#region License Information
/* HeuristicLab
* Copyright (C) 2002-2011 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
*
* This file is part of HeuristicLab.
*
* HeuristicLab is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* HeuristicLab is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with HeuristicLab. If not, see .
*/
#endregion
using System;
using System.Collections.Generic;
using System.IO;
using System.Linq;
using HeuristicLab.Common;
using HeuristicLab.Core;
using HeuristicLab.Data;
using HeuristicLab.Parameters;
using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
using HeuristicLab.PluginInfrastructure;
namespace HeuristicLab.Problems.DataAnalysis {
[Item("DataAnalysisProblemData", "Represents an item containing all data defining a data analysis problem.")]
[StorableClass]
[NonDiscoverableType]
public class DataAnalysisProblemData : ParameterizedNamedItem, IStorableContent {
protected bool suppressEvents = false;
#region IStorableContent Members
public string Filename { get; set; }
#endregion
#region default data
// y = x^4 + x^3 + x^2 + x
private static double[,] kozaF1 = new double[,] {
{2.017885919, -1.449165046},
{1.30060506, -1.344523885},
{1.147134798, -1.317989331},
{0.877182504, -1.266142284},
{0.852562452, -1.261020794},
{0.431095788, -1.158793317},
{0.112586002, -1.050908405},
{0.04594507, -1.021989402},
{0.042572879, -1.020438113},
{-0.074027291, -0.959859562},
{-0.109178553, -0.938094706},
{-0.259721109, -0.803635355},
{-0.272991057, -0.387519561},
{-0.161978191, -0.193611001},
{-0.102489983, -0.114215349},
{-0.01469968, -0.014918985},
{-0.008863365, -0.008942626},
{0.026751057, 0.026054094},
{0.166922436, 0.14309643},
{0.176953808, 0.1504144},
{0.190233418, 0.159916534},
{0.199800708, 0.166635331},
{0.261502822, 0.207600348},
{0.30182879, 0.232370249},
{0.83763905, 0.468046718}
};
#endregion
#region parameter properties
public ValueParameter DatasetParameter {
get { return (ValueParameter)Parameters["Dataset"]; }
}
public IValueParameter TargetVariableParameter {
get { return (IValueParameter)Parameters["TargetVariable"]; }
}
public IValueParameter> InputVariablesParameter {
get { return (IValueParameter>)Parameters["InputVariables"]; }
}
public IValueParameter TrainingSamplesStartParameter {
get { return (IValueParameter)Parameters["TrainingSamplesStart"]; }
}
public IValueParameter TrainingSamplesEndParameter {
get { return (IValueParameter)Parameters["TrainingSamplesEnd"]; }
}
public IValueParameter TestSamplesStartParameter {
get { return (IValueParameter)Parameters["TestSamplesStart"]; }
}
public IValueParameter TestSamplesEndParameter {
get { return (IValueParameter)Parameters["TestSamplesEnd"]; }
}
public IValueParameter ValidationPercentageParameter {
get { return (IValueParameter)Parameters["ValidationPercentage"]; }
}
#endregion
#region properties
public Dataset Dataset {
get { return DatasetParameter.Value; }
set {
if (value != Dataset) {
if (value == null) throw new ArgumentNullException();
DatasetParameter.Value = value;
}
}
}
public StringValue TargetVariable {
get { return TargetVariableParameter.Value; }
set {
if (value != TargetVariableParameter.Value) {
if (value == null) throw new ArgumentNullException();
if (TargetVariable != null) DeregisterStringValueEventHandlers(TargetVariable);
TargetVariableParameter.Value = value;
}
}
}
public ICheckedItemList InputVariables {
get { return InputVariablesParameter.Value; }
set {
if (value != InputVariables) {
if (value == null) throw new ArgumentNullException();
if (InputVariables != null) DeregisterInputVariablesEventHandlers();
InputVariablesParameter.Value = value;
}
}
}
public IntValue TrainingSamplesStart {
get { return TrainingSamplesStartParameter.Value; }
set {
if (value != TrainingSamplesStart) {
if (value == null) throw new ArgumentNullException();
if (TrainingSamplesStart != null) DeregisterValueTypeEventHandlers(TrainingSamplesStart);
TrainingSamplesStartParameter.Value = value;
}
}
}
public IntValue TrainingSamplesEnd {
get { return TrainingSamplesEndParameter.Value; }
set {
if (value != TrainingSamplesEnd) {
if (value == null) throw new ArgumentNullException();
if (TrainingSamplesEnd != null) DeregisterValueTypeEventHandlers(TrainingSamplesEnd);
TrainingSamplesEndParameter.Value = value;
}
}
}
public IntValue TestSamplesStart {
get { return TestSamplesStartParameter.Value; }
set {
if (value != TestSamplesStart) {
if (value == null) throw new ArgumentNullException();
if (TestSamplesStart != null) DeregisterValueTypeEventHandlers(TestSamplesStart);
TestSamplesStartParameter.Value = value;
}
}
}
public IntValue TestSamplesEnd {
get { return TestSamplesEndParameter.Value; }
set {
if (value != TestSamplesEnd) {
if (value == null) throw new ArgumentNullException();
if (TestSamplesEnd != null) DeregisterValueTypeEventHandlers(TestSamplesEnd);
TestSamplesEndParameter.Value = value;
}
}
}
public PercentValue ValidationPercentage {
get { return ValidationPercentageParameter.Value; }
set {
if (value != ValidationPercentage) {
if (value == null) throw new ArgumentNullException();
if (value.Value < 0 || value.Value > 1) throw new ArgumentException("ValidationPercentage must be between 0 and 1.");
if (ValidationPercentage != null) DeregisterValueTypeEventHandlers(ValidationPercentage);
ValidationPercentageParameter.Value = value;
}
}
}
public IEnumerable TrainingIndizes {
get {
return Enumerable.Range(TrainingSamplesStart.Value, TrainingSamplesEnd.Value - TrainingSamplesStart.Value)
.Where(i => i >= 0 && i < Dataset.Rows && (i < TestSamplesStart.Value || TestSamplesEnd.Value <= i));
}
}
public IEnumerable TestIndizes {
get {
return Enumerable.Range(TestSamplesStart.Value, TestSamplesEnd.Value - TestSamplesStart.Value)
.Where(i => i >= 0 && i < Dataset.Rows);
}
}
#endregion
[StorableConstructor]
protected DataAnalysisProblemData(bool deserializing) : base(deserializing) { }
protected DataAnalysisProblemData(DataAnalysisProblemData original, Cloner cloner)
: base(original, cloner) {
RegisterParameterEventHandlers();
RegisterParameterValueEventHandlers();
}
public DataAnalysisProblemData()
: base() {
var inputVariables = new CheckedItemList();
StringValue inputVariable = new StringValue("x");
inputVariables.Add(inputVariable);
StringValue targetVariable = new StringValue("y");
var validTargetVariables = new ItemSet();
validTargetVariables.Add(targetVariable);
Parameters.Add(new ValueParameter("Dataset", new Dataset(new string[] { "y", "x" }, kozaF1)));
Parameters.Add(new ValueParameter>("InputVariables", inputVariables.AsReadOnly()));
Parameters.Add(new ConstrainedValueParameter("TargetVariable", validTargetVariables, targetVariable));
Parameters.Add(new ValueParameter("TrainingSamplesStart", new IntValue(0)));
Parameters.Add(new ValueParameter("TrainingSamplesEnd", new IntValue(15)));
Parameters.Add(new ValueParameter("TestSamplesStart", new IntValue(15)));
Parameters.Add(new ValueParameter("TestSamplesEnd", new IntValue(25)));
Parameters.Add(new ValueParameter("ValidationPercentage", "The relative amount of the training samples that should be used as validation set.", new PercentValue(0.5)));
DatasetParameter.ReactOnValueToStringChangedAndValueItemImageChanged = false;
RegisterParameterEventHandlers();
RegisterParameterValueEventHandlers();
}
public DataAnalysisProblemData(Dataset dataset, IEnumerable inputVariables, string targetVariable,
int trainingSamplesStart, int trainingSamplesEnd, int testSamplesStart, int testSamplesEnd) {
var inputVariablesList = new CheckedItemList(inputVariables.Select(x => new StringValue(x)).ToList());
StringValue targetVariableValue = new StringValue(targetVariable);
var validTargetVariables = new ItemSet();
foreach (var variable in dataset.VariableNames)
if (variable != targetVariable)
validTargetVariables.Add(new StringValue(variable));
validTargetVariables.Add(targetVariableValue);
Parameters.Add(new ValueParameter("Dataset", dataset));
Parameters.Add(new ValueParameter>("InputVariables", inputVariablesList.AsReadOnly()));
Parameters.Add(new ConstrainedValueParameter("TargetVariable", validTargetVariables, targetVariableValue));
Parameters.Add(new ValueParameter("TrainingSamplesStart", new IntValue(trainingSamplesStart)));
Parameters.Add(new ValueParameter("TrainingSamplesEnd", new IntValue(trainingSamplesEnd)));
Parameters.Add(new ValueParameter("TestSamplesStart", new IntValue(testSamplesStart)));
Parameters.Add(new ValueParameter("TestSamplesEnd", new IntValue(testSamplesEnd)));
Parameters.Add(new ValueParameter("ValidationPercentage", "The relative amount of the training samples that should be used as validation set.", new PercentValue(0.5)));
DatasetParameter.ReactOnValueToStringChangedAndValueItemImageChanged = false;
RegisterParameterEventHandlers();
RegisterParameterValueEventHandlers();
}
public override IDeepCloneable Clone(Cloner cloner) {
return new DataAnalysisProblemData(this, cloner);
}
[StorableHook(HookType.AfterDeserialization)]
private void AfterDeserialization() {
if (!Parameters.ContainsKey("ValidationPercentage"))
Parameters.Add(new ValueParameter("ValidationPercentage", "The relative amount of the training samples that should be used as validation set.", new PercentValue(0.5)));
DatasetParameter.ReactOnValueToStringChangedAndValueItemImageChanged = false;
RegisterParameterEventHandlers();
RegisterParameterValueEventHandlers();
}
#region events
public event EventHandler ProblemDataChanged;
protected virtual void OnProblemDataChanged(EventArgs e) {
if (TrainingSamplesStart.Value < 0) TrainingSamplesStart.Value = 0;
else if (TestSamplesStart.Value < 0) TestSamplesStart.Value = 0;
else if (TrainingSamplesEnd.Value > Dataset.Rows - 1) TrainingSamplesEnd.Value = Dataset.Rows - 1;
else if (TestSamplesEnd.Value > Dataset.Rows - 1) TestSamplesEnd.Value = Dataset.Rows - 1;
else if (TrainingSamplesStart.Value > TrainingSamplesEnd.Value) TrainingSamplesStart.Value = TestSamplesEnd.Value;
else if (TestSamplesStart.Value > TestSamplesEnd.Value) TestSamplesStart.Value = TestSamplesEnd.Value;
else if (ValidationPercentage.Value < 0) ValidationPercentage.Value = 0;
else if (ValidationPercentage.Value > 1) ValidationPercentage.Value = 1;
else if (!TrainingIndizes.Any()) throw new ArgumentException("No training samples are available.");
else if (!suppressEvents) {
var listeners = ProblemDataChanged;
if (listeners != null) listeners(this, e);
}
}
private void RegisterParameterEventHandlers() {
DatasetParameter.ValueChanged += new EventHandler(DatasetParameter_ValueChanged);
InputVariablesParameter.ValueChanged += new EventHandler(InputVariablesParameter_ValueChanged);
TargetVariableParameter.ValueChanged += new EventHandler(TargetVariableParameter_ValueChanged);
TrainingSamplesStartParameter.ValueChanged += new EventHandler(TrainingSamplesStartParameter_ValueChanged);
TrainingSamplesEndParameter.ValueChanged += new EventHandler(TrainingSamplesEndParameter_ValueChanged);
TestSamplesStartParameter.ValueChanged += new EventHandler(TestSamplesStartParameter_ValueChanged);
TestSamplesEndParameter.ValueChanged += new EventHandler(TestSamplesEndParameter_ValueChanged);
ValidationPercentageParameter.ValueChanged += new EventHandler(ValidationPercentageParameter_ValueChanged);
}
private void RegisterParameterValueEventHandlers() {
RegisterInputVariablesEventHandlers();
if (TargetVariable != null) RegisterStringValueEventHandlers(TargetVariable);
RegisterValueTypeEventHandlers(TrainingSamplesStart);
RegisterValueTypeEventHandlers(TrainingSamplesEnd);
RegisterValueTypeEventHandlers(TestSamplesStart);
RegisterValueTypeEventHandlers(TestSamplesEnd);
RegisterValueTypeEventHandlers(ValidationPercentage);
}
#region parameter value changed event handlers
private void DatasetParameter_ValueChanged(object sender, EventArgs e) {
OnProblemDataChanged(EventArgs.Empty);
}
private void InputVariablesParameter_ValueChanged(object sender, EventArgs e) {
RegisterInputVariablesEventHandlers();
OnProblemDataChanged(EventArgs.Empty);
}
private void TargetVariableParameter_ValueChanged(object sender, EventArgs e) {
if (TargetVariable != null) {
RegisterStringValueEventHandlers(TargetVariable);
OnProblemDataChanged(EventArgs.Empty);
}
}
private void TrainingSamplesStartParameter_ValueChanged(object sender, EventArgs e) {
RegisterValueTypeEventHandlers(TrainingSamplesStart);
OnProblemDataChanged(EventArgs.Empty);
}
private void TrainingSamplesEndParameter_ValueChanged(object sender, EventArgs e) {
RegisterValueTypeEventHandlers(TrainingSamplesEnd);
OnProblemDataChanged(EventArgs.Empty);
}
private void TestSamplesStartParameter_ValueChanged(object sender, EventArgs e) {
RegisterValueTypeEventHandlers(TestSamplesStart);
OnProblemDataChanged(EventArgs.Empty);
}
private void TestSamplesEndParameter_ValueChanged(object sender, EventArgs e) {
RegisterValueTypeEventHandlers(TestSamplesEnd);
OnProblemDataChanged(EventArgs.Empty);
}
private void ValidationPercentageParameter_ValueChanged(object sender, EventArgs e) {
RegisterValueTypeEventHandlers(ValidationPercentage);
OnProblemDataChanged(EventArgs.Empty);
}
#endregion
private void RegisterInputVariablesEventHandlers() {
InputVariables.CollectionReset += new HeuristicLab.Collections.CollectionItemsChangedEventHandler>(InputVariables_CollectionReset);
InputVariables.ItemsAdded += new HeuristicLab.Collections.CollectionItemsChangedEventHandler>(InputVariables_ItemsAdded);
InputVariables.ItemsRemoved += new HeuristicLab.Collections.CollectionItemsChangedEventHandler>(InputVariables_ItemsRemoved);
InputVariables.CheckedItemsChanged += new HeuristicLab.Collections.CollectionItemsChangedEventHandler>(InputVariables_CheckedItemsChanged);
foreach (var item in InputVariables) {
item.ValueChanged += new EventHandler(InputVariable_ValueChanged);
}
}
private void DeregisterInputVariablesEventHandlers() {
InputVariables.CollectionReset -= new HeuristicLab.Collections.CollectionItemsChangedEventHandler>(InputVariables_CollectionReset);
InputVariables.ItemsAdded -= new HeuristicLab.Collections.CollectionItemsChangedEventHandler>(InputVariables_ItemsAdded);
InputVariables.ItemsRemoved -= new HeuristicLab.Collections.CollectionItemsChangedEventHandler>(InputVariables_ItemsRemoved);
InputVariables.CheckedItemsChanged -= new HeuristicLab.Collections.CollectionItemsChangedEventHandler>(InputVariables_CheckedItemsChanged);
foreach (var item in InputVariables) {
item.ValueChanged -= new EventHandler(InputVariable_ValueChanged);
}
}
private void InputVariables_CheckedItemsChanged(object sender, HeuristicLab.Collections.CollectionItemsChangedEventArgs> e) {
OnProblemDataChanged(e);
}
private void InputVariables_ItemsRemoved(object sender, HeuristicLab.Collections.CollectionItemsChangedEventArgs> e) {
foreach (var indexedItem in e.Items)
indexedItem.Value.ValueChanged -= new EventHandler(InputVariable_ValueChanged);
OnProblemDataChanged(e);
}
private void InputVariables_ItemsAdded(object sender, HeuristicLab.Collections.CollectionItemsChangedEventArgs> e) {
foreach (var indexedItem in e.Items)
indexedItem.Value.ValueChanged += new EventHandler(InputVariable_ValueChanged);
OnProblemDataChanged(e);
}
private void InputVariables_CollectionReset(object sender, HeuristicLab.Collections.CollectionItemsChangedEventArgs> e) {
foreach (var indexedItem in e.OldItems)
indexedItem.Value.ValueChanged -= new EventHandler(InputVariable_ValueChanged);
OnProblemDataChanged(e);
}
private void InputVariable_ValueChanged(object sender, EventArgs e) {
OnProblemDataChanged(e);
}
#region helper
private void RegisterValueTypeEventHandlers(ValueTypeValue value) where T : struct {
value.ValueChanged += new EventHandler(value_ValueChanged);
}
private void DeregisterValueTypeEventHandlers(ValueTypeValue value) where T : struct {
value.ValueChanged -= new EventHandler(value_ValueChanged);
}
private void RegisterStringValueEventHandlers(StringValue value) {
value.ValueChanged += new EventHandler(value_ValueChanged);
}
private void DeregisterStringValueEventHandlers(StringValue value) {
value.ValueChanged -= new EventHandler(value_ValueChanged);
}
private void value_ValueChanged(object sender, EventArgs e) {
OnProblemDataChanged(e);
}
#endregion
#endregion
public virtual void ImportFromFile(string fileName) {
var csvFileParser = new TableFileParser();
csvFileParser.Parse(fileName);
suppressEvents = true;
Name = "Data imported from " + Path.GetFileName(fileName);
Dataset = new Dataset(csvFileParser.VariableNames, csvFileParser.Values);
Dataset.Name = Path.GetFileName(fileName);
var variableNames = Dataset.VariableNames.Select(x => new StringValue(x).AsReadOnly()).ToList();
((ConstrainedValueParameter)TargetVariableParameter).ValidValues.Clear();
foreach (var variableName in variableNames)
((ConstrainedValueParameter)TargetVariableParameter).ValidValues.Add(variableName);
TargetVariable = variableNames.First();
InputVariables = new CheckedItemList(variableNames).AsReadOnly();
InputVariables.SetItemCheckedState(variableNames.First(), false);
int middle = (int)(csvFileParser.Rows * 0.5);
TrainingSamplesEnd = new IntValue(middle);
TrainingSamplesStart = new IntValue(0);
TestSamplesEnd = new IntValue(csvFileParser.Rows);
TestSamplesStart = new IntValue(middle);
suppressEvents = false;
OnProblemDataChanged(EventArgs.Empty);
}
}
}