Mercurial > hg > camir-aes2014
comparison toolboxes/FullBNT-1.0.7/KPMstats/gaussian_prob.m @ 0:e9a9cd732c1e tip
first hg version after svn
author | wolffd |
---|---|
date | Tue, 10 Feb 2015 15:05:51 +0000 |
parents | |
children |
comparison
equal
deleted
inserted
replaced
-1:000000000000 | 0:e9a9cd732c1e |
---|---|
1 function p = gaussian_prob(x, m, C, use_log) | |
2 % GAUSSIAN_PROB Evaluate a multivariate Gaussian density. | |
3 % p = gaussian_prob(X, m, C) | |
4 % p(i) = N(X(:,i), m, C) where C = covariance matrix and each COLUMN of x is a datavector | |
5 | |
6 % p = gaussian_prob(X, m, C, 1) returns log N(X(:,i), m, C) (to prevents underflow). | |
7 % | |
8 % If X has size dxN, then p has size Nx1, where N = number of examples | |
9 | |
10 if nargin < 4, use_log = 0; end | |
11 | |
12 if length(m)==1 % scalar | |
13 x = x(:)'; | |
14 end | |
15 [d N] = size(x); | |
16 %assert(length(m)==d); % slow | |
17 m = m(:); | |
18 M = m*ones(1,N); % replicate the mean across columns | |
19 denom = (2*pi)^(d/2)*sqrt(abs(det(C))); | |
20 mahal = sum(((x-M)'*inv(C)).*(x-M)',2); % Chris Bregler's trick | |
21 if any(mahal<0) | |
22 warning('mahal < 0 => C is not psd') | |
23 end | |
24 if use_log | |
25 p = -0.5*mahal - log(denom); | |
26 else | |
27 p = exp(-0.5*mahal) / (denom+eps); | |
28 end |