[yt-dev] new ART front end (artio)

Matthew Turk matthewturk at gmail.com
Mon Nov 19 05:27:06 PST 2012


Hi Sam,

On Mon, Nov 19, 2012 at 1:58 AM, Sam Leitner <sam.leitner at gmail.com> wrote:
>
>> Great, this is all very useful.  One issue is that spatial chunking is
>> likely to be our preferred chunking method, given the on-disk
>> layout.  I'm hoping we can get something that functions as soon as
>> possible, and use that to learn what what is working and what
>> needs refactoring for performance or flexibility.  Feel free to offload
>> things that need to be implemented or need to be tested in
>> order to make that happen.  Sam seemed to think that the ramses reader was
>> broken for one of the test examples, but he'll need
>> to be more specific and report to the list.
>>
>>
>
> Actually, my issue is not specific to the RAMSES frontend, I just ran
> "yt plot tests/DD0010/moving7_0010t"
> which worked fine for me in 2.4, but fails in the current 3.0.

Hmm, this error actually looks only somewhat related to yt.  This
happens when Cython versions mismatch each other or NumPy.  Can you
try running this inside the 3.0 directory?

python2.7 setup.py build_ext -i -f

That will forcibly rebuild everything.  If you still get it, it might
mean that an older version of yt is getting in the way.

Looking at the tip of the yt branch, the changes to ImageArray (which
directly access "field_info") have broken the plotting method.  I've
gone ahead and run a couple plots and they seem to work for me, and
then I pushed the changes to the main yt 3.0 repository.  Let me know
if there are problems, but this will also require building all the
extensions from scratch.  I've also used some Cython 0.17 features,
which I mention below.

>
> (yt-x86_64)[@lap186 ~/repos/yt-3.0]$  yt plot tests/DD0010/moving7_0010
> Traceback (most recent call last):
>   File "/Users/sleitner/repos/yt-new/yt-x86_64/bin/yt", line 9, in <module>
>     load_entry_point('yt==3.0dev', 'console_scripts', 'yt')()
>   File
> "/Users/sleitner/repos/yt-new/yt-x86_64/lib/python2.7/site-packages/distribute-0.6.21-py2.7.egg/pkg_resources.py",
> line 337, in load_entry_point
>     return get_distribution(dist).load_entry_point(group, name)
>   File
> "/Users/sleitner/repos/yt-new/yt-x86_64/lib/python2.7/site-packages/distribute-0.6.21-py2.7.egg/pkg_resources.py",
> line 2281, in load_entry_point
>     return ep.load()
>   File
> "/Users/sleitner/repos/yt-new/yt-x86_64/lib/python2.7/site-packages/distribute-0.6.21-py2.7.egg/pkg_resources.py",
> line 1991, in load
>     entry = __import__(self.module_name, globals(),globals(), ['__name__'])
>   File "/Users/sleitner/repos/yt-3.0/yt/utilities/command_line.py", line 29,
> in <module>
>     from yt.mods import *
>   File "/Users/sleitner/repos/yt-3.0/yt/mods.py", line 58, in <module>
>     from yt.data_objects.api import \
>   File "/Users/sleitner/repos/yt-3.0/yt/data_objects/api.py", line 31, in
> <module>
>     from grid_patch import \
>   File "/Users/sleitner/repos/yt-3.0/yt/data_objects/grid_patch.py", line
> 35, in <module>
>     from yt.data_objects.data_containers import YTFieldData
>   File "/Users/sleitner/repos/yt-3.0/yt/data_objects/data_containers.py",
> line 51, in <module>
>     import yt.geometry.selection_routines
>   File "oct_container.pxd", line 49, in init yt.geometry.selection_routines
> (yt/geometry/selection_routines.c:15060)
>   File "numpy.pxd", line 155, in init yt.geometry.oct_container
> (yt/geometry/oct_container.c:14580)
> ValueError: numpy.dtype has the wrong size, try recompiling
>
> So I attempted to bash install_script.sh the current 3.0 and that failed as
> well:

I've just changed the install script to match the tip of the main
install script, which includes Cython 0.17 and actually checks out
3.0; before it was only checking out the tip of the main development
branch.  If you wipe the yt-hg directory this will probably go more
smoothly, but you should be able to run the install script as is now.

Sorry about the trouble!  The plotting in particular is a bit
frustrating, since the merges are tricky and I'm trying to keep up to
date.  Getting the plot window tests in line and fixing up the failing
(due to non-implementation) tests in 3.0 will help with that.  I'll
put that high on the agenda.

-Matt

> Installing yt
> ********************************************
>         FAILURE REPORT:
> ********************************************
>
> /var/folders/s_/zwb_tjns1_n4vsv8r45mk9mc0000gn/T//ccfglW2k.s:unknown:FATAL:can't
> create output file:
> build/temp.macosx-10.4-x86_64-2.7/yt/utilities/lib/CICDeposit.o
> yt/utilities/lib/CICDeposit.c: In function ‘__Pyx_GetBuffer’:
> yt/utilities/lib/CICDeposit.c:7976: warning: unused variable
> ‘getbuffer_cobj’
> yt/utilities/lib/CICDeposit.c: In function ‘__Pyx_ReleaseBuffer’:
> yt/utilities/lib/CICDeposit.c:8016: warning: unused variable
> ‘releasebuffer_cobj’
> yt/utilities/lib/CICDeposit.c: At top level:
> /Users/sleitner/repos/yt-3.0/yt-x86_64/lib/python2.7/site-packages/numpy/core/include/numpy/__multiarray_api.h:1533:
> warning: ‘_import_array’ defined but not used
> /Users/sleitner/repos/yt-3.0/yt-x86_64/lib/python2.7/site-packages/numpy/core/include/numpy/__ufunc_api.h:227:
> warning: ‘_import_umath’ defined but not used
> /var/folders/s_/zwb_tjns1_n4vsv8r45mk9mc0000gn/T//ccfglW2k.s:unknown:FATAL:can't
> create output file:
> build/temp.macosx-10.4-x86_64-2.7/yt/utilities/lib/CICDeposit.o
> error: Command "gcc -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes
> -mmacosx-version-min=10.7 -I/opt/local/include -I/opt/local/include
> -I/opt/local/include
> -I/Users/sleitner/repos/yt-3.0/yt-x86_64/lib/python2.7/site-packages/numpy/core/include
> -I/Users/sleitner/repos/yt-3.0/yt-x86_64/include/python2.7 -c
> yt/utilities/lib/CICDeposit.c -o
> build/temp.macosx-10.4-x86_64-2.7/yt/utilities/lib/CICDeposit.o" failed with
> exit status 1
>
> Apologies if I'm doing something stupid here (likely)!
>
> Sam
>
> _______________________________________________
> yt-dev mailing list
> yt-dev at lists.spacepope.org
> http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
>



More information about the yt-dev mailing list