBs like that, so it's a straight forward method.

Thanks a ton for the input!

d.



> but the main obstacle is likely going to be an impedance mismatch
> between the quadtree and the octree, which were developed at different
> times in different ways.  There'll be a visitor this Summer who's
> interested in this type of thing, and I'll see if we can collaborate
> to pull it together.
>
> >
> >>
> >>
> >> In general, when I make FRBs for anything other than imaging, I make
> sure
> >> that the resolution matches the finest cell size, and that it matches
> the
> >> layout of the underlying AMR grids (e.g., no partial overlaps between
> FRB
> >> resolution elements and AMR cells). Something tells me we ought to have
> a
> >> short-hand of some kind for creating an FRB in this way.
> >
> >
> > This is basically a reduced dimensionality covering grid.  I don't think
> we
> > have such a thing yet but I think it would be a worthwhile addition, I
> could
> > see it being useful to not need to think about aligning a FRB with cell
> > boundaries.
> >
> >>
> >>
> >> On Jun 2, 2015, at 5:14 PM, Cameron Hummels <chummels at gmail.com> wrote:
> >>
> >> My script basically did this:
> >>
> >> 1) Make a projection (of a galaxy)
> >> 2) Identify a central point (ie the center of the galaxy i cared about)
> >> 3) Measure the projected distance from the central point to each pixel
> in
> >> the projection.
> >> 4) Bin these projected pixels into radial bins
> >> 5) Take the median value of different projected fields for each radial
> bin
> >> (e.g. column density, density-weighted temperature, etc.)
> >> 6) Plot up each of these fields as they vary for projected radius
> >>
> >> If I were to do it again, I'd probably rely on the new FRB interface as
> it
> >> seems cleaner.  And, it looks like it would better address what you're
> >> trying to do right now.
> >>
> >> Cameron
> >>
> >> On Tue, Jun 2, 2015 at 1:57 PM, David Collins <dcollins4096 at gmail.com>
> >> wrote:
> >>>
> >>>
> >>>
> >>>>
> >>>> Just to clarify, you want to make a phase plot of a projected object?
> >>>> So you want something like how column density varies with
> density-weighted
> >>>> temperature in a 2D plot?  Or something like this?  I only ever did a
> radial
> >>>
> >>>
> >>> That's exactly right.  I'll give the FRB a shot first.  Since the
> >>> ultimate goal is comparison to the sky, FRB is probably just fine.
> >>>
> >>> How different was your code?
> >>>
> >>>
> >>>>
> >>>> profile of projected fields (column density, or metal column density,
> >>>> etc.), but I could try to help you put something together.  The FRB
> idea
> >>>> suggested by John might be the easiest approach depending on what you
> want.
> >>>>
> >>>> On Tue, Jun 2, 2015 at 8:21 AM, David Collins <dcollins4096 at gmail.com
> >
> >>>> wrote:
> >>>>>
> >>>>> Hi, Everybody--
> >>>>>
> >>>>> This was discussed a while back, but I wanted to further clarify, and
> >>>>> possibly mooch a script or two off y'all.
> >>>>>
> >>>>> I'd like a PhasePlot from a projected region.  Earlier FRBs were
> >>>>> discussed; has anyone successfully done the same on a
> YTQuadTreeProj?  I'd
> >>>>> like to retain all the high resolution data, if possible.
> >>>>>
> >>>>> Cameron, back in Jan you mentioned you had a script.  Did you manage
> to
> >>>>> track it down, and/or would you be willing to send it my way?
> >>>>>
> >>>>> Thanks!
> >>>>> d.
> >>>>>
> >>>>>
> >>>>>
> >>>>> On Fri, Jan 16, 2015 at 10:31 PM, John ZuHone <jzuhone at gmail.com>
> >>>>> wrote:
> >>>>>>
> >>>>>> This is a bit of a long way around, but you could write the FRB to a
> >>>>>> FITS file and read it in as a dataset. You would get the coordinate
> system,
> >>>>>> units, etc., and the whole YT machinery.
> >>>>>>
> >>>>>> Use the export_fits method:
> >>>>>>
> >>>>>> my_frb.export_fits("myfile.fits", fields)
> >>>>>>
> >>>>>> Then:
> >>>>>>
> >>>>>> ds = yt.load("my_file.fits")
> >>>>>>
> >>>>>> I'm not at a computer so I don't remember the exact signature, but
> it
> >>>>>> should be in the code and you can check using
> help(my_frb.export_fits). If
> >>>>>> you're interested I can write back with more info later.
> >>>>>>
> >>>>>> The other advantage of this is that you can store the data to disk.
> >>>>>> You'll have to install the AstroPy package.
> >>>>>>
> >>>>>> Kavli Center for Astrophysics and Space Research
> >>>>>> Massachusetts Institute of Technology
> >>>>>> 77 Massachusetts Ave., 37-582G
> >>>>>> Cambridge, MA 02139
> >>>>>> (w) 617-253-2354
> >>>>>> (m) 781-708-5004
> >>>>>> jzuhone at space.mit.edu
> >>>>>> jzuhone at gmail.com
> >>>>>> http://www.jzuhone.com
> >>>>>>
> >>>>>> On Jan 16, 2015, at 8:34 PM, Ben Thompson <bthompson2090 at gmail.com>
> >>>>>> wrote:
> >>>>>>
> >>>>>> Hey guys.
> >>>>>>
> >>>>>> I have a solution together for myself involving the new
> >>>>>> particle_position_relative_[xyz] fields and multiple instances of
> >>>>>> np.histogram and np.linspace which seems to do the trick.
> >>>>>>
> >>>>>> Although I would not complain if a piece of code that would be more
> >>>>>> native to the inner workings of YT existed that made use of the FRB
> data
> >>>>>> objects :). So that would be very nice Cameron. Don't feel the need
> to rush
> >>>>>> with it though, as I made a numpy solution for myself. But I would
> be
> >>>>>> interested to see that code.
> >>>>>>
> >>>>>> Ben
> >>>>>>
> >>>>>> On 17 Jan 2015 01:27, "Britton Smith" <brittonsmith at gmail.com>
> wrote:
> >>>>>>>
> >>>>>>> Hi Ben,
> >>>>>>>
> >>>>>>> I seem to recall people on this list doing similar things in the
> past
> >>>>>>> with their own external code.  Does anyone still have their 2D
> radial
> >>>>>>> profile code around anymore?
> >>>>>>>
> >>>>>>> Would it perhaps work to create a uniform grid dataset from an FRB
> >>>>>>> array?
> >>>>>>>
> >>>>>>> Britton
> >>>>>>>
> >>>>>>> On Tue, Jan 13, 2015 at 12:49 PM, Ben Thompson
> >>>>>>> <bthompson2090 at gmail.com> wrote:
> >>>>>>>>
> >>>>>>>> Hello everyone.
> >>>>>>>>
> >>>>>>>> I was wondering if anyone has had experience with producing a
> >>>>>>>> profile plot from a 2d projection object (FRB object).
> >>>>>>>>
> >>>>>>>> Essentially, what I am trying to do is plot the stellar surface
> >>>>>>>> density of a galaxy as a function of radius.
> >>>>>>>>
> >>>>>>>> This is achieved by doing the following from a disk YT object
> called
> >>>>>>>> cylinder (in which the origional simulation object is called shot)
> >>>>>>>>
> >>>>>>>> center = cylinder.get_field_parameter("center")
> >>>>>>>> normal = cylinder.get_field_parameter("normal")
> >>>>>>>> image_width = (100,"kpc")
> >>>>>>>> three_image_width = YTArray((image_width[0], image_width[0],
> >>>>>>>> image_width[0]),image_width[1])
> >>>>>>>> left = center - image_width
> >>>>>>>> right = center + image_width
> >>>>>>>> region = shot.region(center, left, right)
> >>>>>>>>
> >>>>>>>> proj =
> >>>>>>>>
> yt.ProjectionPlot(cylinder.ds,"z",[("deposit","stars_density")],center=center,width=image_width,data_source=region,axes_unit="kpc")
> >>>>>>>>
> >>>>>>>> the error arrises here
> >>>>>>>>
> >>>>>>>> prof =
> >>>>>>>>
> yt.create_profile(proj,bin_fields="cylindrical_r",fields=[("deposit","stars_density")],n_bins=128,weight_field=None
> >>>>>>>> )
> >>>>>>>>
> >>>>>>>> where I get the error
> >>>>>>>>
> >>>>>>>>
> >>>>>>>> /gpfs/home/........./profiles.pyc in create_profile(data_source,
> >>>>>>>> bin_fields, fields, n_bins, extrema, logs, units, weight_field,
> >>>>>>>> accumulation, fractional)
> >>>>>>>>    1304     else:
> >>>>>>>>    1305         raise NotImplementedError
> >>>>>>>> -> 1306     bin_fields = data_source._determine_fields(bin_fields)
> >>>>>>>>    1307     fields = data_source._determine_fields(fields)
> >>>>>>>>    1308     if units is not None:
> >>>>>>>>
> >>>>>>>> AttributeError: 'FixedResolutionBuffer' object has no attribute
> >>>>>>>> '_determine_fields'
> >>>>>>>>
> >>>>>>>> Any ideas how to get around this error?
> >>>>>>>>
> >>>>>>>>
> >>>>>>>> Also some other things to add as a postscript. Since how the
> >>>>>>>> projection works, if I provide weights=None as a keyword argument
> within the
> >>>>>>>> ProjectionPlot object, I get a surface density (g/cm^2). But also
> a
> >>>>>>>> "cylindrical_r" in cm^2 as well. I *think* the way to get around
> this is to
> >>>>>>>> do another projection where weights="ones", get the radius values
> out of
> >>>>>>>> that profile.. and then in matplotlib, useing the surface density
> array from
> >>>>>>>> the former profile, and the radius bin array from the latter...
> Produce a
> >>>>>>>> plot of the surface density as a function of radius from those
> two arrays (I
> >>>>>>>> might check by hand afterwards to see if this does the trick).
> This seems
> >>>>>>>> kinda convoluted so I am wondering if there is an easier way than
> this.
> >>>>>>>>
> >>>>>>>>
> >>>>>>>>
> >>>>>>>> _______________________________________________
> >>>>>>>> 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
> >>>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>> --
> >>>>> -- Sent from a computer.
> >>>>>
> >>>>> _______________________________________________
> >>>>> yt-users mailing list
> >>>>> yt-users at lists.spacepope.org
> >>>>> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
> >>>>>
> >>>>
> >>>>
> >>>>
> >>>> --
> >>>> Cameron Hummels
> >>>> Postdoctoral Researcher
> >>>> Steward Observatory
> >>>> University of Arizona
> >>>> http://chummels.org
> >>>>
> >>>> _______________________________________________
> >>>> yt-users mailing list
> >>>> yt-users at lists.spacepope.org
> >>>> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
> >>>>
> >>>
> >>>
> >>>
> >>> --
> >>> -- Sent from a computer.
> >>>
> >>> _______________________________________________
> >>> yt-users mailing list
> >>> yt-users at lists.spacepope.org
> >>> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
> >>>
> >>
> >>
> >>
> >> --
> >> Cameron Hummels
> >> Postdoctoral Researcher
> >> Steward Observatory
> >> University of Arizona
> >> http://chummels.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
>



-- 
-- Sent from a computer.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.spacepope.org/pipermail/yt-users-spacepope.org/attachments/20150603/c77c80af/attachment.htm>


More information about the yt-users mailing list