nfda-package: Nonparametric Functional Data Analysis in R

In sipemu/Nonparametric-Functional-Data-Analysis: Nonparametric Functional Data Analysis

Description

C++-functions for different nonparametric kernel estimates. Originally written in R by Ferraty and Vieu (2006) Nonparametric Functional Data Analysis. Original code can be downloaded on their website http://www.math.univ-toulouse.fr/staph/npfda/

Details

Package:	nfda
Type:	Package
Version:	0.2-1
Date:	2011-12-15
License:	GPL-2
LazyLoad:	yes

Author(s)

Maintainer: Simon Mueller <[e-mail]>

References

http://www.math.univ-toulouse.fr/staph/npfda/

Ferraty, F. and Vieu, P. Nonparametric Functional Data Analysis. Springer 2006.

Rachdi, M. and Vieu, P. Nonparametric regression for functional data: automatic smoothing parameter selection. Journal of Statistical Planning and Inference 137, 9 (2007), 2784-2801.

Benhenni, K., Ferraty, F., Rachdi, M., and Vieu, P. Local smoothing regression with functional data. Computational Statistics 22, 3 (2007) 353-369.

Examples

##
#
# an example on nonparametric kernel regression
#
##
# functional data sets
library (fds)

# fat spectrum dataset
Y <- Fatvalues
X <- t(Fatspectrum$y)

# setup semimetric parameters
semimetric.params <- c()
semimetric.params$q <- 2
semimetric.params$nknot  <- 20
semimetric.params$range.grid  <- c (min (Fatspectrum$x), 
                                    max (Fatspectrum$x))

# learn and testsample
learn <- 1:160
test <- 161:215

# parameter estimation and prediction by cross-validation
Learn.Fat.CV <- FuNopaRe (X[learn, ], 
                          Y[learn], 
                          semimetric = "Deriv", 
                          semimetric.params, 
                          bandwidth = "CV")
                          
Predict.Fat.CV <- predict (Learn.Fat.CV, 
                           X[test, ], 
                           method.params = NULL)

plot (Predict.Fat.CV$Prediction, Y[161:215])

sipemu/Nonparametric-Functional-Data-Analysis documentation built on May 29, 2019, 10:10 p.m.