best.wapls {paltran}R Documentation

model selection of best WA-PLS component

Description

Required in MW transfer function - model selection of best WA-PLS component: Not only the RMSEP is used: A combination of R2, min(error), max(error) and RMSEP is used for model election, as the RMSEP alone might be problematic. The component with the highest R2 (lowest min(error),...) gets 2 points, the second highest 1 point (column MW). That component with the highest number of points (MW) will bee chosen for the reconstruction. The RMSEP can bee excluded compleetly from the modelselection.

Usage

best.wapls(error, nro = 4,rmsep.incl=TRUE)

Arguments

error error statistic from a wa-pls.
nro number of components, must be the same as in wapls().
rmsep.incl should the RMSEP included in the model selection?

Value

Error statistic of a WA-PLS run ordered by XR2, XAveBias, XMaxBias , RMSEP. The component with the highest MW value is the best.

Author(s)

Sven Adler, sven.adler2@uni-rostock.de, University Rostock, Institute for Biosciences, General and Systematic Botany, Germany

References

Huebener, T., Dressler,M., Schwarz,A.,Langner, K., Adler,S. 2007. Dynamic adjustment of training sets (`moving windows` reconstruction) by using transfer functions in paleolimnology -a new approache,J. o. Paleolimnology. DOI 10.2007/s10933-007-9145-7.

See Also

mw, wapls, wa and package analogue

Examples

data(train_set.MV)
data(train_env.MV)
data(dud.df)
try1<-wapls(train_set.MV,train_env.MV,dud.df)
try1$performance
best.wapls(try1$performance)

[Package paltran version 1.0-0 Index]