spasm.construct: Experimental sparse smoothers

In mgcv: Mixed GAM Computation Vehicle with Automatic Smoothness Estimation

Experimental sparse smoothers

Description

These are experimental sparse smoothing functions, and should be left well alone!

Usage

spasm.construct(object,data)
spasm.sp(object,sp,w=rep(1,object$nobs),get.trH=TRUE,block=0,centre=FALSE)
spasm.smooth(object,X,residual=FALSE,block=0)

Arguments

object	sparse smooth object
data	data frame
sp	smoothing parameter value
w	optional weights
get.trH	Should (estimated) trace of sparse smoother matrix be returned
block	index of block, 0 for all blocks
centre	should sparse smooth be centred?
X	what to smooth
residual	apply residual operation?

WARNING

It is not recommended to use these yet

Author(s)

Simon N. Wood [e-mail]

mgcv documentation built on May 29, 2024, 4:34 a.m.