Skip to contents

Formulating Constraints

Source: R/constraints.R
Constraints.Rd

Conceptually, constraints work very similar to scores (any score can be put in a constraint). Currently, constraints of the form 'score <=/>= x', 'x <=/>= score' and 'score <=/>= score' are admissible.

Usage

# S4 method for class 'Constraint,TwoStageDesign'
evaluate(s, design, optimization = FALSE, ...)

# S4 method for class 'ConditionalScore,numeric'
e1 <= e2

# S4 method for class 'ConditionalScore,numeric'
e1 >= e2

# S4 method for class 'numeric,ConditionalScore'
e1 <= e2

# S4 method for class 'numeric,ConditionalScore'
e1 >= e2

# S4 method for class 'ConditionalScore,ConditionalScore'
e1 <= e2

# S4 method for class 'ConditionalScore,ConditionalScore'
e1 >= e2

# S4 method for class 'UnconditionalScore,numeric'
e1 <= e2

# S4 method for class 'UnconditionalScore,numeric'
e1 >= e2

# S4 method for class 'numeric,UnconditionalScore'
e1 <= e2

# S4 method for class 'numeric,UnconditionalScore'
e1 >= e2

# S4 method for class 'UnconditionalScore,UnconditionalScore'
e1 <= e2

# S4 method for class 'UnconditionalScore,UnconditionalScore'
e1 >= e2

Arguments

s

Score object

design

object

optimization

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

...

further optional arguments

e1

left hand side (score or numeric)

e2

right hand side (score or numeric)

Value

an object of class Constraint

See also

minimize

Examples

design <- OneStageDesign(50, 1.96)

cp     <- ConditionalPower(Normal(), PointMassPrior(0.4, 1))
pow    <- Power(Normal(), PointMassPrior(0.4, 1))

# unconditional power constraint
constraint1 <- pow >= 0.8
evaluate(constraint1, design)
#> [1] 0.2840466

# conditional power constraint
constraint2 <- cp  >= 0.7
evaluate(constraint2, design, .5)
#> [1] 0.7
constraint3 <- 0.7 <= cp # same as constraint2
evaluate(constraint3, design, .5)
#> [1] 0.7