simulate_scenarios: Simulate multiple scenarios

In emartchenko/mvsimstudy: Multivariate Correlated Data Simulation Tools

Description

Simulates scenarios with varying effect sizes or sample sizes; can accomodate differing numbers of samples per treatment and control groups

Usage

simulate_scenarios(list_effect_sizes = NULL, list_treatment_groups = NULL,
  list_control_groups = NULL, means = NULL, cor_matrix = NULL,
  num_reps = 1000, method = NULL, p_adjust = NULL)

Arguments

list_effect_sizes	A list of effect_size objects.
list_treatment_groups	A list of sample sizes for the group exhibiting effects.
list_control_groups	A list of sample sizes for the group that does not exhibit effects.
means	A vector which contains the mean for each individual feature.
cor_matrix	A matrix which specifies the correlation structure amongst all features.
num_reps	A number that describes how many replications of each scenario there will be.
method	choose between "ofaat", "mv_glm", "lasso"
p_adjust	Used for the internal implementation of p.adjust; takes the same arguments

Value

A dataframe which summarizes the results of each scenario that has been run

Notes

List arguments must be passed as lists, even if only 1 item in the list. If differing lists are not passed for list_treatment_groups and list_control_groups arguments then default equal treatment/control group sizes assigned. However, at least one size list must be passed. Method must be specified.

See Also

analyze for details on method parameter

p.adjust for p value adjustments

Examples

e <- effect_size(c(0.7, 0.2, 0.4))
list_e <- c(e)
group_sizes <- c(20,40,60,80,100)
simulate_scenarios(list_e, group_sizes, method="ofaat")

emartchenko/mvsimstudy documentation built on May 24, 2019, 5:04 a.m.