File: rcond.man

package info (click to toggle)
scilab 2.6-4
  • links: PTS
  • area: non-free
  • in suites: woody
  • size: 54,632 kB
  • ctags: 40,267
  • sloc: ansic: 267,851; fortran: 166,549; sh: 10,005; makefile: 4,119; tcl: 1,070; cpp: 233; csh: 143; asm: 135; perl: 130; java: 39
file content (39 lines) | stat: -rw-r--r-- 717 bytes parent folder | download | duplicates (2) 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
 
.TH rcond 1 "April 1993" "Scilab Group" "Scilab Function"
.so ../sci.an
.SH NAME
rcond -  inverse condition number
.SH CALLING SEQUENCE
.nf
rcond(X) 
.fi
.SH PARAMETERS
.TP
X
: real or complex square matrix
.SH DESCRIPTION
.Vb rcond(X)
is an estimate for the reciprocal of the
condition of \fVX\fR in the 1-norm.
.LP
If \fVX\fR is well conditioned, \fVrcond(X)\fR is close to 1.
If not, \fVrcond(X)\fR is close to 0.
.LP
.Vb [r,z]=rcond(X)  
sets \fVr\fR to \fVrcond(X)\fR and returns \fVz\fR such that
.LP
.Vb norm(X*z,1) = r*norm(X,1)*norm(z,1) 
Thus, if \fVrcond\fR is small \fVz\fR is a vector in the kernel.
.SH EXAMPLE
.nf
A=diag([1:10]);
rcond(A)
A(1,1)=0.000001;
rcond(A)
.fi
.SH SEE ALSO
svd, cond, inv