measures.jl

using Statistics: Statistics

include("faithfulness/faithfulness.jl")
include("plausibility/plausibility.jl")

"""
    validity(ce::CounterfactualExplanation; γ=0.5)

Checks of the counterfactual search has been successful with respect to the probability threshold `γ`. In case multiple counterfactuals were generated, the function returns the proportion of successful counterfactuals.
"""
function validity(ce::CounterfactualExplanation; agg=Statistics.mean, γ=0.5)
    val = agg(CounterfactualExplanations.target_probs(ce) .>= γ)
    val = val isa LinearAlgebra.AbstractMatrix ? vec(val) : val
    return val
end

"""
    validity_strict(ce::CounterfactualExplanation)

Checks if the counterfactual search has been strictly valid in the sense that it has converged with respect to the pre-specified target probability `γ`.
"""
function validity_strict(ce::CounterfactualExplanation)
    return validity(ce; γ=ce.convergence.decision_threshold)
end

"""
    redundancy(ce::CounterfactualExplanation)

Computes the feature redundancy: that is, the number of features that remain unchanged from their original, factual values.
"""
function redundancy(ce::CounterfactualExplanation; agg=Statistics.mean, tol=1e-5)
    x′ = CounterfactualExplanations.counterfactual(ce)
    redundant_x = [
        agg(sum(abs.(x .- ce.x) .< tol) / size(x, 1)) for x in eachslice(x′; dims=ndims(x′))
    ]
    redundant_x = length(redundant_x) == 1 ? redundant_x[1] : redundant_x
    return redundant_x
end