Mercurial > hg > camir-aes2014
comparison toolboxes/FullBNT-1.0.7/KPMstats/dirichlet_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 theta = dirichlet_sample(alpha, N) | |
2 % SAMPLE_DIRICHLET Sample N vectors from Dir(alpha(1), ..., alpha(k)) | |
3 % theta = sample_dirichlet(alpha, N) | |
4 % theta(i,j) = i'th sample of theta_j, where theta ~ Dir | |
5 | |
6 % We use the method from p. 482 of "Bayesian Data Analysis", Gelman et al. | |
7 | |
8 assert(alpha > 0); | |
9 k = length(alpha); | |
10 theta = zeros(N, k); | |
11 scale = 1; % arbitrary | |
12 for i=1:k | |
13 %theta(:,i) = gamrnd(alpha(i), scale, N, 1); | |
14 theta(:,i) = gamma_sample(alpha(i), scale, N, 1); | |
15 end | |
16 %theta = mk_stochastic(theta); | |
17 S = sum(theta,2); | |
18 theta = theta ./ repmat(S, 1, k); |