plot.sPLSDA.scatter

Plots the scatter of two components of a sPLS-DA, as generated by perform.sPLSDA.

plot.sPLSDA.scatter(
  DEprot.sPLSDA.object,
  comp.x = 1,
  comp.y = 2,
  color.column = NULL,
  shape.column = NULL,
  label.column = NULL,
  ellipse = TRUE,
  ellipse.level = 0.95,
  plot.zero.line.x = TRUE,
  plot.zero.line.y = TRUE,
  point.size = 3
)

Arguments

DEprot.sPLSDA.object

An object of class DEprot.sPLSDA.

comp.x

Number indicating which component to display on the x-axis. Default: 1.

comp.y

Number indicating which component to display on the y-axis. Default: 2.

color.column

String indicating the name of the column in the metadata to use as factor for the dot colors. Default: NULL (the column used to fit the model).

shape.column

String indicating the name of the column in the metadata to use as factor for the dot shapes. Default: NULL (all dots).

label.column

String indicating the name of the column in the metadata to use as label of the dots. Default: NULL (no labeling).

ellipse

Logical value indicating whether a confidence ellipse should be drawn around each group of the color.column. Default: TRUE.

ellipse.level

Numeric value (0-1) indicating the confidence level of the ellipses. Default: 0.95.

plot.zero.line.x

Logical value to indicate whether to plot a gray dashed line in correspondence of x=0. Default: TRUE.

plot.zero.line.y

Logical value to indicate whether to plot a gray dashed line in correspondence of y=0. Default: TRUE.

point.size

Numeric value indicating the size of the dots. Default: 3.

Value

A ggplot object.

Details

The ellipses are a reading aid, not a test: a supervised model is fitted to separate the classes, so on the training samples the groups will look tidy even when the cross-validated error rate is poor. The error rates stored in the performance slot are what says whether the separation holds (plot.sPLSDA.performance).

See also

perform.sPLSDA, plot.sPLSDA.scatter.123, plot.sPLSDA.biplot

Author

Sebastian Gregoricchio

Examples

splsda <- perform.sPLSDA(DEprot.object = DEprot::test.toolbox$dpo.imp,
                         group.column = "condition",
                         keepX = 5,
                         validate = FALSE)
#> Warning: The number of 'folds' (5) is larger than the smallest class (4 samples): 'folds' has been set to 4.

plot.sPLSDA.scatter(DEprot.sPLSDA.object = splsda,
                    comp.x = 1,
                    comp.y = 2,
                    shape.column = "replicate")
#> Warning: The following aesthetics were dropped during statistical transformation: shape.
#> ℹ This can happen when ggplot fails to infer the correct grouping structure in
#>   the data.
#> ℹ Did you forget to specify a `group` aesthetic or to convert a numerical
#>   variable into a factor?