#18303: faster comparisons in AA and QQbar
--------------------------------------+------------------------
Reporter: vdelecroix | Owner:
Type: enhancement | Status: new
Priority: major | Milestone: sage-6.7
Component: number fields | Resolution:
Keywords: | Merged in:
Authors: Vincent Delecroix | Reviewers:
Report Upstream: N/A | Work issues:
Branch: u/vdelecroix/18303 | Commit:
Dependencies: | Stopgaps:
--------------------------------------+------------------------
Changes (by {'newvalue': u'Vincent Delecroix', 'oldvalue': ''}):
* branch: => u/vdelecroix/18303
* author: => Vincent Delecroix
Old description:
> As remarked in #17890 the comparisons of elements in `QQbar` can be much
> faster by implementing `_richcmp_` instead of `__cmp__`. A naive
> implementation already gives a great speedup
>
> old timing
> {{{
> sage: a = AA(125/13)
> sage: b = AA(3/5)
> sage: timeit("a < b", repeat=10, number=50000)
> 50000 loops, best of 10: 15.9 µs per loop
> sage: a = AA(2).sqrt()
> sage: timeit("a < b", repeat=10, number=50000)
> 50000 loops, best of 10: 17.9 µs per loop
> }}}
>
> new timing
> {{{
> sage: a = AA(125/13)
> sage: b = AA(3/5)
> sage: timeit("a < b", repeat=10, number=50000)
> 50000 loops, best of 10: 1.02 µs per loop
> sage: a = AA(2).sqrt()
> sage: timeit("a < b", repeat=10, number=50000)
> 50000 loops, best of 10: 7.6 µs per loop
> }}}
>
> see also: #18304 (comparisons involving integers and rationals).
New description:
The comparisons of elements of `AA` can be much faster! We got
old timing
{{{
sage: a = AA(125/13)
sage: b = AA(3/5)
sage: timeit("a < b", repeat=10, number=50000)
50000 loops, best of 10: 15.9 µs per loop
sage: a = AA(2).sqrt()
sage: timeit("a < b", repeat=10, number=50000)
50000 loops, best of 10: 17.9 µs per loop
}}}
new timing
{{{
sage: a = AA(125/13)
sage: b = AA(3/5)
sage: timeit("a < b", repeat=10, number=50000)
50000 loops, best of 10: 680 ns per loop
sage: a = AA(2).sqrt()
sage: timeit("a < b", repeat=10, number=50000)
50000 loops, best of 10: 686 ns per loop
}}}
Also, instead of implementing `__cmp__` we implement `_cmp_` (that avoids
one function call).
see also:
- #17890 (remove `_(rich)_cmp_impl`)
- #18304 (comparisons involving integers and rationals)
--
--
Ticket URL: <http://trac.sagemath.org/ticket/18303#comment:2>
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 unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/sage-trac.
For more options, visit https://groups.google.com/d/optout.
