sopls_plots: Scores, loadings and plots for sopls objects

In khliland/multiblock: Multiblock Data Fusion in Statistics and Machine Learning

Scores, loadings and plots for sopls objects

Description

Extraction of scores and loadings and adaptation of scoreplot, loadingplot and biplot from package pls for sopls objects.

Usage

## S3 method for class 'sopls'
loadings(object, ncomp = "all", block = 1, y = FALSE, ...)

## S3 method for class 'sopls'
scores(object, ncomp = "all", block = 1, y = FALSE, ...)

## S3 method for class 'sopls'
scoreplot(
  object,
  comps = 1:2,
  ncomp = NULL,
  block = 1,
  labels,
  identify = FALSE,
  type = "p",
  xlab,
  ylab,
  ...
)

## S3 method for class 'sopls'
loadingplot(
  object,
  comps = 1:2,
  ncomp = NULL,
  block = 1,
  scatter = TRUE,
  labels,
  identify = FALSE,
  type,
  lty,
  lwd = NULL,
  pch,
  cex = NULL,
  col,
  legendpos,
  xlab,
  ylab,
  pretty.xlabels = TRUE,
  xlim,
  ...
)

## S3 method for class 'sopls'
corrplot(
  object,
  comps = 1:2,
  ncomp = NULL,
  block = 1,
  labels = TRUE,
  col = 1:5,
  plotx = TRUE,
  ploty = TRUE,
  ...
)

## S3 method for class 'sopls'
biplot(
  x,
  comps = 1:2,
  ncomp = "all",
  block = 1,
  which = c("x", "y", "scores", "loadings"),
  var.axes = FALSE,
  xlabs,
  ylabs,
  main,
  ...
)

Arguments

object	sopls object
ncomp	integer vector giving components from all blocks before block (see next argument).
block	integer indicating which block to extract components from.
y	logical extract Y loadings/scores instead of X loadings/scores (default = FALSE).
...	further arguments sent to the underlying plot function(s)
comps	integer vector giving components, within block, to plot (see Details regarding combination of blocks).
labels	character indicating if "names" or "numbers" should be plot symbols (optional).
identify	logical for activating identify to interactively identify points.
type	character for selecting type of plot to make. Defaults to "p" (points) for scatter plots and "l" (lines) for line plots.
xlab	character text for x labels.
ylab	character text for y labels.
scatter	logical indicating if a scatterplot of loadings should be made (default = TRUE).
lty	Vector of line type specifications (see par for details).
lwd	numeric vector of line width specifications.
pch	Vector of point specifications (see points for details).
cex	numeric vector of plot size expansions (see par for details).
col	integer vector of symbol/line colours (see par for details).
legendpos	character indicating legend position (if scatter is FALSE), e.g. legendpos = "topright".
pretty.xlabels	logical indicating if xlabels should be more nicely plotted (default = TRUE).
xlim	numeric vector of length two, with the x limits of the plot (optional).
plotx	locical or integer/character. Whether to plot the X correlation loadings, optionally which block(s). Defaults to TRUE.
ploty	logical. Whether to plot the Y correlation loadings. Defaults to TRUE.
x	sopls object
which	character for selecting type of biplot ("x" = default, "y", "scores", "loadings").
var.axes	logical indicating if second axes of a biplot should have arrows.
xlabs	character vector for labelling first set of biplot points (optional).
ylabs	character vector for labelling second set of biplot points (optional).
main	character for setting the main title of a plot.

Details

If comps is supplied as a list for scoreplot, it is assumed that its elements refer to each of the blocks up to block number block. For instance comps = list(1, 0, 1:2) will select 1 component from the first block, no components from the second block and the first two components from the last block. This must be matched by ncomp, specifying how many components were selected before block number block.

See Also

Overviews of available methods, multiblock, and methods organised by main structure: basic, unsupervised, asca, supervised and complex. Common functions for computation and extraction of results are found in sopls_results.

#' @return The score and loading functions return scores and loadings, while plot functions have no return (except use of 'identify').

Examples

data(potato)
so <- sopls(Sensory ~ Chemical + Compression + NIRraw, data=potato, ncomp=c(5,5,5))

# Loadings
loadings(so, ncomp=c(3), block=2)[, 1:3]

# Scores
scores(so, block=1)[, 1:4]

# Default plot from first block
scoreplot(so)

# Second block with names
scoreplot(so, ncomp=c(3), block=2, labels="names")

# Scatterplot matrix
scoreplot(so, ncomp=c(3,2), block=3, comps=1:3)

# Combination of blocks (see Details)
scoreplot(so, ncomp=c(3,2), block=3, comps=list(1,0,1))

# Default plot from first block
loadingplot(so, scatter=TRUE)

# Second block with names
loadingplot(so, ncomp=c(3), block=2, labels="names", scatter=TRUE)

# Scatterplot matrix
loadingplot(so, ncomp=c(3,2), block=3, comps=1:3, scatter=TRUE)

# Correlation loadings
corrplot(so, block=2, ncomp=1)

# Default plot from first block
biplot(so)

khliland/multiblock documentation built on April 17, 2025, 5:09 p.m.