#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
---------------------------+------------------------------------------------
Comment(by SimonKing):
The crucial question: Does it fix the speed regression?
Here are Jeroen's examples from sage-devel.
'''sage-4.7.2.alpha2'''
{{{
sage: %time D = J0(46).endomorphism_ring().discriminant()
CPU times: user 5.60 s, sys: 0.21 s, total: 5.81 s
Wall time: 5.83 s
sage: %time TestSuite(CrystalOfTableaux(['B',4],shape=[2,1,1,0])).run()
CPU times: user 7.08 s, sys: 0.03 s, total: 7.11 s
Wall time: 7.75 s
sage: W.<z> = CyclotomicField(13)
sage: %time M = Matrix(W, 2, 3, [10^30*(1-z)^13, 1, 2, 3, 4,
z]).echelon_form()
CPU times: user 3.29 s, sys: 0.03 s, total: 3.32 s
Wall time: 3.38 s
sage: %time L = EllipticCurve('960d1').prove_BSD()
CPU times: user 3.73 s, sys: 0.04 s, total: 3.78 s
Wall time: 4.01 s
sage: def test(E):
....: for p in prime_range(10000):
....: if p != 389:
....: G = E.change_ring(GF(p)).abelian_group()
....:
sage: E = EllipticCurve('389a')
sage: %time test(E)
CPU times: user 16.47 s, sys: 0.04 s, total: 16.50 s
Wall time: 16.73 s
sage: %time for E in cremona_curves([11..100]): S =
E.integral_points(both_signs=False)
CPU times: user 13.88 s, sys: 0.06 s, total: 13.94 s
Wall time: 14.01 s
}}}
'''sage-4.7.2.alpha2 plus #9138 and its dependency'''
{{{
sage: %time D = J0(46).endomorphism_ring().discriminant()
CPU times: user 8.80 s, sys: 0.15 s, total: 8.95 s
Wall time: 8.98 s
sage: %time TestSuite(CrystalOfTableaux(['B',4],shape=[2,1,1,0])).run()
CPU times: user 7.11 s, sys: 0.03 s, total: 7.14 s
Wall time: 7.48 s
sage: W.<z> = CyclotomicField(13)
sage: %time M = Matrix(W, 2, 3, [10^30*(1-z)^13, 1, 2, 3, 4,
z]).echelon_form()
CPU times: user 6.07 s, sys: 0.02 s, total: 6.09 s
Wall time: 6.13 s
sage: %time L = EllipticCurve('960d1').prove_BSD()
CPU times: user 10.10 s, sys: 0.07 s, total: 10.17 s
Wall time: 10.55 s
sage: def test(E):
....: for p in prime_range(10000):
....: if p != 389:
....: G = E.change_ring(GF(p)).abelian_group()
....:
sage: E = EllipticCurve('389a')
sage: %time test(E)
CPU times: user 23.03 s, sys: 0.05 s, total: 23.08 s
Wall time: 23.18 s
sage: %time for E in cremona_curves([11..100]): S =
E.integral_points(both_signs=False)
CPU times: user 16.29 s, sys: 0.05 s, total: 16.34 s
Wall time: 16.48 s
}}}
'''sage-4.7.2.alpha2 plus #9138 and its dependency plus the patch from
here'''
{{{
sage: %time D = J0(46).endomorphism_ring().discriminant()
CPU times: user 6.02 s, sys: 0.20 s, total: 6.22 s
Wall time: 6.23 s
sage: %time TestSuite(CrystalOfTableaux(['B',4],shape=[2,1,1,0])).run()
CPU times: user 7.11 s, sys: 0.01 s, total: 7.12 s
Wall time: 7.39 s
sage: W.<z> = CyclotomicField(13)
sage: %time M = Matrix(W, 2, 3, [10^30*(1-z)^13, 1, 2, 3, 4,
z]).echelon_form()
CPU times: user 3.81 s, sys: 0.04 s, total: 3.84 s
Wall time: 3.85 s
sage: %time L = EllipticCurve('960d1').prove_BSD()
CPU times: user 6.11 s, sys: 0.07 s, total: 6.18 s
Wall time: 6.20 s
sage: def test(E):
....: for p in prime_range(10000):
....: if p != 389:
....: G = E.change_ring(GF(p)).abelian_group()
....:
sage: E = EllipticCurve('389a')
sage: %time test(E)
CPU times: user 18.11 s, sys: 0.07 s, total: 18.17 s
Wall time: 18.23 s
sage: %time for E in cremona_curves([11..100]): S =
E.integral_points(both_signs=False)
CPU times: user 13.13 s, sys: 0.05 s, total: 13.18 s
Wall time: 13.22 s
}}}
'''__Conclusion__'''
#9138 is indeed responsible for all but one regression. The only exception
is the test suite for `CrytalOfTableaux`; it has already been argued on
sage-devel that it is actually not a regression, since the test suite
became much longer by #11183.
The patch from here fixes most regressions. Problematic remain the
`EllipticCurve('960d1').prove_BSD()` and the
`E.change_ring(GF(p)).abelian_group()` tests.
--
Ticket URL: <http://trac.sagemath.org/sage_trac/ticket/11900#comment:14>
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.
