public string writeModel(string outModelPath)
{
if (lr == null)
{
buildModel();
}
outPath = outModelPath;
using (System.IO.StreamWriter sw = new System.IO.StreamWriter(outPath))
{
sw.WriteLine(modelTypes.LogisticRegression.ToString());
sw.WriteLine(InTablePath);
sw.WriteLine(String.Join(",", IndependentFieldNames));
sw.WriteLine(String.Join(",", DependentFieldNames));
sw.WriteLine(String.Join(",", ClassFieldNames));
sw.WriteLine(String.Join(",", Categories));
sw.WriteLine(NumberOfIndependentVariables.ToString());
sw.WriteLine(NumberOfCategories.ToString());
sw.WriteLine(LogLikelihood.ToString());
sw.WriteLine(Deviance.ToString());
sw.WriteLine(X2.ToString());
sw.WriteLine(PValue.ToString());
sw.WriteLine(String.Join(",", (from double d in minValues select d.ToString()).ToArray()));
sw.WriteLine(String.Join(",", (from double d in maxValues select d.ToString()).ToArray()));
sw.WriteLine(String.Join(",", (from double d in sumValues select d.ToString()).ToArray()));
sw.WriteLine(String.Join(",", (from double d in Coefficients select d.ToString()).ToArray()));
sw.WriteLine(String.Join(",", (from double d in StandardError select d.ToString()).ToArray()));
sw.WriteLine(String.Join(",", (from Accord.Statistics.Testing.WaldTest d in WaldStatistic select d.Statistic.ToString()).ToArray()));
sw.WriteLine(String.Join(",", (from Accord.Statistics.Testing.WaldTest d in WaldStatistic select d.PValue.ToString()).ToArray()));
sw.Close();
}
return(outPath);
}
public void getReport(double alpha)
{
Forms.RunningProcess.frmRunningProcessDialog rd = new Forms.RunningProcess.frmRunningProcessDialog(false);
rd.Text = "Logistic Regression Results";
rd.TopLevel = true;
rd.pgbProcess.Visible = false;
rd.FormBorderStyle = System.Windows.Forms.FormBorderStyle.Sizable;
rd.addMessage("LogLikelihood = " + LogLikelihood.ToString());
rd.addMessage("Deviance = " + Deviance.ToString());
rd.addMessage("Likelihood Ratio Chi-squared = " + X2.ToString() + " DF = " + DegreesOfFredom.ToString() + " p-value = " + PValue.ToString());
rd.addMessage("\n\nClasses: " + String.Join(", ", Categories));
rd.addMessage("Reference Class = " + Categories[0]);
rd.addMessage("\nParameter coefficients:\n");
rd.addMessage("intercept, " + String.Join(", ", IndependentFieldNames));
rd.addMessage("Coef: " + String.Join(" ", (from double d in Coefficients select d.ToString()).ToArray()));
rd.addMessage("STE: " + String.Join(", ", (from double d in StandardError select d.ToString()).ToArray()));
rd.addMessage("\n\nWald stats for coefficients:\nchi-sq: " + String.Join(", ", (from Accord.Statistics.Testing.WaldTest d in WaldStatistic select d.Statistic.ToString()).ToArray()));
rd.addMessage("p-value: " + String.Join(", ", (from Accord.Statistics.Testing.WaldTest d in WaldStatistic select d.PValue.ToString()).ToArray()));
try
{
if (ModelHelper.chartingAvailable() && System.Windows.Forms.MessageBox.Show("Do you want to build probability graphs?", "Graphs", System.Windows.Forms.MessageBoxButtons.YesNo) == System.Windows.Forms.DialogResult.Yes)
{
createRegChart();
}
}
catch
{
System.Windows.Forms.MessageBox.Show("Cannot create charts");
}
rd.Show();
rd.enableClose();
}
private double[] PredictNextValue(IGaussianMixtureModelState model, IPredictionRequest request, double[][] trainingSet)
{
var N = trainingSet.Length;
var K = trainingSet[0].Length;
var result = new double[K];
var yesterday = trainingSet[N - 1];
var yesterdayLikelihood = LogLikelihood.Calculate(new[] { yesterday }, model.Mixture.Coefficients, GetCovariances(model.Mixture.Components), GetMeans(model.Mixture.Components));
Debug.WriteLine("Yesterday Likelihood : " + new Vector(yesterday) + " : " + yesterdayLikelihood + " ");
var guessess = FindMostSimilarObservations(model, trainingSet, yesterdayLikelihood, request.Tolerance);
var bestGuessPlace = FindBestGuess(request, guessess);
var tomorrow = trainingSet[bestGuessPlace.PlaceInSequence + 1];
var mostSimilar = trainingSet[bestGuessPlace.PlaceInSequence];
for (var k = 0; k < K; k++)
{
if (bestGuessPlace.PlaceInSequence != trainingSet.Length)
{
result[k] = yesterday[k] + (tomorrow[k] - mostSimilar[k]);
}
}
Debug.WriteLine("Predicted : " + new Vector(result) + " : " + LogLikelihood.Calculate(new[] { result }, model.Mixture.Coefficients, GetCovariances(model.Mixture.Components), GetMeans(model.Mixture.Components)));
return(result);
}
public void getReport(double alpha)
{
Forms.RunningProcess.frmRunningProcessDialog rd = new Forms.RunningProcess.frmRunningProcessDialog(false);
rd.Text = "GLM Results";
rd.TopLevel = true;
rd.pgbProcess.Visible = false;
rd.FormBorderStyle = System.Windows.Forms.FormBorderStyle.Sizable;
rd.addMessage("Dependent field = " + DependentFieldNames[0]);
rd.addMessage("Independent fields = " + String.Join(", ", IndependentFieldNames));
rd.addMessage("Sample size = " + SampleSize.ToString());
rd.addMessage("Iteration = " + Iterations.ToString());
rd.addMessage("Delta Convergence " + DeltaC.ToString());
rd.addMessage("Chi-Sqr = " + ChiSquare.ToString() + " p-value = " + PValue.ToString());
rd.addMessage("Deviance = " + Deviance.ToString());
rd.addMessage("Log Likelihood = " + LogLikelihood.ToString());
rd.addMessage("Log Likelihood Ratio = " + LogLikelihoodratio.ToString() + "\n\nCoefficents and standard errors:\n");
rd.addMessage("Param: Intercept, " + String.Join(", ", IndependentFieldNames));
rd.addMessage("Coef: " + string.Join(", ", (from double d in Coefficients select d.ToString()).ToArray()));
rd.addMessage("STE: " + string.Join(", ", (from double d in StdError select d.ToString()).ToArray()) + "\n");
try
{
if (ModelHelper.chartingAvailable() && System.Windows.Forms.MessageBox.Show("Do you want to build distribution graphs?", "Graphs", System.Windows.Forms.MessageBoxButtons.YesNo) == System.Windows.Forms.DialogResult.Yes)
{
createRegChart();
}
}
catch
{
System.Windows.Forms.MessageBox.Show("Cannot create charts.");
}
rd.Show();
rd.enableClose();
}
/// <summary>
/// Fits Distribution parameters to given observation vector
/// </summary>
/// <param name="observations">Vector</param>
/// <param name="weights">Vector</param>
/// <param name="likelihood">Output value</param>
/// <returns>IDistribution</returns>
public override IDistribution Evaluate(double[][] observations, double[] weights, out double likelihood)
{
var mean = Utils.Mean(observations, weights);
var covariance = Utils.Covariance(observations, mean, weights);
var result = new NormalDistribution(mean, covariance);
likelihood = LogLikelihood.Calculate(observations, result.Covariance, result.Mean);
return(result);
}
public void getReport(double alpha)
{
Forms.RunningProcess.frmRunningProcessDialog rd = new Forms.RunningProcess.frmRunningProcessDialog(false);
rd.Text = "PLR Results";
rd.TopLevel = true;
rd.pgbProcess.Visible = false;
rd.FormBorderStyle = System.Windows.Forms.FormBorderStyle.Sizable;
rd.addMessage("Iterations = " + NumberOfIterationsToConverge.ToString());
rd.addMessage("Convergence = " + Delta.ToString());
rd.addMessage("LogLikelihood = " + LogLikelihood.ToString());
rd.addMessage("Deviance = " + Deviance.ToString());
rd.addMessage("Likelihood Ratio Chi-squared = " + X2.ToString() + " DF = " + DegreesOfFredom.ToString() + " p-value = " + PValue.ToString());
rd.addMessage("\n\nClasses: " + String.Join(", ", Categories));
rd.addMessage("Reference Class = " + Categories[0]);
rd.addMessage("\nParameter coefficients:\n");
rd.addMessage("intercept, " + String.Join(", ", IndependentFieldNames));
for (int i = 0; i < Coefficients.Length; i++)
{
double[] c = Coefficients[i];
rd.addMessage(String.Join(" ", (from double d in c select d.ToString()).ToArray()));
}
rd.addMessage("\nParameter standard error:");
for (int i = 0; i < StandardError.Length; i++)
{
double[] c = StandardError[i];
rd.addMessage(String.Join(", ", (from double d in c select d.ToString()).ToArray()));
}
rd.addMessage("\nParameter Wald Chi-Squared:");
for (int i = 0; i < WaldStatistic.Length; i++)
{
double[] c = WaldStatistic[i];
rd.addMessage(String.Join(", ", (from double d in c select d.ToString()).ToArray()));
}
rd.addMessage("\nParameter Wald P-value:");
for (int i = 0; i < WaldPvalue.Length; i++)
{
double[] c = WaldPvalue[i];
rd.addMessage(String.Join(", ", (from double d in c select d.ToString()).ToArray()));
}
try
{
if (ModelHelper.chartingAvailable() && System.Windows.Forms.MessageBox.Show("Do you want to build probability graphs?", "Graphs", System.Windows.Forms.MessageBoxButtons.YesNo) == System.Windows.Forms.DialogResult.Yes)
{
regChart();
//if (mlr == null) mlr = buildModel();
}
}
catch
{
System.Windows.Forms.MessageBox.Show("Cannot create charts");
}
rd.Show();
rd.enableClose();
}
public string writeModel(string outModelPath)
{
if (mlr == null)
{
buildModel();
}
outPath = outModelPath;
using (System.IO.StreamWriter sw = new System.IO.StreamWriter(outPath))
{
sw.WriteLine(modelTypes.PLR);
sw.WriteLine(InTablePath);
sw.WriteLine(String.Join(",", IndependentFieldNames));
sw.WriteLine(String.Join(",", DependentFieldNames));
sw.WriteLine(String.Join(",", ClassFieldNames));
sw.WriteLine(String.Join(",", Categories));
sw.WriteLine(NumberOfIndependentVariables.ToString());
sw.WriteLine(NumberOfCategories.ToString());
sw.WriteLine(Delta.ToString());
sw.WriteLine(NumberOfIterationsToConverge.ToString());
sw.WriteLine(LogLikelihood.ToString());
sw.WriteLine(Deviance.ToString());
sw.WriteLine(X2.ToString());
sw.WriteLine(PValue.ToString());
sw.WriteLine(String.Join(",", (from double d in minValues select d.ToString()).ToArray()));
sw.WriteLine(String.Join(",", (from double d in maxValues select d.ToString()).ToArray()));
sw.WriteLine(String.Join(",", (from double d in sumValues select d.ToString()).ToArray()));
for (int i = 0; i < Coefficients.Length; i++)
{
string[] vlArr = (from double d in Coefficients[i] select d.ToString()).ToArray();
sw.WriteLine(String.Join(" ", vlArr));
}
for (int i = 0; i < StandardError.Length; i++)
{
string[] vlArr = (from double d in StandardError[i] select d.ToString()).ToArray();
sw.WriteLine(String.Join(" ", vlArr));
}
for (int i = 0; i < WaldStatistic.Length; i++)
{
string[] vlArr = (from double d in WaldStatistic[i] select d.ToString()).ToArray();
sw.WriteLine(String.Join(" ", vlArr));
}
for (int i = 0; i < WaldPvalue.Length; i++)
{
string[] vlArr = (from double d in WaldPvalue[i] select d.ToString()).ToArray();
sw.WriteLine(String.Join(" ", vlArr));
}
sw.Close();
}
return(outPath);
}
private double doItemSimilarity(long itemID1, long itemID2, long preferring1, long numUsers)
{
DataModel model = base.getDataModel();
long num = model.getNumUsersWithPreferenceFor(itemID1, itemID2);
if (num == 0L)
{
return(double.NaN);
}
long num2 = model.getNumUsersWithPreferenceFor(itemID2);
double num3 = LogLikelihood.logLikelihoodRatio(num, num2 - num, preferring1 - num, ((numUsers - preferring1) - num2) + num);
return(1.0 - (1.0 / (1.0 + num3)));
}
private double doItemSimilarity(long itemID1, long itemID2, long preferring1, long numUsers)
{
IDataModel dataModel = getDataModel();
long preferring1and2 = dataModel.GetNumUsersWithPreferenceFor(itemID1, itemID2);
if (preferring1and2 == 0)
{
return(Double.NaN);
}
long preferring2 = dataModel.GetNumUsersWithPreferenceFor(itemID2);
double logLikelihood =
LogLikelihood.logLikelihoodRatio(preferring1and2,
preferring2 - preferring1and2,
preferring1 - preferring1and2,
numUsers - preferring1 - preferring2 + preferring1and2);
return(1.0 - 1.0 / (1.0 + logLikelihood));
}
public double userSimilarity(long userID1, long userID2)
{
DataModel model = base.getDataModel();
FastIDSet other = model.getItemIDsFromUser(userID1);
FastIDSet set2 = model.getItemIDsFromUser(userID2);
long num = other.size();
long num2 = set2.size();
long num3 = (num < num2) ? ((long)set2.intersectionSize(other)) : ((long)other.intersectionSize(set2));
if (num3 == 0L)
{
return(double.NaN);
}
long num4 = model.getNumItems();
double num5 = LogLikelihood.logLikelihoodRatio(num3, num2 - num3, num - num3, ((num4 - num) - num2) + num3);
return(1.0 - (1.0 / (1.0 + num5)));
}
public IDistribution Evaluate(double[][] observations, double[] weights, double treshold, out double likelihood)
{
// Calculate Maximization of the Likelihood
_stepsTillConvirgence = 0;
var convirged = false;
var pdf = (IMultivariateDistribution[])_components.Clone();
var K = _coefficients.Length;
var N = observations.Length;
//// Initialize pi, covariance matrix and mean
var pi = (double[])_coefficients.Clone();
weights = weights.Product(N);
likelihood = LogLikelihood.Calculate(observations, _coefficients, GetCovariances(pdf), GetMeans(pdf));
while (!convirged)
{
//// Expectation
var gamma = CalculateGamma(observations, weights, pdf, pi);
//// Maximization
pi = CalculatePi(gamma);
//// For each component train new distribution function
for (var k = 0; k < K; k++)
{
var mean = CalculateMean(gamma, observations, k);
var covariance = CalculateCovariance(gamma, observations, mean, k);
pdf[k] = new NormalDistribution(mean, covariance);
}
//// Check treshold
var newLikelihood = LogLikelihood.Calculate(observations, _coefficients, GetCovariances(pdf), GetMeans(pdf));
if (double.IsNaN(newLikelihood) || double.IsInfinity(newLikelihood))
{
throw new ApplicationException("EM algorithm does not convirged");
}
convirged = (newLikelihood - likelihood) <= treshold;
likelihood = newLikelihood;
_stepsTillConvirgence++;
}
return(new Mixture <IMultivariateDistribution>(pi, pdf));
}
private IList <ObservationWithLikelihood <double[]> > FindMostSimilarObservations(IGaussianMixtureModelState model, double[][] trainingSet, double yesterdayLikelihood, double tolerance)
{
var N = trainingSet.Length;
var guessess = new List <ObservationWithLikelihood <double[]> >();
for (var n = N - 2; n > 0; n--)
{
var x = new[] { trainingSet[n] };
var likelihood = LogLikelihood.Calculate(x, model.Mixture.Coefficients, GetCovariances(model.Mixture.Components), GetMeans(model.Mixture.Components));
//Debug.Write((new Vector(observations[n])).ToString() + " : " + likelihood + " " + Environment.NewLine);
if (Math.Abs(yesterdayLikelihood) - tolerance < Math.Abs(likelihood) && Math.Abs(yesterdayLikelihood) + tolerance > Math.Abs(likelihood))
{
guessess.Add(new ObservationWithLikelihood <double[]>()
{
LogLikelihood = likelihood, Observation = trainingSet[n], PlaceInSequence = n - 1
});
}
}
return(guessess);
}
public double UserSimilarity(long userID1, long userID2)
{
IDataModel dataModel = getDataModel();
FastIDSet prefs1 = dataModel.GetItemIDsFromUser(userID1);
FastIDSet prefs2 = dataModel.GetItemIDsFromUser(userID2);
long prefs1Size = prefs1.Count();
long prefs2Size = prefs2.Count();
long intersectionSize =
prefs1Size < prefs2Size?prefs2.IntersectionSize(prefs1) : prefs1.IntersectionSize(prefs2);
if (intersectionSize == 0)
{
return(Double.NaN);
}
long numItems = dataModel.GetNumItems();
double logLikelihood =
LogLikelihood.logLikelihoodRatio(intersectionSize,
prefs2Size - intersectionSize,
prefs1Size - intersectionSize,
numItems - prefs1Size - prefs2Size + intersectionSize);
return(1.0 - 1.0 / (1.0 + logLikelihood));
}
