Matrix transformations on generalized sequence spaces (Q1063218)

Let \(\Lambda(X)\) and \(\Lambda(Y)\) be two vector-valued sequence spaces (VVSS) formed by the sequences from locally convex spaces \((X,T)\) and \((Y,S)\) respectively. Then a linear map \(Z: \Lambda(X)\to \Lambda(Y)\) is said to be a T-matrix transformation or a matrix transformation if there is a matrix \([Z_{ij}]\) of linear maps \(Z_{ij}: X\to Y\), \(i,j=1,2,3,..\). such that for each \(\bar x=\{x_ i\}\in \Lambda(x)\), the series \(\sum_{j\geq 1}Z_{ij}(x_ j)\) converges to \(y_ i\) in \((Y,T)\) and \(Z(\bar x)=\{y_ i\}.\) We find restrictions on the spaces so as to represent a continuous linear transformation in terms of an infinite matrix of continuous linear operators and vice-versa in the most general setting. We also deal with the transpose of such matrix of linear operators defined in terms of adjoint linear maps. Finally, we give characterizations of nuclear and precompact diagonal operators on VVSS.