[yt-dev] Issues with accessing ActiveDimensions attribute?

[Nathan Goldbaum]

On Fri, May 20, 2016 at 1:36 PM, Rasmi Elasmar <re2300 at columbia.edu> wrote:

> Hi all,
>
> I'm trying to create two derived fields (one for stellar density, and one
> for dark matter density) based on
> <https://bitbucket.org/yt_analysis/yt/src/cee2a160c5351c376afea65bea3462792406154e/yt/frontends/enzo/fields.py?at=yt-2.x&fileviewer=file-view-default#fields.py-350>
> these
> <https://bitbucket.org/yt_analysis/yt/src/cee2a160c5351c376afea65bea3462792406154e/yt/frontends/enzo/fields.py?at=yt-2.x&fileviewer=file-view-default#fields.py-394>
> from yt2. I'm running yt 3.2.3, and the star_density and dm_density fields
> seem to have been removed. So far, the yt2 code has transferred fine, with
> minimal changes:
>
>    - amrutils.CICDeposit_3() -> amrutils.CICDeposit.CICDeposit_3()
>    - np.float64(data['dx']) -> np.float64(np.ravel(data['dx'])[0]
>       - (data['dx'] appears to be a multidimensional array of all the
>       same values of dx, so I'm just taking the first one -- is there a better
>       way to do this?)
>
> Besides those two quirks, the code runs (I'm able to add a field using
> ds.add_field), but I get an error when I try to access the data using
> ds.all_data()['star_density'] (it seems now is when the field is actually
> computed).
>
> I get the following error:
>
>> <ipython-input-5-1af8fbcb718a> in _spdensity(field, data)
>>       1 def _spdensity(field, data):
>> ----> 2     blank = np.zeros(data.ActiveDimensions, dtype='float64')
>>       3     if data['particle_position_x'].size == 0: return blank
>>       4     filter = data['creation_time'] > 0.0
>>       5     if not filter.any(): return blank
>> AttributeError: 'YTRegion' object has no attribute 'ActiveDimensions'
>
>
> All the examples of ActiveDimensions that I saw in the docs and source
> accessed ActiveDimensions as a property of some grid (i.e.
> g.ActiveDimensions). Is there a simple way to access that from the
> data/YTRegion object that I'm missing? I can access ds.index.grids and take
> the first grid there, but ActiveDimensions differs across the grids (And
> when I try to do this from within _spdensity, data doesn't have an
> index.grids object, just fake_index (which can't access the grids)). Is
> there something I'm missing?
>
>
> Thanks,
>
> Rasmi
>

Hi Rasmi,

Rather than directly porting your scripts from yt-2, which will be a bit
painful, it would be much easier to use built-in features of yt-3 to
accomplish this. In particular, you want to use particle filters and
deposited fields.

Here is a worked example where I construct "star" and "dark matter"
particle filters for an Enzo dataset, then plot the deposited density of
these two particle types.

https://gist.github.com/b13bd7a545dda6819972eaaaf81e83f8

Here's some more documentation on particle filters and deposited particle
fields:

http://yt-project.org/docs/dev/analyzing/filtering.html#filtering-particle-fields
http://yt-project.org/docs/dev/analyzing/fields.html#deposited-particle-fields

Hope that's helpful,

Nathan


>
> _______________________________________________
> yt-dev mailing list
> yt-dev at lists.spacepope.org
> http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.spacepope.org/pipermail/yt-dev-spacepope.org/attachments/20160520/1ef24a61/attachment.html>