Skip to main content
updated broken link
Source Link
Christian Stump
  • 3.3k
  • 1
  • 22
  • 32

The promised writeup, proving both conjectures, is now finished. It will be available on the arxiv on Friday, in the meantime it is available on the herearXiv.

The promised writeup, proving both conjectures, is now finished. It will be available on the arxiv on Friday, in the meantime it is available here.

The promised writeup, proving both conjectures, is now finished. It will be available on the arxiv on Friday, in the meantime it is available on the arXiv.

Bounty Awarded with 400 reputation awarded by Mare
provide paper
Source Link
Martin Rubey
  • 6k
  • 2
  • 28
  • 45

UPDATE

The promised writeup, proving both conjectures, is now finished. It will be available on the arxiv on Friday, in the meantime it is available here.

FindStat's original answer

FindStat's original answer

UPDATE

The promised writeup, proving both conjectures, is now finished. It will be available on the arxiv on Friday, in the meantime it is available here.

FindStat's original answer

include outline of proof
Source Link
Martin Rubey
  • 6k
  • 2
  • 28
  • 45

EDIT: We now know how to prove the following and will update the answer with an outline.

FindStat's original answer

Then it appears that the number of $2$-Gorenstein failures is the composition of http://findstat.org/Mp00129 (The Billey-JokuschJockusch-Stanley bijection to 321-avoiding permutations) and http://findstat.org/St000732 (The number of double deficiencies of a permutation).

Outline of proof

HereThroughout the proof, a Dyck path starts at the origin, has north and east steps and never goes below the diagonal $y=x$. We think of the Dyck path as running in an array of columns labelled $1$ to $n$ from left to right, and rows labelled $1$ to $n$ from bottom to top.

Let us first recall the bijection due to Billey, Jockusch and Stanley, sending a Dyck path $D$ to a $321$-avoiding permutation $\pi$. To obtain the (horizontally flipped) permutation matrix of $\pi$ corresponding to $D$, put a cross into each valley (an east step followed by a north step) and fill the remaining slots with an increasing sequence of crosses.

Our claim is an immediate consequence of the Sage code usingfollowing three statements.

Lemma 1. $c_{x+1} = c_x - 1$ if and only if $x$ is a weak deficiency of $\pi$, that is $\pi(x)\leq x$. Equivalently, there is a double east step spanning columns $x$ and $x+1$.

Lemma 2. $c_{x+1} = c_x - 1$ and $c_{x+1} + d_x = d_{x+c_{x+1}}$ if and only if $x$ is a fixed point of $\pi$.

Lemma 3. Let $x$ be a strict deficiency of $\pi$. Then $\pi^{-1}(x)$ is also a strict deficiency if and only if $x+1=a+b$ where $a$ is a descent and $b\in X_a$.

We intend to provide a detailed and illustrated version of this argument within the FindStat interface:next few days.

Sage code used to discover the conjecture

EDIT: We now know how to prove the following and will update the answer with an outline.

Then it appears that the number of $2$-Gorenstein failures is the composition of http://findstat.org/Mp00129 (The Billey-Jokusch-Stanley bijection to 321-avoiding permutations) and http://findstat.org/St000732 (The number of double deficiencies of a permutation).

Here is the Sage code using the FindStat interface:

FindStat's original answer

Then it appears that the number of $2$-Gorenstein failures is the composition of http://findstat.org/Mp00129 (The Billey-Jockusch-Stanley bijection to 321-avoiding permutations) and http://findstat.org/St000732 (The number of double deficiencies of a permutation).

Outline of proof

Throughout the proof, a Dyck path starts at the origin, has north and east steps and never goes below the diagonal $y=x$. We think of the Dyck path as running in an array of columns labelled $1$ to $n$ from left to right, and rows labelled $1$ to $n$ from bottom to top.

Let us first recall the bijection due to Billey, Jockusch and Stanley, sending a Dyck path $D$ to a $321$-avoiding permutation $\pi$. To obtain the (horizontally flipped) permutation matrix of $\pi$ corresponding to $D$, put a cross into each valley (an east step followed by a north step) and fill the remaining slots with an increasing sequence of crosses.

Our claim is an immediate consequence of the following three statements.

Lemma 1. $c_{x+1} = c_x - 1$ if and only if $x$ is a weak deficiency of $\pi$, that is $\pi(x)\leq x$. Equivalently, there is a double east step spanning columns $x$ and $x+1$.

Lemma 2. $c_{x+1} = c_x - 1$ and $c_{x+1} + d_x = d_{x+c_{x+1}}$ if and only if $x$ is a fixed point of $\pi$.

Lemma 3. Let $x$ be a strict deficiency of $\pi$. Then $\pi^{-1}(x)$ is also a strict deficiency if and only if $x+1=a+b$ where $a$ is a descent and $b\in X_a$.

We intend to provide a detailed and illustrated version of this argument within the next few days.

Sage code used to discover the conjecture

added code
Source Link
FindStat
  • 849
  • 4
  • 10
Loading
correct conjecture
Source Link
Martin Rubey
  • 6k
  • 2
  • 28
  • 45
Loading
Source Link
FindStat
  • 849
  • 4
  • 10
Loading