Run null simulations to estimate Type I error rates

Run null simulations to estimate Type I error rates

Usage

run_null_sim(
  n_iter = 200,
  n_samples = 500,
  n_features = 20,
  conf_str = 0.8,
  w_signal = 0.7,
  feat_cor = 0,
  base_seed = 300,
  n_cores = 1,
  alpha = 0.05
)

Arguments

n_iter

Number of replicates. Default 200.

n_samples

Observations per replicate. Default 500.

n_features

Features per replicate. Default 20.

conf_str

Confounding strength delta. Default 0.80.

w_signal

Proxy quality omega. Default 0.70.

feat_cor

Within-module feature correlation. Default 0.

base_seed

Seed offset. Default 300.

n_cores

Parallel workers. Default 1.

alpha

Significance threshold. Default 0.05.

Value

A list with rates (data frame) and raw (full results).

Examples

null <- run_null_sim(n_iter = 2, n_samples = 100, n_features = 5)
null$rates
#>        method type1_error     flag
#> UNADJ   UNADJ         1.0 INFLATED
#> DIRECT DIRECT         1.0 INFLATED
#> COCA     COCA         0.2 INFLATED
#> IV2SLS IV2SLS         0.1       OK
#> PGC       PGC         0.9 INFLATED