#18304: Integer and Rational comparisons
-------------------------------------+-------------------------------------
Reporter: vdelecroix | Owner:
Type: enhancement | Status: needs_work
Priority: major | Milestone: sage-6.9
Component: performance | Resolution:
Keywords: | Merged in:
Authors: Vincent Delecroix | Reviewers:
Report Upstream: N/A | Work issues:
Branch: | Commit:
u/vdelecroix/18304 | e03a26e63dd50f79fc7ada885daefd8c9970c482
Dependencies: | Stopgaps:
-------------------------------------+-------------------------------------
Changes (by vdelecroix):
* status: needs_review => needs_work
Old description:
> Comparing integers is very fast, comparing rationals is very fast but
> comparing integer with rational is about 5x slower:
> {{{
> sage: a = 1; b = 18
> sage: timeit("a < b", number=500000, repeat=100)
> 500000 loops, best of 100: 64.3 ns per loop
> sage: a = QQ(1); b = QQ(18)
> sage: timeit("a < b", number=500000, repeat=100)
> 500000 loops, best of 100: 84 ns per loop
> sage: a = 1; b = QQ(18)
> sage: timeit("a < b", number=500000, repeat=100)
> 500000 loops, best of 100: 405 ns per loop
> }}}
>
> With the branch applied, it gets better
> {{{
> sage: a = 1; b = 18
> sage: timeit("a < b", number=500000, repeat=100)
> 500000 loops, best of 100: 64 ns per loop
> sage: a = QQ(1); b = QQ(18)
> sage: timeit("a < b", number=500000, repeat=100)
> 500000 loops, best of 100: 65.6 ns per loop
> sage: a = 1; b = QQ(18)
> sage: timeit("a < b", number=500000, repeat=100)
> 500000 loops, best of 100: 87.1 ns per loop
> }}}
>
> Upstream mpir pull request for the `mpq_cmp_mpz` function:
>
> [https://github.com/wbhart/mpir/pull/157
> https://github.com/wbhart/mpir/pull/157]
>
> Warning: if you pull the branch and run `make` then you will likely
> rebuild your whole Sage!
New description:
DISCLAIMER: this ticket is waiting from two upstream modifications (GMP
and mpir). Namely the creation of the function `mpq_cmp_z(mpq_t, mpz_t)`.
Comparing integers is very fast, comparing rationals is very fast but
comparing integer with rational is about 5x slower:
{{{
sage: a = 1; b = 18
sage: timeit("a < b", number=500000, repeat=100)
500000 loops, best of 100: 64.3 ns per loop
sage: a = QQ(1); b = QQ(18)
sage: timeit("a < b", number=500000, repeat=100)
500000 loops, best of 100: 84 ns per loop
sage: a = 1; b = QQ(18)
sage: timeit("a < b", number=500000, repeat=100)
500000 loops, best of 100: 405 ns per loop
}}}
With the branch applied, it gets better
{{{
sage: a = 1; b = 18
sage: timeit("a < b", number=500000, repeat=100)
500000 loops, best of 100: 64 ns per loop
sage: a = QQ(1); b = QQ(18)
sage: timeit("a < b", number=500000, repeat=100)
500000 loops, best of 100: 65.6 ns per loop
sage: a = 1; b = QQ(18)
sage: timeit("a < b", number=500000, repeat=100)
500000 loops, best of 100: 87.1 ns per loop
}}}
Upstream mpir pull request for the `mpq_cmp_mpz` function:
[https://github.com/wbhart/mpir/pull/157
https://github.com/wbhart/mpir/pull/157]
Warning: if you pull the branch and run `make` then you will likely
rebuild your whole Sage!
--
Comment:
Replying to [comment:17 mmezzarobba]:
> Replying to [comment:16 vdelecroix]:
> > You are right, I need to also patch gmp in order to make Sage works.
Another option is to wait for upstream upgrading from both GMP and mpir.
What do you think? I mostly wanted to test it.
>
> I'd wait.
Looks more reasonable to me as well (I do not like to patch upstream). So
I will push harder upstream to have these functions in the next releases.
--
Ticket URL: <http://trac.sagemath.org/ticket/18304#comment:18>
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.
