Calibrate a one-stage single-arm ROPE design for a binomial endpoint

Usage

design_singlearm_onestage_rope(
  n_min,
  n_max,
  p0,
  delta,
  gamma_eq,
  gamma_diff = gamma_eq,
  direction = c("equivalence", "noninferiority", "superiority"),
  a = 1,
  b = 1,
  da0,
  db0,
  da1,
  db1,
  calibration = c("Bayesian", "frequentist", "hybrid", "full"),
  dp = NULL,
  target_power = NULL,
  target_type1 = NULL,
  target_pce_h0 = NULL,
  target_freq_power = NULL,
  target_freq_type1 = NULL,
  sustain_n = 1,
  return_grid = TRUE
)

Arguments

n_min: Minimum sample size.
n_max: Maximum sample size.
p0: Benchmark response probability.
delta: ROPE half-width (equivalence), NI margin, or superiority margin.
gamma_eq: Posterior probability threshold for accepting H1.
gamma_diff: Posterior probability threshold for compelling evidence for H0. Defaults to gamma_eq.
direction: Decision type: "equivalence", "noninferiority", or "superiority".
a, b: Analysis prior parameters for Beta(a,b).
da0, db0: Design prior parameters under H0.
da1, db1: Design prior parameters under H1.
calibration: Calibration mode: "Bayesian", "frequentist", "hybrid", or "full".
dp: Point alternative in the favorable H1 region at which frequentist power is computed.
target_power: Target Bayesian predictive power under H1.
target_type1: Target Bayesian predictive type-I error under H0.
target_pce_h0: Optional target for predictive compelling evidence for H0 under H0.
target_freq_power: Target frequentist power at dp.
target_freq_type1: Target worst-case frequentist type-I error at the null boundary.
sustain_n: Number of consecutive feasible sample sizes required.
return_grid: Return the full evaluation grid.

Value

An object of class bfbin2arm_rope_design.