AdaptivePrediction-package: Prediction rules based on an adaptive handling of dependence
In fhebert/AdaptivePrediction: Adaptive Prediction Rules

Description Author(s) See Also Examples

This package provides functions to construct prediction rules using an adaptive handling of dependence.

Florian Hébert, Mathieu Emily, David Causeur

Maintainer: Florian Hébert <florian.hebert@agrocampus-ouest.fr>

AdaptiveReg PredictAdaptiveReg PlotAdaptiveReg

  library(prospectr)
  data(NIRsoil)
  indNA = which(is.na(NIRsoil$Nt))
  NIRNt = list(x=NIRsoil$spc[-indNA,],y=NIRsoil$Nt[-indNA]) #remove NA
  NIRNt = list(x=NIRNt$x[-c(287,316,321,486,641,643),],
               y=NIRNt$y[-c(287,316,321,486,641,643)]) #remove outliers
  ind.test = sample(1:length(NIRNt$y),50)
  fit = AdaptiveReg(NIRNt$x[-ind.test,],NIRNt$y[-ind.test],nvmax=100)
  PlotAdaptiveReg(fit)
  pred = PredictAdaptiveReg(fit,NIRNt$x[ind.test,])
  plot(NIRNt$y[ind.test],pred,pch=16,xlab="Y",ylab=expression(hat(Y)))