Providing initial solution

[ghost]

Dear Robert,

A great thanks for providing such a great optimisation algorithm which I use extensively. From your example it is not clear to me how to include the initial guess in my problem:

`SearchRange = SEARCHRANGE(optim, optionHypix)

InitialGuess = [0.3, 0.1,0.0031593277198464265, 0.5, 3.0, 0.8]

Optimization = BlackBoxOptim.bboptimize(X -> OF_HYPIX(..., X, ...); SearchRange=SearchRange, NumDimensions=optim.NparamOpt, TraceMode=:silent, MaxFuncEvals=paramHypix.opt.NmaxFuncEvals)

X = BlackBoxOptim.best_candidate(Optimization)`

Many thanks for any help you may provide and hoping you are enjoying the warm Sweedish summer,
Joseph

[robertfeldt]

Hmm, my code from the README still works so I don't really see the problem here. You provide an initial guess as the second argument to bboptimize. I replicate my README code in a new Julia 1.8 here and maybe it helps you find where you go wrong:

julia> versioninfo()
Julia Version 1.8.0
Commit 5544a0fab76 (2022-08-17 13:38 UTC)
Platform Info:
  OS: macOS (arm64-apple-darwin21.3.0)
  CPU: 10 × Apple M1 Max
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-13.0.1 (ORCJIT, apple-m1)
  Threads: 8 on 8 virtual cores
Environment:
  JULIA_NUM_THREADS = 8

(@v1.8) pkg> add BlackBoxOptim#master
...

julia> myproblem(x) = (x[1] - 3.14)^2 + (x[2] - 7.2)^4
myproblem (generic function with 1 method)

julia> optimum = [3.14, 7.2]
2-element Vector{Float64}:
 3.14
 7.2

julia> good_guess = [3.0, 7.2]
2-element Vector{Float64}:
 3.0
 7.2

julia> using BlackBoxOptim

julia> res1 = bboptimize(myproblem, good_guess; NumDimensions = 2, SearchRange = (-10.0, 10.0));
...

Best candidate found: [3.14, 7.2]

Fitness: 0.000000000

julia> two_good_guesses = [good_guess, [3.1, 7.3]]
2-element Vector{Vector{Float64}}:

[3.0, 7.2]
 [3.1, 7.3]

julia> res2 = bboptimize(myproblem, two_good_guesses; NumDimensions = 2, SearchRange = (-10.0, 10.0));
...

julia> @assert isapprox(best_fitness(res2), myproblem(optimum); atol = 1e-30)

Please check this code also on your side and then adapt to your own use case. Get back to me if doesn't solve it.