Construct a Maturity object from manually specified maturity points

User-facing helper for hand-specifying a maturity point (or a set of per-group maturity points) without running detect_maturity(). The returned "Maturity" object plugs into any function that consumes a Maturity result – fit_ratio(), fit_loss(), backtest(), and the maturity input dispatcher – by carrying the same row schema as detect_maturity() output (group columns plus ata_from, change, ata_link).

Use this when company-standard or domain-knowledge maturity points are known a priori and you want to override the data-driven detection. Stat columns (mean, cv, f, rse, ...) are set to NA_real_ because manual entry carries no estimates.

Argument syntax mirrors data.frame() / data.table(): named vectors of equal length, one of which must be change (the maturity point, an integer dev index). Any other named arguments are treated as group columns.

Usage

maturity_at(...)

Arguments

...

Named vectors of equal length. Must include change (coercible to integer; the maturity point, i.e. the to-index of the first mature ata link). Any other named arguments are interpreted as group column values (e.g. coverage, channel). With no group columns the result is a pooled (single-row) Maturity.

Value

A data.table with class "Maturity" carrying the same columns as detect_maturity() output: group columns (if any), ata_from = change - 1L, change, ata_link = "<from>-<to>", and the diagnostic stat columns (mean, median, wt, cv, f, f_se, rse, sigma, n_cohorts, n_valid, n_inf, n_nan, valid_ratio) set to NA_real_. attr(., "groups") holds the group column names (possibly character(0)).

See also

detect_maturity(), regime_at()

Examples

if (FALSE) { # \dontrun{
# Single-group manual override
maturity_at(coverage = "surgery", change = 4)

# Multi-group manual override (e.g. company-standard k*)
maturity_at(coverage = c("cancer", "ci", "inpatient", "surgery"),
            change   = c(   9,   10,     7,     4))

# Pooled (no group columns)
maturity_at(change = 5)
} # }