id,summary,reporter,owner,description,type,status,priority,milestone,component,resolution,keywords,cc,merged,author,reviewer,upstream,work_issues,branch,commit,dependencies,stopgaps
10544,LLL reduced kernel bases are not always correct,Rob Beezer,Jason Grout William Stein,"On 4.6.1.alpha 3:
{{{
sage: sage: E=matrix(ZZ, 5, 8,
....: [5, -23, 21, 77, -50, 8, -76, 16, 2,
....: -9, 8, 30, -19, 3, -29, 6, 4, -16, 13,
....: 53, -30, 4, -47, 9, 0, 5, -14, -24,
....: 33, -5, 47, -14, -2, 11, -9, -35, 23,
....: -5, 34, -7])
sage: sage: KK=E.right_kernel(LLL=True)
sage: sage: BB=KK.basis_matrix().change_ring(ZZ)
sage: sage: BB.is_LLL_reduced()
False
sage: sage: LL=BB.LLL()
sage: sage: LL-BB
[0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0]
[1 0 1 0 1 1 0 1]
[0 0 0 0 0 0 0 0]
}}}
I can't say for certain that this is a bug, but it does not look correct to me. This may happen about one time in twenty for random matrices of this size. Any smaller and it seemed to be quite rare. Larger and it was more frequent. Wish I had a better test case to offer.
",defect,closed,major,sage-duplicate/invalid/wontfix,linear algebra,wontfix,days84,wjp,,,Thierry Monteil,N/A,,,,,