Specify a sinusoidal finite-design logit truth

Usage

synthetic.truth.logit(
  amplitude = 1.5,
  target.prevalence = 0.5,
  clip = c(0.05, 0.95)
)

Arguments

amplitude: Sinusoidal log-odds amplitude.
target.prevalence: Target mean pre-clipping probability.
clip: Probability bounds.

Value

A synthetic truth component.

Examples

synthetic.truth.logit(amplitude = 1, target.prevalence = 0.4)
#> $kind
#> [1] "truth"
#> 
#> $family
#> [1] "named"
#> 
#> $version
#> [1] 1
#> 
#> $parameters
#> $parameters$scope
#> [1] "finite.design"
#> 
#> $parameters$estimand
#> [1] "E[Y_i|X_1:n] = clipped expit(alpha + amplitude*sin(pi*x_i1)), with alpha solving the design prevalence"
#> 
#> $parameters$evaluation.coordinates
#> [1] "predictors"
#> 
#> $parameters$parameters
#> $parameters$parameters$evaluator.id
#> [1] "g6.sin.logit.v1"
#> 
#> $parameters$parameters$evaluator.parameters
#> $parameters$parameters$evaluator.parameters$amplitude
#> [1] 1
#> 
#> $parameters$parameters$evaluator.parameters$target.prevalence
#> [1] 0.4
#> 
#> $parameters$parameters$evaluator.parameters$clip
#> [1] 0.05 0.95
#> 
#> 
#> 
#> 
#> attr(,"class")
#> [1] "synthetic_named_truth"     "synthetic_truth_component"
#> [3] "synthetic_truth"           "synthetic_component"