(Conditional) Power of a Design — ConditionalPower-class • adoptr

This score evaluates P[X₂ > c2(design, X₁) | X₁ = x₁]. Note that the distribution of X₂ is the posterior predictive after observing X₁ = x₁.

ConditionalPower(dist, prior, label = "Pr[x2>=c2(x1)|x1]")

Power(dist, prior, label = "Pr[x2>=c2(x1)]")

# S4 method for ConditionalPower,TwoStageDesign
evaluate(s, design, x1, optimization = FALSE, ...)

Arguments

dist: a univariate distribution object
prior: a Prior object
label: object label (string)
s: Score object
design: object
x1: stage-one test statistic
optimization: logical, if TRUE uses a relaxation to real parameters of the underlying design; used for smooth optimization.
...: further optional arguments

See also

Examples

prior <- PointMassPrior(.4, 1)
cp <- ConditionalPower(Normal(), prior)
evaluate(
   cp,
   TwoStageDesign(50, .0, 2.0, 50, 2.0, order = 5L),
   x1 = 1
)
#> [1] 0.5
# these two are equivalent:
expected(cp, Normal(), prior)
#> E[Pr[x2>=c2(x1)|x1]]<Normal<two-armed>;PointMass<0.40>> 
Power(Normal(), prior)
#> Pr[x2>=c2(x1)]