Error on execution of doc example

[pekpuglia]

I tried to run this example and got a giant error at sim!, which begins with:

ERROR: MethodError: no method matching ndims(::Type{PreallocationTools.DiffCache{Vector{Float64}, Vector{Float64}}})

Closest candidates are:
ndims(::Base.Broadcast.Broadcasted)
@ Base broadcast.jl:247
ndims(::ArrayInterface.ArrayIndex{N}) where N
@ ArrayInterface ~/.julia/packages/ArrayInterface/a4Wwt/src/ArrayInterface.jl:851
ndims(::Base.Generator)
@ Base generator.jl:53
...

Stacktrace:
[1] Base.Broadcast.BroadcastStyle(#unused#::Type{PreallocationTools.DiffCache{Vector{Float64}, Vector{Float64}}})
@ Base.Broadcast ./broadcast.jl:103
[2] combine_styles(c::PreallocationTools.DiffCache{Vector{Float64}, Vector{Float64}})
@ Base.Broadcast ./broadcast.jl:448
[3] combine_styles(c1::PreallocationTools.DiffCache{Vector{Float64}, Vector{Float64}}, c2::Base.Broadcast.Broadcasted{Base.Broadcast.DefaultArrayStyle{1}, Nothing, typeof(identity), Tuple{Vector{Float64}}})
@ Base.Broadcast ./broadcast.jl:449
[4] materialize!
@ ./broadcast.jl:881 [inlined]
[5] (::ModelPredictiveControl.var"#inner_solver_f!#24"{DataType, RungeKutta, ModelPredictiveControl.var"#f!#35"{typeof(f)}, PreallocationTools.DiffCache{Vector{Float64}, Vector{Float64}}, PreallocationTools.DiffCache{Vector{Float64}, Vector{Float64}}, PreallocationTools.DiffCache{Vector{Float64}, Vector{Float64}}, PreallocationTools.DiffCache{Vector{Float64}, Vector{Float64}}, PreallocationTools.DiffCache{Vector{Float64}, Vector{Float64}}, Float64})(xnext::Vector{Float64}, x::Vector{Float64}, u::Vector{Float64}, d::Vector{Float64}, p::Vector{Float64})
@ ModelPredictiveControl ~/.julia/packages/ModelPredictiveControl/KRRfW/src/model/solver.jl:56

The error appears to come from this line:

@. xcur = x

I'm using Julia 1.9.

[franckgaga]

Thanks for the report. Probably because 1.9 is no longer officially supported, the new LTS is 1.10. edit: I'll do some test on 1.9 and I'll come back to you

[franckgaga]

I just tried to run the motor exemple in the manual on Julia 1.9, and there is no error. Are you using the last stable version of the package (v1.1.4) ?

[franckgaga]

Weird I tested it today and now I can reproduce. I'll investigate on that.

[franckgaga]

Ok yes this is a compatibility issue. You cannot use the last stable version of ModelPredictiveControl.jl since Julia 1.9 is no longer supported. You have two choices: installing v1.0.1 version of this package:

using Pkg; Pkg.add(name="ModelPredictiveControl", version="1.0.1")

or updating Julia to 1.10 or newer. This can be done easily with juliaup installer.

[franckgaga]

@pekpuglia, is the issue persisting with one of the two workarounds ?

[pekpuglia]

Hello, thank you for your responses, I still haven't had the time to get back to this. Maybe by the end of the week, but this looks good. Will try migrating to julia 1.10.

[pekpuglia]

Hi, switched to julia 1.10, works like a charm. Maybe add a compat line in Project.toml?

[franckgaga]

Yes, I added the line in the last version of the package (v1.2.0).