CI trigger

[ArnoStrouwen]

No description provided.

[ArnoStrouwen]

@ChrisRackauckas This package regressed since 2 weeks ago.
https://github.com/SciML/QuasiMonteCarlo.jl/actions/runs/7330602882/job/19961974828?pr=100#step:6:234

[ChrisRackauckas]

Looks like it's JuliaIntervals/IntervalArithmetic.jl#582. @OlivierHnt what should we do here?

[OlivierHnt]

Indeed, the current stable release of IntervalArithmetic (v0.22) removed the structure IntervalBox; an AbstractVector{<:Interval} will be directly interpreted as a multi-dimensional interval.
After a quick look, try to:

• replace IntervalBox{N,T} with Vector{Interval{T}} (or even AbstractVector{<:Interval}) in type signature (cf.

  QuasiMonteCarlo.jl/test/runtests.jl

  Line 78 in eab1519

  inCloseOpen(X::AbstractVector, Y::IntervalBox{N, T}) where {N, T} = all(inCloseOpen.(X, Y))

  )
• delete the call IntervalBox in line

  QuasiMonteCarlo.jl/test/runtests.jl

  Line 56 in eab1519

  intervals = mince(IntervalBox([interval(zero(T), one(T)) for i in 1:s]),

• do not access the lower and upper bounds of an interval directly, use inf and sup respectively; for instance the line

  QuasiMonteCarlo.jl/test/runtests.jl

  Line 76 in eab1519

  a.lo <= x < a.hi

  should read inf(a) <= x < sup(a)

Hopefully this fixes the CI 🙂; I can provide more pointers if need be.