Power curve for the PP mean estimator

Computes Monte Carlo power values for a grid of labeled sample sizes using simulate_power() given pre-computed variance components.

Usage

power_curve_mean(
  n_grid,
  delta,
  N,
  var_f,
  var_res,
  alpha = 0.05,
  R = 10000,
  seed = NULL
)

Arguments

n_grid

Integer vector of candidate labeled sample sizes.

delta

Effect size \(\theta - \theta_0\).

N

Unlabeled sample size.

var_f

Variance of \(f(X)\).

var_res

Variance of residuals \(Y - f(X)\).

alpha

Two-sided significance level.

R

Number of Monte Carlo replicates passed to simulate_power().

seed

Optional RNG seed for reproducibility.

Value

Data frame with columns n, power_empirical, power_exact, delta, N, alpha, var_f, and var_res ready for plotting power curves.

Examples

power_curve_mean(
  n_grid = seq(50, 200, by = 25),
  delta = 0.2,
  N = 2000,
  var_f = 0.4,
  var_res = 1.0,
  R = 5000,
  seed = 123
)
#>     n power_empirical power_exact delta    N alpha var_f var_res
#> 1  50          0.2906   0.2905906   0.2 2000  0.05   0.4       1
#> 2  75          0.4050   0.4049879   0.2 2000  0.05   0.4       1
#> 3 100          0.5082   0.5081511   0.2 2000  0.05   0.4       1
#> 4 125          0.5982   0.5982060   0.2 2000  0.05   0.4       1
#> 5 150          0.6750   0.6749441   0.2 2000  0.05   0.4       1
#> 6 175          0.7392   0.7391332   0.2 2000  0.05   0.4       1
#> 7 200          0.7920   0.7920460   0.2 2000  0.05   0.4       1