public CaseItem SaveCaseItem(CaseItem item)
{
if (item == null) return item;
if (item.CaseId <= 0) return item;
SetDefaultTime(item);
return CasesData.SaveCaseItems(new List<CaseItem>() { item }).First();
}
public List<CaseItem> LoadCaseItemsByCaseId(int caseId)
{
var items = CasesData.LoadCaseItemsByCaseId(caseId);
items.Sort(CaseItemDisplayOrderComparer.Instance);
LoadCaseItemAttachment(items);
return items;
}
public Cases GetCase(int caseId)
{
Cases cases = CasesData.GetCase(caseId);
if (cases != null)
{
cases.CaseItems = CasesData.LoadCaseItemsByCaseId(caseId);
cases.CaseItems = cases.CaseItems ?? new List<CaseItem>(0);
LoadCaseItemAttachment(cases.CaseItems);
}
return cases;
}
private IList <RegressionPoint> MakeRegressionPointList(CasesData dataSet)
{
IList <RegressionPoint> points = new List <RegressionPoint>();
for (int i = 0; i < dataSet.CasesCount; i++)
{
points.Add(new RegressionPoint(dataSet.GetInput(i).At(0), dataSet.GetIdealOutput(i).At(0)));
}
return(points);
}
public List<Cases> LoadCases(Pagination pagination, CaseType caseType)
{
var cases = CasesData.LoadCases(pagination, caseType);
LoadCaseItems(cases);
if (cases != null && cases.Count > 0)
{
var items = new List<CaseItem>();
cases.ForEach(c => { if (c.CaseItems != null)items.AddRange(c.CaseItems); });
LoadCaseItemAttachment(items);
}
return cases ?? new List<Cases>(0);
}
private IList <ClassificationPoint> MakeClassificationPointList(CasesData dataSet)
{
IList <ClassificationPoint> points = new List <ClassificationPoint>();
for (int i = 0; i < dataSet.CasesCount; i++)
{
points.Add(new ClassificationPoint(dataSet.GetInput(i).At(0),
dataSet.GetInput(i).At(1), (int)dataSet.GetClasificationOutput(i).At(0)));
}
return(points);
}
public Cases SaveCase(Cases cases)
{
SetDefaultTime(cases);
cases = CasesData.SaveCase(cases);
if (cases.CaseItems != null && cases.CaseItems.Count > 0)
{
cases.CaseItems.ForEach(SetDefaultTime);
cases.CaseItems = CasesData.SaveCaseItems(cases.CaseItems);
}
return cases;
}
private void StartButtonClick(object sender, RoutedEventArgs e)
{
var layers = LayersTextBox.Text.Split(new string[] { ",", " ", "-", "_", "." }, StringSplitOptions.RemoveEmptyEntries);
List <int> layersVal = new List <int>();
foreach (var layer in layers)
{
layersVal.Add(int.Parse(layer));
}
YesNo bias = (YesNo)BiasCombobox.SelectedItem;
ActivationFunction activation = (ActivationFunction)ActivationCombobox.SelectedItem;
int iterations = int.Parse(Iterations.Text);
double learningRate = LearningRateSlider.Value;
double momentum = MomentumSlider.Value;
BackpropagationRunMode runMode = (BackpropagationRunMode)RunCombobox.SelectedItem;
ProblemType problem = (ProblemType)ProblemCombobox.SelectedItem;
CasesData trainingCases;
CasesData testCases;
YesNo appendTestFile = (YesNo)AppendTestCombobox.SelectedItem;
int input = testData.First().Count();
int output = trainingData.First().Count() - input;
CasesData.InitializeAllData(trainingData, testData, problem, out trainingCases, out testCases);
layersVal.Insert(0, input);
layersVal.Add((problem == ProblemType.Regression) ? output : trainingCases.ClassCount);
NeuralNetwork network = new NeuralNetwork(
(activation == ActivationFunction.Bipolar) ? (IActivation) new BipolarTanhActivation() : new UnipolarSigmoidActivation(),
bias == YesNo.Yes, layersVal.ToArray());
LearningResult learningResult = BackpropagationManager.Run(network, trainingCases, testCases,
runMode, iterations, learningRate, momentum);
if (appendTestFile == YesNo.Yes)
{
AppendCSVile(testDataPath, testCases);
}
ShowNetworkErrorWindow(learningResult);
if (problem == ProblemType.Regression)
{
ShowRegressionWindow(network, trainingCases, testCases);
}
else if (problem == ProblemType.Classification)
{
ShowClassificationWindow(network, trainingCases, testCases);
}
}
private void AppendCSVile(string path, CasesData data)
{
var output = new List <DenseVector>();
for (int i = 0; i < data.CasesCount; i++)
{
output.Add(
data.ProblemType == ProblemType.Classification ?
data.GetClasificationOutput(i) :
data.GetIdealOutput(i));
}
FileManager.AppendDataToCSV(path, output);
}
private void ShowClassificationWindow(NeuralNetwork neuralNetwork, CasesData trainingSet, CasesData testSet)
{
List <ClassificationPoint> testPoints = new List <ClassificationPoint>();
testPoints.AddRange(MakeClassificationPointList(testSet));
List <ClassificationPoint> idealPoints = new List <ClassificationPoint>();
idealPoints.AddRange(MakeClassificationPointList(trainingSet));
Window classificationWindow = new ClassificationWindow(testPoints, idealPoints, trainingSet.ClassIndexes);
classificationWindow.Show();
}
private void ShowRegressionWindow(NeuralNetwork neuralNetwork, CasesData trainingSet, CasesData testSet)
{
List <RegressionPoint> idealPoints = new List <RegressionPoint>();
idealPoints.AddRange(MakeRegressionPointList(trainingSet));
List <RegressionPoint> testPoints = new List <RegressionPoint>();
testPoints.AddRange(MakeRegressionPointList(testSet));
Window regressionWindow = new RegressionWindow(testPoints, idealPoints); // trainingPoints are "ideal"
regressionWindow.Show();
}
private void LoadCaseItems(List<Cases> caseses)
{
if (caseses == null || caseses.Count == 0) return;
List<int> caseIds = caseses.Select(c => c.CaseId).ToList();
if (caseIds.Count == 0) return;
List<CaseItem> items = CasesData.LoadCaseItemsByCaseIds(caseIds);
var caseIdItems = new Dictionary<int, List<CaseItem>>(caseses.Count);
items.ForEach(item =>
{
if (caseIdItems.ContainsKey(item.CaseId)) caseIdItems[item.CaseId].Add(item);
else caseIdItems.Add(item.CaseId, new List<CaseItem>() { item });
});
var itemComparer = CaseItemDisplayOrderComparer.Instance;
foreach (var c in caseses)
{
if (caseIdItems.ContainsKey(c.CaseId))
{
c.CaseItems = caseIdItems[c.CaseId];
c.CaseItems.Sort(itemComparer);
}
}
}
public bool DeleteCase(int caseId)
{
return CasesData.DeleteCase(caseId);
}
public List<Cases> LoadSimpleCases(Pagination pagination, CaseType caseType)
{
var cases = CasesData.LoadCases(pagination, caseType);
return cases ?? new List<Cases>(0);
}
public bool DeleteCaseItem(int caseItemId)
{
return CasesData.DeleteCaseItem(caseItemId);
}
