1DLACON(1) LAPACK auxiliary routine (version 3.1) DLACON(1)
2
3
4
6 DLACON - the 1-norm of a square, real matrix A
7
9 SUBROUTINE DLACON( N, V, X, ISGN, EST, KASE )
10
11 INTEGER KASE, N
12
13 DOUBLE PRECISION EST
14
15 INTEGER ISGN( * )
16
17 DOUBLE PRECISION V( * ), X( * )
18
20 DLACON estimates the 1-norm of a square, real matrix A. Reverse commu‐
21 nication is used for evaluating matrix-vector products.
22
23
25 N (input) INTEGER
26 The order of the matrix. N >= 1.
27
28 V (workspace) DOUBLE PRECISION array, dimension (N)
29 On the final return, V = A*W, where EST = norm(V)/norm(W) (W
30 is not returned).
31
32 X (input/output) DOUBLE PRECISION array, dimension (N)
33 On an intermediate return, X should be overwritten by A * X,
34 if KASE=1, A' * X, if KASE=2, and DLACON must be re-called with
35 all the other parameters unchanged.
36
37 ISGN (workspace) INTEGER array, dimension (N)
38
39 EST (input/output) DOUBLE PRECISION
40 On entry with KASE = 1 or 2 and JUMP = 3, EST should be
41 unchanged from the previous call to DLACON. On exit, EST is an
42 estimate (a lower bound) for norm(A).
43
44 KASE (input/output) INTEGER
45 On the initial call to DLACON, KASE should be 0. On an interme‐
46 diate return, KASE will be 1 or 2, indicating whether X should
47 be overwritten by A * X or A' * X. On the final return from
48 DLACON, KASE will again be 0.
49
51 Contributed by Nick Higham, University of Manchester.
52 Originally named SONEST, dated March 16, 1988.
53
54 Reference: N.J. Higham, "FORTRAN codes for estimating the one-norm of a
55 real or complex matrix, with applications to condition estimation", ACM
56 Trans. Math. Soft., vol. 14, no. 4, pp. 381-396, December 1988.
57
58
59
60
61 LAPACK auxiliary routine (versionNo3v.e1m)ber 2006 DLACON(1)