RadialBasis: kernel values (linear, cubic, quintic, thin-plate,
polyharmonic, multiquadric family, Gaussian, Wendland).
RbfInterpolate: interpolation or smoothing in any dimension with
polynomial augmentation and optional anisotropic transform.
RbfLoocv: Rippa leave-one-out errors per shape parameter.
RbfMultiscale: multiscale Wendland residual fitting.
RbfGrid: grid evaluation. Identical to the Python arm
morie.fn.rbfsurf.
Usage
RadialBasis(r, kernel = "thin_plate", epsilon = 1, order = 3, support = 1)
RbfInterpolate(
X,
y,
Xnew,
kernel = "thin_plate",
epsilon = 1,
degree = 1,
smoothing = 0,
order = 3,
support = 1,
transform = NULL
)
RbfLoocv(X, y, epsilons, kernel = "gaussian", degree = -1, smoothing = 0)
RbfMultiscale(X, y, Xnew, supports, degree = -1)
RbfGrid(X, y, xs, ys, ...)Arguments
- r
Distances.
- kernel
Kernel name.
- epsilon
Shape parameter.
- order
Polyharmonic order.
- support
Wendland support radius.
- X, Xnew
Data and prediction coordinates (matrices).
- y
Data values.
- degree
Polynomial degree (-1 for none).
- smoothing
Smoothing parameter.
- transform
Coordinate transform matrix.
- epsilons
Candidate shape parameters.
- supports
Decreasing Wendland supports.
- xs, ys
Grid axes.
- ...
Options passed to
RbfInterpolate.
References
Fasshauer, G. E. (2007). Meshfree Approximation Methods with MATLAB. World Scientific.
Rippa, S. (1999). An algorithm for selecting a good value for the parameter c in radial basis function interpolation. Advances in Computational Mathematics 11, 193-210.
Floater, M. S. and Iske, A. (1996). Multistep scattered data interpolation using compactly supported radial basis functions. Journal of Computational and Applied Mathematics 73, 65-78.