Routines callable from Fortran and C are described which implement matrix-matrix multiplication and transposition for a variety of sparse matrix formats. Conversion routines between various formats are provided. © 1993 J.C. Baltzer AG, Science Publishers.
