[Amr] paraview-yt halo finding

Matthew Turk matthewturk at gmail.com
Wed Jun 22 20:00:10 EDT 2011


Hi Jorge,

Sorry for the delay in replying!  I am happy you fixed it.  I think
that down the road when we move to a proper ParaView interface (once
we have finished prototyping with the "Stream" interface) we can make
sure that things like this don't happen.

Thanks for all your hard work, and sorry that I didn't catch this sooner.

-Matt

On Wed, Jun 22, 2011 at 12:02 PM, Jorge Poco <jorge.poco at kitware.com> wrote:
> Hi Matt
> I found the problem.
> pf = stream.StreamStaticOutput(handler)
> pf.h
> pf.field_info['Density'].take_log = True
> pf.field_info['Density'].take_log was False and should be True, because
> Transfer function was
> working in log scale.
>
> Regards
> On Wed, Jun 22, 2011 at 12:22 PM, Matthew Turk <matthewturk at gmail.com>
> wrote:
>>
>> Hi Jorge,
>>
>> Sure -- I'm going to be away from my desk for a little while, but I
>> will get back to you this afternoon (pacific time, unfortunately).
>>
>> -Matt
>>
>> On Wed, Jun 22, 2011 at 9:21 AM, Jorge Poco <jorge.poco at kitware.com>
>> wrote:
>> > Hi Matt
>> >
>> > I have tested the volume rendering using your first script to convert
>> > from
>> > Enzo loader to StreamHandler
>> > and it's giving me bad results. Could you check the results with that
>> > script, please?
>> >
>> > Regards
>> >
>> > On Wed, Jun 22, 2011 at 11:35 AM, Jorge Poco <jorge.poco at kitware.com>
>> > wrote:
>> >>
>> >> Hi Matt
>> >>  images are attached.
>> >>
>> >> On Wed, Jun 22, 2011 at 11:24 AM, Matthew Turk <matthewturk at gmail.com>
>> >> wrote:
>> >>>
>> >>> Hi Jorge,
>> >>>
>> >>> It's not entirely clear to me.  I think what you have done should be
>> >>> immune to a difference in multiplicative factor.  Would it be possible
>> >>> to see the image that comes out of PV and compare to the image
>> >>> directly out of yt?
>> >>>
>> >>> -Matt
>> >>>
>> >>> On Wed, Jun 22, 2011 at 8:07 AM, Jorge Poco <jorge.poco at kitware.com>
>> >>> wrote:
>> >>> > Hi Matt
>> >>> > I have until now:
>> >>> > * Projection plugin: it shows the image in paraview.
>> >>> > * Slice plugin: it's also an image, but it' placed inside the
>> >>> > volume,
>> >>> > so you
>> >>> > can see the image as a common slice in 3D.
>> >>> > * Halo finder: halos information are recovered from YT and
>> >>> > represented
>> >>> > in
>> >>> > Paraview using spheres.
>> >>> > Now, I'm trying to run a simple volume rendering, and display the
>> >>> > image
>> >>> > in
>> >>> > paraview. It runs. but the image is not
>> >>> > the same as loading data directly with yt. I think the problem is
>> >>> > the
>> >>> > conversion factor. because using the
>> >>> > StreamHandler, the data is saved without conversion.
>> >>> > My script is:
>> >>> > pf = stream.StreamStaticOutput(handler)
>> >>> > L = [0.5, 0.2, 0.7]
>> >>> > W = 1.0
>> >>> > c = [0.5, 0.5, 0.5]
>> >>> > Nvec = 512
>> >>> > dd = pf.h.all_data()
>> >>> > mi, ma = dd.quantities["Extrema"]("Density")[0]
>> >>> > tf = ColorTransferFunction((na.log10(mi), na.log10(ma)))
>> >>> > tf.add_layers(6, w=0.02)
>> >>> > vp = pf.h.volume_rendering(L, W, c, Nvec, tf, whole_box=True)
>> >>> > vp.ray_cast()
>> >>> > What would be the best way to solve it?
>> >>> >
>> >>> > Regards
>> >>> >
>> >>> >
>> >>> > Halo Finder is working with paraview, halos are represented in
>> >>> > paraview
>> >>> > using spheres.
>> >>> >
>> >>> > On Fri, Jun 17, 2011 at 1:39 PM, Matthew Turk
>> >>> > <matthewturk at gmail.com>
>> >>> > wrote:
>> >>> >>
>> >>> >> Hi Jorge,
>> >>> >>
>> >>> >> On Fri, Jun 17, 2011 at 9:51 AM, Jorge Poco
>> >>> >> <jorge.poco at kitware.com>
>> >>> >> wrote:
>> >>> >> > Hi Matt
>> >>> >> > I will try to use HaloFinder from YT, the script would be:
>> >>> >>
>> >>> >> Wow, cool!
>> >>> >>
>> >>> >> > # create Paraview Handler
>> >>> >> > pf = stream.StreamStaticOutput(handler)
>> >>> >> > # halos will be added to this polydata
>> >>> >> > spheres = vtkAppendPolyData()
>> >>> >> > halos = HaloFinder(pf)
>> >>> >> > for halo in halos:
>> >>> >> >     # create a vtkSphere using halo information
>> >>> >> >     s = vtkSphereSource()
>> >>> >> >     s.SetCenter(halo.get_sphere().center)
>> >>> >> >     s.SetRadius(halo.get_sphere().radius)
>> >>> >> >     s.Update()
>> >>> >> >     spheres.AddInput(s.GetOutput())
>> >>> >> > spheres.Update()
>> >>> >> >
>> >>> >> > output.SetPoints(s1.GetOutput().GetPoints())
>> >>> >> >
>> >>> >> > output.SetPolys(s1.GetOutput().GetPolys())
>> >>> >> >
>> >>> >> > Using this script we will be able to recover the halos as a set
>> >>> >> > of
>> >>> >> > spheres
>> >>> >> > which
>> >>> >> >
>> >>> >> > could be visualized using paraview.
>> >>> >> >
>> >>> >> > Questions:
>> >>> >> >
>> >>> >> > * HaloFinder needs the particle information right?
>> >>> >>
>> >>> >> Yup, that's right.
>> >>> >>
>> >>> >> >
>> >>> >> > * Current stream.StreamHandler support the particle data?
>> >>> >>
>> >>> >> Yup, it should, although in the future we can make this work more
>> >>> >> seamlessly.  The easiest way to supply this would be to push them
>> >>> >> through the stream IO handler, and supply them as (1D) fields
>> >>> >> called
>> >>> >> "particle_mass", "particle_position_[xyz]",
>> >>> >> "particle_velocity_[xyz]".
>> >>> >>  I think this should work just fine, but there may be other
>> >>> >> hangups.
>> >>> >> If you run into any problems, let me know.
>> >>> >>
>> >>> >> Best,
>> >>> >>
>> >>> >> -Matt
>> >>> >>
>> >>> >> >
>> >>> >> > Regards,
>> >>> >> >
>> >>> >> > Att. Jorge Poco
>> >>> >> >
>> >>> >> > _______________________________________________
>> >>> >> > Amr mailing list
>> >>> >> > Amr at public.kitware.com
>> >>> >> > http://public.kitware.com/cgi-bin/mailman/listinfo/amr
>> >>> >> >
>> >>> >> >
>> >>> >
>> >>> >
>> >>
>> >
>> >
>
>


More information about the Amr mailing list