New issue 927: Field parameter registry
https://bitbucket.org/yt_analysis/yt/issue/927/field-parameter-registry
Nathan Goldbaum:
I think we should have some sort of field parameter registry. Now that field parameters can have units, we need to somehow know what the units are when we mock up the field parameter during field detection.
I think the easiest way to do this would be to explicitly add user-defined field parameters to a `dataset` instance after it is loaded in memory, sort of like particle filters.
Before creating a full YTEP for this, I'm opening this issue to facilitate discussion and to check if people think this is a good idea.
New issue 926: cylindrical_radial_velocity does not look right
https://bitbucket.org/yt_analysis/yt/issue/926/cylindrical_radial_velocity-…
Suoqing Ji:
Recently I just found that ``cylindrical_radial_velocity`` in yt3 does not give me what I expect. To test this, I setup a 3-D Sedov problem with domain ``[-0.5, 0.5]^3`` and explosion center at ``[0, 0, 0]``, and plot the ``cylindrical_radial_velocity`` as the following image. According to the pattern of the velocity streamline, the cylindrical radial velocity should be positive.

I don't understand in the function of ``_cylindrical_radial`` in [yt/fields/vector_operations.py](https://bitbucket.org/yt_analysis/yt/src/10…:
```
#!python
def _cylindrical_radial(field, data):
normal = data.get_field_parameter("normal")
vectors = obtain_rv_vec(data, (xn, yn, zn),
"bulk_%s" % basename)
theta = resize_vector(data["index", 'cylindrical_theta'], vectors)
return get_cyl_r_component(vectors, theta, normal)
```
what does the line of ``resize_vector`` do. If I replace this simply with line ``theta = data['cylindrical_theta']``, it will give the image I expected as the following.

In addition, if I want to plot the same field for a 2-D Sedov dataset, the error of ``yt.utilities.exceptions.YTFieldNotFound: Could not find field '('all', 'cylindrical_radial_velocity')'`` will be raised. However, if changing the line
```
#!python
vectors = obtain_rv_vec(data, (xn, yn, zn),
"bulk_%s" % basename)
```
into
```
#!python
vectors = obtain_rv_vec(data)
```
the correct image will be generated without raising any error. But I'm not very certain what's the correct way to fix this.
Hello everyone,
Yesterday morning, we held the second yt Team Meeting as per the governance
YTEP. We had representatives from each of the 5 subcomponents of the code,
as well as several additional attendees since it was held at the same time
as the yt development workshop at NCSA. I personally think it worked
really well to have so many people in attendance, as we were able to
address issues more fully with a larger group. I would be in favor of
having future meetings open to the full member base and potentially the
full dev base.
Nathan acted as scribe and took notes on what was discussed. You can see
those here:
https://docs.google.com/document/d/1CZgBOkvO3gZQwWUzcBEaCgqynVZlvoHXEiWiEYc…
We had to finish the meeting somewhat abruptly and were unable to schedule
a follow up meeting as well as the coordinator/scribe for that meeting. I
propose that we hold the next meeting a month from now, on Friday 11/14. I
propose that Hilary Egan be the organizer, since she was interested in
hosting the last meeting, but due to a conflict was unable to do so. Are
there any problems with this?
Great job, everyone.
Cameron
--
Cameron Hummels
Postdoctoral Researcher
Steward Observatory
University of Arizona
http://chummels.org
New issue 924: Annotated annotate_grids?
https://bitbucket.org/yt_analysis/yt/issue/924/annotated-annotate_grids
Sam Skillman:
According to the documentation:
http://yt-project.org/doc/visualizing/callbacks.html#overplot-grids
One should be able to annotate the grid lines with the grid id. Looks like this functionality is gone:
```
#!python
In [2]: sl.annotate_grids(annotate=True)
---------------------------------------------------------------------------
TypeError Traceback (most recent call last)
/home/skillman/local/yt-x86_64/src/yt-3.0/yt/mods.py in <module>()
----> 1 sl.annotate_grids(annotate=True)
/home/skillman/local/yt-x86_64/src/yt-3.0/yt/visualization/plot_container.pyc in newfunc(*args, **kwargs)
68 rv = f(*args, **kwargs)
69 args[0]._plot_valid = False
---> 70 args[0]._setup_plots()
71 return rv
72 return newfunc
/home/skillman/local/yt-x86_64/src/yt-3.0/yt/visualization/plot_window.pyc in _setup_plots(self)
861 label.set_fontproperties(fp)
862
--> 863 self.run_callbacks(f)
864
865 if draw_axes is False:
/home/skillman/local/yt-x86_64/src/yt-3.0/yt/visualization/plot_container.pyc in run_callbacks(self, f)
312 cbw = CallbackWrapper(self, self.plots[f], self.frb, f)
313 CallbackMaker = callback_registry[name]
--> 314 callback = CallbackMaker(*args[1:], **kwargs)
315 callback(cbw)
316 for key in self.frb.keys():
TypeError: __init__() got an unexpected keyword argument 'annotate'
```
GridBoundaryCallback has an init like this now:
```
#!python
392 def __init__(self, alpha=0.7, min_pix=1, min_pix_ids=20, draw_ids=False,
393 periodic=True, min_level=None, max_level=None,
394 cmap='B-W LINEAR_r', edgecolors=None, linewidth=1.0):
```
Hi all,
I just submitted PR #1264 which adds a data_source parameter to all YTSelectionContainer classes (except those in construction_data_containers.py). This is apparently a long-standing desired feature request (Issue #521), so please test it and see if it works for you.
An example script that shows a slice through a sphere object can be downloaded here:
http://paste.yt-project.org/show/5167/
Douglas Rudd
Scientific Computing Consultant
Research Computing Center
drudd(a)uchicago.edu
New issue 923: to_pw should raise an error if it would produce a plot with an empty field list
https://bitbucket.org/yt_analysis/yt/issue/923/to_pw-should-raise-an-error-…
Nathan Goldbaum:
This can happen if a user does not preload a field on a data source and then calls `to_pw` with no arguments.
Responsible: ngoldbaum