Can't launch a basic example with Evolution.jl

New to Julia
1

I’m trying to run a genetic algorithm optimization, and to reuse Evolution.jl package instead of writing the implementation on my own.

The docs don’t show any basic example, so I tried this code in tests as a pattern, studied the arguments one by one and tried to construct the simplest possible code to run.

I use import instead of using, because other packages like GeoDataFrames redefine union and other important funcitons, and they clash in the global scope.

The basic example I’m trying to run is: generate a vector of 20 floats and take a sum of them.

import StatsBase as sb
import Evolutionary as ev
import StableRNGs as sr

P = 100
G = 20
N = 30

result = ev.optimize(
    (v1) -> -sum(v1),
    () -> [rand(G) for i in 1:P],
        ev.GA(
        populationSize = P,
        selection = (fitness) -> sortperm(fitness, rev=true)[1:N],
        crossover = (v1, v2) -> sb.shuffle(collect(union(v1, v2)))[1:G],
        crossoverRate = 0.8,
        mutation = (v1) -> rand(G),
        mutationRate = 0.06,
    ),
    ev.Options(successive_f_tol=30)
    # another variant: ev.Options(rng=sr.StableRNG(42), successive_f_tol=30)
);

As I see, somewhere deep in the code it tries to call zero with my gene type (Vector{Float64}), but fails.

Does anyone have a simplest example of this lib to try out?

MethodError: no method matching zero(::Type{Vector{Float64}})
Closest candidates are:
  zero(::Union{Type{P}, P}) where P<:Dates.Period at /snap/julia/62/share/julia/stdlib/v1.8/Dates/src/periods.jl:53
  zero(::ForwardDiff.Dual) at ~/.julia/packages/ForwardDiff/pDtsf/src/dual.jl:355
  zero(::LinearAlgebra.Diagonal{T, StaticArraysCore.SVector{N, T}}) where {N, T} at ~/.julia/packages/StaticArrays/PUoe1/src/SDiagonal.jl:41
  ...

Stacktrace:
 [1] default_values(x::Vector{Vector{Float64}})
   @ Evolutionary ~/.julia/packages/Evolutionary/65hL6/src/api/utilities.jl:130
 [2] Evolutionary.EvolutionaryObjective(f::var"#47#53", x::Vector{Vector{Float64}}, F::Vector{Float64}; eval::Symbol)
   @ Evolutionary ~/.julia/packages/Evolutionary/65hL6/src/api/objective.jl:19
 [3] optimize(f::var"#47#53", constraints::Evolutionary.NoConstraints, method::Evolutionary.GA{var"#50#56", var"#51#57", var"#52#58"}, population::Vector{Vector{Vector{Float64}}}, opts::Evolutionary.Options{Nothing, Random.TaskLocalRNG})
   @ Evolutionary ~/.julia/packages/Evolutionary/65hL6/src/api/optimize.jl:54
 [4] optimize(f::var"#47#53", constraints::Evolutionary.NoConstraints, individual::Function, method::Evolutionary.GA{var"#50#56", var"#51#57", var"#52#58"}, opts::Evolutionary.Options{Nothing, Random.TaskLocalRNG})
   @ Evolutionary ~/.julia/packages/Evolutionary/65hL6/src/api/optimize.jl:42
 [5] optimize(f::var"#47#53", individual::Function, method::Evolutionary.GA{var"#50#56", var"#51#57", var"#52#58"}, opts::Evolutionary.Options{Nothing, Random.TaskLocalRNG})
   @ Evolutionary ~/.julia/packages/Evolutionary/65hL6/src/api/optimize.jl:15
 [6] top-level scope
   @ In[16]:6
 [7] eval
   @ ./boot.jl:368 [inlined]
 [8] include_string(mapexpr::typeof(REPL.softscope), mod::Module, code::String, filename::String)
   @ Base ./loading.jl:1428