gstbmm routines multiply a symmetric Toeplitz block diagonal matrix, i.e., as used by stbmm routines, but with some sets of rows and corresponding columns removed, by an arbitrary data matrix. Such an effective matrix, referred hereafter as a gstb (a gappy, symmetric, Toeplitz block matrix), in general will not be made of only Toeplitz blocks anymore. (See the example below). On the input this matrix is however still represented as a corresponding stb matrix complemented by a list of column (or row) which are to be neglected (or effectively excised). The operation is then construed also as a stbmm operation, which the sbt matrix representating the gstb and, and the data matrix with the corresponding rows set to zero. Note that the data matrix is assumed to have all the rows on the input including the ones to be neglected.

On the output the routine will produce a matrix of the same type as the input one with the values in the rows to be neglected set to zero.

Example.