adoptr (version 0.2.2)

Scores: Scores

Description

In adoptr scores are used to assess the performance of a design. This can be done either conditionally on the observed stage-one outcome or unconditionally. Consequently, score objects are either of class ConditionalScore or UnconditionalScore.

Usage

# S4 method for Score
show(object)
expected(s, data_distribution, prior, ...)
# S4 method for ConditionalScore
expected(s, data_distribution, prior, ...)
evaluate(s, design, ...)
# S4 method for IntegralScore,TwoStageDesign
evaluate(s, design,
  optimization = FALSE, subdivisions = 10000L, ...)

Arguments

object

object to show

s

Score object

data_distribution

DataDistribution object

prior

a Prior object

...

further optional arguments

design

object

optimization

logical, if TRUE uses a relaxation to real parameters of the underlying design; used for smooth optimization.

subdivisions

maximal number of subdivisions when evaluating an integral score using adaptive quadrature (optimization = FALSE)

Details

All scores can be evaluated on a design using the evaluate method. Note that evaluate requires a third argument x1 for conditional scores (observed stage-one outcome). Any ConditionalScore can be converted to a UnconditionalScore by forming its expected value using expected. The returned unconditional score is of class IntegralScore.

See Also

ConditionalPower, ConditionalSampleSize, composite

Examples

# NOT RUN {
design <- TwoStageDesign(
  n1    = 25,
  c1f   = 0,
  c1e   = 2.5,
  n2    = 50,
  c2    = 1.96,
  order = 7L
)
prior <- PointMassPrior(.3, 1)

# conditional
cp <- ConditionalPower(Normal(), prior)
expected(cp, Normal(), prior)
evaluate(cp, design, x1 = .5)

# unconditional
power <- Power(Normal(), prior)
evaluate(power, design)
evaluate(power, design, optimization = TRUE) # use non-adaptive quadrature

# }