Prediction weights: private API

PortfolioOptimisers.fold_label — Function
fold_label(ts::Nothing, k::Integer)
fold_label(ts::AbstractVector, k::Integer)

The label of fold k of a path: the first timestamp of the fold, or "fold k" when the fold carries no timestamps.

Arguments

  • ts: The timestamps of the fold, or nothing.
  • k: The position of the fold in the path.

Returns

  • label::String: The label.

Related

source
PortfolioOptimisers.prediction_members — Function
prediction_members(pred::PredictionResult)
prediction_members(pred::MultiPeriodPredictionResult)
prediction_members(ppred::PopulationPredictionResult)

The number of members of a prediction: the number of portfolios that its optimiser gives at each fold.

An optimiser that gives one portfolio has one member. An optimiser that gives many portfolios, such as a frontier, holds one weight vector for each member. The folds of a path and the paths of a population come from one optimiser, so the first fold gives the count.

Arguments

  • pred, ppred: The prediction result.

Validation

  • A population holds at least one path. Otherwise the function raises an IsEmptyError.

Returns

  • n::Int: The number of members.

Related

source