SDP Constraints

A member of RatioFormulation implements the verb below. The private page documents the other names of this topic.

PortfolioOptimisers.sdp_objective_lift_flag — Function
sdp_objective_lift_flag(model::JuMP.Model, ratio_alg::RatioFormulation,
                        settings::RiskMeasureSettings)

Return true when a lifted variance states its objective term on the lift with a constant corner, set_sdp_objective_lift!.

The term needs that lift only when it is a positive term of the objective of a MaximumRatio: ratio_alg is SquareRootRatio, the head registered sdp_corner with set_sdp_lift_corner!, settings.rke is true and settings.scale > 0. Under every other objective $k$ is the constant $1$, and the two lifts are one matrix. A bound needs degree one, so it reads the lift with the corner $k$. A term with a negative scale is maximised, and only a bound holds a lifted matrix from above. LinearRatio gives false.

Arguments

  • model::JuMP.Model: The JuMP optimisation model.
  • ratio_alg: The ratio_alg of the measure.
  • settings: The settings of the measure.

Returns

  • flag::Bool.

Related

source