Add more tests and robust file check (#127)

* Add more tests and robust file check

* loose

Author: tmigot

src/NLPModelsIpopt.jl

@@ -239,35 +239,45 @@ function SolverCore.solve!(
   real_time = time()
   status = IpoptSolve(problem)
   real_time = time() - real_time
-  ipopt_output = readlines(ipopt_log_file)
-
-  Δt = 0.0
-  dual_feas = primal_feas = Inf
-  iter = -1
-  for line in ipopt_output
-    if occursin("Total seconds", line)
-      Δt += Meta.parse(split(line, "=")[2])
-    elseif occursin("Dual infeasibility", line)
-      dual_feas = Meta.parse(split(line)[4])
-    elseif occursin("Constraint violation", line)
-      primal_feas = Meta.parse(split(line)[4])
-    elseif occursin("Number of Iterations....", line)
-      iter = Meta.parse(split(line)[4])
-    end
-  end
 
   set_status!(stats, get(ipopt_statuses, status, :unknown))
   set_solution!(stats, problem.x)
   set_objective!(stats, problem.obj_val)
-  set_residuals!(stats, primal_feas, dual_feas)
-  set_iter!(stats, iter)
-  set_time!(stats, Δt)
   set_constraint_multipliers!(stats, problem.mult_g)
   if has_bounds(nlp)
     set_bounds_multipliers!(stats, problem.mult_x_L, problem.mult_x_U)
   end
   set_solver_specific!(stats, :internal_msg, Ipopt._STATUS_CODES[status])
   set_solver_specific!(stats, :real_time, real_time)
+
+  try
+    ipopt_output = readlines(ipopt_log_file)
+
+    Δt = 0.0
+    dual_feas = primal_feas = Inf
+    iter = -1
+    for line in ipopt_output
+      if occursin("Total seconds", line)
+        Δt += Meta.parse(split(line, "=")[2])
+      elseif occursin("Dual infeasibility", line)
+        dual_feas = Meta.parse(split(line)[4])
+      elseif occursin("Constraint violation", line)
+        primal_feas = Meta.parse(split(line)[4])
+      elseif occursin("Number of Iterations....", line)
+        iter = Meta.parse(split(line)[4])
+      end
+    end
+    set_residuals!(stats, primal_feas, dual_feas)
+    set_iter!(stats, iter)
+    set_time!(stats, Δt)
+  catch e
+    @warn("could not parse Ipopt log file. $e")
+    stats.primal_residual_reliable = false  
+    stats.dual_residual_reliable = false
+    stats.iter_reliable = false
+    stats.time_reliable = false
+  end
+
   stats
 end
 

test/runtests.jl

@@ -7,32 +7,50 @@ using ADNLPModels, NLPModelsIpopt, NLPModels, Ipopt, SolverCore, Test
   stats = solve!(solver, nlp, stats, print_level = 0)
   @test isapprox(stats.solution, [1.0; 1.0], rtol = 1e-6)
   @test stats.status == :first_order
+  @test stats.iter == 21
+  @test stats.elapsed_time > 0
+  @test stats.primal_feas ≈ 0.0
+  @test stats.dual_feas ≈ 0.0 atol = 1.49e-8
+
   nlp = ADNLPModel(x -> (x[1])^2 + 100 * (x[2] - x[1]^2)^2, [-1.2; 1.0])
   reset!(solver, nlp)
   stats = solve!(solver, nlp, stats, print_level = 0)
   @test isapprox(stats.solution, [0.0; 0.0], atol = 1e-6)
   @test stats.status == :first_order
+  @test stats.iter == 16
+  @test stats.elapsed_time > 0
+  @test stats.primal_feas ≈ 0.0
+  @test stats.dual_feas ≈ 0.0 atol = 1.49e-8
 end
 
-function tests()
+@testset "Unit tests NLPModelsIpopt" begin
   nlp = ADNLPModel(x -> (x[1] - 1)^2 + 100 * (x[2] - x[1]^2)^2, [-1.2; 1.0])
   stats = ipopt(nlp, print_level = 0)
   @test isapprox(stats.solution, [1.0; 1.0], rtol = 1e-6)
   @test stats.status == :first_order
   @test stats.elapsed_time > 0
+  @test stats.iter == 21
+  @test stats.primal_feas ≈ 0.0
+  @test stats.dual_feas ≈ 0.0 atol = 1.49e-8
 
   nlp = ADNLPModel(x -> (x[1] - 1)^2 + 100 * (x[2] - x[1]^2)^2, [-1.2; 1.0])
   stats = ipopt(nlp, tol = 1e-12, print_level = 0)
   @test isapprox(stats.solution, [1.0; 1.0], rtol = 1e-6)
   @test stats.status == :first_order
   @test stats.elapsed_time > 0
+  @test stats.iter == 22
+  @test stats.primal_feas ≈ 0.0
+  @test stats.dual_feas ≈ 0.0
 
   # solve again from solution
   x0 = copy(stats.solution)
   stats = ipopt(nlp, x0 = x0, tol = 1e-12, print_level = 0)
   @test isapprox(stats.solution, x0, rtol = 1e-6)
   @test stats.status == :first_order
   @test stats.iter == 0
+  @test stats.elapsed_time >= 0
+  @test stats.primal_feas ≈ 0.0
+  @test stats.dual_feas ≈ 0.0
 
   function callback(alg_mod, iter_count, args...)
     return iter_count < 1
@@ -43,6 +61,8 @@ function tests()
   @test first(stats.solver_specific[:internal_msg]) == :User_Requested_Stop
   @test stats.iter == 1
   @test stats.elapsed_time > 0
+  @test stats.primal_feas ≈ 0.0
+  # @test stats.dual_feas ≈ 4.63
 
   nlp =
     ADNLPModel(x -> (x[1] - 1)^2 + 4 * (x[2] - 3)^2, zeros(2), x -> [sum(x) - 1.0], [0.0], [0.0])
@@ -51,6 +71,8 @@ function tests()
   @test stats.iter == 1
   @test stats.status == :first_order
   @test stats.elapsed_time > 0
+  @test stats.primal_feas ≈ 0.0
+  @test stats.dual_feas ≈ 0.0 atol = 1.49e-8
 
   # solve constrained problem again from solution
   x0 = copy(stats.solution)
@@ -65,8 +87,11 @@ function tests()
     @test isapprox(stats.multipliers_L, zL, rtol = 1e-6)
     @test isapprox(stats.multipliers_U, zU, rtol = 1e-6)
   end
+  @test stats.elapsed_time >= 0
   @test stats.iter == 0
   @test stats.status == :first_order
+  @test stats.primal_feas ≈ 0.0
+  @test stats.dual_feas ≈ 0.0 atol = 1.49e-8
 
   x0, f = rand(1), x -> x[1]
   nlp = ADNLPModel(f, x0, zeros(1), ones(1), minimize = false)
@@ -78,6 +103,7 @@ function tests()
   @test isapprox(stats.multipliers_U, -ones(1), rtol = 1e-6)
   @test stats.status == :first_order
   @test stats.elapsed_time > 0
+  @test stats.iter == 5
+  @test stats.primal_feas ≈ 0.0
+  @test stats.dual_feas ≈ 0.0 atol = 1.49e-8
 end
-
-tests()