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, 15 insertions, 0 deletions

diff --git a/sourcecodes/bnt-master/KPMtools/logsumexpv.m b/sourcecodes/bnt-master/KPMtools/logsumexpv.m
new file mode 100644
index 00000000..d8e4d8a7
--- /dev/null
+++ b/sourcecodes/bnt-master/KPMtools/logsumexpv.m
@@ -0,0 +1,15 @@
+function s = logsumexpv(a)
+% Returns log(sum(exp(a)))  while avoiding numerical underflow.
+%
+% e.g., log(e^a1 + e^a2) = a1 + log(1 + e^(a2-a1)) if a1>a2
+% If a1 ~ a2, and a1>a2, then e^(a2-a1) is exp(small negative number),
+% which can be computed without underflow.
+
+% Same as logsumexp, except we assume a is a vector.
+% This avoids a call to repmat, which takes 50% of the time!
+
+a = a(:)'; % make row vector
+m = max(a);
+b = a - m*ones(1,length(a));
+s = m + log(sum(exp(b)));
+