Skip to main content

Crate rbf3

Crate rbf3 

Source
Expand description

A library for multidimensional interpolation.

Re-exports§

pub use nalgebra as na;

Structs§

Gaussian
HMapScatter
Scatter2
ThinPlateSpline
Gaussian basis function: φ(r) = exp(−(ε*r).pow(2)) where ε = 1/R0 R0 = R / 1000 The main benefit of this function is that it is compact - it is small at distances about 3·R0 from the center. At distances equal to 6·R0 or larger this function can be considered zero. As result, we have to solve linear system with sparse matrix. However, significant drawback is that we have one parameter to tune - R0. Too small R0 will make model sharp and non-smooth, and too large one will result in system with ill-conditioned matrix. Gaussian(f32),

Traits§

BasisFunction

Type Aliases§

Normal
Points
Row2
Row3
Scatter
Vec1
Vec2
Vec3
XY