public HiddenMarkovModel CloneStructure()
{
HiddenMarkovModel hmm;
if (IsDiscrete)
{
hmm = new HiddenMarkovModel(StateCount, _items);
}
else
{
hmm = new HiddenMarkovModel(StateCount);
}
return hmm;
}
object ICloneable.Clone()
{
var result = new HiddenMarkovModel(StateCount);
EngineArray.ArrayCopy(Pi, result.Pi);
EngineArray.ArrayCopy(TransitionProbability, result.TransitionProbability);
for (int i = 0; i < StateCount;i++)
{
result.StateDistributions[i] = StateDistributions[i].Clone();
}
return result;
}
static HiddenMarkovModel BuildDiscHMM()
{
HiddenMarkovModel hmm =
new HiddenMarkovModel(2, 2);
hmm.Pi[0] = 0.95;
hmm.Pi[1] = 0.05;
hmm.StateDistributions[0] = new DiscreteDistribution(new double[][] { new[] { 0.95, 0.05 } });
hmm.StateDistributions[1] = new DiscreteDistribution(new double[][] { new[] { 0.20, 0.80 } });
hmm.TransitionProbability[0][1] = 0.05;
hmm.TransitionProbability[0][0] = 0.95;
hmm.TransitionProbability[1][0] = 0.10;
hmm.TransitionProbability[1][1] = 0.90;
return hmm;
}
public void Validate(HiddenMarkovModel result, HiddenMarkovModel source)
{
KullbackLeiblerDistanceCalculator klc =
new KullbackLeiblerDistanceCalculator();
double e = klc.Distance(result, source);
Assert.IsTrue(e < 0.01);
}
static HiddenMarkovModel BuildContHMM()
{
double [] mean1 = {0.25, -0.25};
double [][] covariance1 = { new[] {1.0, 2.0}, new[] {1.0, 4.0} };
double [] mean2 = {0.5, 0.25};
double [][] covariance2 = { new[] {4.0, 2.0}, new[] {3.0, 4.0} };
HiddenMarkovModel hmm = new HiddenMarkovModel(2);
hmm.Pi[0] = 0.8;
hmm.Pi[1] = 0.2;
hmm.StateDistributions[0] = new ContinousDistribution(mean1,covariance1);
hmm.StateDistributions[1] = new ContinousDistribution(mean2,covariance2);
hmm.TransitionProbability[0][1] = 0.05;
hmm.TransitionProbability[0][0] = 0.95;
hmm.TransitionProbability[1][0] = 0.10;
hmm.TransitionProbability[1][1] = 0.90;
return hmm;
}
/* Initial guess for the Baum-Welch algorithm */
private static HiddenMarkovModel BuildDiscInitHMM()
{
var hmm = new HiddenMarkovModel(2, 2);
hmm.Pi[0] = 0.50;
hmm.Pi[1] = 0.50;
hmm.StateDistributions[0] = new DiscreteDistribution(new[] {new[] {0.8, 0.2}});
hmm.StateDistributions[1] = new DiscreteDistribution(new[] {new[] {0.1, 0.9}});
hmm.TransitionProbability[0][1] = 0.2;
hmm.TransitionProbability[0][0] = 0.8;
hmm.TransitionProbability[1][0] = 0.2;
hmm.TransitionProbability[1][1] = 0.8;
return hmm;
}
private static HiddenMarkovModel BuildContInitHMM()
{
double[] mean1 = {0.20, -0.20};
double[][] covariance1 = {new[] {1.3, 2.2}, new[] {1.3, 4.3}};
double[] mean2 = {0.5, 0.25};
double[][] covariance2 = {new[] {4.1, 2.1}, new[] {3.2, 4.4}};
var hmm = new HiddenMarkovModel(2);
hmm.Pi[0] = 0.9;
hmm.Pi[1] = 0.1;
hmm.StateDistributions[0] = new ContinousDistribution(mean1, covariance1);
hmm.StateDistributions[1] = new ContinousDistribution(mean2, covariance2);
hmm.TransitionProbability[0][1] = 0.10;
hmm.TransitionProbability[0][0] = 0.90;
hmm.TransitionProbability[1][0] = 0.15;
hmm.TransitionProbability[1][1] = 0.85;
return hmm;
}
/// <inheritdoc/>
public Object Read(Stream istream)
{
int states = 0;
int[] items;
double[] pi = null;
Matrix transitionProbability = null;
IDictionary<String, String> properties = null;
IList<IStateDistribution> distributions = new List<IStateDistribution>();
EncogReadHelper reader = new EncogReadHelper(istream);
EncogFileSection section;
while ((section = reader.ReadNextSection()) != null)
{
if (section.SectionName.Equals("HMM")
&& section.SubSectionName.Equals("PARAMS"))
{
properties = section.ParseParams();
}
if (section.SectionName.Equals("HMM")
&& section.SubSectionName.Equals("CONFIG"))
{
IDictionary<String, String> p = section.ParseParams();
states = EncogFileSection.ParseInt(p, HiddenMarkovModel.TAG_STATES);
if (p.ContainsKey(HiddenMarkovModel.TAG_ITEMS))
{
items = EncogFileSection.ParseIntArray(p, HiddenMarkovModel.TAG_ITEMS);
}
pi = section.ParseDoubleArray(p, HiddenMarkovModel.TAG_PI);
transitionProbability = EncogFileSection.ParseMatrix(p, HiddenMarkovModel.TAG_TRANSITION);
}
else if (section.SectionName.Equals("HMM")
&& section.SubSectionName.StartsWith("DISTRIBUTION-"))
{
IDictionary<String, String> p = section.ParseParams();
String t = p[HiddenMarkovModel.TAG_DIST_TYPE];
if ("ContinousDistribution".Equals(t))
{
double[] mean = section.ParseDoubleArray(p, HiddenMarkovModel.TAG_MEAN);
Matrix cova = EncogFileSection.ParseMatrix(p, HiddenMarkovModel.TAG_COVARIANCE);
ContinousDistribution dist = new ContinousDistribution(mean, cova.Data);
distributions.Add(dist);
}
else if ("DiscreteDistribution".Equals(t))
{
Matrix prob = EncogFileSection.ParseMatrix(p, HiddenMarkovModel.TAG_PROBABILITIES);
DiscreteDistribution dist = new DiscreteDistribution(prob.Data);
distributions.Add(dist);
}
}
}
HiddenMarkovModel result = new HiddenMarkovModel(states);
EngineArray.PutAll(properties, result.Properties);
result.TransitionProbability = transitionProbability.Data;
result.Pi = pi;
int index = 0;
foreach (IStateDistribution dist in distributions)
{
result.StateDistributions[index++] = dist;
}
return result;
}
/// <inheritdoc/>
public Object Read(Stream istream)
{
int states = 0;
int[] items;
double[] pi = null;
Matrix transitionProbability = null;
IDictionary <String, String> properties = null;
IList <IStateDistribution> distributions = new List <IStateDistribution>();
EncogReadHelper reader = new EncogReadHelper(istream);
EncogFileSection section;
while ((section = reader.ReadNextSection()) != null)
{
if (section.SectionName.Equals("HMM") &&
section.SubSectionName.Equals("PARAMS"))
{
properties = section.ParseParams();
}
if (section.SectionName.Equals("HMM") &&
section.SubSectionName.Equals("CONFIG"))
{
IDictionary <String, String> p = section.ParseParams();
states = EncogFileSection.ParseInt(p, HiddenMarkovModel.TAG_STATES);
if (p.ContainsKey(HiddenMarkovModel.TAG_ITEMS))
{
items = EncogFileSection.ParseIntArray(p, HiddenMarkovModel.TAG_ITEMS);
}
pi = section.ParseDoubleArray(p, HiddenMarkovModel.TAG_PI);
transitionProbability = EncogFileSection.ParseMatrix(p, HiddenMarkovModel.TAG_TRANSITION);
}
else if (section.SectionName.Equals("HMM") &&
section.SubSectionName.StartsWith("DISTRIBUTION-"))
{
IDictionary <String, String> p = section.ParseParams();
String t = p[HiddenMarkovModel.TAG_DIST_TYPE];
if ("ContinousDistribution".Equals(t))
{
double[] mean = section.ParseDoubleArray(p, HiddenMarkovModel.TAG_MEAN);
Matrix cova = EncogFileSection.ParseMatrix(p, HiddenMarkovModel.TAG_COVARIANCE);
ContinousDistribution dist = new ContinousDistribution(mean, cova.Data);
distributions.Add(dist);
}
else if ("DiscreteDistribution".Equals(t))
{
Matrix prob = EncogFileSection.ParseMatrix(p, HiddenMarkovModel.TAG_PROBABILITIES);
DiscreteDistribution dist = new DiscreteDistribution(prob.Data);
distributions.Add(dist);
}
}
}
HiddenMarkovModel result = new HiddenMarkovModel(states);
EngineArray.PutAll(properties, result.Properties);
result.TransitionProbability = transitionProbability.Data;
result.Pi = pi;
int index = 0;
foreach (IStateDistribution dist in distributions)
{
result.StateDistributions[index++] = dist;
}
return(result);
}
