Plot a loss ratio fit

Visualise an object of class "RatioFit".

The plotted metric is the cross-product of metric and cell_type:

• metric = "ratio", cell_type = "cumulative": cumulative loss ratio (default).

• metric = "incr_ratio" (i.e., cell_type = "incremental"): per-period loss ratio.

• metric = "loss" / "premium": same split – cumulative or per-period amounts.

Confidence bands are drawn only for cumulative metrics (cell_type = "cumulative"), since the fit output does not carry SE columns for incremental projections.

Usage

# S3 method for class 'RatioFit'
plot(
  x,
  metric = c("ratio", "loss", "premium"),
  cell_type = c("cumulative", "incremental"),
  per_group = NULL,
  ask = grDevices::dev.interactive(),
  conf_level = 0.95,
  show_interval = TRUE,
  amount_divisor = "auto",
  scales = c("fixed", "free_y", "free_x", "free"),
  theme = c("view", "save", "shiny"),
  nrow = NULL,
  ncol = NULL,
  ...
)

Arguments

x

An object of class "RatioFit".

metric

Metric to plot. One of "ratio" (default), "loss", "premium".

cell_type

Aggregation. One of "cumulative" (default) or "incremental".

per_group

Logical or NULL. When TRUE (auto for multi-group fits), produce one ggplot per group and print them sequentially with devAskNewPage() – mirrors base R's plot.lm() pattern of stepping through related diagnostic plots. Returns the list of plots invisibly. When FALSE (auto for single-group fits), facets every (group, cohort) combination in a single ggplot.

ask

Passed to devAskNewPage() when per_group = TRUE. Defaults to dev.interactive().

conf_level

Confidence level. Default is 0.95.

show_interval

Logical. Default is TRUE.

amount_divisor

Numeric. Default is 1e8.

scales

Facet scale argument.

theme

Theme string.

nrow, ncol

Facet dimensions.

...

Additional arguments.

Value

A ggplot object.