NonlinearCurve.FillNumericalPartialDerivatives Method

public void FillNumericalPartialDerivatives(
	double x,
	DenseVector<double> derivatives
)

Public Sub FillNumericalPartialDerivatives ( 
	x As Double,
	derivatives As DenseVector(Of Double)
)

public:
void FillNumericalPartialDerivatives(
	double x, 
	DenseVector<double>^ derivatives
)

member FillNumericalPartialDerivatives : 
        x : float * 
        derivatives : DenseVector<float> -> unit 

Parameters

x  Double

The value at which the partial derivatives should be evaluated.

derivatives  DenseVector<Double>

A dense vector that is to hold the partial derivatives. The length of this vector must be at least the number of parameters of the curve.

Use this method to validate partial derivatives calculated directly using an overridden implementation of FillPartialDerivatives(Double, DenseVector<Double>). In most cases, the numerical derivatives should be within approximately SqrtEpsilon of the actual values. If the difference is significantly greater, then most likely the direct calculation of the partial derivatives contains an error.

Reference