plot.gslcca: Plot Results from a GSLCCA Analysis
In gslcca: Generalized Semi-Linear Canonical Correlation Analysis

Description Usage Arguments Details Author(s) See Also Examples

Plot ‘signatures’ (loadings of the Y matrix) or (projected) observed and fitted values (Y and X scores) from an “gslcca” object.

    ## S3 method for class 'gslcca'
plot(x, type = "signature", series = x$treatment,
        mean = FALSE, overlay = length(agrep(type, "signature")), 
        ask = dev.interactive(), lattice = !length(agrep(type, "signature")), 
        main = NULL, xlab = NULL, ylab = NULL, col = NULL, lty = 1, lwd = 1.5, 
        pch = NULL, legend.x = "topright", space = "bottom", corner = NULL, 
        columns = 2, ...)

`x`	an object of class `"gslcca"`.
`type`	the type of plot: either `"signature"`, `"fitted"` or `"scores"`.
`series`	if `x$treatment` is `NULL` and there are multiple time series for each subject, a factor to differentiate these series.
`mean`	logical: if `TRUE` values are averaged over subjects.
`overlay`	for `type = \"signature\"` only, a logical value specifying whether to overlay signatures for different subjects on the same plot or create a separate plot for each subject.
`ask`	logical: if `TRUE`, the user is asked before each plot, see `par(ask=.)`. Ignored when only one plot is produced.
`lattice`	logical: if `TRUE` the values for each subject are displayed in separate panels of a trellis plot.
`main`	an overall title for the plot.
`xlab`	a title for the x axis.
`ylab`	a title for the y axis.
`col`	a vector specifying a colour palette to use for the lines/points. By default the palette is specified by `rainbow_hcl` if available, otherwise the standard plotting colours are used.
`lty`	a vector of line types, see `par`.
`lwd`	a vector of line widths, see `par`.
`pch`	a vector of plotting characters, see `points`.
`legend.x`	a character vector to use in the legend.
`space`	when `type = "fitted"` and `lattice = TRUE`, the space outside the plot region in which to put the legend, see `xyplot`. Overrides the `corner` argument.
`corner`	when `type = "fitted"` and `lattice = TRUE`, the corner inside the plot region in which to put the legend, see `xyplot`.
`columns`	when `type = "fitted"` and `lattice = TRUE`, how many columns to use to format the legend, see `xyplot`.
`...`	arguments passed on to the plotting functions (`plot` for single curves, `xyplot` for trellis graphics, else `matplot`)

For type = "signature", the loadings of the Y matrix from the GSLCCA analysis are displayed for each subject or their average is displayed, as specified by individual. When both individual and overlay are TRUE, the arguments col, lty and legend.x control the line style and legend text for each subject.

For type = "fitted", the X scores are displayed with lines. If there are multiple X scores for each time point, then a line is displayed for each times series specified by series (usually the treatment factor used in GSLCCA) and the arguments col, lty, pch and legend.x control the display for each series. By default, grey or black is used to display values corresponding to the reference treatment (specified by x$ref).

For type = "scores" the X scores are displayed as for type = "fitted" and points are added for the Y scores, again coloured by series.

Heather Turner

gslcca

data(clonidine)

### Separate critical exponential model for each treatment
result <- gslcca(spectra, "Critical Exponential",
    time = Time, treatment = Treatment, subject = Rat,
    subject.smooth = 4, data = clonidine)

## plot of individual signatures
plot(result, type = "signature", overlay = FALSE)

## plot of individual projected observed and fitted values
plot(result, type = "scores", lattice = FALSE)

## plot of individual projected fitted values
plot(result, type = "fitted", lattice = FALSE)

## plot of individual signatures - overlaid
plot(result, type = "signature")

## plot of individual signatures - lattice
plot(result, type = "signature", overlay = FALSE, lattice = TRUE)

## plot of individual projected observed and fitted values - lattice
plot(result, type = "scores")

## plot of individual projected fitted values - lattice
plot(result, type = "fitted")

## plot of mean signature
plot(result, type = "signature", mean = TRUE)

## plot of mean fitted
plot(result, type = "fitted", mean = TRUE)