[BUG]: Cannot predict with MultitargetSRRegressor, ERROR: MethodError: no method matching zero(::@NamedTuple{a::Vector{Float64}, b::Vector{Float64}, c::Vector{Float64}}) The function zero exists, but no method is defined for this combination of argument types.

[Klipp-Linding-Lab]

What happened?

Dear all, we cannot execute this code nor our own for a similar purpose; this code is copy/paste from the API doc.

using MLJ
MultitargetSRRegressor = @load MultitargetSRRegressor pkg=SymbolicRegression
X = (a=rand(100), b=rand(100), c=rand(100))
Y = (y1=(@. cos(X.c) * 2.1 - 0.9), y2=(@. X.a * X.b + X.c))
model = MultitargetSRRegressor(binary_operators=[+, -, *], unary_operators=[exp], niterations=100)
mach = machine(model, X, Y)
fit!(mach)
y_hat = predict(mach, X)

View the equations used:

r = report(mach)
for (output_index, (eq, i)) in enumerate(zip(r.equation_strings, r.best_idx))
println("Equation used for ", output_index, ": ", eq[i])
end

Version

1.11.2

Operating System

Linux

Interface

Julia REPL

Relevant log output

julia> y_hat = predict(mach, X)
       # View the equations used:
ERROR: MethodError: no method matching zero(::@NamedTuple{a::Vector{Float64}, b::Vector{Float64}, c::Vector{Float64}})
The function `zero` exists, but no method is defined for this combination of argument types.

Closest candidates are:
  zero(::Type{Union{}}, Any...)
   @ Base number.jl:310
  zero(::Type{MutableArithmetics.Zero})
   @ MutableArithmetics ~/.julia/packages/MutableArithmetics/BLlgj/src/rewrite.jl:35
  zero(::Type{Dates.Date})
   @ Dates /usr/share/julia/stdlib/v1.11/Dates/src/types.jl:459
  ...

Stacktrace:
 [1] alg_cache(alg::Vern7{…}, u::@NamedTuple{…}, rate_prototype::@NamedTuple{…}, ::Type{…}, ::Type{…}, ::Type{…}, uprev::@NamedTuple{…}, uprev2::@NamedTuple{…}, f::ODEFunction{…}, t::Float64, dt::Float64, reltol::Float64, p::Machine{…}, calck::Bool, ::Val{…})
   @ OrdinaryDiffEqVerner ~/.julia/packages/OrdinaryDiffEqVerner/m0ylj/src/verner_caches.jl:96
 [2] __init(prob::ODEProblem{…}, alg::Vern7{…}, timeseries_init::Tuple{}, ts_init::Tuple{}, ks_init::Tuple{}, recompile::Type{…}; saveat::Vector{…}, tstops::Tuple{}, d_discontinuities::Tuple{}, save_idxs::Nothing, save_everystep::Bool, save_on::Bool, save_start::Bool, save_end::Nothing, callback::Nothing, dense::Bool, calck::Bool, dt::Float64, dtmin::Float64, dtmax::Float64, force_dtmin::Bool, adaptive::Bool, gamma::Rational{…}, abstol::Float64, reltol::Float64, qmin::Rational{…}, qmax::Int64, qsteady_min::Int64, qsteady_max::Int64, beta1::Nothing, beta2::Nothing, qoldinit::Rational{…}, controller::Nothing, fullnormalize::Bool, failfactor::Int64, maxiters::Int64, internalnorm::typeof(DiffEqBase.ODE_DEFAULT_NORM), internalopnorm::typeof(opnorm), isoutofdomain::typeof(DiffEqBase.ODE_DEFAULT_ISOUTOFDOMAIN), unstable_check::typeof(DiffEqBase.ODE_DEFAULT_UNSTABLE_CHECK), verbose::Bool, timeseries_errors::Bool, dense_errors::Bool, advance_to_tstop::Bool, stop_at_next_tstop::Bool, initialize_save::Bool, progress::Bool, progress_steps::Int64, progress_name::String, progress_message::typeof(DiffEqBase.ODE_DEFAULT_PROG_MESSAGE), progress_id::Symbol, userdata::Nothing, allow_extrapolation::Bool, initialize_integrator::Bool, alias_u0::Bool, alias_du0::Bool, initializealg::OrdinaryDiffEqCore.DefaultInit, kwargs::@Kwargs{})
   @ OrdinaryDiffEqCore ~/.julia/packages/OrdinaryDiffEqCore/H25Bn/src/solve.jl:360
 [3] __solve(::ODEProblem{…}, ::Vern7{…}; kwargs::@Kwargs{…})
   @ OrdinaryDiffEqCore ~/.julia/packages/OrdinaryDiffEqCore/H25Bn/src/solve.jl:6
 [4] solve_call(_prob::ODEProblem{…}, args::Vern7{…}; merge_callbacks::Bool, kwargshandle::Nothing, kwargs::@Kwargs{…})
   @ DiffEqBase ~/.julia/packages/DiffEqBase/HW4ge/src/solve.jl:632
 [5] solve_up(prob::ODEProblem{…}, sensealg::QuadratureAdjoint{…}, u0::@NamedTuple{…}, p::Machine{…}, args::Vern7{…}; kwargs::@Kwargs{…})
   @ DiffEqBase ~/.julia/packages/DiffEqBase/HW4ge/src/solve.jl:1120
 [6] solve(prob::ODEProblem{…}, args::Vern7{…}; sensealg::QuadratureAdjoint{…}, u0::Nothing, p::Nothing, wrap::Val{…}, kwargs::@Kwargs{…})
   @ DiffEqBase ~/.julia/packages/DiffEqBase/HW4ge/src/solve.jl:1036
 [7] predict(θ::Machine{…}, X::@NamedTuple{…}, T::Vector{…})
   @ Main ~/JuliaCode/Phospho_work_multi.jl:79
 [8] predict(θ::Machine{…}, X::@NamedTuple{…})
   @ Main ~/JuliaCode/Phospho_work_multi.jl:78
 [9] top-level scope
   @ REPL[57]:1
Some type information was truncated. Use `show(err)` to see complete types.

julia> r = report(mach)
(best_idx = [12, 3],
 equations = Vector{Expression{Float64, DynamicExpressions.NodeModule.Node{Float64}, @NamedTuple{operators::DynamicExpressions.OperatorEnumModule.OperatorEnum{Tuple{typeof(+), typeof(-), typeof(*)}, Tuple{typeof(exp)}}, variable_names::Vector{String}}}}[[0.881348865751391, 1.3756828183029748 - c, (c * -0.9719064318724064) + 1.3617952137309295, (exp(c) * -0.5878680910661597) + 1.8809092039980162, ((c * -0.9809376280537772) * c) + 1.1931951498025708, (((c * -0.906876499286865) - 0.07757180782560917) * c) + 1.2079970263208184, ((c * (-1.0995189493322781 - (c * -0.13000357499675755))) * c) + 1.20103431524694, ((c * c) * (-1.1287057437085772 - (exp(c) * -0.05985652480674373))) + 1.2003258712015616, ((c * (-1.048886141802872 - (c * (c * -0.0838688737571194)))) * c) + 1.1999494876771033, ((c * (-1.0519066959939267 - (((c * -0.07901736953345984) + -0.007620780644438593) * c))) * c) + 1.2000182742759145, (c * ((-1.0520028270192574 - ((((c * -0.07945100368209045) + -0.007917907493167769) * c) * 0.9926495499429491)) * c)) + 1.2000205805924693, ((c * (-1.0495894647085022 - (c * ((exp(exp(c)) * 0.00024345834052550238) + (c * -0.08785656137392589))))) * c) + 1.1999906355972207, (((c * -1.0467501249516413) - (((exp(c + c) * 0.002105260850929585) + (c * (c * -0.09691448575689704))) * c)) * c) + 1.1999846245675594], [c, c + 0.2483967687211474, c + (a * b)]],
 equation_strings = [["0.881348865751391", "1.3756828183029748 - c", "(c * -0.9719064318724064) + 1.3617952137309295", "(exp(c) * -0.5878680910661597) + 1.8809092039980162", "((c * -0.9809376280537772) * c) + 1.1931951498025708", "(((c * -0.906876499286865) - 0.07757180782560917) * c) + 1.2079970263208184", "((c * (-1.0995189493322781 - (c * -0.13000357499675755))) * c) + 1.20103431524694", "((c * c) * (-1.1287057437085772 - (exp(c) * -0.05985652480674373))) + 1.2003258712015616", "((c * (-1.048886141802872 - (c * (c * -0.0838688737571194)))) * c) + 1.1999494876771033", "((c * (-1.0519066959939267 - (((c * -0.07901736953345984) + -0.007620780644438593) * c))) * c) + 1.2000182742759145", "(c * ((-1.0520028270192574 - ((((c * -0.07945100368209045) + -0.007917907493167769) * c) * 0.9926495499429491)) * c)) + 1.2000205805924693", "((c * (-1.0495894647085022 - (c * ((exp(exp(c)) * 0.00024345834052550238) + (c * -0.08785656137392589))))) * c) + 1.1999906355972207", "(((c * -1.0467501249516413) - (((exp(c + c) * 0.002105260850929585) + (c * (c * -0.09691448575689704))) * c)) * c) + 1.1999846245675594"], ["c", "c + 0.2483967687211474", "c + (a * b)"]],
 losses = [[0.07312460305530863, 0.003716344408720649, 0.0036583028508886477, 0.0007973156443596637, 3.198335137167272e-5, 6.113176053502374e-6, 3.9148436746139914e-7, 2.0797653690360404e-8, 1.9244315475448397e-9, 1.6390605075791748e-10, 1.576166129196469e-10, 4.662930288906626e-11, 4.380522004124344e-11], [0.11494901030186083, 0.053248055568929614, 0.0]],
 complexities = [[1, 3, 5, 6, 7, 9, 11, 12, 13, 15, 17, 19, 22], [1, 3, 5]],
 scores = [[36.04365338911715, 1.4897121889644036, 0.007870580688856953, 1.5235039759207538, 3.216035143211375, 0.8273845236461245, 1.3741280723447382, 2.9351054077812564, 2.380209552843019, 1.2315462503456258, 0.019563909324637165, 0.6089682114621419, 0.020825390521924444], [36.04365338911715, 0.384763676286292, 18.021826694558577]],)

julia> for (output_index, (eq, i)) in enumerate(zip(r.equation_strings, r.best_idx))
           println("Equation used for ", output_index, ": ", eq[i])
       end
Equation used for 1: ((c * (-1.0495894647085022 - (c * ((exp(exp(c)) * 0.00024345834052550238) + (c * -0.08785656137392589))))) * c) + 1.1999906355972207
Equation used for 2: c + (a * b)

Extra Info

our own code gives:
ERROR: MethodError: no method matching zero(::@NamedTuple{data::@NamedTuple{u1::Vector{…}, u2::Vector{…}, u3::Vector{…}, u4::Vector{…}}, idx::Vector{Int64}})
The function zero exists, but no method is defined for this combination of argument types.

Closest candidates are:
zero(::Type{Union{}}, Any...)
@ Base number.jl:310
zero(::Type{MutableArithmetics.Zero})
@ MutableArithmetics ~/.julia/packages/MutableArithmetics/BLlgj/src/rewrite.jl:35
zero(::Type{Dates.Date})
@ Dates /usr/share/julia/stdlib/v1.11/Dates/src/types.jl:459

Even if we try something like:

predict(mach,(data=X1x, idx=[4,4,5,3]))
ERROR: MethodError: no method matching zero(::@NamedTuple{data::@NamedTuple{u1::Vector{…}, u2::Vector{…}, u3::Vector{…}, u4::Vector{…}}, idx::Vector{Int64}})
The function zero exists, but no method is defined for this combination of argument types.

Our environment is:
(@v1.11) pkg> status
Status ~/.julia/environments/v1.11/Project.toml
⌃ [8ce10254] Bumper v0.6.0
[b0b7db55] ComponentArrays v0.15.19
[2445eb08] DataDrivenDiffEq v1.5.0
[5b588203] DataDrivenSparse v0.1.2
[0c46a032] DifferentialEquations v7.15.0
[98e50ef6] JuliaFormatter v1.0.62
[23fbe1c1] Latexify v0.16.5
[d3d80556] LineSearches v7.3.0
[b2108857] Lux v1.4.1
[add582a8] MLJ v0.20.7
[961ee093] ModelingToolkit v9.57.0
[7f7a1694] Optimization v4.0.5
[36348300] OptimizationOptimJL v0.4.1
[42dfb2eb] OptimizationOptimisers v0.3.6
[1dea7af3] OrdinaryDiffEq v6.90.1
[91a5bcdd] Plots v1.40.9
[1ed8b502] SciMLSensitivity v7.71.2
[860ef19b] StableRNGs v1.0.2
[8254be44] SymbolicRegression v1.3.0
[d1185830] SymbolicUtils v3.7.2
[e88e6eb3] Zygote v0.6.73

[MilesCranmer]

It looks like you defined a function predict in your script? The one SymbolicRegression.jl uses is from MLJ. Your traceback looks to have some OrdinaryDiffEqCore stuff - which SymbolicRegression.jl doesn't depend on.

[Klipp-Linding-Lab]

Dear Miles

Ofcourse; how embarrassing ;) this explains a lot. Many thanks and sorry for bug report.

best wishes

[MilesCranmer]

No worries :)