/// <summary>
/// Returns a collation of all unassigned, training, validation, and test data points.
/// </summary>
/// <returns></returns>
public DataPoint[] GetWholeSet()
{
DataPoint[] set = new DataPoint[unassigned.Count + TrainingSet.Length +
ValidationSet.Length + TestSet.Length];
TrainingSet.CopyTo(set, 0);
ValidationSet.CopyTo(set, TrainingSet.Length);
TestSet.CopyTo(set, TrainingSet.Length + ValidationSet.Length);
unassigned.CopyTo(set, set.Length - unassigned.Count);
return(set);
}
public void CopyTo()
{
// Setup
var sequence = InitializeSequence();
var set = new TestSet <SimpleType>(sequence);
var array = new SimpleType[4];
// Execute
set.CopyTo(array, 0);
// Assert
Assert.IsTrue(array.All(i => i != null));
Assert.AreSame(sequence[0], array[0]);
Assert.AreSame(sequence[1], array[1]);
Assert.AreSame(sequence[2], array[2]);
Assert.AreSame(sequence[3], array[3]);
}
/// <summary>
/// Randomly assign any unassigned points to the training, validation, and test sets with
/// the given ratio of probabilities.
/// </summary>
/// <param name="trainingRatio"></param>
/// <param name="validationRatio"></param>
/// <param name="testRatio"></param>
public void AssignDataPoints(double trainingRatio, double validationRatio,
double testRatio)
{
// Calculate probability cutoffs for assignment
double cutoff1 = trainingRatio / (trainingRatio + validationRatio + testRatio);
double cutoff2 = (trainingRatio + validationRatio) /
(trainingRatio + validationRatio + testRatio);
List <DataPoint> training = new List <DataPoint>();
List <DataPoint> validation = new List <DataPoint>();
List <DataPoint> test = new List <DataPoint>();
while (unassigned.Count > 0)
{
double randomDouble = random.NextDouble();
if (randomDouble < cutoff1)
{
training.Add(unassigned.Dequeue());
}
else if (randomDouble < cutoff2)
{
validation.Add(unassigned.Dequeue());
}
else
{
test.Add(unassigned.Dequeue());
}
}
// Construct new arrays
DataPoint[] trainingBuffer = new DataPoint[TrainingSet.Length + training.Count];
TrainingSet.CopyTo(trainingBuffer, training.Count);
training.CopyTo(trainingBuffer, 0);
TrainingSet = trainingBuffer;
DataPoint[] validationBuffer = new DataPoint[ValidationSet.Length + validation.Count];
ValidationSet.CopyTo(validationBuffer, validation.Count);
validation.CopyTo(validationBuffer, 0);
ValidationSet = validationBuffer;
DataPoint[] testBuffer = new DataPoint[TestSet.Length + test.Count];
TestSet.CopyTo(testBuffer, test.Count);
test.CopyTo(testBuffer, 0);
TestSet = testBuffer;
}
public void CopyToAtNonZeroArrayIndex()
{
// Setup
var sequence = InitializeSequence();
var set = new TestSet <SimpleType>(sequence);
var array = new SimpleType[5];
// Execute
set.CopyTo(array, 1);
// Assert
Assert.IsNull(array[0]);
Assert.IsTrue(array.Skip(1).All(i => i != null));
Assert.AreSame(sequence[0], array[1]);
Assert.AreSame(sequence[1], array[2]);
Assert.AreSame(sequence[2], array[3]);
Assert.AreSame(sequence[3], array[4]);
}
public void CopyToAtNonZeroArrayIndex()
{
// Setup
var sequence = InitializeSequence();
var set = new TestSet<SimpleType>(sequence);
var array = new SimpleType[5];
// Execute
set.CopyTo(array, 1);
// Assert
Assert.IsNull(array[0]);
Assert.IsTrue(array.Skip(1).All(i => i != null));
Assert.AreSame(sequence[0], array[1]);
Assert.AreSame(sequence[1], array[2]);
Assert.AreSame(sequence[2], array[3]);
Assert.AreSame(sequence[3], array[4]);
}
public void CopyTo()
{
// Setup
var sequence = InitializeSequence();
var set = new TestSet<SimpleType>(sequence);
var array = new SimpleType[4];
// Execute
set.CopyTo(array, 0);
// Assert
Assert.IsTrue(array.All(i => i != null));
Assert.AreSame(sequence[0], array[0]);
Assert.AreSame(sequence[1], array[1]);
Assert.AreSame(sequence[2], array[2]);
Assert.AreSame(sequence[3], array[3]);
}