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, 36 insertions, 0 deletions

diff --git a/sourcecodes/bnt-master/netlab3.3/gpunpak.m b/sourcecodes/bnt-master/netlab3.3/gpunpak.m
new file mode 100644
index 00000000..8e1058e4
--- /dev/null
+++ b/sourcecodes/bnt-master/netlab3.3/gpunpak.m
@@ -0,0 +1,36 @@
+function net = gpunpak(net, hp)
+%GPUNPAK Separates hyperparameter vector into components. 
+%
+%	Description
+%	NET = GPUNPAK(NET, HP) takes an Gaussian Process data structure NET
+%	and  a hyperparameter vector HP, and returns a Gaussian Process data
+%	structure  identical to the input model, except that the covariance
+%	bias BIAS, output noise NOISE, the input weight vector INWEIGHTS and
+%	the vector of covariance function specific parameters  FPAR have all
+%	been set to the corresponding elements of HP.
+%
+%	See also
+%	GP, GPPAK, GPFWD, GPERR, GPGRAD
+%
+
+%	Copyright (c) Ian T Nabney (1996-2001)
+
+% Check arguments for consistency
+errstring = consist(net, 'gp');
+if ~isempty(errstring);
+  error(errstring);
+end
+if net.nwts ~= length(hp)
+  error('Invalid weight vector length');
+end
+
+net.bias = hp(1);
+net.noise = hp(2);
+
+% Unpack input weights
+mark1 = 2 + net.nin;
+net.inweights = hp(3:mark1);
+
+% Unpack function specific parameters
+net.fpar = hp(mark1 + 1:size(hp, 2));
+