varySmooth: Vary the Level of Smoothing used Prior to GSLCCA
In gslcca: Generalized Semi-Linear Canonical Correlation Analysis

Description Usage Arguments Details Value Author(s) See Also Examples

Re-run GSLCCA with different amounts of smoothing at the subject level prior to analysis.

1	varySmooth(x, subject.smooth = 1:10, ...)

`x`	a `"gslcca"` object as returned by `gslcca`.
`subject.smooth`	a vector of integers specifying the number of SVD roots use when smoothing the data
`...`	currently ignored

This function can be used to investigate the effect of smoothing on GSLCCA.

When the subject.smooth argument of gslcca is an integer, the data matrix for each subject is approximated using the corresponding number of SVD roots.

An object of class "varySmooth" which is a list of "gslcca" objects obtained by re-running the original GSLCCA given by x with each value of subject.smooth

Heather Turner

gslcca, plot.varySmooth

data(clonidine)

### Smoothed data - automatically select number of roots
result <- gslcca(spectra, "Critical Exponential",
    time = Time, treatment = Treatment, subject.smooth = TRUE,
    data = clonidine, subset = Rat == "42")

### Vary number of roots 
multiRoots <- varySmooth(result, 2:15)

## plot optimised value
plot(multiRoots, "opt")

## plot fitted values
plot(multiRoots, "fitted")

## plot signature
plot(multiRoots, "signature")