Plot LDA results

Visualize LDA results with discrimination plots, loading plots, or histograms (for binary classification).

Usage

# S3 method for class 'stat_lda'
plot(
  x,
  type = c("scores", "loadings"),
  lds = c(1, 2),
  color = NULL,
  labels = NULL,
  chull = TRUE,
  cex = 0.7,
  legend = TRUE,
  ...
)

Arguments

x

A stat_lda object

type

Character. Type of plot:

• "scores": LD score plot (default) - scatter for 2+ LDs, boxplot for 1 LD

• "loadings": Loading plot for first two LDs

lds

Integer vector of length 2. Which LDs to plot for score/loading plots. Default is c(1, 2). For binary classification (1 LD), only first element used.

color

Column name (bare or quoted) for coloring points. Default uses the response variable.

labels

Column name (bare or quoted) for text labels in score plot. For loadings, logical: should variable names be shown? Default TRUE.

chull

Logical. Draw convex hulls around groups? Only works when color is a factor. Default TRUE.

cex

Numeric. Character expansion for labels. Default is 0.7.

legend

Logical. Show legend for colors? Default TRUE.

...

Additional arguments (reserved for future use)

Value

NULL (invisibly). Draws plot as side effect.

Examples

if (FALSE) { # \dontrun{
lda <- boteft %>% stat_lda(type)

# Score plot (default)
plot(lda)
plot(lda, color = type)
plot(lda, labels = pred)
plot(lda, color = type, chull = TRUE)

# Different LDs
plot(lda, lds = c(2, 3))

# Loading plot
plot(lda, type = "loadings")
plot(lda, type = "loadings", labels = FALSE)
} # }