1SSTEVX(1)             LAPACK driver routine (version 3.2)            SSTEVX(1)
2
3
4

NAME

6       SSTEVX - computes selected eigenvalues and, optionally, eigenvectors of
7       a real symmetric tridiagonal matrix A
8

SYNOPSIS

10       SUBROUTINE SSTEVX( JOBZ, RANGE, N, D, E, VL, VU, IL, IU, ABSTOL, M,  W,
11                          Z, LDZ, WORK, IWORK, IFAIL, INFO )
12
13           CHARACTER      JOBZ, RANGE
14
15           INTEGER        IL, INFO, IU, LDZ, M, N
16
17           REAL           ABSTOL, VL, VU
18
19           INTEGER        IFAIL( * ), IWORK( * )
20
21           REAL           D( * ), E( * ), W( * ), WORK( * ), Z( LDZ, * )
22

PURPOSE

24       SSTEVX computes selected eigenvalues and, optionally, eigenvectors of a
25       real symmetric tridiagonal matrix A.  Eigenvalues and eigenvectors  can
26       be  selected  by  specifying  either  a  range  of values or a range of
27       indices for the desired eigenvalues.
28

ARGUMENTS

30       JOBZ    (input) CHARACTER*1
31               = 'N':  Compute eigenvalues only;
32               = 'V':  Compute eigenvalues and eigenvectors.
33
34       RANGE   (input) CHARACTER*1
35               = 'A': all eigenvalues will be found.
36               = 'V': all eigenvalues in the half-open interval  (VL,VU]  will
37               be  found.   = 'I': the IL-th through IU-th eigenvalues will be
38               found.
39
40       N       (input) INTEGER
41               The order of the matrix.  N >= 0.
42
43       D       (input/output) REAL array, dimension (N)
44               On entry, the n diagonal elements of the tridiagonal matrix  A.
45               On  exit,  D  may  be multiplied by a constant factor chosen to
46               avoid over/underflow in computing the eigenvalues.
47
48       E       (input/output) REAL array, dimension (max(1,N-1))
49               On entry, the (n-1) subdiagonal  elements  of  the  tridiagonal
50               matrix  A  in elements 1 to N-1 of E.  On exit, E may be multi‐
51               plied by a constant factor chosen to  avoid  over/underflow  in
52               computing the eigenvalues.
53
54       VL      (input) REAL
55               VU       (input)  REAL If RANGE='V', the lower and upper bounds
56               of the interval to be searched for eigenvalues. VL <  VU.   Not
57               referenced if RANGE = 'A' or 'I'.
58
59       IL      (input) INTEGER
60               IU      (input) INTEGER If RANGE='I', the indices (in ascending
61               order) of the smallest and largest eigenvalues to be  returned.
62               1  <= IL <= IU <= N, if N > 0; IL = 1 and IU = 0 if N = 0.  Not
63               referenced if RANGE = 'A' or 'V'.
64
65       ABSTOL  (input) REAL
66               The absolute error tolerance for the eigenvalues.  An  approxi‐
67               mate  eigenvalue is accepted as converged when it is determined
68               to lie in an interval [a,b] of width  less  than  or  equal  to
69               ABSTOL + EPS *   max( |a|,|b| ) , where EPS is the machine pre‐
70               cision.  If ABSTOL is less than or equal to zero, then  EPS*|T|
71               will  be  used  in  its  place,  where |T| is the 1-norm of the
72               tridiagonal matrix.  Eigenvalues will be  computed  most  accu‐
73               rately  when  ABSTOL  is  set  to twice the underflow threshold
74               2*SLAMCH('S'), not zero.  If this routine returns with  INFO>0,
75               indicating that some eigenvectors did not converge, try setting
76               ABSTOL to 2*SLAMCH('S').  See "Computing Small Singular  Values
77               of Bidiagonal Matrices with Guaranteed High Relative Accuracy,"
78               by Demmel and Kahan, LAPACK Working Note #3.
79
80       M       (output) INTEGER
81               The total number of eigenvalues found.  0 <= M <= N.  If  RANGE
82               = 'A', M = N, and if RANGE = 'I', M = IU-IL+1.
83
84       W       (output) REAL array, dimension (N)
85               The  first  M  elements  contain  the  selected  eigenvalues in
86               ascending order.
87
88       Z       (output) REAL array, dimension (LDZ, max(1,M) )
89               If JOBZ = 'V', then if INFO = 0, the first M columns of Z  con‐
90               tain the orthonormal eigenvectors of the matrix A corresponding
91               to the selected eigenvalues, with the i-th column of Z  holding
92               the  eigenvector associated with W(i).  If an eigenvector fails
93               to converge (INFO > 0), then that column of Z contains the lat‐
94               est  approximation  to  the  eigenvector,  and the index of the
95               eigenvector is returned in IFAIL.  If JOBZ = 'N', then Z is not
96               referenced.   Note: the user must ensure that at least max(1,M)
97               columns are supplied in the array Z; if RANGE = 'V', the  exact
98               value  of  M is not known in advance and an upper bound must be
99               used.
100
101       LDZ     (input) INTEGER
102               The leading dimension of the array Z.  LDZ >= 1, and if JOBZ  =
103               'V', LDZ >= max(1,N).
104
105       WORK    (workspace) REAL array, dimension (5*N)
106
107       IWORK   (workspace) INTEGER array, dimension (5*N)
108
109       IFAIL   (output) INTEGER array, dimension (N)
110               If  JOBZ = 'V', then if INFO = 0, the first M elements of IFAIL
111               are zero.  If INFO > 0, then IFAIL contains the indices of  the
112               eigenvectors  that  failed  to  converge.   If JOBZ = 'N', then
113               IFAIL is not referenced.
114
115       INFO    (output) INTEGER
116               = 0:  successful exit
117               < 0:  if INFO = -i, the i-th argument had an illegal value
118               > 0:  if INFO = i, then  i  eigenvectors  failed  to  converge.
119               Their indices are stored in array IFAIL.
120
121
122
123 LAPACK driver routine (version 3.N2o)vember 2008                       SSTEVX(1)
Impressum