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, 59 insertions, 0 deletions

diff --git a/sourcecodes/bnt-master/docs/cellarray.html b/sourcecodes/bnt-master/docs/cellarray.html
new file mode 100644
index 00000000..19a8da21
--- /dev/null
+++ b/sourcecodes/bnt-master/docs/cellarray.html
@@ -0,0 +1,59 @@
+Cell arrays are a little tricky in Matlab.
+Consider this example.
+
+C=num2cell(rand(2,3))
+C = 
+    [0.4565]    [0.8214]    [0.6154]
+    [0.0185]    [0.4447]    [0.7919]
+
+C{1,2}  % this is the contents of this cell (could be a vector or a string)
+ans =
+    0.8214
+
+C{1:2,2} % this is the contents of these cells - returns multiple
+answers!
+ans =
+    0.8214
+ans =
+    0.4447
+
+A = C(1:2,2)  % this is a slice of the cell array
+ans = 
+    [0.8214]
+    [0.4447]
+
+A{1}  % A is itself a cell array
+ans =
+    0.8214
+
+
+
+>> C(1:2,2)=0 % can't assign a scalar to a cell array
+C(1:2,2)=0
+??? Conversion to cell from double is not possible.
+
+
+>> C(1:2,2)={0;0} % can assign a cell array to a cell array
+C(1:2,2)={0;0}
+C = 
+    [0.4565]    [0]    [0.6154]
+    [0.0185]    [0]    [0.7919]
+
+
+BTW, I use cell arrays for evidence for 2 reasons:
+
+1. [] indicates missing values
+2. it can easily represent vector-valued nodes
+
+The following example  makes this clear
+
+C{1,1} = []
+C{2,1} = rand(3,1)
+
+C = 
+              []    [0]    [0.6154]
+    [3x1 double]    [0]    [0.7919]
+
+
+Hope this helps,
+Kevin