InterpolantSettings

ferreus_rbf.interpolant_config.InterpolantSettings(kernel_type, spheroidal_order=None, drift=None, nugget=None, base_range=None, total_sill=None, fitting_accuracy=None)

Holds the configuration parameters for an RBF kernel.

The only required input is the RBFKernelType.

If no additional options are provided, defaults are applied:

• Drift is set to the minimum valid choice for the kernel.
• For spheroidal kernels, base_range and total_sill default to 1.0.
• The nugget defaults to 0.0 but may be specified for any kernel.
• fitting_accuracy: FittingAccuracy(tolerance=1E-6, tolerance_type=FittingAccuracyType.Relative)

Parameters:

Name	Type	Description	Default
kernel_type	RBFKernelType	The RBF kernel to use for interpolation.	required
spheroidal_order	Optional[SpheroidalOrder]	The order (alpha) of spheroidal kernel. If None, SpheroidalOrder.Three will be used.	None
drift	Drift	The polynomial drift term added to the RBF system. If None, the implementation will infer a default based on kernel_type.	None
nugget	float	Optional smoothing parameter. A value of 0.0 (default) enforces an exact fit to all input data. Larger values soften the fit, which can reduce sensitivity to noisy data.	None
base_range	float	Controls how quickly the interpolant decays with distance from each point. Smaller values restrict influence to a local neighborhood, while larger values produce smoother, broader effects. Typically chosen based on the spacing of your data. Only used in spheroidal kernels.	None
total_sill	float	Sets the overall strength of influence each point exerts. Higher values give points more weight and stronger local effects. Lower values yield smoother, less pronounced variation. Works in combination with base_range and the spheroidal_order. Only used in spheroidal kernels.	None
fitting_accuracy	FittingAccuracy	Desired fitting accuracy and tolerance criteria.	None