support Julia 1.7

[dehann]

Likely issue on complex numbers and log.

[Affie]

I just had a quick look at the CI log and the first failure is from the log of a negative number called from KDE.
Stacktrace:

DomainError with -2.3542838229142713e306:
log will only return a complex result if called with a complex argument. Try log(Complex(x)).
Stacktrace:
  [1] throw_complex_domainerror(f::Symbol, x::Float64)
    @ Base.Math ./math.jl:33
  [2] _log(x::Float64, base::Val{:ℯ}, func::Symbol)
    @ Base.Math ./special/log.jl:304
  [3] log
    @ ./special/log.jl:269 [inlined]
  [4] _broadcast_getindex_evalf
    @ ./broadcast.jl:670 [inlined]
  [5] _broadcast_getindex
    @ ./broadcast.jl:643 [inlined]
  [6] getindex
    @ ./broadcast.jl:597 [inlined]
  [7] macro expansion
    @ ./broadcast.jl:1005 [inlined]
  [8] macro expansion
    @ ./simdloop.jl:77 [inlined]
  [9] copyto!
    @ ./broadcast.jl:1004 [inlined]
 [10] copyto!
    @ ./broadcast.jl:957 [inlined]
 [11] copy
    @ ./broadcast.jl:929 [inlined]
 [12] materialize
    @ ./broadcast.jl:904 [inlined]
 [13] evalAvgLogL(bd1::BallTreeDensity, bd2::BallTreeDensity, addop::Tuple{typeof(+)}, diffop::Tuple{typeof(-)})
    @ KernelDensityEstimate ~/.julia/dev/KernelDensityEstimate/src/DualTree01.jl:468
 [14] entropy
    @ ~/.julia/dev/KernelDensityEstimate/src/DualTree01.jl:507 [inlined]
 [15] nLOO_LL(alpha::Float64, bd::BallTreeDensity, addop::Tuple{typeof(+)}, diffop::Tuple{typeof(-)})
    @ KernelDensityEstimate ~/.julia/dev/KernelDensityEstimate/src/CrossValidation.jl:20
 [16] golden(bd::BallTreeDensity, ax::Float64, bx::Float64, cx::Float64, tol::Float64, addop::Tuple{typeof(+)}, diffop::Tuple{typeof(-)})
    @ KernelDensityEstimate ~/.julia/dev/KernelDensityEstimate/src/CrossValidation.jl:83
 [17] ksize(bd::BallTreeDensity, addop::Tuple{typeof(+)}, diffop::Tuple{typeof(-)})
    @ KernelDensityEstimate ~/.julia/dev/KernelDensityEstimate/src/CrossValidation.jl:116
 [18] kde!(points::Matrix{Float64}, addop::Tuple{typeof(+)}, diffop::Tuple{typeof(-)})
    @ KernelDensityEstimate ~/.julia/dev/KernelDensityEstimate/src/KDE01.jl:19
 [19] kde! (repeats 2 times)
    @ ~/.julia/dev/KernelDensityEstimate/src/KDE01.jl:31 [inlined]
 [20] getKDEManifoldBandwidths(pts::Matrix{Float64}, manif::Tuple{Symbol, Symbol, Symbol})
    @ ApproxManifoldProducts ~/.julia/packages/ApproxManifoldProducts/oopxJ/src/Legacy.jl:83
 [21] ManifoldKernelDensity(M::GroupManifold{ℝ, ProductManifold{ℝ, Tuple{TranslationGroup{Tuple{2}, ℝ}, SpecialOrthogonal{2}}}, Manifolds.SemidirectProductOperation{RotationAction{TranslationGroup{Tuple{2}, ℝ}, SpecialOrthogonal{2}, LeftAction}}}, vecP::Vector{ProductRepr{Tuple{Vector{Float64}, Matrix{Float64}}}}, u0::ProductRepr{Tuple{Vector{Float64}, Matrix{Float64}}}; partial::Nothing, infoPerCoord::Vector{Float64}, dims::Int64, bw::Nothing)
    @ ApproxManifoldProducts ~/.julia/packages/ApproxManifoldProducts/oopxJ/src/services/ManifoldKernelDensity.jl:64
 [22] #manikde!#37
    @ ~/.julia/packages/ApproxManifoldProducts/oopxJ/src/services/ManifoldKernelDensity.jl:81 [inlined]
 [23] approxConvBelief(dfg::LightDFG{SolverParams, DFGVariable, DFGFactor}, fc::DFGFactor{CommonConvWrapper{ManifoldPrior{GroupManifold{ℝ, ProductManifold{ℝ, Tuple{TranslationGroup{Tuple{2}, ℝ}, SpecialOrthogonal{2}}}, Manifolds.SemidirectProductOperation{RotationAction{TranslationGroup{Tuple{2}, ℝ}, SpecialOrthogonal{2}, LeftAction}}}, ZeroMeanDiagNormal{Tuple{Base.OneTo{Int64}}}, ProductRepr{Tuple{MVector{2, Float64}, MMatrix{2, 2, Float64, 4}}}, ManifoldsBase.VeeOrthogonalBasis{ℝ}}, Nothing, Vector{Int64}, NamedTuple{(:x0,), Tuple{Vector{Any}}}, Nothing, ProductRepr{Tuple{MVector{2, Float64}, MMatrix{2, 2, Float64, 4}}}}, 1}, target::Symbol, measurement::Vector{Tuple}; solveKey::Symbol, N::Int64, skipSolve::Bool)
    @ IncrementalInference ~/.julia/dev/IncrementalInference/src/services/ApproxConv.jl:26
 [24] #calcProposalBelief#320
    @ ~/.julia/dev/IncrementalInference/src/services/ApproxConv.jl:161 [inlined]
 [25] proposalbeliefs!(dfg::LightDFG{SolverParams, DFGVariable, DFGFactor}, destlbl::Symbol, factors::Vector{DFGFactor}, dens::Vector{ManifoldKernelDensity}, measurement::Vector{Tuple}; solveKey::Symbol, N::Int64, dbg::Bool)
    @ IncrementalInference ~/.julia/dev/IncrementalInference/src/services/ApproxConv.jl:229
 [26] propagateBelief(dfg::LightDFG{SolverParams, DFGVariable, DFGFactor}, destvar::DFGVariable{SpecialEuclidean2}, factors::Vector{DFGFactor}; solveKey::Symbol, dens::Vector{ManifoldKernelDensity}, N::Int64, needFreshMeasurements::Bool, dbg::Bool, logger::Base.CoreLogging.SimpleLogger)
    @ IncrementalInference ~/.julia/dev/IncrementalInference/src/GraphProductOperations.jl:30
 [27] predictbelief(dfg::LightDFG{SolverParams, DFGVariable, DFGFactor}, destvert::DFGVariable{SpecialEuclidean2}, factors::Vector{DFGFactor}; asPartial::Bool, kw::Base.Pairs{Symbol, Any, NTuple{4, Symbol}, NamedTuple{(:needFreshMeasurements, :N, :dbg, :logger), Tuple{Bool, Int64, Bool, Base.CoreLogging.SimpleLogger}}})
    @ IncrementalInference ~/.julia/dev/IncrementalInference/src/GraphProductOperations.jl:67
 [28] predictbelief(dfg::LightDFG{SolverParams, DFGVariable, DFGFactor}, destlbl::Symbol, fctlbls::Vector{Symbol}; kw::Base.Pairs{Symbol, Any, NTuple{4, Symbol}, NamedTuple{(:needFreshMeasurements, :N, :dbg, :logger), Tuple{Bool, Int64, Bool, Base.CoreLogging.SimpleLogger}}})
    @ IncrementalInference ~/.julia/dev/IncrementalInference/src/GraphProductOperations.jl:76
 [29] #predictbelief#332
    @ ~/.julia/dev/IncrementalInference/src/GraphProductOperations.jl:82 [inlined]
 [30] doFMCIteration(fgl::LightDFG{SolverParams, DFGVariable, DFGFactor}, vsym::Symbol, solveKey::Symbol, cliq::IncrementalInference.TreeClique, fmsgs::Vector{LikelihoodMessage}, N::Int64, dbg::Bool, needFreshMeasurements::Bool, logger::Base.CoreLogging.SimpleLogger)
    @ IncrementalInference ~/.julia/dev/IncrementalInference/src/SolveTree.jl:65
 [31] fmcmc!(fgl::LightDFG{SolverParams, DFGVariable, DFGFactor}, cliq::IncrementalInference.TreeClique, fmsgs::Vector{LikelihoodMessage}, lbls::Vector{Symbol}, solveKey::Symbol, N::Int64, MCMCIter::Int64, dbg::Bool, logger::Base.CoreLogging.SimpleLogger, multithreaded::Bool)
    @ IncrementalInference ~/.julia/dev/IncrementalInference/src/SolveTree.jl:116
 [32] upGibbsCliqueDensity(dfg::LightDFG{SolverParams, DFGVariable, DFGFactor}, cliq::IncrementalInference.TreeClique, solveKey::Symbol, inmsgs::Vector{LikelihoodMessage}, N::Int64, dbg::Bool, iters::Int64, logger::Base.CoreLogging.SimpleLogger)
    @ IncrementalInference ~/.julia/dev/IncrementalInference/src/SolveTree.jl:186
 [33] approxCliqMarginalUp!(csmc::CliqStateMachineContainer{BayesTreeNodeData, LightDFG{SolverParams, DFGVariable, DFGFactor}, LightDFG{SolverParams, DFGVariable, DFGFactor}, MetaBayesTree}, childmsgs::Vector{LikelihoodMessage}; N::Int64, dbg::Bool, multiproc::Bool, logger::Base.CoreLogging.SimpleLogger, iters::Int64, drawpdf::Bool)
    @ IncrementalInference ~/.julia/dev/IncrementalInference/src/CliqStateMachineUtils.jl:387
 [34] __doCliqUpSolveInitialized!(csmc::CliqStateMachineContainer{BayesTreeNodeData, LightDFG{SolverParams, DFGVariable, DFGFactor}, LightDFG{SolverParams, DFGVariable, DFGFactor}, MetaBayesTree})
    @ IncrementalInference ~/.julia/dev/IncrementalInference/src/CliqStateMachineUtils.jl:32
 [35] solveUp_StateMachine(csmc::CliqStateMachineContainer{BayesTreeNodeData, LightDFG{SolverParams, DFGVariable, DFGFactor}, LightDFG{SolverParams, DFGVariable, DFGFactor}, MetaBayesTree})
    @ IncrementalInference ~/.julia/dev/IncrementalInference/src/CliqueStateMachine.jl:367
 [36] (::StateMachine{CliqStateMachineContainer})(userdata::CliqStateMachineContainer{BayesTreeNodeData, LightDFG{SolverParams, DFGVariable, DFGFactor}, LightDFG{SolverParams, DFGVariable, DFGFactor}, MetaBayesTree}, timeout::Nothing; pollinterval::Float64, breakafter::Function, verbose::Bool, verbosefid::Base.TTY, verboseXtra::IncrementalInference.CliqStatus, iterlimit::Int64, injectDelayBefore::Nothing, recordhistory::Bool, housekeeping_cb::IncrementalInference.var"#445#447"{IncrementalInference.TreeClique, CliqStateMachineContainer{BayesTreeNodeData, LightDFG{SolverParams, DFGVariable, DFGFactor}, LightDFG{SolverParams, DFGVariable, DFGFactor}, MetaBayesTree}})
    @ FunctionalStateMachine ~/.julia/packages/FunctionalStateMachine/2JZFG/src/StateMachine.jl:82
 [37] initStartCliqStateMachine!(dfg::LightDFG{SolverParams, DFGVariable, DFGFactor}, tree::MetaBayesTree, cliq::IncrementalInference.TreeClique, timeout::Nothing; oldcliqdata::BayesTreeNodeData, verbose::Bool, verbosefid::Base.TTY, drawtree::Bool, show::Bool, incremental::Bool, limititers::Int64, upsolve::Bool, downsolve::Bool, recordhistory::Bool, delay::Bool, logger::Base.CoreLogging.SimpleLogger, solve_progressbar::Nothing, algorithm::Symbol, solveKey::Symbol)
    @ IncrementalInference ~/.julia/dev/IncrementalInference/src/CliqueStateMachine.jl:64
 [38] tryCliqStateMachineSolve!(dfg::LightDFG{SolverParams, DFGVariable, DFGFactor}, treel::MetaBayesTree, cliqKey::Int64, timeout::Nothing; oldtree::MetaBayesTree, verbose::Bool, verbosefid::Base.TTY, drawtree::Bool, limititers::Int64, downsolve::Bool, incremental::Bool, delaycliqs::Vector{Symbol}, recordcliqs::Vector{Symbol}, solve_progressbar::Nothing, algorithm::Symbol, solveKey::Symbol)
    @ IncrementalInference ~/.julia/dev/IncrementalInference/src/SolverAPI.jl:112
 [39] (::IncrementalInference.var"#496#499"{MetaBayesTree, Bool, Bool, Base.TTY, Bool, Bool, Vector{Symbol}, Vector{Symbol}, Symbol, Symbol, LightDFG{SolverParams, DFGVariable, DFGFactor}, MetaBayesTree, Nothing, Nothing, Int64, Int64})()
    @ IncrementalInference ./task.jl:411

[Affie]

see discourse: https://discourse.julialang.org/t/whats-going-on-with-exp-and-math-mode-fast/64619/27

Function causing issues is called here:
https://github.com/JuliaRobotics/KernelDensityEstimate.jl/blob/fda7f5d54447cbdabbccb8828971c2ad83d960e9/src/DualTree01.jl#L44

[dehann]

xref JuliaLang/julia#42747

[Affie]

The issue above is now fixed in the julia nightly build. It will be included in v1.7.0/v1.7.0-RC3.

[Affie]

The nightly tests now shows a different upstream issue:

Test loading and saving of Bayes (Junction) tree: Error During Test at /home/runner/work/IncrementalInference.jl/IncrementalInference.jl/test/testTreeSaveLoad.jl:6
  Got exception outside of a @test
  type DataType has no field mutable
  Stacktrace:
    [1] getproperty
      @ ./Base.jl:33 [inlined]
    [2] hasfielddata(T::Any, encounteredtypes::Vector{DataType})
      @ JLD2 ~/.julia/packages/JLD2/qncOK/src/data/writing_datatypes.jl:24
    [3] hasdata(T::DataType, encounteredtypes::Vector{DataType})
      @ JLD2 ~/.julia/packages/JLD2/qncOK/src/data/writing_datatypes.jl:32
    [4] hasdata
      @ ~/.julia/packages/JLD2/qncOK/src/data/writing_datatypes.jl:30 [inlined]
    [5] odr(#unused#::Type{MetaBayesTree})
      @ JLD2 ~/.julia/packages/JLD2/qncOK/src/data/writing_datatypes.jl:563
    [6] objodr
      @ ~/.julia/packages/JLD2/qncOK/src/data/writing_datatypes.jl:109 [inlined]
    [7] write_dataset
      @ ~/.julia/packages/JLD2/qncOK/src/datasets.jl:541 [inlined]
    [8] write(g::JLD2.Group{JLD2.JLDFile{JLD2.MmapIO}}, name::String, obj::MetaBayesTree, wsession::JLD2.JLDWriteSession{Dict{UInt64, JLD2.RelOffset}})
      @ JLD2 ~/.julia/packages/JLD2/qncOK/src/groups.jl:119
    [9] write(f::JLD2.JLDFile{JLD2.MmapIO}, name::String, obj::MetaBayesTree, wsession::JLD2.JLDWriteSession{Dict{UInt64, JLD2.RelOffset}})
      @ JLD2 ~/.julia/packages/JLD2/qncOK/src/JLD2.jl:347
   [10] macro expansion
      @ ~/.julia/packages/JLD2/qncOK/src/loadsave.jl:66 [inlined]
   [11] saveTree(treel::MetaBayesTree, filepath::String)
      @ IncrementalInference ~/work/IncrementalInference.jl/IncrementalInference.jl/src/JunctionTree.jl:1636
   [12] saveTree(treel::MetaBayesTree)
      @ IncrementalInference ~/work/IncrementalInference.jl/IncrementalInference.jl/src/JunctionTree.jl:1629
   [13] macro expansion
      @ ~/work/IncrementalInference.jl/IncrementalInference.jl/test/testTreeSaveLoad.jl:12 [inlined]
   [14] macro expansion
      @ /opt/hostedtoolcache/julia/nightly/x64/share/julia/stdlib/v1.8/Test/src/Test.jl:1321 [inlined]
   [15] top-level scope
      @ ~/work/IncrementalInference.jl/IncrementalInference.jl/test/testTreeSaveLoad.jl:8
   [16] include(fname::String)
      @ Base.MainInclude ./client.jl:451
   [17] top-level scope
      @ ~/work/IncrementalInference.jl/IncrementalInference.jl/test/runtests.jl:30
   [18] include(fname::String)
      @ Base.MainInclude ./client.jl:451
   [19] top-level scope
      @ none:6
   [20] eval
      @ ./boot.jl:368 [inlined]
   [21] exec_options(opts::Base.JLOptions)
      @ Base ./client.jl:271
   [22] _start()
      @ Base ./client.jl:497

[dehann]

Looks like it's the test for saveTree. How far do the tests get if you suppress just that JLD2 save/load tree test?

[Affie]

Lets see: #1438

[Affie]

Yes, It's the only issue on the nightly build.
I'm leaning towards fixing together with the change of LightGraphs (MetaGraphs) to Graphs and when the next julia rc or stable version is released.