BlockSolverInterface module (dev)

set_constrs_decomposition!(s::AbstractMathProgSolver, data::Array)

sends to the solver s in which subproblems are the constraints. Each element of the data array is a Tuple containing (constr_name::Symbol, constr_id::Tuple, sp_type::Symbol, sp_id::Tuple). constr_name and constr_id are the name and the index of the constraint in the JuMP model. sp_type and sp_id are the type and the index of the subproblem to which the constraint is assigned.

set_vars_decomposition!(s::AbstractMathProgSolver, data::Array)

sends to the solver s in which subproblems are the variables. Each element of the data array is a Tuple containing (var_name::Symbol, var_id::Tuple, sp_type::Symbol, sp_id::Tuple). var_name and var_id are the name and the index of the variable in the JuMP model. sp_type and sp_id are the type and the index of the subproblem to which the variable is assigned.

set_oracles!(s::AbstractMathProgSolver, oracles::Array)

sends to the solver s the list with subproblems and oracles functions. Each element of the oracles array is a Tuple containing (sp_id::Tuple, sp_type::Symbol, oracle::Function) with oracle, the function defined by the user.

set_sp_mult!(s::AbstractMathProgSolver, multiplicities::Array)

sends to the solver s the multiplicity of each subproblem. Each element of the multiplicities array is a Tuple containing (sp_id::Tuple, sp_type::Symbol, mult_lb, mult_ub) with mult_lb the lower bound of the multiplicity and mult_ub the upper bound of the multiplicity.

set_sp_prio!(s::AbstractMathProgSolver, priorities::Array)

sends to the solver s the list of subproblem priorities. Each element of the data array is a Tuple containing (sp_id::Tuple, sp_type::Symbol, sp_prio).

set_var_branching_prio!(s::AbstractMathProgSolver, priorities::Array)

sends to the solver s the list of variables branching priorities. The number stored at the row i is the branching priority of the variable stored at the column i in the JuMP model.

set_branching_rules!(s::AbstractMathProgSolver, rules::Dict{Symbol, Any})

sends to the solver s a Dictionary rules. The key is the name of the branching rule and the content is an array of branching instances. The array contains Tuple of variables name and parameters. Parameters are stored in an array of tuple (name_of_parameter, value_of_parameter).

For instance,

rules = (:branching_rule_name => [(:x, [(:priority, 1)]), (:y, [(:priority, 1)])])

Names of branching rules depend on solvers.

set_objective_bounds_and_magnitude!(s::AbstractMathProgSolver, magn, lb, ub)

sends to the solver s the magnitude magn, the lower bound lb and the upper bound ub of the objective function.

getcurrentcost(m::AbstractMathProgModel, vcol::Integer)

returns the current cost of the vcol ${}^{th}$ variable.

getdisaggregatedvalueofvariable(m::AbstractMathProgModel, vcol::Integer)

returns the disaggregated value of the vcol ${}^{th}$ variable.

set_oraclesolution_solution(o::OracleSolverData, x::JuMP.Variable, v::Real)

Set the value of the variable x to v in the oracle solver solution stored in the OracleSolverData object o.

set_oraclesolution_objval(o::OracleSolverData, v::Real)

Sets the objective value stored in the OracleSolverData object o to v.

set_oraclesolution_newsolution(o::OracleSolverData)

creates a new solution in the oracle solver solution. It is usefull, if the user wants to return several solutions.

get_oracle_phaseofstageapproach(o::OracleSolverData)

Returns the phase of stage approach.