multiblock (version 0.8.2)

mbpls: Multiblock Partial Least Squares - MB-PLS

Description

A function computing MB-PLS scores, loadings, etc. on the super-level and block-level.

Usage

mbpls(
  formula,
  data,
  subset,
  na.action,
  X = NULL,
  Y = NULL,
  ncomp = 1,
  scale = FALSE,
  ...
)

Value

multiblock, mvr object with super-scores, super-loadings, block-scores and block-loading, and the underlying mvr (PLS) object for the super model, with all its result and plot possibilities. Relevant plotting functions: multiblock_plots

and result functions: multiblock_results.

Arguments

formula

Model formula accepting a single response (block) and predictor block names separated by + signs.

data

The data set to analyse.

subset

Expression for subsetting the data before modelling.

na.action

How to handle NAs (no action implemented).

X

list of input blocks. If X is supplied, the formula interface is skipped.

Y

matrix of responses.

ncomp

integer number of PLS components.

scale

logical for autoscaling inputs (default = FALSE).

...

additional arguments to pls::plsr.

Details

MB-PLS is the prototypical component based supervised multiblock method. It was originally formulated as a two-level method with a block-level and a super-level, but it was later discovered that it could be expressed as an ordinary PLS on concatenated weighted X blocks followed by a simple loop for calculating block-level loading weights, loadings and scores. This implementation uses the plsr function on the scaled input blocks (1/sqrt(ncol)) enabling all summaries and plots from the pls package.

References

• Wangen, L.E. and Kowalski, B.R. (1988). A multiblock partial least squares algorithm for investigating complex chemical systems. Journal of Chemometrics, 3, 3–20.

• Westerhuis, J.A., Kourti, T., and MacGregor,J.F. (1998). Analysis of multiblock and hierarchical PCA and PLS models. Journal of Chemometrics, 12, 301–321.

See Also

Overviews of available methods, multiblock, and methods organised by main structure: basic, unsupervised, asca, supervised and complex.

Examples

data(potato)
# Formula interface
mb <- mbpls(Sensory ~ Chemical+Compression, data=potato, ncomp = 5)
# ... or X and Y
mb.XY <- mbpls(X=potato[c('Chemical','Compression')], Y=potato[['Sensory']], ncomp = 5)
identical(mb$scores, mb.XY$scores)
print(mb)
scoreplot(mb, labels="names") # Exploiting mvr object structure from pls package