Click or drag to resize

QuickNonlinearRegression Properties

The QuickNonlinearRegression type exposes the following members.

Properties
 NameDescription
Public propertyAccuracyOrderOfNumericalDerivatives Gets or sets the accuracy order of numerical derivatives (1..6). Default is 1.
Public propertyDoNotAllowUsingDerivativesOfFitFunction Gets or sets a value indicating whether or not it is allowed to use the derivatives of fit function.
Public propertyFunctionTolerance The stopping threshold for the function value or L2 norm of the residuals.
Public propertyGradientTolerance The stopping threshold for infinity norm of the relative gradient value. The relative gradient is the gradient divided by the parameter value.
Public propertyInitialMu Gets or sets the initial mu value (sometimes also named lambda).
Public propertyMaximumNumberOfIterations The maximum number of iterations.
Public propertyMinimalRSSImprovement Gets or sets the minimal RSS (Chi²) improvement [0, 1). If after 8 iterations the Chi² improvement is smaller than this value, the evaluation is stopped.
Public propertyStepTolerance The stopping threshold for L2 norm of the change of the parameters.
Public propertyUseMethodOfTransformedParameters Gets or sets a value indicating which nonlinear optimization method with constrained parameters is used: if false, the method with clamped steps is used; if true, the method with transformed parameters is used. If no parameter constraints are given, both methods should perform equally well.
Top
See Also