fixup! test: clean up NeuralAdapter tests and use res.u instead of re…

…s.minimizer
SciML · Mar 4, 2024 · abd8c87 · abd8c87
1 parent 08bd81a
commit abd8c87
Show file tree

Hide file tree

Showing 2 changed files with 8 additions and 10 deletions.
diff --git a/lib/NeuralPDELogging/test/adaptive_loss_log_tests.jl b/lib/NeuralPDELogging/test/adaptive_loss_log_tests.jl
@@ -88,7 +88,7 @@ function test_2d_poisson_equation_adaptive_loss(adaptive_loss, run, outdir, hasl
     res = Optimization.solve(prob, OptimizationOptimisers.Adam(0.03); maxiters = maxiters,
                              callback = callback)
 
-    u_predict = reshape([first(phi([x, y], res.minimizer)) for x in xs for y in ys],
+    u_predict = reshape([first(phi([x, y], res.u)) for x in xs for y in ys],
                         (length(xs), length(ys)))
     diff_u = abs.(u_predict .- u_real)
     total_diff = sum(diff_u)

diff --git a/test/neural_adapter_tests.jl b/test/neural_adapter_tests.jl
@@ -164,11 +164,9 @@ end
     for i in 1:count_decomp
         println("decomposition $i")
         domains_ = domains_map[i]
-        phi_in(cord) = phis[i - 1](cord, reses[i - 1].minimizer)
-        # phi_bound(x,y) = if (x isa Matrix)  phi_in(vcat(x, fill(y,size(x)))) else  phi_in(vcat(fill(x,size(y)),y)) end
+        phi_in(cord) = phis[i - 1](cord, reses[i - 1].u)
         phi_bound(x, y) = phi_in(vcat(x, y))
         @register_symbolic phi_bound(x, y)
-        global phi_bound
         Base.Broadcast.broadcasted(::typeof(phi_bound), x, y) = phi_bound(x, y)
         bcs_ = create_bcs(domains_[1].domain, phi_bound)
         @named pde_system_ = PDESystem(eq, bcs_, domains_, [x, y], [u(x, y)])
@@ -177,7 +175,7 @@ end
         discretization = PhysicsInformedNN(chains[i], strategy;  init_params = init_params[i])
         prob = discretize(pde_system_, discretization)
         @time res_ = Optimization.solve(prob, OptimizationOptimisers.Adam(5e-3), maxiters = 10000)
-        @show res_.minimum
+        @show res_.objective
         phi = discretization.phi
         push!(reses, res_)
         push!(phis, phi)
@@ -198,7 +196,7 @@ end
         end
         for x_ in xs
             i = index_of_interval(x_)
-            u_predict_sub = [first(phis[i]([x_, y], reses[i].minimizer)) for y in ys]
+            u_predict_sub = [first(phis[i]([x_, y], reses[i].u)) for y in ys]
             u_real_sub = [analytic_sol_func(x_, y) for y in ys]
             diff_u_sub = u_predict_sub .- u_real_sub
             append!(u_predict_array, u_predict_sub)
@@ -235,15 +233,15 @@ end
     prob_ = NeuralPDE.neural_adapter(losses, init_params2, pde_system_map,
                                     GridTraining([0.1 / count_decomp, 0.1]))
     @time res_ = solve(prob_, OptimizationOptimisers.Adam(5e-3); maxiters = 5000)
-    @show res_.minimum
-    prob_ = NeuralPDE.neural_adapter(losses, res_.minimizer, pde_system_map,
+    @show res_.objective
+    prob_ = NeuralPDE.neural_adapter(losses, res_.u, pde_system_map,
                                     GridTraining(0.01))
     @time res_ = solve(prob_, OptimizationOptimisers.Adam(5e-3); maxiters = 5000)
-    @show res_.minimum
+    @show res_.objective
 
     phi_ = NeuralPDE.Phi(chain2)
     xs, ys = [infimum(d.domain):dx:supremum(d.domain) for d in domains]
-    u_predict_ = reshape([first(phi_([x, y], res_.minimizer)) for x in xs for y in ys],
+    u_predict_ = reshape([first(phi_([x, y], res_.u)) for x in xs for y in ys],
                         (length(xs), length(ys)))
     u_real = reshape([analytic_sol_func(x, y) for x in xs for y in ys],
                     (length(xs), length(ys)))