funcy (version 1.0.1)

fpcCtrl-class: Class "fpcCtrl"

Description

Hyperparameters for functional principal component analysis. See fpca.

Objects from the class

Objects can be created by calls of the form new("fpcCtrl"). In addition, named lists can be coerced to fpcCtrl objects, names are completed if unique.

Slots

Objects of class fpcCtrl have the following slots:

select:

Character string, one of "automatic" or "manual" specifying whether the bandwidth for smoothing is given or should be calculated based on the data.

h1Dim:

If select="manual", bandwidth for smoothing the mean.

h2Dim:

If select="manual", bandwidth for smoothing the covariance.

sm1Dim:

Character string, name of the mean smoothing function. One of "sm.regression" or "sm1".

sm2Dim:

Character string, name of the covariance smoothing function. One of "sm.regression" or "sm2".

coeffsCalc:

Character string, specifying how to calculate the coefficients. One of "estimate" or "integrate".

nrMaxTime:

Maximum number of evaluation time points for the covariance matrix.

average:

If TRUE, matrix calculation is used and speeds up the calculation if curves have many evaluation time points in common.

Details

"sm.regression" is a nonparametric regression estimate from the R-package sm. "sm1" and "sm2" are kernel smoothers defined by Chiou2007. The coefficients for the basis functions can be computed by numerical integration (coeffCalc="integrate") or by a sparse estimation defined by Mueller2005 (coeffCalc="estimate").

References

Chiou J-M and Li P-L. "Functional clustering and identifying substructures of longitudinal data". Journal of the Royal Statistical Society: Series B. 69 (4). pp. 679--699. 2007

F. Yao and H-G. Mueller and J-L. Wang. Functional Data Analysis for Sparse Longitudinal Data. Journal of the American Statistical Association. 100 (470). 577--590. 2005

Examples

# NOT RUN {
##Have a look at the defaults
new("fpcCtrl")

##Coerce a list
mycont = list(select="automatic", sm1Dim="sm1", sm2Dim="sm2", nrMaxTime=20)
as(mycont, "fpcCtrl")

##Default values for coefficients calculation procedure
new("fpcCtrl")@coeffsCalc 
# }