DSXT1  computes the difference between a set of eigenvalues.
 The result is returned as the function value.
 IJOB = 1:   Computes   max { min | D1(i)-D2(j) | }
                         i     j
 IJOB = 2:   Computes   max { min | D1(i)-D2(j) | /
                         i     j
                              ( ABSTOL + |D1(i)|*ULP ) }

IJOB

          IJOB is INTEGER
          Specifies the type of tests to be performed.  (See above.)

D1

          D1 is DOUBLE PRECISION array, dimension (N1)
          The first array.  D1 should be in increasing order, i.e.,
          D1(j) <= D1(j+1).

N1

          N1 is INTEGER
          The length of D1.

D2

          D2 is DOUBLE PRECISION array, dimension (N2)
          The second array.  D2 should be in increasing order, i.e.,
          D2(j) <= D2(j+1).

N2

          N2 is INTEGER
          The length of D2.

ABSTOL

          ABSTOL is DOUBLE PRECISION
          The absolute tolerance, used as a measure of the error.

ULP

          ULP is DOUBLE PRECISION
          Machine precision.

UNFL

          UNFL is DOUBLE PRECISION
          The smallest positive number whose reciprocal does not
          overflow.

Univ. of Tennessee

Univ. of California Berkeley

Univ. of Colorado Denver

NAG Ltd.