New issue 619: user-defined outputs with Athena
https://bitbucket.org/yt_analysis/yt/issue/619/user-defined-outputs-with-at…
Anonymous:
Dear Developers,
I would like to visualize/analyze some streaming-nstability simulation outputs created with Athena with yt. The output is a user-defined variable, such as the particle density enhancement. The (2-d) data is written to a .vtk file which I tried to read with yt using the load command. I Immediately get errors.
Here is the session information:
pf = load("id0/Streaming2d.0000.difdpar.vtk")
yt : [INFO ] 2013-08-05 15:30:17,851 Temporarily setting domain_right_edge = -domain_left_edge. This will be corrected automatically if it is not the case.
yt : [INFO ] 2013-08-05 15:30:17,855 Parameters: current_time = 0
yt : [INFO ] 2013-08-05 15:30:17,855 Parameters: domain_dimensions = [64 64 1]
yt : [INFO ] 2013-08-05 15:30:17,856 Parameters: domain_left_edge = [-1. -1. 0.]
yt : [INFO ] 2013-08-05 15:30:17,857 Parameters: domain_right_edge = [ 1. 1. -0.]
yt : [INFO ] 2013-08-05 15:30:17,858 Parameters: cosmological_simulation = 0.0
yt : [WARNING ] 2013-08-05 15:30:17,860 Dimensionality less than 3: reverting to overlap_proj
---------------------------------------------------------------------------
KeyError Traceback (most recent call last)
<ipython-input-3-fa5200178679> in <module>()
----> 1 pf = load("id0/Streaming2d.0000.difdpar.vtk")
/home/hlcapelo/.local/lib/python2.7/site-packages/yt-2.6dev-py2.7-linux-x86_64.egg/yt/convenience.pyc in load(*args, **kwargs)
75 if c._is_valid(*args, **kwargs): candidates.append(n)
76 if len(candidates) == 1:
---> 77 return output_type_registry[candidates[0]](*args, **kwargs)
78 if len(candidates) == 0:
79 if ytcfg.get("yt", "enzo_db") != '' \
/home/hlcapelo/.local/lib/python2.7/site-packages/yt-2.6dev-py2.7-linux-x86_64.egg/yt/frontends/athena/data_structures.pyc in __init__(self, filename, data_style, storage_filename, parameters)
353 # instantiated so that we can make sure we have the correct left
354 # and right domain edges.
--> 355 self.h
356
357 def _set_units(self):
/home/hlcapelo/.local/lib/python2.7/site-packages/yt-2.6dev-py2.7-linux-x86_64.egg/yt/data_objects/static_output.pyc in hierarchy(self)
201 raise RuntimeError("You should not instantiate StaticOutput.")
202 self._instantiated_hierarchy = self._hierarchy_class(
--> 203 self, data_style=self.data_style)
204 return self._instantiated_hierarchy
205 h = hierarchy # alias
/home/hlcapelo/.local/lib/python2.7/site-packages/yt-2.6dev-py2.7-linux-x86_64.egg/yt/frontends/athena/data_structures.pyc in __init__(self, pf, data_style)
129 #self.directory = os.path.dirname(self.hierarchy_filename)
130 self._fhandle = file(self.hierarchy_filename,'rb')
--> 131 AMRHierarchy.__init__(self, pf, data_style)
132
133 self._fhandle.close()
/home/hlcapelo/.local/lib/python2.7/site-packages/yt-2.6dev-py2.7-linux-x86_64.egg/yt/data_objects/hierarchy.pyc in __init__(self, pf, data_style)
70
71 mylog.debug("Parsing hierarchy.")
---> 72 self._parse_hierarchy()
73
74 mylog.debug("Constructing grid objects.")
/home/hlcapelo/.local/lib/python2.7/site-packages/yt-2.6dev-py2.7-linux-x86_64.egg/yt/frontends/athena/data_structures.pyc in _parse_hierarchy(self)
251 line = f.readline()
252 f.close()
--> 253 levels[j] = gridread['level']
254 glis[j,0] = gridread['left_edge'][0]
255 glis[j,1] = gridread['left_edge'][1]
KeyError: 'level'
New issue 618: Install script fails when compiling with gcc provided by MESA sdk
https://bitbucket.org/yt_analysis/yt/issue/618/install-script-fails-when-co…
Nathan Goldbaum:
It seems a lot of people have the gcc provided by the [MESA](http://mesa.sourceforge.net/) sdk in their `$PATH`. The install script currently fails while trying to compile mysql using this compiler. I've only seen this on a mac, and while building yt with xcode's gcc still works, it's a bit confusing for new users who aren't familiar with POSIX semantics.
Fixing this will require testing with a MESA install on a mac. Hopefully there's some way to determine that we're using the MESA sdk compilers so we can either warn users or adjust the defaults automagically.
New issue 616: Loading vtk files from Athena joined with their join_vtk.c executable
https://bitbucket.org/yt_analysis/yt/issue/616/loading-vtk-files-from-athen…
Anonymous:
It seems that when loading Athena vtk files, if the file was created by the Athena supplied executable join_vtk (for stitching together vtk files created from the different processors in an mpi run) then the loaded data is missing all the fields except the 'density' field.
Hi all,
Just a heads up, I will be away until the 15th or so. Don't feel that PRs
need to wait for me to return unless they are something that definitely
needs my attention.
Matt
New issue 615: plots look grainy and not so nice under matplotlib 1.3.0
https://bitbucket.org/yt_analysis/yt/issue/615/plots-look-grainy-and-not-so…
Nathan Goldbaum:
Something must have changed internally in matplotlib, as png plots are now significantly lower resolution by default.
Responsible: ngoldbaum
New issue 614: yt's libraries are not compatible with CentOS binaries
https://bitbucket.org/yt_analysis/yt/issue/614/yts-libraries-are-not-compat…
Nathan Goldbaum:
I was trying to use `eog` to look at some images on a supercomputer. Under the yt environment, I get a seg fault when I try to use `eog`, but after deactivating and trying again, everything works correctly. `eog` is dynamically loading yt's freetype and libz. I've pasted a shell session with information about this issue below:
```
#!bash
[goldbaum@hyades ~]$ eog
Segmentation fault
[goldbaum@hyades ~]$ which eog
/usr/bin/eog
[goldbaum@hyades ~]$ ldd /usr/bin/eog | grep yt
libfreetype.so.6 => /home/goldbaum/yt-x86_64/lib/libfreetype.so.6 (0x00002b9c0759f000)
libz.so.1 => /home/goldbaum/yt-x86_64/lib/libz.so.1 (0x00002b9c0787b000)
[goldbaum@hyades ~]$ deactivate
[goldbaum@hyades ~]$
[goldbaum@hyades ~]$ eog
[goldbaum@hyades ~]$ ldd /usr/bin/eog | grep 'freetype\|libz'
libfreetype.so.6 => /usr/lib64/libfreetype.so.6 (0x00000037af800000)
libz.so.1 => /lib64/libz.so.1 (0x00000037ab800000)
[goldbaum@hyades ~]$ cat /etc/centos-release
CentOS release 6.3 (Final)
```
It's not clear to me why yt's freetype and/or libz are incompatible with the CentOS libraries. @xarthisius, do you have any idea what's going on here?
It's probably not worth spending too much time on this, but this issue can at least serve as a reference point. The workaround for this issue is just to deactivate the yt environment.