z_A4._Define_bivariate_Spline_: Bivariate spline
In scpm: An R Package for Spatial Smoothing

Description Usage Arguments Details Author(s)

Define a bivariate spline using tensor products or thin plate splines.

s2D(data = NULL, penalty = c("none",  "cs",  "ps",  "tps"),
        is.X = c("none",  "tensor",  "tps"), intercept = TRUE,
        ps.order = 2, aniso.angle = 0, aniso.ratio = 1,
        env = .GlobalEnv, ...)

`data`	sss object. Data of class sss generated by as.sss or create.sss.
`penalty`	character. Type of spline to use and penalty to define. One of cs or ps for cubic splines based on tensor products, tps for thin plate splines or none. cs define the penalty based on roughness matrices of natural cubic splines while ps define the penalty based on differences of order `ps.order`. See scp. If `penalty="none"` then no spline nor penalty are defined and the model for the spatial surface is defined by `is.X`.
`is.X`	character. Model for the spatial surface. One of tensor, tps or none. Only required if `penalty="none"`. See details.
`intercept`	logical. Define whether to include an intercept or not. Default to `TRUE`.
`ps.order`	integer. Order for differences if `penalty = "ps"`.
`aniso.angle`	numeric. Angle for geometric anisotropy.
`aniso.ratio`	numeric. Ratio between [0,1] for geometric anisotropy.
`env`	environment. Where to search for data if `data=NULL`.
`...`	additional arguments. Not required.

Note that is.X is only needed if penalty="none". By defining is.X="none" it only define an intercept (if intercept=TRUE), is.X="tps" defines an intercept, coordinate 1, and coordinate 2 as covariates, while is.X="tensor" defines also the interaction coordinate 1*coordinate 2. See scp.

Mario A. Martinez Araya, r@marioma.me

scpm documentation built on Feb. 17, 2020, 5:08 p.m.