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, 25 insertions, 0 deletions

diff --git a/sourcecodes/bnt-master/BNT/potentials/@upot/upot_to_opt_policy.m b/sourcecodes/bnt-master/BNT/potentials/@upot/upot_to_opt_policy.m
new file mode 100644
index 00000000..20f8d2ec
--- /dev/null
+++ b/sourcecodes/bnt-master/BNT/potentials/@upot/upot_to_opt_policy.m
@@ -0,0 +1,25 @@
+function [policy, EU] = upot_to_opt_policy(pot)
+% UPOT_TO_OPT_POLICY Compute an optimal deterministic policy given a utility potential
+% [policy, EU] = upot_to_opt_policy(pot)
+%
+% policy(a,b, ..., z) = P(do z | a, b, ..), which will be a delta function
+% EU is the contraction of this potential, i.e., P .* U
+
+sz = pot.sizes; % mysize(pot.p);
+if isempty(sz)
+  EU = pot.u;
+  policy = [];
+  return;
+end
+
+parent_size = prod(sz(1:end-1));
+self_size = sz(end); 
+C = pot.p .* pot.u; % contraction
+C = reshape(C, parent_size, self_size);
+policy = zeros(parent_size, self_size);
+for i=1:parent_size
+  act = argmax(C(i,:));
+  policy(i, act) = 1;
+end
+policy = myreshape(policy, sz);
+EU = sum(C(:));