--- title: Para-Krawtchouk Polynomials Overview url: https://www.emergentmind.com/topics/para-krawtchouk-polynomials type: topic --- # Para-Krawtchouk Polynomials Overview Para-Krawtchouk polynomials are a finite family of orthogonal polynomials defined on a bi-lattice, introduced by Vinet and Zhedanov in the analysis of quantum spin chains with perfect state transfer and later identified as a special case of a complete bilattice-classical classification of orthogonal polynomial systems. Their defining feature is that the spectral support is not a single uniform lattice but the union of two interlaced arithmetic progressions; in this sense they are a bilattice analogue of the classical Krawtchouk family, while retaining explicit recurrence relations, orthogonality, difference equations, and algebraic structure [1110.6475] [2311.05636]. ## 1. Bilattice setting and defining spectral data The bilattice framework used for para-Krawtchouk polynomials is based on \[ x(s)=s+\gamma\bigl(1+(-1)^s\bigr), \qquad s\in\mathbb C, \] with \(\gamma\in\mathbb C\). This is a two-component lattice in which even and odd values of \(s\) are shifted differently by the parameter \(\gamma\). In the finite spin-chain realization, the spectrum is written as \[ x_s = s + \frac{\gamma}{2}(1-\delta_s), \qquad s=0,1,\ldots,N, \] where \(N\) is odd and \(0<\gamma<2\). Equivalently, the support splits into the two uniform sublattices \[ x_{2s}=2s,\qquad x_{2s+1}=2s+\gamma. \] When \(\gamma=1\), the two sublattices merge into the uniform lattice \(x_s=s\), which recovers the classical Krawtchouk situation [2311.05636] [1110.6475]. This bilattice geometry is the structural feature that distinguishes the family from standard discrete orthogonal polynomials on a linear lattice. It also explains why para-Krawtchouk polynomials arise naturally in inverse spectral problems for Jacobi matrices: the finite spectrum is prescribed as a union of two arithmetic progressions, and the orthogonal polynomials are then determined by the associated tridiagonal matrix. ## 2. Recurrence relations and orthogonality In the classification paper, the para-Krawtchouk family is presented as the monic orthogonal polynomial sequence \((P_n(\cdot;M,N))_{n\ge 0}\) satisfying \[ B_n=-\frac{N-1}{2}, \qquad C_{n+1} = \frac{(n+1)(n-N)(2n+1-N-p)(2n+1-N+p)}{4(2n-N)(2n-N+2)}, \] with \(00\). Their orthogonality relation is \[ \sum_{s=0}^N P_n(x_s)P_m(x_s)w_s = h_n\,\delta_{nm}, \] with \(h_n=U_1U_2\cdots U_n\). For the finite bi-lattice spectrum, the weights are given explicitly by \[ w_{2s}=2^{-N}\left(1-\frac{\gamma}{2}\right)^J \frac{(-J)_s\left(-\frac{\gamma}{2}-J\right)_s}{\left(\frac12\right)_s\,s!\left(1-\frac{\gamma}{2}\right)_s}, \] and \[ w_{2s+1}=2^{-N}\left(1+\frac{\gamma}{2}\right)^J \frac{(-J)_s\left(\frac{\gamma}{2}-J\right)_s}{\left(\frac12\right)_s\,s!\left(1+\frac{\gamma}{2}\right)_s}, \] where \(J=(N-1)/2\), and they satisfy the normalization \[ \sum_{s=0}^N w_s=1. \] The recurrence data are equally explicit: \[ B_n = \frac{N-1+\gamma}{2}, \] \[ U_n=J_n^2=\frac{n(N+1-n)\big((N+1-2n)^2-\gamma^2\big)}{4(N-2n)(N+2-2n)}, \] together with the factorization \[ U_n=A_{n-1}C_n,\qquad B_n=-A_n-C_n, \] where \[ A_n=\frac{(N-n)(N+1-2n+\gamma)}{2(2n-N)}, \qquad C_n=\frac{n(N+1-2n-\gamma)}{2(2n-N)}. \] These formulas are central in applications because they determine the Jacobi matrix and hence the finite orthogonality measure [1110.6475]. ## 3. Bilattice-classical characterization The modern structural description of para-Krawtchouk polynomials is formulated in terms of the bilattice analogue of the classical functional equation \[ D_s(\varphi u)=S_s(\psi u), \] where \(u\in\mathcal P^*\) is a linear functional, \(\varphi\in\mathcal P_2\), \(\psi\in\mathcal P_1\), and \[ D_sf(x(s))=\frac{f(x(s+1))-f(x(s-1))}{2},\qquad S_sf(x(s))=\frac{f(x(s+1))+f(x(s-1))}{2}. \] For \[ \varphi(z)=az^2+bz+c,\qquad \psi(z)=dz+e, \] and \[ d_n:=an+d,\qquad e_n:=bn+e, \] the functional \(u\) is regular if and only if \[ d_n\neq 0,\qquad e_n-(-1)^s\gamma\bigl(1-(-1)^n\bigr)\neq 0 \] for each \(n=0,1,2,\dots\). In this setting, regularity means that \(u\) admits an orthogonal polynomial sequence \((P_n)\) such that \[ (u,P_nP_k)=h_n\delta_{nk},\qquad h_n\neq 0. \] The admissibility of \((\varphi,\psi)\) is encoded by these nonvanishing conditions [2311.05636]. A major consequence is the bilattice Rodrigues structure. If \(u\) is regular and satisfies the bilattice functional equation, then the monic orthogonal polynomial sequence \((P_n)\) associated with \(u\) satisfies \[ P_nu = k_n D_s^n[u_n],\qquad k_n:=(-1)^n\prod_{j=1}^n d_{n+j-2}. \] Equivalently, for the auxiliary sequence \((R_n)\), \[ R_nu = D_s^n u[n]. \] The iterative transformed functionals \(u[k]\) remain regular for all \(k\ge 1\), and the orthogonal polynomial sequence associated with \(u[k]\) is again an orthogonal polynomial system. This places para-Krawtchouk polynomials among the families that are classical in the bilattice sense rather than merely semi-classical [2311.05636]. A recurrent misconception is to regard para-Krawtchouk polynomials as an isolated ad hoc deformation. The bilattice functional-equation theory shows instead that they arise naturally as one admissible solution of a general classical problem on a bi-lattice. ## 4. Classification and relation to the Krawtchouk family The classification theorem states that, up to an affine change of variable, the only classical orthogonal polynomial systems on the bilattice are the two families \[ (H_n(\,\cdot\,;a,b))_{n\ge0} \quad\text{and}\quad (Q_n(\,\cdot\,;a,b,c))_{n\ge0}. \] The standard classical discrete families on the linear lattice then appear as specializations of the bilattice theory. In particular, the paper recovers Meixner, Charlier, Krawtchouk, Hahn, and para-Krawtchouk polynomials by choosing special \(\varphi\) and \(\psi\) in the bilattice functional equation. The organizing principle is: - if \(\deg\varphi=0\), one gets a Charlier-type system; - if \(\deg\varphi=1\), one gets a Meixner/Krawtchouk-type system; - if \(\deg\varphi=2\), one obtains Hahn-type and para-Krawtchouk-type systems. The explicit identifications include \[ C_n(z,a)=(-2)^{-n}H_n(-2(z-a);0,4a), \] \[ K_n(z,p,N)=\text{an affine transform of }H_n, \] \[ L_n(z;a,b,N)=\text{an affine transform of }Q_n, \] and \[ P_n(z;M,N) = Q_n\!\left( z;\,\frac{1-N}{2},\,\frac12,\,0 \right). \] Within this scheme, para-Krawtchouk polynomials are not outside the classical hierarchy; they occupy a specific point in the \(Q_n\)-family [2311.05636]. The classical Krawtchouk family provides the immediate backdrop. For parameters \[ n,m,s\in \mathbb{N},\qquad m\le n,\qquad s\ge 2, \] the Krawtchouk polynomial is defined by \[ K_{m,n,s}(X)=\sum_{j=0}^{m}(-1)^j\,P_j(X)\,P_{m-j}(n-X)\,(s-1)^{m-j}, \] where \[ P_0(X)=1,\qquad P_j(X)=X(X-1)\cdots(X-j+1)\quad (j\ge 1). \] With respect to the inner product \[ (A,B)=\sum_{i=0}^{n}\binom{n}{i}(s-1)^i\,A(i)B(i), \] these polynomials satisfy \[ (K_k,K_\ell)= \begin{cases} 0, & k\ne \ell,\\[4pt] s^n (s-1)^k \binom{n}{k}, & k=\ell. \end{cases} \] They obey the three-term recurrence \[ (m+1)K_{m+1} = \bigl(m+(s-1)(n-m)-sX\bigr)K_m -(s-1)(n-m+1)K_{m-1}, \] and for \(1\le m\le n\) have \(m\) distinct real roots in \((0,n)\), with interlacing of consecutive zeros. In the binary case \(s=2\), \[ K_{m,n,2}(n-x)=(-1)^m K_{m,n,2}(x). \] This provides the classical finite orthogonal template from which para-Krawtchouk systems depart by replacing the single lattice with a bi-lattice [1101.1798]. ## 5. Spin-chain origin, bispectral structure, and algebraic realization Para-Krawtchouk polynomials were introduced in the study of the XX Hamiltonian \[ H=\frac12\sum_{l=0}^{N-1}J_{l+1}\big(\sigma_l^x\sigma_{l+1}^x+\sigma_l^y\sigma_{l+1}^y\big)+\frac12\sum_{l=0}^N B_l(\sigma_l^z+1), \] whose one-excitation restriction is a Jacobi matrix \(J\). Perfect state transfer occurs when the spectrum satisfies the odd-spacing condition \[ x_{s+1}-x_s=\frac{\pi}{T}M_s, \qquad M_s\in 2\mathbb{Z}+1, \] and the chain is mirror-symmetric: \[ RJR=J. \] For the para-Krawtchouk family, the recurrence coefficients supply explicit couplings and fields, and the perfect-transfer amplitude is \[ e^{-iHT}|0\rangle = e^{i\phi}|N\rangle. \] Under an affine rescaling \[ \widetilde{x}_s = ax_s+\beta, \] the recurrence coefficients transform as \[ \widetilde{B}_n=aB_n+\beta,\qquad \widetilde{U}_n=a^2U_n, \] and the perfect-state-transfer normalization imposes \[ \gamma=\frac{M_1}{M_2}, \] where \(M_1,M_2\) are positive coprime integers and \(M_1\) is odd [1110.6475]. The family also satisfies a second-order difference equation \[ E(x)P_n(x+2)+F(x)P_n(x-2)-\big(E(x)+F(x)\big)P_n(x)=2n(n-N)P_n(x), \] with \[ E(x)=(x-N+1)(x-N+1-\gamma),\qquad F(x)=x(x-\gamma). \] Because the shifts \(x\mapsto x\pm2\) preserve each sublattice, this is interpreted as a fourth-order relation on the underlying bi-lattice. When \(\gamma=1\), one recovers the classical symmetric Krawtchouk case, \[ B_n=0,\qquad U_n=\frac{n(N+1-n)}{4}, \] and the difference equation becomes the square of the Krawtchouk eigenvalue equation. The explicit hypergeometric realization is obtained through complementary Bannai–Ito polynomials, and the characteristic polynomial factorizes as \[ P_{N+1}(x)=\text{const}\,(x-\gamma)_{M+1}(x)_{M+1}, \] so that its roots are exactly the bi-lattice points [1110.6475]. The algebraic structure behind the family is the quadratic Hahn algebra. With operators \[ Xf(x)=xf(x), \] \[ Yf(x)=E(x)f(x+2)+F(x)f(x-2)-\big(E(x)+F(x)\big)f(x), \] and \[ Z=[X,Y]=XY-YX, \] the commutation relations are \[ [Y,Z]=-4\{X,Y\}+C_1X+GY+M_1, \qquad [Z,X]=-4X^2+C_2Y+GX+M_2, \] with \[ C_1=4(1-N^2),\qquad C_2=-4,\qquad G=4(N+\gamma-1), \] \[ M_1=2(N+\gamma-1)(N^2-1),\qquad M_2=2(1-N)(N+\gamma-1). \] The Casimir operator acts as a scalar, and the eigenbasis of \(Y\) is precisely the para-Krawtchouk basis: \[ YT_n=\lambda_nT_n,\qquad \lambda_n=2n(n-N). \] This gives a direct algebraic explanation of the recurrence structure [1110.6475]. ## 6. \(q\)-generalizations and later algebraic developments A \(q\)-generalization of the para-Krawtchouk family is obtained from a singular truncation of the Askey–Wilson polynomials. The truncation condition is \[ 1-abcd\,q^{N-1}=0. \] After regularization and passage to a finite family \(R_n(x;a,c,\alpha\mid q)\) of \(q\)-para-Racah polynomials, a scaling limit produces the \(q\)-para-Krawtchouk polynomials. Writing \[ \theta=ac,\qquad \Delta=\frac{a}{c}, \] and rescaling by \[ x=\frac{\theta}{2a}y,\qquad Q_n(y)=\left(\frac{\theta}{2a}\right)^{-n}R_n(x), \] the support becomes the exponential bi-lattice \[ y_{2s}=\Delta q^s,\qquad y_{2s+1}=q^s. \] The resulting family satisfies \[ yQ_n(y)=Q_{n+1}(y)+\tilde b_nQ_n(y)+\tilde u_nQ_{n-1}(y), \] and is orthogonal on that exponential bi-lattice. In this construction, the family is bispectral, but the eigenvalues on the \(q\)-difference side are doubly degenerate after finite truncation, so it is not classical in the strict Leonard-pair sense [1708.03368]. A further development gives an algebraic interpretation through S-Heun operators on linear and \(q\)-linear grids. In the discrete linear case, truncation of the Continuous Hahn family occurs when \[ 1-(a+b+c+d)=N, \] and yields the para-Krawtchouk polynomials. In the \(q\)-linear case, truncation of Big \(q\)-Jacobi occurs when \[ \sqrt{abcd}=q^{1-N}, \] and yields the \(q\)-para-Krawtchouk family. The central claim is that these polynomial systems are finite-dimensional representation bases of Sklyanin-like algebras: \(Skl_4\) in the discrete linear case and \(U_q(sl_2)\) in the \(q\)-linear case. This resolves what the paper describes as a missing algebraic interpretation of para-Krawtchouk and \(q\)-para-Krawtchouk polynomials. The same framework also explains the subtle dimension count: the degree-\(N+1\) polynomial annihilated by the raising operator is the characteristic polynomial of the upper block of the truncated Jacobi matrix, vanishes on the orthogonality grid, and therefore represents a null vector; the true representation dimension is \(N+1\) [2008.03266]. Taken together, these developments place para-Krawtchouk polynomials at the intersection of finite orthogonal polynomial theory, inverse spectral problems for Jacobi matrices, exact spin-chain design, and nontrivial algebraic structures on discrete and \(q\)-discrete grids. The classification on bi-lattices shows that they belong to the classical orthogonal-polynomial landscape in a precise bilattice sense, while the spin-chain and Sklyanin-like perspectives explain why they continue to recur in finite-dimensional spectral models. Source: https://www.emergentmind.com/topics/para-krawtchouk-polynomials