#11900: Serious regression caused by #9138
---------------------------+------------------------------------------------
Reporter: SimonKing | Owner: tbd
Type: defect | Status: needs_work
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 SimonKing):
* cc: malb (added)
Comment:
I managed to implement the new coercion model for libsingular rings,
therefore cc to Martin.
One trick to make that work: One should acknowledge that polynomial rings
are unique parents (unless one destroys uniqueness on purpose), thus, the
hash can be `id(self)` and `__cmp__` can be identity of objects.
What has hash and cmp to do with the new coercion model? There was a
custom `__richcmp__` in the code. It somehow managed to trigger a test for
the old coercion model. And unfortunately, if replaces `__cmp__` by
`_element_constructor_` then the test for the ''old'' coercion model
raises an exception, even though `_element_constructor_` is supposed to be
a tool of the ''new'' model.
I think that in a different ticket, one can build on top of that and
replace the current (and always repeated!) coercion tests on elements by
tests that are done ''once'' for the parent of these elements. That would
improve the performance.
Perhaps that was not a very clear explanation. Anyway.
With the new coercion, it becomes possible to avoid the construction of a
coercion from the base ring during initialisation.
In addition, I am working on a simplification of the choice of category
during `PolynomialRing_generic.__init__`. It uses
`sage.rings.categories.Category.join`, but could actually directly
construct a `JoinCategory`, since all the additional work in
`sage.rings.categories.Category.join` is not needed. Since `join()` eats a
lot of time in the benchmarks, I am confident that we will see an
improvement.
That's the plan, now I have to do tests...
--
Ticket URL: <http://trac.sagemath.org/sage_trac/ticket/11900#comment:26>
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.
