Mercurial > hg > camir-aes2014
comparison toolboxes/FullBNT-1.0.7/KPMstats/multinomial_sample.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 Y = sample_cond_multinomial(X, M) | |
2 % SAMPLE_MULTINOMIAL Sample Y(i) ~ M(X(i), :) | |
3 % function Y = sample_multinomial(X, M) | |
4 % | |
5 % X(i) = i'th sample | |
6 % M(i,j) = P(Y=j | X=i) = noisy channel model | |
7 % | |
8 % e.g., if X is a binary image, | |
9 % Y = sample_multinomial(softeye(2, 0.9), X) | |
10 % will create a noisy version of X, where bits are flipped with probability 0.1 | |
11 | |
12 if any(X(:)==0) | |
13 error('data must only contain positive integers') | |
14 end | |
15 | |
16 Y = zeros(size(X)); | |
17 for i=min(X(:)):max(X(:)) | |
18 ndx = find(X==i); | |
19 Y(ndx) = sample_discrete(M(i,:), length(ndx), 1); | |
20 end | |
21 | |
22 |