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
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
|
<HTML>
<HEAD>
<TITLE>TSP (libtsp/MA) - MAfTpSolve</TITLE>
</HEAD>
<BODY BGCOLOR="#FFFACD">
<H2>MAfTpSolve</H2>
<HR>
<H4>Routine</H4>
<DL>
<DT>
int MAfTpSolve (const float R[], const float g[], float c[], int N)
</DL>
<H4>Purpose</H4>
<DL>
<DT>
Solve a Toeplitz set of equations
</DL>
<H4>Description</H4>
This routine solves the set of matrix equations
<P>
<PRE>
R c = g
</PRE>
<P>
where R is an N by N symmetric Toeplitz matrix and c and g are N element
column vectors. The Toeplitz matrix R has elements which are equal along the
diagonals, i.e. R(i,j)=r(abs(i-j)).
<P>
This routine uses Levinson's method to calculate the solution. In the
case that either numerical instability or an inappropriate set of matrix
coefficients results in a matrix which is not positive definite or singular,
an error code is returned.
<P>
This routine uses
<PRE>
2N-1 divides, and
2N^2-3N multiplies and adds.
</PRE>
<P>
<DL>
<DT>
Reference:
<DD>
E. A. Robinson, "Multichannel time series analysis with digital computer
programs", Holden-Day, 1967, p. 44.
</DL>
<H4>Parameters</H4>
<DL>
<DT>
<- int MAfTpSolve
<DD>
Error flag, zero for no error. The return value is set to 1 if the
matrix not positive definite.
<DT>
-> const float R[]
<DD>
Vector specifying the first column of the Toeplitz matrix
<DT>
-> const float g[]
<DD>
Righthand side vector
<DT>
<- float c[]
<DD>
Solution vector
<DT>
-> int N
<DD>
Number of equations
</DL>
<H4>Author / revision</H4>
P. Kabal Copyright (C) 1996
/ Revision 1.4 1996/05/31
<H4>See Also</H4>
<A HREF="MAfChSolve.html">MAfChSolve</A>
<P>
<HR>
Main Index <A HREF="../libtsp.html">libtsp</A>
</BODY>
</HTML>
|
