Bug Gol-Lazar-Belta on Julia 1

[adrienbanse]

@blegat Seems like the Gol-Lazar-Belta bugs only on Julia 1 (not 1.6). Locally I have no problem.
If I re-run the error can disappear.. Any idea? Seems related to Pavito





Depth: 11: Test Failed at /home/runner/work/Dionysos.jl/Dionysos.jl/test/examples/test_gol_lazar_belta.jl:31
Expression: MOI.get(optimizer, MOI.TerminationStatus()) in [MOI.OPTIMAL, MOI.LOCALLY_SOLVED]
Evaluated: MathOptInterface.ALMOST_LOCALLY_SOLVED in MathOptInterface.TerminationStatusCode[MathOptInterface.OPTIMAL, MathOptInterface.LOCALLY_SOLVED]

Stacktrace:
[1] macro expansion
@ /opt/hostedtoolcache/julia/1.9.1/x64/share/julia/stdlib/v1.9/Test/src/Test.jl:478 [inlined]
[2] _test(algo::MathOptInterface.OptimizerWithAttributes, N::Int64, q0::Int64, x0::Vector{Float64}, x_expected::Vector{Vector{Float64}}, u_expected::Vector{Vector{Float64}}, obj_expected::Float64, zero_cost::Bool, mi::Bool; kws::Base.Pairs{Symbol, Union{}, Tuple{}, NamedTuple{(), Tuple{}}})
@ Main.TestMain ~/work/Dionysos.jl/Dionysos.jl/test/examples/test_gol_lazar_belta.jl:31
Depth: 11: Error During Test at /home/runner/work/Dionysos.jl/Dionysos.jl/test/examples/test_gol_lazar_belta.jl:151
Got exception outside of a @test
Result index of attribute MathOptInterface.VariablePrimal(1) out of bounds. There are currently 0 solution(s) in the model.
Stacktrace:
[1] check_result_index_bounds
@ ~/.julia/packages/MathOptInterface/BlCD1/src/attributes.jl:207 [inlined]
[2] get
@ ~/.julia/packages/Pavito/bhsCM/src/MOI_wrapper.jl:312 [inlined]
[3] get(b::MathOptInterface.Bridges.LazyBridgeOptimizer{Pavito.Optimizer}, attr::MathOptInterface.VariablePrimal, index::MathOptInterface.VariableIndex)
@ MathOptInterface.Bridges ~/.julia/packages/MathOptInterface/BlCD1/src/Bridges/bridge_optimizer.jl:1195
[4] get(model::MathOptInterface.Utilities.CachingOptimizer{MathOptInterface.AbstractOptimizer, MathOptInterface.Utilities.UniversalFallback{MathOptInterface.Utilities.Model{Float64}}}, attr::MathOptInterface.VariablePrimal, index::MathOptInterface.VariableIndex)
@ MathOptInterface.Utilities ~/.julia/packages/MathOptInterface/BlCD1/src/Utilities/cachingoptimizer.jl:911
[5] _value
@ ~/work/Dionysos.jl/Dionysos.jl/src/optim/bemporad_morari.jl:110 [inlined]
[6] _broadcast_getindex_evalf
@ ./broadcast.jl:683 [inlined]
[7] _broadcast_getindex
@ ./broadcast.jl:656 [inlined]
[8] getindex
@ ./broadcast.jl:610 [inlined]
[9] copy
@ ./broadcast.jl:912 [inlined]
[10] materialize(bc::Base.Broadcast.Broadcasted{Base.Broadcast.DefaultArrayStyle{2}, Nothing, typeof(Dionysos.Optim.BemporadMorari._value), Tuple{Base.RefValue{MathOptInterface.Utilities.CachingOptimizer{MathOptInterface.AbstractOptimizer, MathOptInterface.Utilities.UniversalFallback{MathOptInterface.Utilities.Model{Float64}}}}, Matrix{MathOptInterface.VariableIndex}}})
@ Base.Broadcast ./broadcast.jl:873
[11] get(optimizer::Dionysos.Optim.BemporadMorari.Optimizer{Float64}, #unused#::Dionysos.Control.ContinuousTrajectoryAttribute)
@ Dionysos.Optim.BemporadMorari ~/work/Dionysos.jl/Dionysos.jl/src/optim/bemporad_morari.jl:349
[12] _test(algo::MathOptInterface.OptimizerWithAttributes, N::Int64, q0::Int64, x0::Vector{Float64}, x_expected::Vector{Vector{Float64}}, u_expected::Vector{Vector{Float64}}, obj_expected::Float64, zero_cost::Bool, mi::Bool; kws::Base.Pairs{Symbol, Union{}, Tuple{}, NamedTuple{(), Tuple{}}})
@ Main.TestMain ~/work/Dionysos.jl/Dionysos.jl/test/examples/test_gol_lazar_belta.jl:32
[13] _test
@ ~/work/Dionysos.jl/Dionysos.jl/test/examples/test_gol_lazar_belta.jl:21 [inlined]
[14] #_test11#4
@ ~/work/Dionysos.jl/Dionysos.jl/test/examples/test_gol_lazar_belta.jl:119 [inlined]
[15] _test11
@ ~/work/Dionysos.jl/Dionysos.jl/test/examples/test_gol_lazar_belta.jl:118 [inlined]
[16] macro expansion
@ ~/work/Dionysos.jl/Dionysos.jl/test/examples/test_gol_lazar_belta.jl:152 [inlined]
[17] macro expansion
@ /opt/hostedtoolcache/julia/1.9.1/x64/share/julia/stdlib/v1.9/Test/src/Test.jl:1498 [inlined]
[18] macro expansion
@ ~/work/Dionysos.jl/Dionysos.jl/test/examples/test_gol_lazar_belta.jl:152 [inlined]
[19] macro expansion
@ /opt/hostedtoolcache/julia/1.9.1/x64/share/julia/stdlib/v1.9/Test/src/Test.jl:1586 [inlined]
[20] (::Main.TestMain.var"#tests#5")(qp_solver::MathOptInterface.OptimizerWithAttributes, miqp_solver::MathOptInterface.OptimizerWithAttributes)
@ Main.TestMain ~/work/Dionysos.jl/Dionysos.jl/test/examples/test_gol_lazar_belta.jl:127
[21] macro expansion
@ ~/work/Dionysos.jl/Dionysos.jl/test/examples/test_gol_lazar_belta.jl:184 [inlined]
[22] macro expansion
@ /opt/hostedtoolcache/julia/1.9.1/x64/share/julia/stdlib/v1.9/Test/src/Test.jl:1498 [inlined]
[23] top-level scope
@ ~/work/Dionysos.jl/Dionysos.jl/test/examples/test_gol_lazar_belta.jl:125
[24] include(fname::String)
@ Base.MainInclude ./client.jl:478
[25] top-level scope
@ ~/work/Dionysos.jl/Dionysos.jl/test/runtests.jl:33
[26] include(fname::String)
@ Base.MainInclude ./client.jl:478
[27] top-level scope
@ none:6
[28] eval
@ ./boot.jl:370 [inlined]
[29] exec_options(opts::Base.JLOptions)
@ Base ./client.jl:280
[30] _start()
@ Base ./client.jl:522

[adrienbanse]

Maybe related to the switch to HiGHS #209

[blegat]

Maybe we can allow ALMOST_LOCALLY_SOLVED as well. Otherwise, we could try to play with the tolerance but the test would remain flaky

[adrienbanse]

Thanks for your reply.
Any idea where this is coming from? Is there something random in the solver?

[blegat]

Not in Pavito itself but maybe in one of the inner solvers.