On a family of conjectures of Joel Lewis on alternating permutations (Q2014704)

The word permutation is used in its classical sense -- for a string of designated symbols (here integers \(1, 2, \dots\)) without repetitions. A permutation \(p=p_1p_2\dots p_n\) of length \(n\) avoids a permutation \(q=q_1q_2\dots q_k\) of length \(k\) if there are no \(k\) indices \(i_1<i_2<\cdots<i_k\) so that \(p_{i_a}<p_{i_b}\) iff \(q_a<q_b\) \((a,b\in\{1,2,\dots,k\})\); \(p=p_1p_2\dots p_n\) is alternating if \(p_i<p_{i+1}\) iff \(i\) is odd. \(A_n(q)\) denotes the number of alternating permutations of length \(n\) that avoid a permutation \(q\). Theorem 1: Let \(k\geq3\) be an integer, and let \(n\) be an even positive integer. Then we have \(A_n(1,2,\dots,k-2,k-1,k)=A_n(1,2,\dots,k-2,k,k-1)\). This theorem generalizes a conjecture of \textit{J. B. Lewis} [Electron. J. Comb. 19, No. 1, Research Paper P21, 21 p. (2012; Zbl 1243.05011)] that, for all positive integers \(n\), \(A_{2n}(1234)=A_{2n}(1243)\) and \(A_{2n}(12345)=A_{2n}(12354)\). From \S 3: ``Finally, we use a slightly modified version of our argument to prove \(\dots\) Theorem 2: Let \(n\) be any positive integer. Then, for all \(k\), we have \(A_n(1,2,3,\dots,k)=A_n(2,1,3,\dots,k)\).'' (Unlike Theorem 1, this theorem has no parity restriction on \(n\).)