ME50: Sparse unsymmetric system: conventional direct method

These subroutines are for the solution of a general sparse \(m \times n\) system of complex linear equations (the most usual case being square, \(m = n\)), stored by columns. No block triangularization, iterative refinement, or error estimation is included.

If the user requires a more convenient data interface, the ME48 package should be used. The ME48 subroutines call the ME50 subroutines after checking and sorting the user’s input data and optionally using MC21 and MC13 to permute the matrix to block triangular form.