Generalized Hermite formula for the two-sided Lagrange-Sylvester interpolation (Q1765265)

The two-sided Lagrange-Sylvester interpolation (TSLSI) can be formulated as follows: find all matrix-valued functions (or matrix polynomials) \(F(z)\) such that \(z_iF(c_i)=w_i\) for \(i=1,\ldots,N\); \(F(d_j)y_j=v_j\) for \(j=1,\ldots,M\); \(z_iF(\xi_{ij})y_i=\rho_{ij}\) for all \(i,j\) such that \(\xi_{ij}=c_i=d_j\), where \(z_i \in C^{1 \times p}\), \(w_i \in C^{1 \times q}\), \(c_i,d_j \in C\), \(y_j \in C^{p \times 1}\), \(\rho_{ij} \in C\). The authors consider the matrix polynomial case and give a generalization of the classical Hermite formula for polynomial interpolation to the case of TSLSI with a complete interpolation data; a parametrization of all matrix polynomial solutions to the TSLSI on the basis of the generalized Hermite formula is given.