Hi all,
Registration is now open for the Learn yt Workshop at the University of
Illinois Urbana-Champaign, October 10-12, 2016. This workshop will
introduce attendees to the yt project (yt-project.org), a python toolkit
for the analysis and visualization of volumetric data.
We have create a website for the workshop with more information:
http://yt-project.org/workshop2016/
In addition, if you wold like to register for the workshop, please fill out
this online form:
https://goo.gl/forms/c6oIzNQywU1YWgOe2
The workshop will cover basic usage of yt, including the yt data model,
yt's field system, low-level data inspection, basic visualization
workflows, and data analysis and reduction tasks. We will also cover more
advanced usages like generating simulated observations, halo finding and
halo analysis, volume rendering, advanced 3D visualizations, and advanced
data analysis and reduction tasks. Finally, we will cover how to modify and
extend yt, as well as the development and contribution process.
In addition, there will be time set aside for exploring data you bring to
the workshop, along with opportunities to work directly with yt developers
to load and explore your data.
The workshop will take place at the National Center for Supercomputing
Applications building on the north end of the UIUC campus. The NCSA
building is about a block away from the conference hotel and is next door
to a parking structure that offers metered all-day parking. There are a
number of food trucks nearby, a university-run cafeteria about 2 blocks
away, and a university business district with many good lunch restaurants
about a half a mile away.
We are planning to offer funding for hotel and travel for those requesting
support. If you request funding, you will be notified of available funds
by September 15. Travel awards will be made in the form of arranged
lodging and airfare, with reservations being made by the conference
organizers.
We hope to see you there.
On behalf of the organizing committee,
Matt Turk
Nathan Goldbaum
Jill Naiman
John ZuHone
Kandace Turner
New issue 1268: enzo embedded parallel python/yt problem with gradient field
https://bitbucket.org/yt_analysis/yt/issues/1268/enzo-embedded-parallel-p...
Bodo Schwabe:
I currently try to run enzo with embedded parallelized python/yt (Version = 3.4-dev
Changeset = c941aa457bc3 (yt) tip) with the following user_script.py < http://paste.yt-project.org/show/6794/ > . It works fine on a single processor (mpirun -n 1 ...) but gives me the following error message for multiple processors <http://paste.yt-project.org/show/6795/ > . Regular derived_fields are working well, but gradient_fields aren't. Thank you very much for any suggestions.
New issue 1267: Profile datasets don't save profile.field_info dict data
https://bitbucket.org/yt_analysis/yt/issues/1267/profile-datasets-dont-save
Nathan Goldbaum:
This is illustrated by the following script:
```
import yt
ds = yt.load('IsolatedGalaxy/galaxy0030/galaxy0030')
ad = ds.all_data()
prof = yt.create_profile(ad, [('gas', 'H_number_density'), ('gas', 'temperature')\
], fields='cell_mass')
print (prof.field_info)
prof.save_as_dataset()
prof_ds = yt.load('galaxy0030_Profile2D.h5')
print (prof_ds.profile.field_info)
plot = yt.PhasePlot(prof_ds.data, "H_number_density", "temperature", "cell_mass",\
weight_field=None)
```
On my machine this prints:
```
{('gas', 'temperature'): Alias Field for "('enzo', 'Temperature')" (gas, temperature): (units: K), ('gas', 'H_number_density'): Derived Field (gas, H_number_density): (units: cm**(-3)), ('gas', 'cell_mass'): Derived Field (gas, cell_mass): (units: g)}
{}
```
And then crashes with the following traceback:
```
Traceback (most recent call last):
File "test.py", line 11, in <module>
weight_field=None)
File "/Users/goldbaum/Documents/yt-hg/yt/visualization/profile_plotter.py", line 745, in __init__
if isinstance(data_source.ds, YTProfileDataset):
File "/Users/goldbaum/Documents/yt-hg/yt/visualization/profile_plotter.py", line 775, in _initialize_instance
obj._setup_plots()
File "/Users/goldbaum/Documents/yt-hg/yt/visualization/profile_plotter.py", line 856, in _setup_plots
x_scale, y_scale, z_scale = self._get_field_log(f, self.profile)
File "/Users/goldbaum/Documents/yt-hg/yt/visualization/profile_plotter.py", line 815, in _get_field_log
zfi = profile.field_info[field_z]
KeyError: ('data', 'cell_mass')
```
@brittonsmith any chance you can take a look at this?
New issue 1266: ds.add_gradient_fields not working properly on 2D data
https://bitbucket.org/yt_analysis/yt/issues/1266/dsadd_gradient_fields-no...
Dominik Derigs:
I want to plot the divergence of the magnetic field. For this, I think, the right approach is to use the function add_gradient_field. However, I have been unable to successfully generate any gradient field on my data (2D, FLASH).
There is no suitable public dataset available that could be used to show my problem (2D, periodic, magnetic fields), so I uploaded an exemplary data file [here](https://hera.ph1.uni-koeln.de/~derigs/rotor_mhd_2d_hdf5_chk_0003_cleaning).
MWE:
```
#!python
import yt
ds = yt.load("rotor_mhd_2d_hdf5_chk_0003_cleaning")
slc = yt.SlicePlot(ds, 'z', ["pressure", "pressure_gradient_x", "pressure_gradient_y"])
slc.save()
```
**While the pressure plot is correct, the gradient plots are just empty.**
pressure:
pressure_gradient_x:
pressure_gradient_y:
New issue 1264: Ambiguity when converting between code units of two datasets
https://bitbucket.org/yt_analysis/yt/issues/1264/ambiguity-when-convertin...
Nathan Goldbaum:
This is illustrated by the following test script:
```
from yt.testing import fake_random_ds
ds1 = fake_random_ds(32, length_unit=1)
ds2 = fake_random_ds(32, length_unit=3)
l1 = ds1.quan(0.3, 'unitary')
l2 = ds2.quan(l1, 'unitary')
print l1.in_cgs()
print l2.in_cgs()
print ds1.quan(0.3, 'unitary').in_cgs()
print ds2.quan(0.3, 'unitary').in_cgs()
```
Right now this prints:
```
0.9 cm
0.9 cm
0.3 cm
0.9 cm
```
I would expect it to print:
```
0.3 cm
0.9 cm
0.3 cm
0.9 cm
```
I noticed @brittonsmith was working around this, but I don't think it should be necessary to do that.
Responsible: ngoldbaum
Hi folks,
A while ago, yt-project.org was converted to a "generated" site. Templates
are used to generate each page, so it's much easier to modify. I've also
moved the calculation of contributors into a function, and the gallery and
the members list are in .yaml files, so they can be updated without writing
raw HTML.
I took some time this week to experiment a bit with the website. My fork
is here:
https://bitbucket.org/MatthewTurk/yt.website
and a build is here:
http://matt-website.hub.yt/
The big changes:
* Light instead of dark background ("cyborg" theme to "flatly")
* Made the text more normal-sized most places
* Changed the masthead on the front page to *not* be a carousel and have
the new logo, along with a more punchy intro
* Added a set of "examples", which come from a .yaml file.
I've been experimenting with the examples and the theme in particular, and
I personally like it a lot more.
If you fork my repo and make changes (especially if you want to add
examples, which can take *either* text output *or* images) as soon as the
changes go in, it auto-builds on matt-website.hub.yt . Feedback here or on
slack would be very much appreciated, and if folks like it, I'd like to
deploy it.
A few other things it would need before it's ready:
* Add an extensions page, with extension information drawn from .yaml file
* Make the blog look like this thing, and maybe figure out if the blog
could be part of the repo
* Update the text and images wherever appropriate (i.e., update references
to the hub)
* Streamline navigation a little bit
* Fill out the examples, maybe even have them link to the hub
-Matt
Hi folks,
I'm updating the website with examples (see the thread from last week
about this) and I wanted to publicly ask for your favorite examples of
doing or making something cool in yt with just a few lines.
Especially if it's something pretty cool or special, it'd be great to
include.
Reply to me off list and I'll add them to the examples!
Thanks,
Matt
