1SLARRR(1)           LAPACK auxiliary routine (version 3.1)           SLARRR(1)
2
3
4

NAME

6       SLARRR  - to decide whether the symmetric tridiagonal matrix T warrants
7       expensive computations which guarantee high relative  accuracy  in  the
8       eigenvalues
9

SYNOPSIS

11       SUBROUTINE SLARRR( N, D, E, INFO )
12
13           INTEGER        N, INFO
14
15           REAL           D( * ), E( * )
16

PURPOSE

18       Perform tests to decide whether the symmetric tridiagonal matrix T war‐
19       rants expensive computations which guarantee high relative accuracy  in
20       the eigenvalues.
21
22

ARGUMENTS

24       N       (input) INTEGER
25               The order of the matrix. N > 0.
26
27       D       (input) REAL             array, dimension (N)
28               The N diagonal elements of the tridiagonal matrix T.
29
30       E       (input/output) REAL             array, dimension (N)
31               On  entry, the first (N-1) entries contain the subdiagonal ele‐
32               ments of the tridiagonal matrix T; E(N) is set to ZERO.
33
34       INFO    (output) INTEGER
35               INFO = 0(default) : the matrix warrants computations preserving
36               relative  accuracy.   INFO  =  1          : the matrix warrants
37               computations guaranteeing only absolute accuracy.
38

FURTHER DETAILS

40       Based on contributions by
41          Beresford Parlett, University of California, Berkeley, USA
42          Jim Demmel, University of California, Berkeley, USA
43          Inderjit Dhillon, University of Texas, Austin, USA
44          Osni Marques, LBNL/NERSC, USA
45          Christof Voemel, University of California, Berkeley, USA
46
47
48
49
50 LAPACK auxiliary routine (versionNo3v.e1m)ber 2006                       SLARRR(1)