simulations_data {mlpwr} | R Documentation |
Display info on the generated data
Description
Output a data frame that, for each set of design parameters, includes the cost, the estimated power and SE, the surrogate model estimated power and SE, and the number of performed evaluations.
Usage
simulations_data(ds)
Arguments
ds |
Object of class designresult as created by the find.design function |
Details
The raw power is estimated using the ratio of significant hypothesis tests among the performed evaluations. The raw SE is estimated using p(1-p)/n
with p
being the raw power and n
being the number of performed evaluations.
For the estimation of SE using a surrogate model, GPR is employed because it is capable of variance estimation, unlike the other surrogate models.
Value
A data frame
Examples
# Load a simulation function
simfun <- example.simfun('ttest')
# Perform the search
ds <- find.design(simfun = simfun, boundaries = c(100,300), power = .95)
# Display more info on the collected data
simulations_data(ds)
[Package mlpwr version 1.1.0 Index]