Semimetric: Semimetric

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

Description

Semimetric is a function that calculates a distance matrix for functional data. A Rcpp port of the semimetric functions of Ferraty and Vieu.

Usage

Semimetric(Data1, Data2, semimetric, semimetric.params)

Arguments

Data1	Matrix with functional data (curves) each row one; n rows
Data2	Matrix with a second set of functional data (curves) each row one; m rows
semimetric	A string of choosing the semimetric; allowed are: "Deriv", "PCA", and "PLS"
semimetric.params	Parameters for the semimetric function.

Value

Semimetric returns a list with: semimetric as the n x m - distance-matrix ... reusable calculations, e.g. PCA: the q eigen vectors

Author(s)

Simon Mueller [e-mail]

References

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

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

See Also

SemimetricDeriv, SemimetricPCA, SemimetricPLS

Examples

# functional data sets
require (fds)

# fat spectrum dataset
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))

# calculate distance matrix
Dist <- Semimetric (X, 
                    X, 
                    semimetric = "SemimetricDeriv", 
                    semimetric.params)

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