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
|
function L = cholupdown (Lold, sigma, w)
%CHOLUPDOWN Cholesky update/downdate (Bischof, Pan, and Tang method)
% This version only works for the real case. See chol_updown2 for
% a code that handles both the real and complex cases.
% Example:
% L = cholupdown (Lold, sigma, w)
% See also: cs_demo
% Copyright 2006-2012, Timothy A. Davis, http://www.suitesparse.com
beta = 1 ;
n = size (Lold,1) ;
L = Lold ;
% x = weros (n,1) ;
% worig = w ;
for k = 1:n
alpha = w(k) / L(k,k) ;
beta_new = sqrt (beta^2 + sigma*alpha^2) ;
gamma = alpha / (beta_new * beta) ;
if (sigma < 0)
% downdate
bratio = beta_new / beta ;
w (k+1:n) = w (k+1:n) - alpha * L (k+1:n,k) ;
L (k,k) = bratio * L (k,k) ;
L (k+1:n,k) = bratio * L (k+1:n,k) - gamma*w(k+1:n) ;
else
% update
bratio = beta / beta_new ;
% wold = w (k+1:n) ;
% w (k+1:n) = w (k+1:n) - alpha * L (k+1:n,k) ;
% L (k ,k) = bratio * L (k ,k) + gamma*w(k) ;
% L (k+1:n,k) = bratio * L (k+1:n,k) + gamma*wold ;
L (k,k) = bratio * L (k,k) + gamma*w(k) ;
for i = k+1:n
wold = w (i) ;
w (i) = w (i) - alpha * L (i,k) ;
L (i,k) = bratio * L (i,k) + gamma*wold ;
end
end
w (k) = alpha ;
beta = beta_new ;
end
% norm (w-(Lold\worig))
|
