Hi,
Is there any way in yt to get a script to delete (or ignore) all the .yt
and .harray files and clear its memory? I run into errors when I'm
re-doing a plot or I've re-run the simulation and over-written the last
data set I looked at with yt. I have to exit yt, delete the above files
by hand and restart, which is a bit of a pain.
Elizabeth
Hi,
I've added PotentialField to the data that gets outputted. I'd like to
read this into yt, but how does this work when yt won't be expecting
PotentialField to be one of the data sets?
Elizabeth
Hi All,
I have just pushed to the development branch of yt a modification to the clump finder(*) which should make the identification of gravitationally bound clumps run much more quickly. Previously, the gravitational potential energy of clumps was calculated using the full double sum that iterates over all unique pairs of cells in the clump. This is, of course, accurate, but it is quite slow because it is an O(N^2) operation.
A generally much faster and very accurate (but not perfectly accurate) method of calculating gravitational potentials is the treecode method of Barnes & Hut [1] which reduces the operation down to O(N log N). Using the Octree that Matt added to yt some time ago (and had never used) I implemented a treecode on top of it. There may be some further optimizations to the code, but especially for clumps larger than 100,000 cells, the treecode is substantially faster, up to several times for very large clumps. For intermediate clumps of a few tens of thousands of cells, the treecode is actually a bit slower, but no worse than 50%. This is due to likely both the extra cost of setting up the treecode hierarchy, and perhaps some optimizations I have yet to catch. At any rate, for now the time gains from the very large clumps should offset the time losses for smaller clumps.
Unfortunately, the treecode will not help unigrid datasets because the levels of the octree mirror the levels in the dataset, and a treecode needs more than one level to work. But it is my impression that people do not do clump finding on unigrid datasets. Please correct me if my impression is wrong! Also, the octree is x2 refinement-specific.
In the course of this work I noticed that periodic clumps would not be processed correctly, so I have added some machinery that should fix that using either the N^2 or treecode method.
The treecode is now on by default in the development branch. Let me know if you have any questions, or problems with the code. Thanks!
[1] http://adsabs.harvard.edu/abs/1986Natur.324..446B
(*) To be precise, the modifications are to the "IsBound" derived quantity, but that is most often called from within the clump finder. As a point of fact, I have actually not touched the clump finder at all.
Stephen Skory
stephenskory(a)yahoo.com
http://stephenskory.com/
510.621.3687 (google voice)
Hi guys,
I would like to plot a graph of mass in stars against time. In addition to
the normal y axis (total stellar mass in solar masses) I would like a second
y axis on the right side which translates this stellar mass into a fraction
of the original gas mass. Thus the line is the same it is just the two y
axes are scaled versions of each other. Does that make sense? Can anyone
suggest a way to do this?
Thanks,
Libby
--
Elizabeth Harper-Clark MA MSci
PhD Candidate, Canadian Institute for Theoretical Astrophysics, UofT
Sciences and Engineering Coordinator, Teaching Assistants' Training Program,
UofT
www.astro.utoronto.ca/~h-clark <http://www.astro.utoronto.ca/%7Eh-clark>
h-clark(a)cita.utoronto.ca
Astronomy office phone: +1-416-978-5759
Hi,
I'm sure I've seen this before, but possibly in the old version of the
yt online docs .... is there a list of fields, derived and raw, that you
can call via data["FieldIWant"]? I'd really like it to be linked from
the "Analyzing Data" page here:
http://yt.enzotools.org/doc/analyzing/index.html
Perhaps right before "Derived Quantities" since I keep hopefully looking
there automatically .....? (That's assuming that the list has gone awol
and isn't somewhere equally obvious).
Hoping doc comments are ultimately helpful and not just annoying,
Elizabeth
Hi,
How do I set a y-range on add_profile_sphere?
As a suggestion, perhaps we could add this to one of the examples in the
cookbook? Also, in the cookbook, these links no longer work:
"The latest version of this recipe can be downloaded here:
http://hg.enzotools.org/cookbook/raw-file/tip/recipes/simple_slice.py ."
Elizabeth
Hi all,
I'd like to create a new yt field from a yt vector field where I take its magnitude. Any thoughts on a good NumPy-ish way to do this?
Best,
John ZuHone
Hi,
I've had to reinstall yt due to an unconnected python problem but now I
have a runtime error:
/home/taskere/yt/yt-2/lib/python2.7/site-packages/matplotlib/backends/backend_agg.py
in _get_agg_font(self, prop)
219 font = self._fontd.get(fname)
220 if font is None:
--> 221 font = FT2Font(str(fname))
222 self._fontd[fname] = font
223 self._fontd[key] = font
RuntimeError: Could not open facefile
/home/taskere/yt/yt-2/lib/python2.6/site-packages/matplotlib/mpl-data/fonts/ttf/Vera.ttf;
Cannotpen_Resource
I ... forget whether this happened before :\
Elizabeth
Hi all,
I would like to create a multi plot equivalent to the procedure
described in the Cookbook (Multi plot 3x2) *BUT* with different datasets
(1 slice for 1 dataset).
The first issue I am encountering is that the module raven is not detected:
#############################################################
Traceback (most recent call last):
File "/rpod2/jcpassy/Enzo/python/MultiPlot.py", line 20, in <module>
fig, axes, colorbars = raven.get_multi_plot( 2, 4, colorbar=orient,
bw = 4)
NameError: name 'raven' is not defined
#############################################################
Then, is there a specific description somewhere of how to do a multi
plot with different datasets ?
Thanks a lot for your help,
JC
Hi All,
We'd like some feedback on your use and your opinions about the future
of yt. It's really short, and no questions are mandatory, so feel free
to fill in as many or as few as you'd like. Please visit
https://spreadsheets.google.com/viewform?formkey=dGlFazdOcUtFbHJLVU80a2NGM3…
and answer away.
The last user survey was really useful for us to help make yt-2.0 what
it was. The future of yt is looking really bright, and we'd like to
ensure it is as useful as we can make it. If you'd like to get
involved at any point in the development process, feel free to email
yt-dev, or any member of the dev team individually.
Thanks, and we look forward to hearing from you!
Jeff
