Opened 8 years ago
Closed 8 years ago
#14617 closed enhancement (fixed)
Update scipy to 0.12
Reported by: | jpflori | Owned by: | jdemeyer |
---|---|---|---|
Priority: | major | Milestone: | sage-5.10 |
Component: | packages: standard | Keywords: | scipy spkg |
Cc: | Merged in: | sage-5.10.beta5 | |
Authors: | Jean-Pierre Flori | Reviewers: | Volker Braun |
Report Upstream: | N/A | Work issues: | |
Branch: | Commit: | ||
Dependencies: | Stopgaps: |
Description (last modified by )
Change History (15)
comment:1 Changed 8 years ago by
- Description modified (diff)
comment:2 Changed 8 years ago by
My, bad, it is. I read the wrong README in the tarball and got a wrong feeling.
comment:4 Changed 8 years ago by
There is one here as well:
Testing it right now :)
comment:5 Changed 8 years ago by
The way I read it, scipy's arpack has been renamed to arpack-ng. Nothing has been replaced, but there are a bunch of bugfixes to arpack in the new release.
comment:6 Changed 8 years ago by
Hum, anyway arpack-ng is just a fork of arpack, so would basically fit with "rename and fix bugs" because upstream was not active anymore. See:
comment:7 Changed 8 years ago by
And my scipy spkg failed on top of 5.10.beta3 plus #10508.
comment:8 Changed 8 years ago by
- Description modified (diff)
I've reposted an spkg which now compiles (the only patch we used was now upstream). I've not tested Volker's one, whence the change in the ticket description, but if anyone (including Volker :)) feels it is better, feel free to change it back.
comment:9 Changed 8 years ago by
I had removed the upsteamed patch, too. Doctests pass for me.
comment:10 Changed 8 years ago by
Great, so I'll have a look at your spkg and give it positive review when my tests finish.
comment:11 Changed 8 years ago by
- Reviewers set to Volker Braun
- Status changed from new to needs_review
comment:12 Changed 8 years ago by
- Status changed from needs_review to positive_review
comment:13 Changed 8 years ago by
On hawk
, importing scipy
fails with
sage -t --long devel/sage/sage/plot/plot3d/list_plot3d.py ********************************************************************** File "devel/sage/sage/plot/plot3d/list_plot3d.py", line 87, in sage.plot.plot3d.list_plot3d.list_plot3d Failed example: list_plot3d(m, texture='yellow', interpolation_type='nn',frame_aspect_ratio=[1,1,1/3]) Exception raised: Traceback (most recent call last): File "/export/home/buildbot/build/sage/hawk-1/hawk_full/build/sage-5.10.beta5/local/lib/python2.7/site-packages/sage/doctest/forker.py", line 466, in _run self.execute(example, compiled, test.globs) File "/export/home/buildbot/build/sage/hawk-1/hawk_full/build/sage-5.10.beta5/local/lib/python2.7/site-packages/sage/doctest/forker.py", line 825, in execute exec compiled in globs File "<doctest sage.plot.plot3d.list_plot3d.list_plot3d[5]>", line 1, in <module> list_plot3d(m, texture='yellow', interpolation_type='nn',frame_aspect_ratio=[Integer(1),Integer(1),Integer(1)/Integer(3)]) File "/export/home/buildbot/build/sage/hawk-1/hawk_full/build/sage-5.10.beta5/local/lib/python2.7/site-packages/sage/plot/plot3d/list_plot3d.py", line 174, in list_plot3d return list_plot3d_tuples(l,interpolation_type,texture,**kwds) File "/export/home/buildbot/build/sage/hawk-1/hawk_full/build/sage-5.10.beta5/local/lib/python2.7/site-packages/sage/plot/plot3d/list_plot3d.py", line 356, in list_plot3d_tuples from scipy import interpolate File "/export/home/buildbot/build/sage/hawk-1/hawk_full/build/sage-5.10.beta5/local/lib/python2.7/site-packages/scipy/interpolate/__init__.py", line 160, in <module> from .ndgriddata import * File "/export/home/buildbot/build/sage/hawk-1/hawk_full/build/sage-5.10.beta5/local/lib/python2.7/site-packages/scipy/interpolate/ndgriddata.py", line 10, in <module> from .interpnd import LinearNDInterpolator, NDInterpolatorBase, \ ImportError: ld.so.1: python2.7: fatal: relocation error: file /export/home/buildbot/build/sage/hawk-1/hawk_full/build/sage-5.10.beta5/local/lib/python2.7/site-packages/scipy/interpolate/interpnd.so: symbol __sync_fetch_and_add_4: referenced symbol not found **********************************************************************
This seems to come from GCC, so I'll try again with Sage's GCC (the error above was with GCC-4.4.3)
comment:14 Changed 8 years ago by
Indeed. Some pointers:
and a bunch of similar ones on other systems alos closed as invalid, or
For GCC 4.4.x you should explicitely pass -march=something-that-works to get that function, whereas for more recent GCCs it seems to be implied automatically from the value of target.
comment:15 Changed 8 years ago by
- Merged in set to sage-5.10.beta5
- Resolution set to fixed
- Status changed from positive_review to closed
This won't fix the problem with updated ATLAS as ARPACK-NG is not in 0.12 unfortunately, but in the git repo only: https://github.com/scipy/scipy/pull/386