Allow continuous solvers to solve relaxations of MIPs

src/solvers.jl

@@ -26,8 +26,8 @@ type ProblemTraits
     sos::Bool  # has an SOS constraint
     conic::Bool  # has an SDP or SOC constraint
 end
-function ProblemTraits(m::Model)
-    int = any(c-> !(c == :Cont || c == :Fixed), m.colCat)
+function ProblemTraits(m::Model; relaxation=false)
+    int = !relaxation && any(c-> !(c == :Cont || c == :Fixed), m.colCat)
     qp = !isempty(m.obj.qvars1)
     qc = !isempty(m.quadconstr)
     nlp = m.nlpdata !== nothing
@@ -128,10 +128,10 @@ function solve(m::Model; suppress_warnings=false,
     unset = m.solver == UnsetSolver()
 
     # Analyze the problems traits to determine what solvers we can use
-    traits = ProblemTraits(m)
+    traits = ProblemTraits(m, relaxation=relaxation)
 
     # Build the MathProgBase model from the JuMP model
-    build(m, traits, suppress_warnings=suppress_warnings, relaxation=relaxation)
+    build(m, traits=traits, suppress_warnings=suppress_warnings, relaxation=relaxation)
 
     # If the model is a general nonlinear, use different logic in
     # nlp.jl to solve the problem
@@ -303,8 +303,7 @@ end
 
 # Converts the JuMP Model into a MathProgBase model based on the
 # traits of the model
-function build(m::Model, traits=ProblemTraits(m);
-               suppress_warnings=false, relaxation=false)
+function build(m::Model; suppress_warnings=false, relaxation=false, traits=ProblemTraits(m,relaxation=relaxation))
     # Set solver based on the model's traits if it hasn't provided
     if isa(m.solver, UnsetSolver)
         m.solver = default_solver(traits)

test/model.jl

@@ -890,3 +890,24 @@ facts("[model] Unrecognized keyword argument to solve") do
     m = Model()
     @fact_throws solve(m, this_should_throw=true)
 end
+
+facts("[model] Solve MIP relaxation with continuous solvers") do
+for solver in lp_solvers
+context("With solver $(typeof(solver))") do
+    m = Model(solver=solver)
+    @variable(m, x, Bin)
+    @constraint(m, x >= 0.5)
+    @objective(m, Min, x)
+    @fact solve(m, relaxation=true) --> :Optimal
+    @fact getvalue(x) --> roughly(0.5, TOL)
+end; end
+for solver in nlp_solvers
+context("With solver $(typeof(solver))") do
+    m = Model(solver=solver)
+    @variable(m, x, Bin)
+    @objective(m, Min, x)
+    @NLconstraint(m, x >= 0.5)
+    @fact solve(m, relaxation=true) --> :Optimal
+    @fact getvalue(x) --> roughly(0.5, TOL)
+end; end
+end