Skip to contents

Build age-to-age (ata) factors from Triangle data

Source: R/ata.R
build_ata.Rd

Construct age-to-age development factors from an object of class Triangle, typically produced by build_triangle(). Age-to-age factors are calculated from cumulative values such as closs or crp.

Age-to-age is defined as: $$ata_{k \to k+1} = value_{k+1} / value_k$$

where value_k is the selected cumulative metric at development period k.

Usage

build_ata(
  x,
  value_var = "closs",
  weight_var = NULL,
  min_denom = 0,
  drop_invalid = FALSE
)

Arguments

x

An object of class Triangle.

value_var

A single cumulative metric used to compute age-to-age. Must be one of "closs", "crp", or "clr".

weight_var

An optional single cumulative metric used as weights in WLS estimation via summary.ATA(). Must be one of "closs", "crp", or "clr", and must differ from value_var. Typical use is weight_var = "crp" when value_var = "clr", since clr values are ratios and carry no exposure information. When NULL (default), value_from is used as the WLS weight, which corresponds to the standard volume-weighted chain ladder.

min_denom

Minimum denominator required to compute age-to-age. If value_from <= min_denom, ata is set to NA. This is useful when early development periods contain structural zeros (for example, waiting periods in life insurance).

drop_invalid

Logical; if TRUE, rows with invalid (non-finite) age-to-age factors are dropped. Useful for clean output when passing to summary.ATA() or find_ata_maturity(). When FALSE (default), all rows are retained, preserving the full triangle structure for diagnostic visualisation via plot_triangle.ATA().

Value

A data.frame with class "ATA" containing:

ata_from

Current development period.

ata_to

Next development period.

value_from

Current cumulative value.

value_to

Next cumulative value.

ata

Age-to-age factor (value_to / value_from), or NA when value_from <= min_denom.

weight

Weight values at ata_from derived from weight_var. Only present when weight_var is supplied.

The returned object carries the following attributes: group_var, cohort_var, dev_var, value_var, and weight_var.

See also

build_triangle(), summary.ATA(), fit_ata()

Examples

if (FALSE) { # \dontrun{
d <- build_triangle(df, group_var = cv_nm)

ata1 <- build_ata(d, "closs")
ata2 <- build_ata(d, "crp")
ata3 <- build_ata(d, "clr", weight_var = "crp")

head(ata1)
attr(ata1, "value_var")
attr(ata1, "weight_var")
} # }