Diagnostics Plot for Compositional Diagnostics — plot.diagnostics • multilevelcoda

Make a scatterplot matrix of diagnostics results from a complr object, with density and rug plots on the diagonal.

Usage

# S3 method for class 'diagnostics'
plot(x, transform = c("clr", "raw"), ...)

Arguments

x: A diagnostics class object created by diagnostics.complr.
transform: Character string indicating whether to plot the raw composition values in x$x or centered logratio transformed values. Defaults to "clr".
...: Currently unused.

Value

A ggplot graph object showing pairwise scatterplots for each composition part, with point colour indicating extreme value status. Diagonal panels show marginal densities and rug marks for observations flagged as extreme values.

Examples

data(mcompd)
data(sbp)

ids <- unique(mcompd$ID)[1:20]
cilr <- complr(
  data = mcompd[ID %in% ids, .SD[1:3], by = ID],
  sbp = sbp,
  parts = c("TST", "WAKE", "MVPA", "LPA", "SB"),
  idvar = "ID",
  total = 1440
)

between_diag <- diagnostics(cilr, level = "between", ev.perc = .05)

# Plot centered logratio transformed composition values
plot(between_diag)


# Plot on the raw composition scale
plot(between_diag, transform = "raw")