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, 55 insertions, 0 deletions

diff --git a/sourcecodes/bnt-master/netlab3.3/eigdec.m b/sourcecodes/bnt-master/netlab3.3/eigdec.m
new file mode 100644
index 00000000..f5901d42
--- /dev/null
+++ b/sourcecodes/bnt-master/netlab3.3/eigdec.m
@@ -0,0 +1,55 @@
+function [evals, evec] = eigdec(x, N)
+%EIGDEC	Sorted eigendecomposition
+%
+%	Description
+%	 EVALS = EIGDEC(X, N computes the largest N eigenvalues of the
+%	matrix X in descending order.  [EVALS, EVEC] = EIGDEC(X, N) also
+%	computes the corresponding eigenvectors.
+%
+%	See also
+%	PCA, PPCA
+%
+
+%	Copyright (c) Ian T Nabney (1996-2001)
+
+if nargout == 1
+   evals_only = logical(1);
+else
+   evals_only = logical(0);
+end
+
+if N ~= round(N) | N < 1 | N > size(x, 2)
+   error('Number of PCs must be integer, >0, < dim');
+end
+
+% Find the eigenvalues of the data covariance matrix
+if evals_only
+   % Use eig function as always more efficient than eigs here
+   temp_evals = eig(x);
+else
+   % Use eig function unless fraction of eigenvalues required is tiny
+   if (N/size(x, 2)) > 0.04
+      [temp_evec, temp_evals] = eig(x);
+   else
+      options.disp = 0;
+      [temp_evec, temp_evals] = eigs(x, N, 'LM', options);
+   end
+   temp_evals = diag(temp_evals);
+end
+
+% Eigenvalues nearly always returned in descending order, but just
+% to make sure.....
+[evals perm] = sort(-temp_evals);
+evals = -evals(1:N);
+if ~evals_only
+   if evals == temp_evals(1:N)
+      % Originals were in order
+      evec = temp_evec(:, 1:N);
+      return
+   else
+      % Need to reorder the eigenvectors
+      for i=1:N
+         evec(:,i) = temp_evec(:,perm(i));
+      end
+   end
+end
\ No newline at end of file