/// <summary>
/// Prepare the left side and determine the cellshape.
/// </summary>
/// <param name="left"></param>
/// <param name="right"></param>
private CalculationArguments PrepareVariables(AType left, AType right)
{
// Error if the arguments:
// - are boxes
// - or not of the same type and:
// - not numbers
// - or one of them is a Null
if (left.IsBox || right.IsBox ||
(left.Type != right.Type &&
!(Utils.DifferentNumberType(left, right) || left.Type == ATypes.ANull || right.Type == ATypes.ANull)
))
{
throw new Error.Type(TypeErrorText);
}
CalculationArguments arguments = new CalculationArguments()
{
Interval = left,
CellShape = (left.Rank > 1 && left.Length > 0) ? left[0].Shape : new List <int>()
};
if (right.Rank < arguments.CellShape.Count)
{
throw new Error.Rank(RankErrorText);
}
return(arguments);
}
/// <summary>
/// Prepare the left side and determine the cellshape.
/// </summary>
/// <param name="left"></param>
/// <param name="right"></param>
private CalculationArguments PrepareVariables(AType left, AType right)
{
// Error if the arguments:
// - are boxes
// - or not of the same type and:
// - not numbers
// - or one of them is a Null
if (left.IsBox || right.IsBox ||
(left.Type != right.Type &&
!(Utils.DifferentNumberType(left, right) || left.Type == ATypes.ANull || right.Type == ATypes.ANull)
))
{
throw new Error.Type(TypeErrorText);
}
CalculationArguments arguments = new CalculationArguments()
{
Interval = left,
CellShape = (left.Rank > 1 && left.Length > 0) ? left[0].Shape : new List<int>()
};
if (right.Rank < arguments.CellShape.Count)
{
throw new Error.Rank(RankErrorText);
}
return arguments;
}
public void Learn(
SampleList samples, CalculationArguments arguments)
// samples = yjks
{
arguments.reporter?.WriteStart($"Learning the network using a subset of {samples.Count} random samples...");
Stopwatch timer = new Stopwatch();
timer.Start();
int nSamples = samples.Count; // number of sample rows
int nCoefficients = CoefficientCount();
// Current biasses and weights of the neurons in this network:
Single1D coefficients = new Single1D(nCoefficients);
// The derivatives of the cost with respect to the biasses and weights:
Single1D derivatives = new Single1D(nCoefficients);
Single1D velocities = new Single1D(nCoefficients);
velocities.Clear();
MeasurementList measurements = new MeasurementList(nSamples, Last.Count);
GetCoefficients(coefficients, 0);
Minimization minimization = new Minimization()
{
MaxIter = arguments.settings.MaxIter,
Eps = arguments.settings.Epsilon,
Tol = arguments.settings.Tolerance,
};
float finalCost = minimization.MomentumBasedGradientDescent(coefficients, derivatives, velocities,
(iter) =>
{
SetCoefficients(coefficients, 0);
arguments.reporter?.ReportCoefficients(coefficients);
float cost = GetCostAndDerivatives(samples, derivatives, measurements, arguments);
arguments.reporter?.ReportCostAndDerivatives(cost, derivatives, measurements);
return(cost);
}, arguments.settings.LearningRate, arguments.settings.MomentumCoefficient);
arguments.reporter?.WriteEnd($"The network has learned in {timer.Elapsed.TotalSeconds} s, and the final cost value is {finalCost:F4}.");
}
// ----------------------------------------------------------------------------------------
#region Network
public float GetCostAndDerivatives(
SampleList samples, Single1D derivatives, MeasurementList measurements,
CalculationArguments arguments)
{
CostFunctionEnum costFunction = arguments.settings.CostFunction;
float lambda = arguments.settings.Lambda;
int nSamples = samples.Count;
int nCoeffs = derivatives.Count;
float cost = 0f;
Layer last = Last;
for (int i = 0; i < nCoeffs; i++)
{
derivatives[i] = 0f;
}
for (int iSample = 0; iSample < nSamples; iSample++)
{
arguments.ThrowIfCancellationRequested();
Sample sample = samples[iSample];
Single1D measurement = measurements[iSample];
Input.SetActivations(sample.Inputs, 0);
FeedForward(true);
last.GetActivations(measurement, 0);
cost += CostFunction(measurement, sample.Requirements, costFunction);
int weightCount = CountWeight();
cost += 0.5f * lambda * SumWeightSqr() / weightCount; // regularization
last.CalculateDeltas(sample.Requirements, costFunction);
FeedBackward(true);
AddDerivatives(derivatives, 0, lambda / weightCount);
arguments.reporter?.ReportProgress(iSample, nSamples);
}
arguments.reporter?.ReportProgress(0, nSamples);
cost /= nSamples;
for (int i = 0; i < nCoeffs; i++)
{
derivatives[i] /= nSamples;
}
return(cost);
}
/// <summary>
/// Classify element to the corresponding group.
/// </summary>
/// <param name="element"></param>
/// <param name="arguments">Instead of class variables.</param>
/// <returns></returns>
private AType Classify(AType element, CalculationArguments arguments)
{
int index;
if (arguments.Interval.IsArray)
{
AType intervalArray = arguments.Interval;
for (index = 0; index < intervalArray.Length; index++)
{
if (element.CompareTo(intervalArray[index]) <= 0)
{
break;
}
}
}
else
{
index = (element.CompareTo(arguments.Interval) <= 0) ? 0 : arguments.Interval.Length;
}
return(AInteger.Create(index));
}
/// <summary>
/// If the cell rank > 1, first step is check the shape of actual cell equal to determined cell shape.
/// If different we throw Length error. The second step is classify the cell.
/// </summary>
/// <param name="cell"></param>
/// <param name="arguments">Instead of class variables.</param>
/// <returns></returns>
private AType MultipleItemsWalking(AType cell, CalculationArguments arguments)
{
if (arguments.CellShape.Count == cell.Shape.Count)
{
if (!arguments.CellShape.SequenceEqual(cell.Shape))
{
throw new Error.Length(LengthErrorText);
}
return(Classify(cell, arguments));
}
else
{
AType result = AArray.Create(ATypes.AInteger);
foreach (AType item in cell)
{
result.AddWithNoUpdate(MultipleItemsWalking(item, arguments));
}
result.UpdateInfo();
return(result);
}
}
public override AType Execute(AType right, AType left, Aplus environment = null)
{
CalculationArguments arguments = PrepareVariables(left, right);
return(MultipleItemsWalking(right, arguments));
}
/// <summary>
/// Classify element to the corresponding group.
/// </summary>
/// <param name="element"></param>
/// <param name="arguments">Instead of class variables.</param>
/// <returns></returns>
private AType Classify(AType element, CalculationArguments arguments)
{
int index;
if (arguments.Interval.IsArray)
{
AType intervalArray = arguments.Interval;
for (index = 0; index < intervalArray.Length; index++)
{
if (element.CompareTo(intervalArray[index]) <= 0)
{
break;
}
}
}
else
{
index = (element.CompareTo(arguments.Interval) <= 0) ? 0 : arguments.Interval.Length;
}
return AInteger.Create(index);
}
/// <summary>
/// If the cell rank > 1, first step is check the shape of actual cell equal to determined cell shape.
/// If different we throw Length error. The second step is classify the cell.
/// </summary>
/// <param name="cell"></param>
/// <param name="arguments">Instead of class variables.</param>
/// <returns></returns>
private AType MultipleItemsWalking(AType cell, CalculationArguments arguments)
{
if (arguments.CellShape.Count == cell.Shape.Count)
{
if (!arguments.CellShape.SequenceEqual(cell.Shape))
{
throw new Error.Length(LengthErrorText);
}
return Classify(cell, arguments);
}
else
{
AType result = AArray.Create(ATypes.AInteger);
foreach (AType item in cell)
{
result.AddWithNoUpdate(MultipleItemsWalking(item, arguments));
}
result.UpdateInfo();
return result;
}
}
