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, 24 insertions, 0 deletions

diff --git a/sourcecodes/bnt-master/BNT/potentials/@scgpot/combine_pots.m b/sourcecodes/bnt-master/BNT/potentials/@scgpot/combine_pots.m
new file mode 100644
index 00000000..2315176a
--- /dev/null
+++ b/sourcecodes/bnt-master/BNT/potentials/@scgpot/combine_pots.m
@@ -0,0 +1,24 @@
+function pot = combine_pots(pot1, pot2)
+% COMBINE_POTS combine two potentials 
+% pot = combine_pots(pot1, pot2)
+
+% Reduce both potentials before trying to combine them. 
+% Cf. "Stable Local computation with Conditional Gaussian Distributions", page 9
+% Consider again two potentials with minimal tail
+
+% Guarantee minimal tails. If pot1 or pot2 are minimal, they are not changed
+pot1 = reduce_pot(pot1);
+pot2 = reduce_pot(pot2);
+
+%if the intersect set of these two potentials' head conts. combination is undifined
+if ~isempty( myintersect(pot1.cheaddom, pot2.cheaddom) )
+    return;
+end
+
+if  isempty( myintersect(pot1.domain, pot2.cheaddom) ) | isempty( myintersect(pot2.domain, pot1.cheaddom))
+    % if satisfy the condition of directed combine
+    pot = direct_combine_pots(pot1, pot2);
+else
+    % perform recursive combine
+    pot = recursive_combine_pots(pot1, pot2);
+end