Transform a diagnostic using predictable weights, either through the linear
e-value localization 1 + w_t (E_t - 1) or through power tempering.
Usage
localize_predictive_diagnostic(
diagnostic,
weights,
localization_name,
mode = c("linear", "power"),
alpha = diagnostic$alpha,
metadata = list()
)
Arguments
- diagnostic
A predictive_e_diagnostic object.
- weights
Numeric weights in [0, 1], either scalar or one per
observation.
- localization_name
Label for the localized process.
- mode
Either "linear" or "power".
- alpha
Monitoring level.
- metadata
Optional metadata list.
Value
A localized predictive_e_diagnostic object.
