API Reference

Stores and validates the values of a deviation for discrete probabilities p.

Arguments

• n::Int: Length of discrete probabilities
• p::Vector{Float64}: Vector of discrete probabilities
• d⁻::Vector{Float64}: Vector of lower bounds
• d⁺::Vector{Float64}: Vector of uppper bounds
• ϵ::Float64: Radius of uncertainty

The arguments should satisfy the following conditions:

n ≥ 1
all(p .≥ 0)
isapprox(sum(p), 1)
length(p) == length(d⁻) == length(d⁺) == n
all(d⁻ .≥ -p)
all(d⁺ .≤ 1.0.-p)
0 ≤ ϵ ≤ 1

Examples

Arguments

n = 2
p = [0.4, 0.6]
d⁻ = [-0.2, -0.2]
d⁺ = [0.2, 0.2]
ϵ = 0.1

Constructs a deviation.

Examples

Deviation(p, d⁻, d⁺, ϵ)

Constructs a deviation.

Deviation(p, d⁻, d⁺)

Constructs a deviation.

Deviation(p, ϵ)

Computes the optimal cross-assignment recursively.

Compute cross-assignment.

Computes the optimal cross assignment from deviation.

Examples

dev = Deviation(...)
cross_assignment(dev)

Polyhedral uncertainty

Polyhedral uncertainty set.