Use of Subroutine VFILL

The vector U should contain an initial approximation to the solution of the linear system before NSPCG is called. If the user has no information for making such a guess, then the zero vector may be used as the starting vector. The subroutine VFILL can be used to fill a vector with a constant. For example,

       CALL VFILL (N,U,VAL)

fills the first N locations of the vector U with value VAL.