Power at a grid of sample sizes

Evaluates a planning criterion's power at a fixed grid of candidate sample sizes, instead of searching for the smallest one that reaches a target (as design does). Useful for seeing how power changes with n, e.g. via plot() on the result. The methods available depend on the class of params. See the method pages listed under Details.

Usage

power_curve(params, method = c("DPIR", "BFDA", "BSDA"), ...)

Arguments

params

A bgm_elicited object, as returned by elicit_prior.

method

Which planning method to evaluate, "DPIR" (default), "BFDA", or "BSDA". One method runs per call.

...

Method-specific arguments passed to the method (e.g., n, the candidate sample sizes to evaluate).

Value

A bgm_power_curve object. print and plot methods are available.

Details

Methods are available for the following classes:

ggm_elicited

Gaussian graphical models

See also

Other sample size planning: bsda_control(), design(), design.ggm_elicited(), power_curve.ggm_elicited(), validate(), validate.ggm_design()