Mercurial > hg > camir-aes2014
annotate toolboxes/FullBNT-1.0.7/KPMtools/compute_counts.m @ 0:e9a9cd732c1e tip
first hg version after svn
author | wolffd |
---|---|
date | Tue, 10 Feb 2015 15:05:51 +0000 |
parents | |
children |
rev | line source |
---|---|
wolffd@0 | 1 function count = compute_counts(data, sz) |
wolffd@0 | 2 % COMPUTE_COUNTS Count the number of times each combination of discrete assignments occurs |
wolffd@0 | 3 % count = compute_counts(data, sz) |
wolffd@0 | 4 % |
wolffd@0 | 5 % data(i,t) is the value of variable i in case t |
wolffd@0 | 6 % sz(i) : values for variable i are assumed to be in [1:sz(i)] |
wolffd@0 | 7 % |
wolffd@0 | 8 % Example: to compute a transition matrix for an HMM from a sequence of labeled states: |
wolffd@0 | 9 % transmat = mk_stochastic(compute_counts([seq(1:end-1); seq(2:end)], [nstates nstates])); |
wolffd@0 | 10 |
wolffd@0 | 11 assert(length(sz) == size(data, 1)); |
wolffd@0 | 12 P = prod(sz); |
wolffd@0 | 13 indices = subv2ind(sz, data'); % each row of data' is a case |
wolffd@0 | 14 %count = histc(indices, 1:P); |
wolffd@0 | 15 count = hist(indices, 1:P); |
wolffd@0 | 16 count = myreshape(count, sz); |
wolffd@0 | 17 |