Specify a finite-design occupation-mixture truth

The analytic Gaussian-mixture density is transformed to Bernoulli occupation probabilities by probability.maximum * density^gamma / max(density^gamma).

Usage

synthetic.truth.occupation.mixture(
  centers,
  covariances,
  weights,
  gamma = 1,
  probability.maximum = 0.65,
  normalization = "design.maximum"
)

Arguments

centers

Component centers, one per row.

covariances

Positive-definite covariance matrices.

weights

Nonnegative mixture weights.

gamma

Positive density-shape exponent.

probability.maximum

Maximum realized occupation probability.

normalization

Currently "design.maximum".

Value

A finite-design synthetic truth component.

Examples

synthetic.truth.occupation.mixture(
  centers = rbind(c(-1, 0), c(1, 0)),
  covariances = list(diag(2), diag(2)), weights = c(1, 1)
)
#> $kind
#> [1] "truth"
#> 
#> $family
#> [1] "occupation.mixture"
#> 
#> $version
#> [1] 1
#> 
#> $parameters
#> $parameters$scope
#> [1] "finite.design"
#> 
#> $parameters$estimand
#> [1] "Bernoulli occupation probability obtained by design-maximum normalization of a Gaussian-mixture density"
#> 
#> $parameters$evaluation.coordinates
#> [1] "predictors"
#> 
#> $parameters$parameters
#> $parameters$parameters$centers
#>      [,1] [,2]
#> [1,]   -1    0
#> [2,]    1    0
#> 
#> $parameters$parameters$scales
#> $parameters$parameters$scales[[1]]
#>      [,1] [,2]
#> [1,]    1    0
#> [2,]    0    1
#> 
#> $parameters$parameters$scales[[2]]
#>      [,1] [,2]
#> [1,]    1    0
#> [2,]    0    1
#> 
#> 
#> $parameters$parameters$weights
#> [1] 0.5 0.5
#> 
#> $parameters$parameters$normalize
#> [1] "none"
#> 
#> $parameters$parameters$gamma
#> [1] 1
#> 
#> $parameters$parameters$probability.maximum
#> [1] 0.65
#> 
#> $parameters$parameters$normalization
#> [1] "design.maximum"
#> 
#> 
#> 
#> attr(,"class")
#> [1] "synthetic_occupation_mixture_truth" "synthetic_truth_component"         
#> [3] "synthetic_truth"                    "synthetic_component"