#16964: Speed up comparisons in QQbar
----------------------------------------------+------------------------
Reporter: gagern | Owner:
Type: defect | Status: new
Priority: critical | Milestone: sage-6.5
Component: number fields | Resolution:
Keywords: variety qqbar cmp singular | Merged in:
Authors: | Reviewers:
Report Upstream: N/A | Work issues:
Branch: | Commit:
Dependencies: | Stopgaps:
----------------------------------------------+------------------------
Changes (by jdemeyer):
* milestone: sage-6.4 => sage-6.5
Old description:
> I just spent 8 hours in an attempt to solve a polynomial system of
> equations using ideal.variety(). After these 8 hours I was left with the
> following error message:
>
> {{{
> File "sage/rings/polynomial/multi_polynomial_ideal.py", line 604, in
> __call__
> return self.f(self._instance, *args, **kwds)
> File "sage/rings/polynomial/multi_polynomial_ideal.py", line 2671, in
> variety
> V.sort()
> File "sage/rings/qqbar.py", line 3889, in __cmp__
> rcmp = cmp(self.real(), other.real())
> File "sage/rings/qqbar.py", line 4529, in __cmp__
> return self._sub_(other).sign()
> File "sage/rings/qqbar.py", line 4864, in sign
> return self.sign()
> File "sage/rings/qqbar.py", line 4867, in sign
> self.exactify()
> File "sage/rings/qqbar.py", line 3600, in exactify
> self._set_descr(self._descr.exactify())
> File "sage/rings/qqbar.py", line 7849, in exactify
> left.exactify()
> File "sage/rings/qqbar.py", line 3600, in exactify
> self._set_descr(self._descr.exactify())
> File "sage/rings/qqbar.py", line 7594, in exactify
> rv.exactify()
> File "sage/rings/qqbar.py", line 3600, in exactify
> self._set_descr(self._descr.exactify())
> File "sage/rings/qqbar.py", line 7849, in exactify
> left.exactify()
> File "sage/rings/qqbar.py", line 3600, in exactify
> self._set_descr(self._descr.exactify())
> File "sage/rings/qqbar.py", line 7851, in exactify
> gen = left._exact_field().union(right._exact_field())
> File "sage/rings/qqbar.py", line 2362, in union
> newpol, self_pol, k = pari_nf.rnfequation(my_factor, 1)
> File "gen.pyx", line 7454, in sage.libs.pari.gen.gen.rnfequation
> (build/cythonized/sage/libs/pari/gen.c:37964)
> File "handle_error.pyx", line 90, in
> sage.libs.pari.handle_error._pari_handle_exception
> (build/cythonized/sage/libs/pari/handle_error.c:1181)
> sage.libs.pari.gen.PariError: not enough precomputed primes
> }}}
>
> This is extremely annoying, since apparently the result of the
> computation was available at that point, and it was only sorting the
> result which failed. Neither the unsorted result nor the triangular
> decomposition had been cached, throwing away 8 hours worth of
> computation. So I suggest
>
> 1. wrap that sorting into a try-except block to return the result
> unsorted if there is any problem sorting things.
> 2. perhaps provide a switch to disable sorting up front, since repeated
> exactification might cost time.
> 3. Try to work out why exactly the comparison fails here, and avoid this
> error in the first place.
>
> The system of equations in this exact form was written by someone else,
> so I'll have to either reproduce this with my own (equivalent) system or
> request permission to post that code. In any case, it's rather lengthy.
>
> I'll also try to run the code on a modified sage which simply disables
> the sort in order to get my hands on the bare QQbar descriptions behind
> the failed computation.
New description:
I just spent 8 hours in an attempt to solve a polynomial system of
equations using ideal.variety(). After these 8 hours I was left with the
following error message:
{{{
File "sage/rings/polynomial/multi_polynomial_ideal.py", line 604, in
__call__
return self.f(self._instance, *args, **kwds)
File "sage/rings/polynomial/multi_polynomial_ideal.py", line 2671, in
variety
V.sort()
File "sage/rings/qqbar.py", line 3889, in __cmp__
rcmp = cmp(self.real(), other.real())
File "sage/rings/qqbar.py", line 4529, in __cmp__
return self._sub_(other).sign()
File "sage/rings/qqbar.py", line 4864, in sign
return self.sign()
File "sage/rings/qqbar.py", line 4867, in sign
self.exactify()
File "sage/rings/qqbar.py", line 3600, in exactify
self._set_descr(self._descr.exactify())
File "sage/rings/qqbar.py", line 7849, in exactify
left.exactify()
File "sage/rings/qqbar.py", line 3600, in exactify
self._set_descr(self._descr.exactify())
File "sage/rings/qqbar.py", line 7594, in exactify
rv.exactify()
File "sage/rings/qqbar.py", line 3600, in exactify
self._set_descr(self._descr.exactify())
File "sage/rings/qqbar.py", line 7849, in exactify
left.exactify()
File "sage/rings/qqbar.py", line 3600, in exactify
self._set_descr(self._descr.exactify())
File "sage/rings/qqbar.py", line 7851, in exactify
gen = left._exact_field().union(right._exact_field())
File "sage/rings/qqbar.py", line 2362, in union
newpol, self_pol, k = pari_nf.rnfequation(my_factor, 1)
File "gen.pyx", line 7454, in sage.libs.pari.gen.gen.rnfequation
(build/cythonized/sage/libs/pari/gen.c:37964)
File "handle_error.pyx", line 90, in
sage.libs.pari.handle_error._pari_handle_exception
(build/cythonized/sage/libs/pari/handle_error.c:1181)
sage.libs.pari.gen.PariError: not enough precomputed primes
}}}
This is extremely annoying, since apparently the result of the computation
was available at that point, and it was only sorting the result which
failed.
--
--
Ticket URL: <http://trac.sagemath.org/ticket/16964#comment:7>
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.
