Plot a cohort regime detection result

Visualise an object of class "Regime" as a PCA scatter of cohort trajectories coloured by detected regime. Points are underwriting cohorts, axes are the first two principal components of the cohort feature matrix (development-period trajectories), and ellipses indicate the 90% contour per regime. Arrows show the loadings of the original development-period features on PC1/PC2.

For a multi-group Regime, plots are faceted by group: each group's PCA is rendered in its own panel using its own feature matrix and loadings (PCA cannot be meaningfully shared across groups with different window-period bases or scale, so per-group PCA is the correct representation).

Usage

# S3 method for class 'Regime'
plot(
  x,
  show_arrow = TRUE,
  show_label = TRUE,
  show_ellipse = TRUE,
  show_mean = TRUE,
  show_median = TRUE,
  alpha = 0.5,
  palette = "Set1",
  theme = c("view", "save", "shiny"),
  ...
)

Arguments

x

An object of class "Regime".

show_arrow

Logical; draw loading arrows. Default TRUE.

show_label

Logical; label arrows with development-period index. Default TRUE.

show_ellipse

Logical; draw 90% ellipse per regime. Default TRUE.

show_mean, show_median

Logical; draw per-regime mean / median point. Defaults TRUE.

alpha

Numeric; point alpha. Default 0.5.

palette

Brewer palette name for discrete regimes. Default "Set1".

theme

Theme string passed to .switch_theme().

...

Additional arguments passed to .switch_theme().

Value

A ggplot object (single-group), or a named list of ggplot objects (multi-group; one entry per group).

See also

detect_regime()