BNW using Octave instead of Matlab.

This version of BNW should perform the same as the original version. The only difference is that it uses Octave instead of Matlab when running BayesNet Toolbox during parameter learning.

I am calling this BNW_1.02. It can be accessed at:
compbio.uthsc.edu/BNW_1.02

1 files changed, 38 insertions, 0 deletions

diff --git a/sourcecodes/bnt-master/BNT/general/mk_mutilated_samples.m b/sourcecodes/bnt-master/BNT/general/mk_mutilated_samples.m
new file mode 100644
index 00000000..21a91063
--- /dev/null
+++ b/sourcecodes/bnt-master/BNT/general/mk_mutilated_samples.m
@@ -0,0 +1,38 @@
+function [data, clamped] = mk_mutilated_samples(bnet, ncases, max_clamp, usecell)
+% GEN_MUTILATED_SAMPLES Do random interventions and then draw random samples
+% [data, clamped] = gen_mutilated_samples(bnet, ncases, max_clamp, usecell)
+%
+% At each step, we pick a random subset of size 0 .. max_clamp, and 
+% clamp these nodes to random values.
+%
+% data(i,m) is the value of node i in case m.
+% clamped(i,m) = 1 if node i in case m was set by intervention.
+
+if nargin < 4, usecell = 1; end
+
+ns = bnet.node_sizes;
+n = length(bnet.dag);
+if usecell
+  data = cell(n, ncases);
+else
+  data = zeros(n, ncases);
+end
+clamped = zeros(n, ncases);
+
+csubsets = subsets(1:n, max_clamp, 0); % includes the empty set
+distrib_cset = normalise(ones(1, length(csubsets)));
+
+for m=1:ncases
+  cset = csubsets{sample_discrete(distrib_cset)};
+  nvals = prod(ns(cset));
+  distrib_cvals = normalise(ones(1, nvals));
+  cvals = ind2subv(ns(cset), sample_discrete(distrib_cvals));
+  mutilated_bnet = do_intervention(bnet, cset, cvals);
+  ev = sample_bnet(mutilated_bnet);
+  if usecell
+    data(:,m) = ev;
+  else
+    data(:,m) = cell2num(ev);
+  end
+  clamped(cset,m) = 1;
+end