NAME

lasq4 - lasq4: dqds step

SYNOPSIS

Functions

subroutine dlasq4 (i0, n0, z, pp, n0in, dmin, dmin1, dmin2, dn, dn1, dn2, tau, ttype, g)
DLASQ4 computes an approximation to the smallest eigenvalue using values of d from the previous transform. Used by sbdsqr. subroutine slasq4 (i0, n0, z, pp, n0in, dmin, dmin1, dmin2, dn, dn1, dn2, tau, ttype, g)
SLASQ4 computes an approximation to the smallest eigenvalue using values of d from the previous transform. Used by sbdsqr.

Detailed Description

Function Documentation

subroutine dlasq4 (integer i0, integer n0, double precision, dimension( * ) z, integer pp, integer n0in, double precision dmin, double precision dmin1, double precision dmin2, double precision dn, double precision dn1, double precision dn2, double precision tau, integer ttype, double precision g)

DLASQ4 computes an approximation to the smallest eigenvalue using values of d from the previous transform. Used by sbdsqr.

Purpose:



 DLASQ4 computes an approximation TAU to the smallest eigenvalue


 using values of d from the previous transform.

Parameters



          I0 is INTEGER


        First index.



          N0 is INTEGER


        Last index.



          Z is DOUBLE PRECISION array, dimension ( 4*N0 )


        Z holds the qd array.



          PP is INTEGER


        PP=0 for ping, PP=1 for pong.

N0IN



          N0IN is INTEGER


        The value of N0 at start of EIGTEST.

DMIN



          DMIN is DOUBLE PRECISION


        Minimum value of d.

DMIN1



          DMIN1 is DOUBLE PRECISION


        Minimum value of d, excluding D( N0 ).

DMIN2



          DMIN2 is DOUBLE PRECISION


        Minimum value of d, excluding D( N0 ) and D( N0-1 ).



          DN is DOUBLE PRECISION


        d(N)

DN1



          DN1 is DOUBLE PRECISION


        d(N-1)

DN2



          DN2 is DOUBLE PRECISION


        d(N-2)

TAU



          TAU is DOUBLE PRECISION


        This is the shift.

TTYPE



          TTYPE is INTEGER


        Shift type.



          G is DOUBLE PRECISION


        G is passed as an argument in order to save its value between


        calls to DLASQ4.

Author

Univ. of Tennessee

Univ. of California Berkeley

Univ. of Colorado Denver

NAG Ltd.

Further Details:



  CNST1 = 9/16

Definition at line 149 of file dlasq4.f.

subroutine slasq4 (integer i0, integer n0, real, dimension( * ) z, integer pp, integer n0in, real dmin, real dmin1, real dmin2, real dn, real dn1, real dn2, real tau, integer ttype, real g)

SLASQ4 computes an approximation to the smallest eigenvalue using values of d from the previous transform. Used by sbdsqr.