[yt-users] Deposited Particle Data

Jason Galyardt jason.galyardt at gmail.com
Mon May 11 12:36:02 PDT 2015


It seems that the problem is in the evaluation of 'end', which gets set to
0. I don't know how the  ds.index._particle_indices container (which p_ind
is set to) is supposed to look, but it's got an awful lot of zeros packed
in front. In fact, of the 8017 elements, 5655 of them are exactly 0.

Looking at line 76 with the debugger, p_ind[g2.id - g2._id_offset + 1] =
p_ind[1000 - 1 + 1] = p_ind[1000] = 0.

Is this normal? Doesn't make sense to me, since there are 9000+ particles
in this time step.

Jason

On Mon, May 11, 2015 at 1:17 PM, Nathan Goldbaum <nathan12343 at gmail.com>
wrote:

> If I had access to the dataset, I would put a breakpoint before line 76 in
> your traceback here:
>
> /home/jeg/yt/yt/frontends/flash/io.pyc in _read_particle_coords(self,
> chunks, ptf)
>      74                 start = p_ind[g1.id - g1._id_offset]
>      75                 end = p_ind[g2.id - g2._id_offset + 1]
> ---> 76                 x = np.asarray(p_fields[start:end, px],
> dtype="=f8")
>      77                 y = np.asarray(p_fields[start:end, py],
> dtype="=f8")
>      78                 z = np.asarray(p_fields[start:end, pz],
> dtype="=f8")
>
> I would try to print out start and end, as well as g1.id, g1._id_offset,
> g2.id,and g2._id_offset.
>
> This is the last frame before we leave yt and enter h5py.  For some reason
> start = end here, and we'd need to look at how the dataset is being read in
> to figure out why.
>
> On Mon, May 11, 2015 at 6:38 AM, Jason Galyardt <jason.galyardt at gmail.com>
> wrote:
>
>> Hi all,
>>
>> Per Matt's suggestion, I've tried accessing the FLASH particles via
>> ['all', 'particle_mass'], to no avail. The traceback can be found here:
>>
>> http://paste.yt-project.org/show/5541/
>>
>> I get the same zero-length selection exception when I use ['io',
>> 'particle_position_x'] or ['all', 'particle_position_x'], as Yuan suggests.
>>
>> To further check the input particle file format, I used the low-level
>> h5py interface. I was able to read both particle position and mass. This
>> script doesn't know about the grid, however.
>>
>> Any other ideas?
>>
>> Thanks,
>> Jason
>>
>> On Sat, May 9, 2015 at 9:59 PM, Yuan Li <bear0980 at gmail.com> wrote:
>>
>>> I tried ['io', 'particle_position_x'] on my Flash data, and it seemed to
>>> work just fine. I use the latest yt.
>>>
>>> Yuan
>>>
>>> On Sat, May 9, 2015 at 7:51 PM, Nathan Goldbaum <nathan12343 at gmail.com>
>>> wrote:
>>>
>>>> I'm not able to reproduce this using the Orbit dataset on
>>>> yt-project.org/data. Unfortunately I don't think we have any other
>>>> flash test datasets that use particles.
>>>>
>>>> It would also help to see the full error traceback you're getting.
>>>>
>>>> On Sat, May 9, 2015 at 4:32 PM, Matthew Turk <matthewturk at gmail.com>
>>>> wrote:
>>>>
>>>>> Hi Jason,
>>>>>
>>>>> Don't worry about posting the data -- that's completely fine, and I
>>>>> think in general we should do all we can to reproduce without the
>>>>> dataset!  I'm wondering if perhaps there's something funny about the
>>>>> "io" particle field.  If you try "all" instead of "io", does it work
>>>>> any better?
>>>>>
>>>>> -Matt
>>>>>
>>>>> On Fri, May 8, 2015 at 1:01 PM, Jason Galyardt <
>>>>> jason.galyardt at gmail.com> wrote:
>>>>> > Hi all,
>>>>> >
>>>>> > Has anyone had success accessing deposited particle data generated
>>>>> with
>>>>> > FLASH? It appears in the dataset derived field list, but when I go
>>>>> to access
>>>>> > it (via ds.all_data()['deposit', 'io_mass']), I get an error
>>>>> concerning
>>>>> > zero-length selection. I get the same error when I try to access the
>>>>> > 'standard' particle data (from ds.field_list), such as
>>>>> ds.all_data()['io',
>>>>> > 'particle_mass'].
>>>>> >
>>>>> > Does this ring any bells? This very well could be user error, so
>>>>> please let
>>>>> > me know if that seems to be the case. I've pasted a basic script to
>>>>> > reproduce the error here:
>>>>> >
>>>>> > http://paste.yt-project.org/show/5538/
>>>>> >
>>>>> > I'd rather not post the dataset publicly, but I can send you a link,
>>>>> if
>>>>> > you're interested.
>>>>> >
>>>>> > Cheers,
>>>>> > Jason Galyardt
>>>>> >
>>>>> >
>>>>> >
>>>>> > _______________________________________________
>>>>> > 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
>>
>>
>
> _______________________________________________
> 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/20150511/588b946b/attachment.html>
-------------- next part --------------
_______________________________________________
yt-users mailing list
yt-users at lists.spacepope.org
http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org


More information about the yt-users mailing list