Julia: Implement more missing methods from JuMPSolver; test CPLEX

src/julia/Manifest.toml

@@ -26,6 +26,12 @@ git-tree-sha1 = "62847acab40e6855a9b5905ccb99c2b5cf6b3ebb"
 uuid = "fa961155-64e5-5f13-b03f-caf6b980ea82"
 version = "0.2.0"
 
+[[CPLEX]]
+deps = ["Libdl", "LinearAlgebra", "MathOptInterface", "MathProgBase", "SparseArrays"]
+git-tree-sha1 = "f8dac98fbff2f7d7fe58fa1fcdbefaaaf29a1f59"
+uuid = "a076750e-1247-5638-91d2-ce28b192dca0"
+version = "0.6.5"
+
 [[CPLEXW]]
 deps = ["CEnum", "Libdl"]
 git-tree-sha1 = "ebad297748ee2a12cc13b5fb07f9bbfa8a900494"

src/julia/Project.toml

@@ -4,6 +4,7 @@ authors = ["Alinson S Xavier <git@axavier.org>"]
 version = "0.1.0"
 
 [deps]
+CPLEX = "a076750e-1247-5638-91d2-ce28b192dca0"
 CPLEXW = "cfecb002-79c2-11e9-35be-cb59aa640f85"
 Gurobi = "2e9cd046-0924-5485-92f1-d5272153d98b"
 JuMP = "4076af6c-e467-56ae-b986-b466b2749572"

src/julia/src/MIPLearn.jl

@@ -6,7 +6,6 @@ __precompile__(false)
 module MIPLearn
 
 using JuMP
-using Gurobi
 using PyCall
 using MathOptInterface
 const MOI = MathOptInterface
@@ -17,26 +16,44 @@ LearningSolver = miplearn.LearningSolver
 InternalSolver = miplearn.solvers.internal.InternalSolver
 
 @pydef mutable struct JuMPSolver <: InternalSolver
+    function __init__(self; optimizer=CPLEX.Optimizer)
+        self.optimizer = optimizer
+    end
+
     function add_constraint(self, constraint)
+        @error "Not implemented"
     end
 
     function set_warm_start(self, solution)
+        for (varname, value) in solution
+            var = JuMP.variable_by_name(self.model, varname)
+            JuMP.set_start_value(var, value)
+        end
     end
 
     function clear_warm_start(self)
+        @error "Not implemented"
     end
 
     function fix(self, solution)
+        for (varname, value) in solution
+            var = JuMP.variable_by_name(self.model, varname)
+            JuMP.fix(var, value, force=true)
+        end
     end
 
     function set_instance(self, instance, model)
         self.instance = instance
         self.model = model
+        self.bin_vars = [var
+                         for var in JuMP.all_variables(self.model)
+                         if JuMP.is_binary(var)]
+        JuMP.set_optimizer(self.model, self.optimizer)
     end
 
     function solve(self; tee=false)
-        JuMP.set_optimizer(self.model, Gurobi.Optimizer)
         JuMP.optimize!(self.model)
+        self._update_solution()
 
         primal_bound = JuMP.objective_value(self.model)
         dual_bound = JuMP.objective_bound(self.model)
@@ -55,27 +72,54 @@ InternalSolver = miplearn.solvers.internal.InternalSolver
 
         return Dict("Lower bound" => lower_bound,
                     "Upper bound" => upper_bound,
-                    "Sense" => sense)
+                    "Sense" => sense,
+                    "Wallclock time" => JuMP.solve_time(self.model),
+                    "Nodes" => 1,
+                    "Log" => nothing,
+                    "Warm start value" => nothing)
     end
 
     function solve_lp(self; tee=false)
+        for var in self.bin_vars
+            JuMP.unset_binary(var)
+            JuMP.set_upper_bound(var, 1.0)
+            JuMP.set_lower_bound(var, 0.0)
+        end
+
+        JuMP.optimize!(self.model)
+        obj_value = JuMP.objective_value(self.model)
+        self._update_solution()
+
+        for var in self.bin_vars
+            JuMP.set_binary(var)
+        end
+
+        return Dict("Optimal value" => obj_value)
     end
 
     function get_solution(self)
-        return Dict(JuMP.name(var) => JuMP.value(var)
+        return self.solution
-                    for var in JuMP.all_variables(self.model))
+    end
+
+    function _update_solution(self)
+        self.solution =  Dict(JuMP.name(var) => JuMP.value(var)
+                              for var in JuMP.all_variables(self.model))
     end
 
     function set_gap_tolerance(self, gap_tolerance)
+        @error "Not implemented"
     end
 
     function set_node_limit(self)
+        @error "Not implemented"
     end
 
     function set_threads(self, threads)
+        @error "Not implemented"
     end
 
     function set_time_limit(self, time_limit)
+        JuMP.set_time_limit_sec(self.model, time_limit)
     end
 end
 

src/julia/test/jump_solver.jl

@@ -0,0 +1,58 @@
+#  MIPLearn: Extensible Framework for Learning-Enhanced Mixed-Integer Optimization
+#  Copyright (C) 2020, UChicago Argonne, LLC. All rights reserved.
+#  Released under the modified BSD license. See COPYING.md for more details.
+
+using Test
+using MIPLearn
+using CPLEX
+using Gurobi
+
+@testset "JuMPSolver" begin
+    for optimizer in [CPLEX.Optimizer, Gurobi.Optimizer]
+        instance = KnapsackInstance([23., 26., 20., 18.],
+                                    [505., 352., 458., 220.],
+                                    67.0)
+        model = instance.to_model()
+
+        solver = JuMPSolver(optimizer=optimizer)
+        solver.set_instance(instance, model)
+        solver.set_time_limit(30)
+        solver.set_warm_start(Dict(
+            "x[1]" => 1.0,
+            "x[2]" => 0.0,
+            "x[3]" => 0.0,
+            "x[4]" => 1.0,
+        ))
+        stats = solver.solve()
+
+        @test stats["Lower bound"] == 1183.0
+        @test stats["Upper bound"] == 1183.0
+        @test stats["Sense"] == "max"
+        @test stats["Wallclock time"] > 0
+
+        solution = solver.get_solution()
+        @test solution["x[1]"] == 1.0
+        @test solution["x[2]"] == 0.0
+        @test solution["x[3]"] == 1.0
+        @test solution["x[4]"] == 1.0
+
+        stats = solver.solve_lp()
+        @test round(stats["Optimal value"], digits=3) == 1287.923
+
+        solution = solver.get_solution()
+        @test round(solution["x[1]"], digits=3) == 1.000
+        @test round(solution["x[2]"], digits=3) == 0.923
+        @test round(solution["x[3]"], digits=3) == 1.000
+        @test round(solution["x[4]"], digits=3) == 0.000
+
+        solver.fix(Dict(
+            "x[1]" => 1.0,
+            "x[2]" => 0.0,
+            "x[3]" => 0.0,
+            "x[4]" => 1.0,
+        ))
+        stats = solver.solve()
+        @test stats["Lower bound"] == 725.0
+        @test stats["Upper bound"] == 725.0
+    end
+end

src/julia/test/runtests.jl

@@ -4,36 +4,9 @@
 
 using Test
 using MIPLearn
+using CPLEX
+using Gurobi
 
 @testset "MIPLearn" begin
-    instance = KnapsackInstance([23., 26., 20., 18.],
+    include("jump_solver.jl")
-                                [505., 352., 458., 220.],
-                                67.0)
-    model = instance.to_model()
-
-    solver = JuMPSolver()
-    solver.set_instance(instance, model)
-    stats = solver.solve()
-
-    # assert len(stats["Log"]) > 100
-    @test stats["Lower bound"] == 1183.0
-    @test stats["Upper bound"] == 1183.0
-    @test stats["Sense"] == "max"
-    # @test isinstance(stats["Wallclock time"], float)
-    # @test isinstance(stats["Nodes"], int)
-
-    solution = solver.get_solution()
-    @test solution["x[1]"] == 1.0
-    @test solution["x[2]"] == 0.0
-    @test solution["x[3]"] == 1.0
-    @test solution["x[4]"] == 1.0
-
-    # stats = solver.solve_lp()
-    # @test round(stats["Optimal value"], 3) == 1287.923
-    #
-    # solution = solver.get_solution()
-    # @test round(solution["x"][0], 3) == 1.000
-    # @test round(solution["x"][1], 3) == 0.923
-    # @test round(solution["x"][2], 3) == 1.000
-    # @test round(solution["x"][3], 3) == 0.000
 end