Hankel determinants of sums of consecutive Motzkin numbers (Q616406): Difference between revisions

The Motzkin numbers~\(m_0,m_1,\dots\) are defined by \[ \frac{1-z-\sqrt{1-2z-3z^2}}{2z^2}=\sum_{k=0}^\infty\, m_kz^k. \] Given a real number \(t\), let \(m^t_k\) be the total weight of all paths from~\((0,0)\) to~\((k,0)\) that stay above the \(x\)-axis and use ``up steps'' \((1,1)\) of weight one, ``down steps'' \((1,-1)\) of weight one, and ``level steps'' \((1,0)\) of weight \(t\). (The weight of a path is the product of weights of all its steps.) Call \(m^t_0,m^t_1,\dots\) the \(t\)-Motzkin numbers. Now \(m^1_k=m_k\), \(m^0_{2k}=c_k\) and \(m^2_k=c_{k+1}\), where \(c_k=\frac{1}{k+1}{2k\choose k}\), the \(k\)th Catalan number. Let \(r\) be a nonnegative integer. The authors study the determinant \[ \det(m^t_{i+j+r}+m^t_{i+j+r+1})_{0\leq i,j\leq n-1}. \] Their main results are a recursive formula (with respect to \(n\)) in the case of \(r=0\) and an explicit formula (applying Chebychev polnomials of the second kind) in the case of \(r=1\).