merge

docs/solvers.rst

@@ -53,3 +53,8 @@ If we wish to increase the maximum number of iterations for ECOS or SCS, we can
 	solve!(p, ECOSSolver(maxit=10000))
 	using SCS
 	solve!(p, SCSSolver(max_iters=10000))
+
+To turn off the problem status warning issued by Convex when a solver is not able to solve a problem to optimality, use the keyword argument `verbose=false` of the solve method, along with any desired solver parameters:
+::
+
+	solve!(p, SCSSolver(verbose=false), verbose=false)

src/solution.jl

@@ -14,7 +14,9 @@ function solve!(problem::Problem,
 end
 
 function solve!(problem::Problem;
-                warmstart=false, check_vexity=true)
+                warmstart=false, 
+                check_vexity=true,
+                verbose=true)
 
   if check_vexity
     vex = vexity(problem)
@@ -34,7 +36,7 @@ function solve!(problem::Problem;
   # populate the status, the primal (and possibly dual) solution
   # and the primal (and possibly dual) variables with values
   populate_solution!(m, problem, var_to_ranges, conic_constraints)
-  if !(problem.status==:Optimal)
+  if !(problem.status==:Optimal) && verbose
     warn("Problem status $(problem.status); solution may be inaccurate.")
   end
 
@@ -95,14 +97,30 @@ function populate_solution!(m::MathProgBase.AbstractConicModel,
                         problem::Problem,
                         var_to_ranges,
                         conic_constraints)
-  try
-    dual = MathProgBase.getdual(m)
-    problem.solution = Solution(MathProgBase.getsolution(m), dual,
-                                MathProgBase.status(m), MathProgBase.getobjval(m))
+  dual = try 
+    MathProgBase.getdual(m)
   catch
-    problem.solution = Solution(MathProgBase.getsolution(m),
-                                MathProgBase.status(m), MathProgBase.getobjval(m))
+    fill(NaN, MathProgBase.numconstr(m))
   end
+
+  solution = try 
+    MathProgBase.getsolution(m)
+  catch
+    fill(NaN, MathProgBase.numvar(m))
+  end
+
+  objective = try
+    MathProgBase.getobjval(m)
+  catch
+    NaN
+  end
+
+  if any(isnan(dual))
+    problem.solution = Solution(solution, MathProgBase.status(m), objective)
+  else
+    problem.solution = Solution(solution, dual, MathProgBase.status(m), objective)
+  end
+
   populate_variables!(problem, var_to_ranges)
 
   if problem.solution.has_dual