PortfolioOptimisers.jl

Appearance

13

Randomised search cross validation ​

PortfolioOptimisers.RandomisedSearchCrossValidation Type
julia
struct RandomisedSearchCrossValidation{__T_p, __T_cv, __T_r, __T_scorer, __T_ex, __T_n_iter, __T_rng, __T_seed, __T_train_score, __T_kwargs} <: AbstractSearchCrossValidationEstimator

Randomised search cross-validation estimator for portfolio optimisation. Samples parameter sets from distributions or vectors, applies cross-validation splits, fits and scores each configuration, and selects the optimal parameters using the provided scoring strategy.

Fields

  • p: Hyperparameter search grid.

  • cv: Cross-validation estimator.

  • r: Risk measure or vector of risk measures.

  • scorer: Scoring function.

  • ex: Parallel execution strategy.

  • n_iter: Number of random iterations.

  • rng: Random number generator.

  • seed: Seed for the random number generator.

  • train_score: Whether to also compute the training set score.

  • kwargs: Additional keyword arguments.

Constructors

julia
RandomisedSearchCrossValidation(
    p::Union{AbstractVector{<:Pair{<:AbstractString, <:Any}},
             AbstractVector{<:AbstractVector{<:Pair{<:AbstractString,
                                                    <:Any}}},
             AbstractDict{<:AbstractString, <:Any},
             AbstractVector{<:AbstractDict{<:AbstractString,
                                           <:Any}}};
    cv::SearchCV = KFold(),
    r::AbstractBaseRiskMeasure = ConditionalValueatRisk(),
    scorer::CrossValSearchScorer = HighestMeanScore(),
    ex::FLoops.Transducers.Executor = FLoops.ThreadedEx(),
    n_iter::Integer = 10,
    rng::Random.AbstractRNG = Random.default_rng(),
    seed::Option{<:Integer} = nothing,
    train_score::Bool = false,
    kwargs::NamedTuple = (;),
) -> RandomisedSearchCrossValidation

Keyword arguments correspond to the struct's fields.

Validation

  • !isempty(p).

  • If p is a vector of parameter sets: each element must not be empty.

  • All keys in p must be of type GSCVKey (i.e. String, Symbol, or Integer).

  • All values in p must be of type RSCVVal (i.e. an AbstractVector or Distributions.Distribution).

  • n_iter > 0 and finite.

Examples

julia
julia> RandomisedSearchCrossValidation(Dict("alpha" => [0.1, 0.2, 0.3],
                                            "beta" => Normal(1.0, 0.5)))
RandomisedSearchCrossValidation
            p ┼ Dict{String, Any}: Dict{String, Any}("alpha" => [0.1, 0.2, 0.3], "beta" => Distributions.Normal{Float64}(μ=1.0, σ=0.5))
           cv ┼ KFold
              │              n ┼ Int64: 5
              │    purged_size ┼ Int64: 0
              │   embargo_size ┴ Int64: 0
            r ┼ ConditionalValueatRisk
              │   settings ┼ RiskMeasureSettings
              │            │   scale ┼ Float64: 1.0
              │            │      ub ┼ nothing
              │            │     rke ┴ Bool: true
              │      alpha ┼ Float64: 0.05
              │          w ┴ nothing
       scorer ┼ HighestMeanScore()
           ex ┼ Transducers.ThreadedEx{@NamedTuple{}}: Transducers.ThreadedEx()
       n_iter ┼ Int64: 10
          rng ┼ Random.TaskLocalRNG: Random.TaskLocalRNG()
         seed ┼ nothing
  train_score ┼ Bool: false
       kwargs ┴ @NamedTuple{}: NamedTuple()

Related

source
PortfolioOptimisers.search_cross_validation Method
julia
search_cross_validation(opt::NonFiniteAllocationOptimisationEstimator,
                       rscv::RandomisedSearchCrossValidation,
                       rd::ReturnsResult)

Performs randomised search cross-validation for portfolio optimisation estimators. Samples parameter sets from distributions or vectors, applies cross-validation splits, fits and scores each configuration, and selects the optimal parameters using the provided scoring strategy.

Arguments

  • opt: Portfolio optimisation estimator to be tuned.

  • rscv: Randomised search cross-validation estimator specifying parameter grid, CV splitter, risk measure, scorer, execution strategy, number of iterations, RNG, and options.

  • rd: Returns result containing asset returns data.

Returns

  • SearchCrossValidationResult: Result type containing the optimal estimator, test and train scores, parameter grid, and selected index.

Details

  • Samples parameter sets from vectors or distributions.

  • Applies cross-validation splits to the returns data.

  • Fits the estimator for each sampled parameter set and split.

  • Scores each configuration using the specified risk measure and scoring function.

  • Selects the optimal parameter set based on cross-validation scores.

  • Returns a result object encapsulating the optimal estimator and score matrices.

Related

source
PortfolioOptimisers.make_p_grid Function
julia
make_p_grid(p, n_iter, rng, ...)

Build a randomised parameter grid from a specification.

Samples n_iter random parameter combinations from the provided distributions or ranges, for use in randomised search cross-validation.

Arguments

  • p: Parameter specification (pair, vector of pairs, or dictionary).

  • n_iter: Number of random parameter combinations to sample.

  • rng: Random number generator.

  • Additional arguments.

Returns

  • Vector of sampled parameter combinations.

Related

source