#11900: Serious regression caused by #9138
---------------------------+------------------------------------------------
Reporter: SimonKing | Owner: tbd
Type: defect | Status: new
Priority: blocker | Milestone: sage-4.7.2
Component: performance | Keywords: categories regression
Work_issues: | Upstream: N/A
Reviewer: | Author: Simon King
Merged: | Dependencies: #9138
---------------------------+------------------------------------------------
Changes (by newvalueoldvalue):
* dependencies: => #9138
* author: => Simon King
Comment:
I have attached a preliminary patch, if someone already wants to have a
look. I need to do more tests (and run doctests), but up to now it seems
that it fixes the regression.
'''__Changes for Matrix Spaces__'''
Matrix spaces will not make use of the category framework by default. But
I introduced a new method that makes them use the category framework after
initialisation:
{{{
sage: MS = MatrixSpace(QQ,8)
sage: TestSuite(MS).run()
Failure in _test_category:
Traceback (most recent call last):
...
AssertionError: category of self improperly initialized
------------------------------------------------------------
The following tests failed: _test_category
sage: type(MS)
<class 'sage.matrix.matrix_space.MatrixSpace_generic'>
sage: MS.full_category_initialisation()
sage: TestSuite(MS).run()
sage: type(MS)
<class
'sage.matrix.matrix_space.MatrixSpace_generic_with_category'>
}}}
Minor change: I made `__matrix_class` a lazy attribute.
'''__Changes for categories__'''
I turn `hom_category` in a cached method - the result is cached anyway,
and thus it makes sense to cache as early as possible, in order to reduce
the overhead, and in order to benefit from #11115.
I also cache the class method `category_of` in
sage.categories.covariant_functorial_construction. According to a
commented-out line, caching has been attempted anyway, but since it needs
to be a class method, caching has not been possible by simply using a
decorator.
'''__Vectors__'''
I added custom `list()` methods for both integral and rational dense
vectors. Here is the speed difference (the timing is with patch attached):
{{{
sage: v = vector(ZZ,range(100))
sage: %timeit L = v.list()
625 loops, best of 3: 11 µs per loop
# was 17 µs per loop in sage-4.7.2.alpha2
sage: v = vector(QQ,range(100))
sage: %timeit L = v.list()
625 loops, best of 3: 24 µs per loop
# was 27.7 µs per loop in sage-4.7.2.alpha2
}}}
The difference is small, but one should benefit from it.
'''__Doc formatting__'''
I corrected some wrong formatting in the documentation of dense integral
and rational vectors.
--
Ticket URL: <http://trac.sagemath.org/sage_trac/ticket/11900#comment:13>
Sage <http://www.sagemath.org>
Sage: Creating a Viable Open Source Alternative to Magma, Maple, Mathematica,
and MATLAB
--
You received this message because you are subscribed to the Google Groups
"sage-trac" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/sage-trac?hl=en.
