Introduction:
-------------

BNJ stands for Bayesian Network tools in JAVA. It is an open-source 
software development toolkit for research in probabilistic learning and 
inference using Bayesian networks. BNJ is developed by the probabilistic 
reasoning group of KDD Lab in the Computing and Information Sciences (CIS) 
Department at Kansas State University. The release of BNJ is governed by 
the GNU Public License.

A Bayesian Network, or Bayesian Belief Network (BBN), is a concise 
representation of a joint probability distribution defined on a finite set 
of random variables. It is a directed acyclic graph (DAG) in which nodes 
represent random variables and arcs represent probabilistic dependencies 
among the variables. A conditional probability distribution is associated 
with each node and describes the dependency between the node and its 
parents. The networks are most often used in expert systems that reason 
under uncertainty.

There are two main research problems in probabilistic reasoning using 
Bayesian networks: learning and inference. Learning the Bayesian network 
from data is automatically constructing the network from data using some 
learning algorithms such as K2. Bayesian network inference invovles 
computing the posterior marginal probabilities of some query nodes, 
P(Q|E), and computing the most probable explanation (MPE) given the values 
of some observed query nodes. Both bayesian network learning and inference 
have been proven to be NP-hard in general. 

BNJ aims to provide researchers and developers a useful toolkit for 
Bayesian network representation, learning, and inference. Several popular 
Bayesian network learning and inference algorithms have been implmented 
and included into BNJ. The first release consists of some core classes for 
representing main data structures, a graphic Bayesian network editor for 
loading and manipulating the network, a learning algorithm K2, an exact 
inference algorithm (the clique-tree propagation algorithm by Lauritzen 
and Spiegelhalter 1988), several stochastic sampling algorithms, and some 
other useful utilities including a network format converter, data 
generator to simulate the network, and a simple DAG layout method. 


Contents:
---------

The current version of BNJ, 1.0 alpha (04 May 2002 build), contains the 
following modules for working with Bayesian networks with discrete chance 
nodes:
 
- Bayesian network core classes that model the structure of a Bayesian 
network 
- XML Bayesian Network Interchange Format (XMLBIF) converter that converts 
networks of other formats to and from XMLBIF (1.0 beta) 
- Graphical editor for editting the network and interchanging formats 
- K2 algorithm for learning the structure of a Bayesian network from data
- Lauritzen-Spiegelhalter (LS) algorithm for exact inference on a Bayesian 
network 
- Importance sampling algorithms evaluated using exact root mean squared 
error including logic sampling, likelihood weighting, self-importance 
sampling, adaptive importance sampling 
- GASLEAK, a research genetic algorithm (GA)-based wrapper that searches 
for inferential loss-minimizing orderings of variables in Bayesian network 
structure learning 


Sample Networks:
----------------

Two sample networks are used in this manual to demonstrate the usage of 
BNJ modules: Asia and ALARM. Asia is an eight-node binary network. ALARM 
has 37 nodes and 46 edges. The default file format is xmlbif format. 
Followed is the content of asia.xml.

===============================asia.xml============================================

<?xml version="1.0"?>

<!-- DTD for the XMLBIF 0.3 format -->
<!DOCTYPE BIF [
	<!ELEMENT BIF ( NETWORK )*>
		<!ATTLIST BIF VERSION CDATA #REQUIRED>
	<!ELEMENT NETWORK ( NAME, ( PROPERTY | VARIABLE | DEFINITION )* )>
	<!ELEMENT NAME (#PCDATA)>
	<!ELEMENT VARIABLE ( NAME, ( OUTCOME | PROPERTY )* ) >
		<!ATTLIST VARIABLE TYPE (nature|decision|utility) 
"nature">
	<!ELEMENT OUTCOME (#PCDATA)>
	<!ELEMENT DEFINITION ( FOR | GIVEN | TABLE | PROPERTY )* >
	<!ELEMENT FOR (#PCDATA)>
	<!ELEMENT GIVEN (#PCDATA)>
	<!ELEMENT TABLE (#PCDATA)>
	<!ELEMENT PROPERTY (#PCDATA)>
]>

<BIF VERSION="0.3">
	<NETWORK>
	<NAME>bayesiannetwork</NAME>
		<!-- Variables -->
		<VARIABLE TYPE="nature">
			<NAME>VisitAsia</NAME>
			<OUTCOME>Visit</OUTCOME>
			<OUTCOME>No_Visit</OUTCOME>
			<PROPERTY>position = (166, 79)</PROPERTY>
		</VARIABLE>
		<VARIABLE TYPE="nature">
			<NAME>Tuberculosis</NAME>
			<OUTCOME>Present</OUTCOME>
			<OUTCOME>Absent</OUTCOME>
			<PROPERTY>position = (166, 175)</PROPERTY>
		</VARIABLE>
		<VARIABLE TYPE="nature">
			<NAME>Smoking</NAME>
			<OUTCOME>Smoker</OUTCOME>
			<OUTCOME>NonSmoker</OUTCOME>
			<PROPERTY>position = (520, 79)</PROPERTY>
		</VARIABLE>
		<VARIABLE TYPE="nature">
			<NAME>Cancer</NAME>
			<OUTCOME>Present</OUTCOME>
			<OUTCOME>Absent</OUTCOME>
			<PROPERTY>position = (418, 175)</PROPERTY>
		</VARIABLE>
		<VARIABLE TYPE="nature">
			<NAME>TbOrCa</NAME>
			<OUTCOME>True</OUTCOME>
			<OUTCOME>False</OUTCOME>
			<PROPERTY>position = (310, 271)</PROPERTY>
		</VARIABLE>
		<VARIABLE TYPE="nature">
			<NAME>XRay</NAME>
			<OUTCOME>Abnormal</OUTCOME>
			<OUTCOME>Normal</OUTCOME>
			<PROPERTY>position = (178, 361)</PROPERTY>
		</VARIABLE>
		<VARIABLE TYPE="nature">
			<NAME>Bronchitis</NAME>
			<OUTCOME>Present</OUTCOME>
			<OUTCOME>Absent</OUTCOME>
			<PROPERTY>position = (658, 175)</PROPERTY>
		</VARIABLE>
		<VARIABLE TYPE="nature">
			<NAME>Dyspnea</NAME>
			<OUTCOME>Present</OUTCOME>
			<OUTCOME>Absent</OUTCOME>
			<PROPERTY>position = (454, 361)</PROPERTY>
		</VARIABLE>
		<!-- Probability Distributions -->
		<DEFINITION>
			<FOR>VisitAsia</FOR>
			<TABLE>0.01 0.99 </TABLE>
		</DEFINITION>
		<DEFINITION>
			<FOR>Tuberculosis</FOR>
			<GIVEN>VisitAsia</GIVEN>
			<TABLE>0.05 0.01 0.95 0.99 </TABLE>
		</DEFINITION>
		<DEFINITION>
			<FOR>Smoking</FOR>
			<TABLE>0.5 0.5 </TABLE>
		</DEFINITION>
		<DEFINITION>
			<FOR>Cancer</FOR>
			<GIVEN>Smoking</GIVEN>
			<TABLE>0.1 0.01 0.9 0.99 </TABLE>
		</DEFINITION>
		<DEFINITION>
			<FOR>TbOrCa</FOR>
			<GIVEN>Tuberculosis</GIVEN>
			<GIVEN>Cancer</GIVEN>
			<TABLE>1.0 1.0 1.0 0.0 0.0 0.0 0.0 1.0 </TABLE>
		</DEFINITION>
		<DEFINITION>
			<FOR>XRay</FOR>
			<GIVEN>TbOrCa</GIVEN>
			<TABLE>0.98 0.05 0.02 0.95 </TABLE>
		</DEFINITION>
		<DEFINITION>
			<FOR>Bronchitis</FOR>
			<GIVEN>Smoking</GIVEN>
			<TABLE>0.6 0.3 0.4 0.7 </TABLE>
		</DEFINITION>
		<DEFINITION>
			<FOR>Dyspnea</FOR>
			<GIVEN>TbOrCa</GIVEN>
			<GIVEN>Bronchitis</GIVEN>
			<TABLE>0.9 0.7 0.8 0.1 0.1 0.3 0.2 0.9 </TABLE>
		</DEFINITION>
	</NETWORK>
</BIF>
===============================asia.xml============================================


Modules:
--------

*********
*  BBN  *
*********

Description:

The package bbn consists of core classes for representing the data 
structure of a Bayesian network. BBN.java and Node.java are used to 
represent a Bayesian network. NodeManager.java helps handle evidence for 
inference. Some useful utilities are also included in this package such as 
datageneator.

Examples:

(example 1) Load the network from xmlbif file and print out the 
information of each node.

The following example shows how to load a Bayesian network from file, and 
how to access its node information. To run this example, create the file 
BBNExample.java as follows and make sure the network file asia.xml is 
under the same directory.

==============================BBNExample.java===================================== 
import bbn.*;
import java.util.Vector;

public class BBNExample {
  
  public static void main(String[] args){

    System.out.println("load the network from file ... ");
    BBN network = new BBN();
    network.loadFromXML("asia.xml");

    System.out.println("printing out information of each node... ");
    for (int i = 0; i < network.size(); i++){
	System.out.println("i="+i);
	network.getNodeAt(i).print();
	network.getNodeAt(i).printStateNames();
        if(network.getNodeAt(i).getParents().size()>0){
	 for(int j=0;j<network.getNodeAt(i).getParents().size();j++)	
	 System.out.println("parents --> 
"+network.getNodeAt(i).getParentName(j));
	} 
	System.out.println("probabilities in CPT --> 
"+network.getNodeAt(i).getProbabilities());
	System.out.println("--------");
    }

  }
}

=============================Result of 
BBNExample========================================
D:\bnj>javac BBNExample.java

D:\bnj>java BBNExample
load the network from file ...
printing out information of each node...
i=0
For node VisitAsia:
  Visit
  No_Visit
probabilities in CPT --> [0.01, 0.99]
--------
i=1
For node Tuberculosis:
  Present
  Absent
parents --> VisitAsia
probabilities in CPT --> [0.05, 0.01, 0.95, 0.99]
--------
i=2
For node Smoking:
  Smoker
  NonSmoker
probabilities in CPT --> [0.5, 0.5]
--------
i=3
For node Cancer:
  Present
  Absent
parents --> Smoking
probabilities in CPT --> [0.1, 0.01, 0.9, 0.99]
--------
i=4
For node TbOrCa:
  True
  False
parents --> Tuberculosis
parents --> Cancer
probabilities in CPT --> [1.0, 1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 1.0]
--------
i=5
For node XRay:
  Abnormal
  Normal
parents --> TbOrCa
probabilities in CPT --> [0.98, 0.05, 0.02, 0.95]
--------
i=6
For node Bronchitis:
  Present
  Absent
parents --> Smoking
probabilities in CPT --> [0.6, 0.3, 0.4, 0.7]
--------
i=7
For node Dyspnea:
  Present
  Absent
parents --> TbOrCa
parents --> Bronchitis
probabilities in CPT --> [0.9, 0.7, 0.8, 0.1, 0.1, 0.3, 0.2, 0.9]
--------

================================================================


******************
*  DataGenerator *
******************

Description:

DataGenerator can be used to simulate the network and generate a data file 
from the network. The output data file can be used as the training dataset 
of any learning algorithm. The simulation process is done by forward 
sampling, i.e., first simulating root nodes, then simulating their 
parents, and so on. It is included in package bbn. DataGenerator2.java is 
an optimized version of DataGenerator.java. The output data file is stored 
in an xml file format.  
DataGenerator is also used to generate the evidence file if the number of 
samples is set to zero.

Examples:

(example 2) Simulate the network and generate a data file from the 
network.

The following example shows how to simulate the network and generate a 
data file with 20 samples, named asia20.xml, from the network.

=============================Example of using 
DataGenerator========================================

D:\bnj>java bbn/DataGenerator2
 * * * DataGenerator2 * * *
Usage: java DataGenerator2 input.xml output.xml number_of_samples (0 will 
generat
e an evdience file)

D:\bnj>java bbn/DataGenerator2 asia.xml asia20.xml 20
 * * * DataGenerator * * *
Usage: java DataGenerator input.xml output.xml number_of_samples (0 will 
generat
e an evdience file)
Success!

D:\bnj>

=============================Result of BBNExample: 
asia20.xml=======================================
<?xml version="1.0" encoding="US-ASCII"?>


<!--
	Bayesian network in XMLBIF v0.3 (BayesNet Interchange Format)
	Produced by JavaBayes (http://www.cs.cmu.edu/~javabayes/
	Output created Tue May 02 16:32:10 CDT 2000
-->



<!-- DTD for the XMLBIF 0.3 format -->
<!DOCTYPE BIF [
	<!ELEMENT BIF ( NETWORK )*>
		<!ATTLIST BIF VERSION CDATA #REQUIRED>
	<!ELEMENT NETWORK ( NAME, ( PROPERTY | VARIABLE | DEFINITION )*)>
	<!ELEMENT NAME (#PCDATA)>
	<!ELEMENT VARIABLE ( NAME, ( OUTCOME |  PROPERTY )* ) >
		<!ATTLIST VARIABLE TYPE (nature|decision|utility) 
"nature">
	<!ELEMENT OUTCOME (#PCDATA)>
	<!ELEMENT DEFINITION ( FOR | GIVEN | TABLE | PROPERTY )* >
	<!ELEMENT FOR (#PCDATA)>
	<!ELEMENT GIVEN (#PCDATA)>
	<!ELEMENT TABLE (#PCDATA)>
	<!ELEMENT PROPERTY (#PCDATA)>
]>


<BIF VERSION="0.3">
<NETWORK>
<NAME>InternalNetwork</NAME>

<!-- Variables -->
<VARIABLE TYPE="nature">
	<NAME>VisitAsia</NAME>
	<OUTCOME>Visit</OUTCOME>
	<OUTCOME>No_Visit</OUTCOME>
	<PROPERTY>position = (166, 79)</PROPERTY>
</VARIABLE>

<VARIABLE TYPE="nature">
	<NAME>Smoking</NAME>
	<OUTCOME>Smoker</OUTCOME>
	<OUTCOME>NonSmoker</OUTCOME>
	<PROPERTY>position = (520, 79)</PROPERTY>
</VARIABLE>

<VARIABLE TYPE="nature">
	<NAME>Tuberculosis</NAME>
	<OUTCOME>Present</OUTCOME>
	<OUTCOME>Absent</OUTCOME>
	<PROPERTY>position = (166, 175)</PROPERTY>
</VARIABLE>

<VARIABLE TYPE="nature">
	<NAME>Cancer</NAME>
	<OUTCOME>Present</OUTCOME>
	<OUTCOME>Absent</OUTCOME>
	<PROPERTY>position = (418, 175)</PROPERTY>
</VARIABLE>

<VARIABLE TYPE="nature">
	<NAME>TbOrCa</NAME>
	<OUTCOME>True</OUTCOME>
	<OUTCOME>False</OUTCOME>
	<PROPERTY>position = (310, 271)</PROPERTY>
</VARIABLE>

<VARIABLE TYPE="nature">
	<NAME>XRay</NAME>
	<OUTCOME>Abnormal</OUTCOME>
	<OUTCOME>Normal</OUTCOME>
	<PROPERTY>position = (178, 361)</PROPERTY>
</VARIABLE>

<VARIABLE TYPE="nature">
	<NAME>Bronchitis</NAME>
	<OUTCOME>Present</OUTCOME>
	<OUTCOME>Absent</OUTCOME>
	<PROPERTY>position = (658, 175)</PROPERTY>
</VARIABLE>

<VARIABLE TYPE="nature">
	<NAME>Dyspnea</NAME>
	<OUTCOME>Present</OUTCOME>
	<OUTCOME>Absent</OUTCOME>
	<PROPERTY>position = (454, 361)</PROPERTY>
</VARIABLE>



</NETWORK>
<SAMPLES>
	<NUMBER>20</NUMBER>
	<SET>No_Visit Smoker Absent Absent False Abnormal Absent 
Absent</SET>
	<SET>No_Visit Smoker Absent Present True Abnormal Absent 
Absent</SET>
	<SET>No_Visit NonSmoker Absent Absent False Normal Absent 
Absent</SET>
	<SET>No_Visit Smoker Absent Absent False Normal Present 
Present</SET>
	<SET>No_Visit Smoker Absent Absent False Normal Present 
Present</SET>
	<SET>No_Visit NonSmoker Absent Absent False Normal Absent 
Absent</SET>
	<SET>No_Visit Smoker Absent Absent False Normal Present 
Absent</SET>
	<SET>Visit Smoker Absent Absent False Normal Absent Absent</SET>
	<SET>No_Visit NonSmoker Absent Absent False Normal Absent 
Absent</SET>
	<SET>No_Visit NonSmoker Absent Absent False Normal Present 
Present</SET>
	<SET>No_Visit Smoker Absent Absent False Normal Present 
Present</SET>
	<SET>No_Visit NonSmoker Absent Absent False Normal Absent 
Present</SET>
	<SET>No_Visit Smoker Absent Absent False Normal Absent 
Absent</SET>
	<SET>No_Visit Smoker Absent Absent False Normal Present 
Present</SET>
	<SET>No_Visit NonSmoker Absent Absent False Normal Present 
Present</SET>
	<SET>No_Visit NonSmoker Absent Absent False Normal Absent 
Absent</SET>
	<SET>No_Visit Smoker Absent Absent False Normal Present 
Present</SET>
	<SET>No_Visit NonSmoker Absent Absent False Abnormal Present 
Present</SET>
	<SET>No_Visit Smoker Absent Absent False Abnormal Present 
Present</SET>
	<SET>No_Visit NonSmoker Absent Absent False Normal Absent 
Absent</SET>
</SAMPLES>
</BIF>

=============================End of BBNExample: 
asia20.xml========================================

The following example shows how to generate an evidence file using 
DataGenerator. The evidence file is useful for inference.

(example 3) use dataGenerator to generate evidence file

=============================Example of using 
DataGenerator========================================


D:\bnj>java bbn/DataGenerator2 asia.xml asia_evidence.xml 0
 * * * DataGenerator2 * * *
Usage: java DataGenerator2 input.xml output.xml number_of_samples (0 will 
genera
te an evdience file)
Success!

D:\bnj>

=============================Result evidence file : 
asia_evidence.xml===================
0 0 0 0 0 0 0 0
VisitAsia|Smoking|Tuberculosis|Cancer|TbOrCa|XRay|Bronchitis|Dyspnea
Visit NonSmoker Absent Present False Normal Present Present
=============================end of evidence 
file=======================================

You can modify the evidence bits at the first line to set evidence nodes(0 
means query, 1 means evidence).



*********
*  K2   *
*********

Description:

K2 is a score based greedy search algorithm for learning Bayesian networks 
from data. It was published in Cooper, G. and Herskovits, E. (1992). A 
Bayesian method for the induction of probabilistic networks from data. 
Machine Learning 9, 309-347 (1992) K2 uses a Bayesian score, P(Bs, D), to 
rank different structures and it uses a greedy search algorithm to maxmize 
P(Bs, D). 
The main classes in package k2 are K2Manager.jaav and K2.java. Class 
K2Manager loads the input training data. Class K2 implements the learning 
procedure.

Inputs to K2 include: 

1 a set of nodes, 
2 an ordering on the nodes, 
3 an upper bound u on the number of parents a node may have 
4 a database D containing m cases of training data 

Output of K2: for each node, K2 returns its most probable parents given 
the training data D. 

In our implementation, the set of nodes and the database D are stored in 
the input training data file in xml format. The upper bound u is set to a 
fixed value 5 in K2.java. The required ordering can be input from command 
line. If no ordering is given at command line, 0 - n is assumed according 
to the node ordering in the input data file.
The output of K2 is the learned network in XMLBIF format. You can use the 
network editor to open and see its structure.


Examples:

(example 4) use K2 to learn a Bayesian network from training data.

The following example shows how to use K2 to learn a network from training 
data. To run K2, first we need to prepare the input training data. In this 
example, we first use DataGenerator to generate the training data, then we 
run K2 on this training dataset to recover(learn) the network from data.
The 
=============================Example of 
K2========================================

D:\bnj>java bbn/DataGenerator2 asia.xml asia1500.xml 1500
 * * * DataGenerator2 * * *
Usage: java DataGenerator2 input.xml output.xml number_of_samples (0 will 
genera
te an evdience file)
Success!

D:\bnj>java k2/K2 asia1500.xml learned.xml
Usage: java k2.K2 input.xml output.xml [order]
 (if no order is given, 0 - n is assumed).
[loaded network and training samples]
[Network learned.]
[Finished.]

D:\bnj>

==========================Result of K2, the learned network: 
learned.xml=============

<?xml version="1.0" encoding="US-ASCII"?>


<!--
	Bayesian network in XMLBIF v0.3 (BayesNet Interchange Format)
	Produced by JavaBayes (http://www.cs.cmu.edu/~javabayes/
	Output created Tue May 02 16:32:10 CDT 2000
-->



<!-- DTD for the XMLBIF 0.3 format -->
<!DOCTYPE BIF [
	<!ELEMENT BIF ( NETWORK )*>
		<!ATTLIST BIF VERSION CDATA #REQUIRED>
	<!ELEMENT NETWORK ( NAME, ( PROPERTY | VARIABLE | DEFINITION )*)>
	<!ELEMENT NAME (#PCDATA)>
	<!ELEMENT VARIABLE ( NAME, ( OUTCOME |  PROPERTY )* ) >
		<!ATTLIST VARIABLE TYPE (nature|decision|utility) 
"nature">
	<!ELEMENT OUTCOME (#PCDATA)>
	<!ELEMENT DEFINITION ( FOR | GIVEN | TABLE | PROPERTY )* >
	<!ELEMENT FOR (#PCDATA)>
	<!ELEMENT GIVEN (#PCDATA)>
	<!ELEMENT TABLE (#PCDATA)>
	<!ELEMENT PROPERTY (#PCDATA)>
]>


<BIF VERSION="0.3">
<NETWORK>
<NAME>InternalNetwork</NAME>

<!-- Variables -->
<VARIABLE TYPE="nature">
	<NAME>VisitAsia</NAME>
	<OUTCOME>Visit</OUTCOME>
	<OUTCOME>No_Visit</OUTCOME>
	<PROPERTY>position = (166, 79)</PROPERTY>
</VARIABLE>

<VARIABLE TYPE="nature">
	<NAME>Smoking</NAME>
	<OUTCOME>Smoker</OUTCOME>
	<OUTCOME>NonSmoker</OUTCOME>
	<PROPERTY>position = (520, 79)</PROPERTY>
</VARIABLE>

<VARIABLE TYPE="nature">
	<NAME>Tuberculosis</NAME>
	<OUTCOME>Present</OUTCOME>
	<OUTCOME>Absent</OUTCOME>
	<PROPERTY>position = (166, 175)</PROPERTY>
</VARIABLE>

<VARIABLE TYPE="nature">
	<NAME>Cancer</NAME>
	<OUTCOME>Present</OUTCOME>
	<OUTCOME>Absent</OUTCOME>
	<PROPERTY>position = (418, 175)</PROPERTY>
</VARIABLE>

<VARIABLE TYPE="nature">
	<NAME>TbOrCa</NAME>
	<OUTCOME>True</OUTCOME>
	<OUTCOME>False</OUTCOME>
	<PROPERTY>position = (310, 271)</PROPERTY>
</VARIABLE>

<VARIABLE TYPE="nature">
	<NAME>XRay</NAME>
	<OUTCOME>Abnormal</OUTCOME>
	<OUTCOME>Normal</OUTCOME>
	<PROPERTY>position = (178, 361)</PROPERTY>
</VARIABLE>

<VARIABLE TYPE="nature">
	<NAME>Bronchitis</NAME>
	<OUTCOME>Present</OUTCOME>
	<OUTCOME>Absent</OUTCOME>
	<PROPERTY>position = (658, 175)</PROPERTY>
</VARIABLE>

<VARIABLE TYPE="nature">
	<NAME>Dyspnea</NAME>
	<OUTCOME>Present</OUTCOME>
	<OUTCOME>Absent</OUTCOME>
	<PROPERTY>position = (454, 361)</PROPERTY>
</VARIABLE>


<!-- Probabilitydistributions-->
<DEFINITION>
	<FOR>VisitAsia</FOR>
	<TABLE>0.012 0.988 </TABLE>
</DEFINITION>

<DEFINITION>
	<FOR>Smoking</FOR>
	<TABLE>0.5026666666666667 0.49733333333333335 </TABLE>
</DEFINITION>

<DEFINITION>
	<FOR>Tuberculosis</FOR>
	<GIVEN>VisitAsia</GIVEN>
	<TABLE>0.1111111111111111 0.008097165991902834 0.8888888888888888 
0.9919028340080972 </TABLE>
</DEFINITION>

<DEFINITION>
	<FOR>Cancer</FOR>
	<GIVEN>Smoking</GIVEN>
	<TABLE>0.11538461538461539 0.00804289544235925 0.8846153846153846 
0.9919571045576407 </TABLE>
</DEFINITION>

<DEFINITION>
	<FOR>TbOrCa</FOR>
	<GIVEN>Cancer</GIVEN>
	<GIVEN>Tuberculosis</GIVEN>
	<TABLE>0.0 1.0 1.0 0.0 0.0 0.0 0.0 1.0 </TABLE>
</DEFINITION>

<DEFINITION>
	<FOR>XRay</FOR>
	<GIVEN>TbOrCa</GIVEN>
	<TABLE>0.9532710280373832 0.04737975592246949 0.04672897196261682 
0.9526202440775305 </TABLE>
</DEFINITION>

<DEFINITION>
	<FOR>Bronchitis</FOR>
	<GIVEN>Smoking</GIVEN>
	<GIVEN>VisitAsia</GIVEN>
	<TABLE>0.4 0.581989247311828 0.625 0.3089430894308943 0.6 
0.41801075268817206 0.375 0.6910569105691057 </TABLE>
</DEFINITION>

<DEFINITION>
	<FOR>Dyspnea</FOR>
	<GIVEN>Bronchitis</GIVEN>
	<GIVEN>TbOrCa</GIVEN>
	<TABLE>0.875 0.7805280528052805 0.6744186046511628 
0.09911054637865312 0.125 0.21947194719471946 0.32558139534883723 
0.9008894536213469 </TABLE>
</DEFINITION>

</NETWORK>
</BIF>

==========================end of 
learned.xml====================================================================



********************
*  Network Editor  *
********************

Network editor is a graphical Bayesian network editor. Users can use it to 
open Bayesian networks of up to 9 file formats including (.bif, .xml, 
.net, .xbn, .dsc, .dsl, .dnet, .ENT, .ideal). After viewing and editing 
the network, users can save the network into 3 different file format: 
.xml, .bif, and .net. 

BIF .bif --- Bayesian networks Interchange Format 
(http://www-2.cs.cmu.edu/~fgcozman/Research/InterchangeFormat/Old/xmlbif02.html)
XMLBIF .xml --- XML Belief Network Interchange File Format 
(http://www-2.cs.cmu.edu/afs/cs/user/fgcozman/www/Research/InterchangeFormat/)
Microsoft XBN .xbn --- XBN File Format 
(http://www.research.microsoft.com/research/dtg/bnformat/default.htm)
Microsoft .dsc --- MSBN Page 
(http://www.research.microsoft.com/adapt/MSBNx/)
Hugin format .net --- the net language 
(http://developer.hugin.com/documentation/net/Net_Language.article)
Genie format .dsl --- Genie home 
(http://www.kddresearch.org/Groups/Probabilistic-Reasoning/www2.sis.pitt.edu/~genie)
Ergo format .ENT --- Ergo by Noetic Systems 
(http://www.noeticsystems.com/ergo.shtml)
Netica format .dnet --- Netica by Norsys (http://www.norsys.com/)
IDEAL format .ideal --- IDEAL format by Rockwell 
(http://www.rpal.rockwell.com/ideal.html)

The main IO functions are implemented in FileIO.java file, i.e. 
FileIO.load() method and FileIO.Save() method. 

The following example shows how to start up the network editor. 

Examples:

(example 5)

==========================EditorExample.java====================================================================

import bbn.*;

public class EditorExample {

  public static void main(String[] args) {
    Debug.println("xbneditor...");
    XBN editor = new XBN();
  }
}

==========================end of 
EditorExample.java====================================================================



*******************
*    Converter    *
*******************

Network converter can convert Bayesian network files from 9 different 
formats(.bif, .xml, .net, .xbn, .dsc, .dsl, .dnet, .ENT, .ideal) to 3 
formats(.xml, .bif, .net). Last section shows how to convert network 
formats using the network editor. This section gives examples of 
converting network formats at command line.

Usage: java xbneditor/BBNConvertor [sourcefilename] [targetfilename]

The following example shows how to convert asia.xml to asia.net.

==========================BBNConvertor 
Example====================================================================

D:\bnj>java xbneditor/BBNConvertor asia.xml asia.net
informat = xml
outformat = net
load XMLBIF .xml file ...
load variables ...
stateList ******* [Visit, No_Visit]
stateList ******* [Present, Absent]
stateList ******* [Smoker, NonSmoker]
stateList ******* [Present, Absent]
stateList ******* [True, False]
stateList ******* [Abnormal, Normal]
stateList ******* [Present, Absent]
stateList ******* [Present, Absent]
load ProbabilityDistributions ...
blockName = "VisitAsia"
parentList = []
probabilitiesList = 0.01 0.99
count = 2
blockName = "Tuberculosis"
parentList = [VisitAsia]
probabilitiesList = 0.05 0.01 0.95 0.99
count = 4
blockName = "Smoking"
parentList = []
probabilitiesList = 0.5 0.5
count = 2
blockName = "Cancer"
parentList = [Smoking]
probabilitiesList = 0.1 0.01 0.9 0.99
count = 4
blockName = "TbOrCa"
parentList = [Tuberculosis, Cancer]
probabilitiesList = 1.0 1.0 1.0 0.0 0.0 0.0 0.0 1.0
count = 8
blockName = "XRay"
parentList = [TbOrCa]
probabilitiesList = 0.98 0.05 0.02 0.95
count = 4
blockName = "Bronchitis"
parentList = [Smoking]
probabilitiesList = 0.6 0.3 0.4 0.7
count = 4
blockName = "Dyspnea"
parentList = [TbOrCa, Bronchitis]
probabilitiesList = 0.9 0.7 0.8 0.1 0.1 0.3 0.2 0.9
count = 8
outputfiletype =net
convert to Hugin .net format!

D:\bnj>
==========================end of BBNConvertor 
Example====================================================================