Reference

Contents

Index

OptimizationProblems.metaConstant

OptimizationProblems.meta A composite type that represents the main features of the optimization problem. optimize obj(x) subject to lvar ≤ x ≤ uvar lcon ≤ cons(x) ≤ ucon –- The following keys are valid:

  • nvar::Int: number of variables
  • variable_nvar::Bool: true if we can modify the number of variables
  • ncon::Int: number of general constraints
  • variable_ncon::Bool: true if we can modify the number of constraints
  • minimize::Bool: true if optimize == minimize
  • name::String: problem name
  • has_equalities_only::Bool: true if the problem has constraints, and all are equality constraints (doesn't include bounds)
  • has_inequalities_only::Bool: true if the problem has constraints, and all are inequality constraints (doesn't include bounds)
  • has_bounds::Bool: true if the problem has bound constraints
  • has_fixed_variables::Bool: true if it has fixed variables
  • objtype::Symbol: type of objective, in [:none, :constant, :linear, :quadratic, :sumofsquares, :other]
  • contype::Symbol: type of constraint, in [:unconstrained, :linear, :quadratic, :general]
  • best_known_lower_bound::Real: lower bound on the global optimal value (default: -Inf for minimization problem, f(x0) for maximization problem if x0 is feasible, -Inf otherwise)
  • best_known_upper_bound::Real: upper bound on the global optimal value (default: Inf for maximization problem, f(x0) for minimization problem if x0 is feasible, Inf otherwise)
  • is_feasible::Union{Bool, Missing}: true if problem is feasible
  • origin::Symbol: origin of the problem, in [:academic, :modelling, :real, :unknown]
source