Construction & inference (hybrid) in C#

// --------------------------------------------------------------------------------------------------------------------
// <copyright file="CLGaussianExample.cs" company="Bayes Server">
//   Copyright (C) Bayes Server.  All rights reserved.
// </copyright>
// --------------------------------------------------------------------------------------------------------------------

namespace BayesServer.HelpSamples
{
    using System;

    using BayesServer.Inference.RelevanceTree;

    public static class CLGaussianExample
    {
        public static void Main()
        {
            // In this example we programatically create a simple Bayesian network
            // that has some continuous variables.
            // Note that you can automatically define nodes from data using
            // classes in BayesServer.Data.Discovery, 
            // and you can automatically learn the parameters using classes in
            // BayesServer.Learning.Parameters,
            // however here we build a Bayesian network from scratch.

            var network = new Network("Demo");

            // add the nodes (variables)

            var dFalse = new State("False");
            var dTrue = new State("True");
            var d = new Variable("D", dFalse, dTrue);

            var c1 = new Variable("C1", VariableValueType.Continuous);
            var c2 = new Variable("C2", VariableValueType.Continuous);

            network.Nodes.Add(new Node(d));
            network.Nodes.Add(new Node(c1));
            network.Nodes.Add(new Node(c2));

            // add some directed links

            network.Links.Add(new Link(d.Node, c2.Node));
            network.Links.Add(new Link(c1.Node, c2.Node));

            // at this point we have fully specified the structural (graphical) specification of the Bayesian Network.

            // We must define the necessary probability distributions for each node.

            // Each node in a Bayesian Network requires a probability distribution conditioned on it's parents.

            // NewDistribution() can be called on a Node to create the appropriate probability distribution for a node
            // or it can be created manually.

            // The interface IDistribution has been designed to represent both discrete and continuous variables,

            var tableD = d.Node.NewDistribution().Table;     // access the table property of the Distribution
           
            // IMPORTANT
            // Note that calling Node.NewDistribution() does NOT assign the distribution to the node.
            // A distribution cannot be assigned to a node until it is correctly specified.
            // If a distribution becomes invalid  (e.g. a parent node is added), it is automatically set to null.

            tableD[dFalse] = 0.2;
            tableD[dTrue] = 0.8;

            // now tableD is correctly specified we can assign it to Node D;
            d.Node.Distribution = tableD;


            var gaussianC1 = (CLGaussian)c1.Node.NewDistribution();
            gaussianC1.SetMean(c1, 3.5);
            gaussianC1.SetVariance(c1, 12.2);
            c1.Node.Distribution = gaussianC1;

            var clgaussianC2 = (CLGaussian)c2.Node.NewDistribution();
            clgaussianC2.SetMean(c2, 22.2, dFalse);
            clgaussianC2.SetMean(c2, 54.6, dTrue);
            clgaussianC2.SetVariance(c2, 14.6, dFalse);
            clgaussianC2.SetVariance(c2, 3.2, dTrue);
            clgaussianC2.SetWeight(c2, c1, 1.2, dFalse);
            clgaussianC2.SetWeight(c2, c1, -2.1, dTrue);
            c2.Node.Distribution = clgaussianC2;


            // The network is now fully specified

            // If required the network can be saved...

            if (false)   // change this to true to save the network
            {
                // network.Save("fileName.bayes");  // replace 'fileName.bayes' with your own path and uncomment start of line
            }

            // Now we will calculate P(C2|C1=4)

            // use the factory design pattern to create the necessary inference related objects
            var factory = new RelevanceTreeInferenceFactory();
            var inference = factory.CreateInferenceEngine(network);
            var queryOptions = factory.CreateQueryOptions();
            var queryOutput = factory.CreateQueryOutput();

            inference.Evidence.Set(c1, 4.0);    // set c1 = 4

            var queryC2 = new CLGaussian(c2);
            inference.QueryDistributions.Add(queryC2);
            inference.Query(queryOptions, queryOutput); // note that this can raise an exception (see help for details)

            Console.WriteLine("P(C2|C1=4) = (mean " + queryC2.GetMean(c2) + ", variance " + queryC2.GetVariance(c2) + ")");

            // Expected output ...
            // P(C2|C1=4) = (mean 42.36, variance 64.46240000000002)

            // Note that we could also calculate other queries such as P(C1,C2|D=True) or P(C2|C1,D=True) or P(C2,D|C1=4), etc...

        }
    }
}