Closed darijgr closed 9 years ago
Branch pushed to git repo; I updated commit sha1. New commits:
cda0403 | more sensible doc, and another doctest |
Could we use some for loops?
sage: p = lambda t, k: t.bender_knuth_involution(k).bender_knuth_involution(k + 1)
sage: t = Tableau([[1,2,2],[3,4]])
sage: x = t
sage: for i in range(6): x = p(x, 2)
sage: x
[[1, 2, 3], [2, 4]]
sage: x == t
False
If you agree and make these changes, you can set a positive review on my behalf.
Branch pushed to git repo; I updated commit sha1. New commits:
abbfda2 | Rewrite doctests following Travis's suggestion |
Branch pushed to git repo; I updated commit sha1. New commits:
7deac5c | Fix a typo |
Reviewer: Travis Scrimshaw
Changed branch from u/darij/benderknuth to 7deac5c
I claimed that
(s_i s_{i+1})^6 = id
, wheres_k
means thek
-th Bender-Knuth involution on semistandard skew tableaux. I even made a doctest that unfortunately used a hook shape, which renders the Bender-Knuth involutions rather boring (though maybe it would make a nice exercise to check it in this case -- although I don't know if it is always true there).The claim is false. There is a counterexample for skew tableaux and
i = 1
(or straight tableaux andi = 2
). The source where I had it from, a paper by Berenstein and Kirillov, only claimed it for straight tableaux andi = 1
(and that is indeed correct).Thanks to Pavel Galashin for finding a counterexample!
CC: @tscrim @anneschilling @sagetrac-sage-combinat @nthiery @sagetrac-jkeitel @sagetrac-jpswanson @MariaMonks
Component: combinatorics
Keywords: tableaux, bender-knuth involutions, sage-combinat
Author: Darij Grinberg
Branch/Commit:
7deac5c
Reviewer: Travis Scrimshaw
Issue created by migration from https://trac.sagemath.org/ticket/18686