formatting

JuliaNonconvex · Nov 15, 2023 · 842becd · 842becd
1 parent a8ecbc6
commit 842becd
Show file tree

Hide file tree

Showing 2 changed files with 75 additions and 56 deletions.
diff --git a/src/NonconvexTOBS.jl b/src/NonconvexTOBS.jl
@@ -15,17 +15,26 @@ struct TOBSOptions
     nt::NamedTuple # one of the fields of the nt is milp_options
 end
 
-function TOBSOptions(
-    ;movelimit::Real = 0.1, # move limit parameter
+function TOBSOptions(;
+    movelimit::Real = 0.1, # move limit parameter
     pastN::Int = 20, # number of past iterations for moving average calculation
     convParam::Real = 0.001, # convergence parameter (upper bound of error)
     constrRelax::Real = 0.1, # constraint relaxation parameter
     timeLimit::Real = 1.0,
     optimizer = Cbc.Optimizer,
     maxiter::Int = 200,
-    timeStable::Bool = true
+    timeStable::Bool = true,
 )
-    return TOBSOptions((; movelimit, pastN, convParam, constrRelax, timeLimit, optimizer, maxiter, timeStable))
+    return TOBSOptions((;
+        movelimit,
+        pastN,
+        convParam,
+        constrRelax,
+        timeLimit,
+        optimizer,
+        maxiter,
+        timeStable,
+    ))
 end
 
 @params mutable struct TOBSWorkspace <: Workspace
@@ -34,20 +43,29 @@ end
     options::TOBSOptions
 end
 function TOBSWorkspace(
-    model::VecModel, x0::AbstractVector = NonconvexCore.getinit(model);
-    options = TOBSOptions(), kwargs...,
+    model::VecModel,
+    x0::AbstractVector = NonconvexCore.getinit(model);
+    options = TOBSOptions(),
+    kwargs...,
 )
     return TOBSWorkspace(model, copy(x0), options)
 end
 @params struct TOBSResult <: AbstractResult
-    minimizer
-    minimum
-    error
+    minimizer::Any
+    minimum::Any
+    error::Any
 end
 
 function optimize!(workspace::TOBSWorkspace)
     @unpack model, x0, options = workspace
-    @unpack movelimit, pastN, convParam, constrRelax, timeLimit, optimizer, maxiter, timeStable = options.nt
+    @unpack movelimit,
+    pastN,
+    convParam,
+    constrRelax,
+    timeLimit,
+    optimizer,
+    maxiter,
+    timeStable = options.nt
     milp_solver = JuMP.optimizer_with_attributes(optimizer)
     numVars = length(NonconvexCore.getinit(model))
     count = 1 # iteration counter
@@ -70,7 +88,8 @@ function optimize!(workspace::TOBSWorkspace)
         JuMP.set_optimizer_attribute(m, "seconds", timeLimit)
         if !skip_step || count == 1
             if count > 1
-                currentConstr, jacConstr = NonconvexCore.value_jacobian(model.ineq_constraints, x)
+                currentConstr, jacConstr =
+                    NonconvexCore.value_jacobian(model.ineq_constraints, x)
             end
             violation = norm(currentConstr)
             if (violation <= best_sol[4] - 1e-8 || violation < 1e-8 && objval < best_sol[2])
@@ -88,7 +107,7 @@ function optimize!(workspace::TOBSWorkspace)
         JuMP.@constraint(m, deltaX .<= absdeltaX)
         JuMP.@constraint(m, .-deltaX .<= absdeltaX)
         # Constrain amount of change per iteration
-        JuMP.@constraint(m, sum(absdeltaX) <= movelimit*numVars)
+        JuMP.@constraint(m, sum(absdeltaX) <= movelimit * numVars)
         # Constraint relaxation
         Δ = map(currentConstr) do c
             abs(c) < constrRelax ? -c : -constrRelax * c
@@ -122,12 +141,12 @@ function optimize!(workspace::TOBSWorkspace)
             objval, objgrad = NonconvexCore.value_gradient(getobjective(model), x)
             # Apply "time stabilization"
             if timeStable
-                objgrad = (objgrad + pastGrad)/2
+                objgrad = (objgrad + pastGrad) / 2
                 pastGrad = copy(objgrad)
             end
             objHist[end] = objval
             if count > pastN
-                er = abs(sum([objHist[i] - objHist[i - 1] for i in 2:pastN])) / sum(objHist)
+                er = abs(sum([objHist[i] - objHist[i-1] for i = 2:pastN])) / sum(objHist)
             end
             @info "iter = $count, obj = $(round.(objHist[end]; digits=3)), constr_vio_norm = $(round(norm(currentConstr), digits=3)), er = $(round(er, digits=3))"
         end
@@ -136,8 +155,8 @@ function optimize!(workspace::TOBSWorkspace)
     return TOBSResult(best_sol[1], best_sol[2], er)
 end
 
-function Workspace(model::VecModel, optimizer::TOBSAlg, x0::AbstractVector; kwargs...,)
+function Workspace(model::VecModel, optimizer::TOBSAlg, x0::AbstractVector; kwargs...)
     return TOBSWorkspace(model, x0; kwargs...)
 end
 
-end
+end
diff --git a/test/runtests.jl b/test/runtests.jl
@@ -2,44 +2,44 @@ using NonconvexTOBS, TopOpt, Test
 
 @testset "Example" begin
 
-  # Recreating cantilever problem from original paper
-
-  E = 1.0 # Young’s modulus
-  v = 0.3 # Poisson’s ratio
-  f = 1.0 # downward force
-  rmin = 6.0 # filter radius
-  xmin = 0.001 # minimum density
-  V = 0.5 # maximum volume fraction
-  p = 3.0 # topological optimization penalty
-
-  # Define FEA problem
-  problem_size = (160, 100) # size of rectangular mesh
-  x0 = fill(1.0, prod(problem_size)) # initial design
-  problem = PointLoadCantilever(Val{:Linear}, problem_size, (1.0, 1.0), E, v, f)
-
-  # define FEA solver and auxiliary functions
-  solver = FEASolver(Direct, problem; xmin=xmin)
-  cheqfilter = DensityFilter(solver; rmin=rmin) # filter function
-  comp = TopOpt.Compliance(problem, solver) # compliance function
-
-  obj(x) = comp(cheqfilter(x)) # compliance objective
-  constr(x) = sum(cheqfilter(x)) / length(x) - V # volume fraction constraint
-
-  # Optimization setup
-  using NonconvexUtils
-  tobj = TraceFunction(obj, on_grad = true)
-  m = Model(tobj) # create optimization model
-  addvar!(m, zeros(length(x0)), ones(length(x0))) # setup optimization variables
-  Nonconvex.add_ineq_constraint!(m, constr) # setup volume inequality constraint
-  options = TOBSOptions() # optimization options with default values
-  TopOpt.setpenalty!(solver, p)
-
-  # Perform TOBS optimization
-  @time r = Nonconvex.optimize(m, TOBSAlg(), x0; options=options)
-
-  # Results
-  @show obj(r.minimizer)
-  @show constr(r.minimizer)
-  topology = r.minimizer
-  
+    # Recreating cantilever problem from original paper
+
+    E = 1.0 # Young’s modulus
+    v = 0.3 # Poisson’s ratio
+    f = 1.0 # downward force
+    rmin = 6.0 # filter radius
+    xmin = 0.001 # minimum density
+    V = 0.5 # maximum volume fraction
+    p = 3.0 # topological optimization penalty
+
+    # Define FEA problem
+    problem_size = (160, 100) # size of rectangular mesh
+    x0 = fill(1.0, prod(problem_size)) # initial design
+    problem = PointLoadCantilever(Val{:Linear}, problem_size, (1.0, 1.0), E, v, f)
+
+    # define FEA solver and auxiliary functions
+    solver = FEASolver(Direct, problem; xmin = xmin)
+    cheqfilter = DensityFilter(solver; rmin = rmin) # filter function
+    comp = TopOpt.Compliance(problem, solver) # compliance function
+
+    obj(x) = comp(cheqfilter(x)) # compliance objective
+    constr(x) = sum(cheqfilter(x)) / length(x) - V # volume fraction constraint
+
+    # Optimization setup
+    using NonconvexUtils
+    tobj = TraceFunction(obj, on_grad = true)
+    m = Model(tobj) # create optimization model
+    addvar!(m, zeros(length(x0)), ones(length(x0))) # setup optimization variables
+    Nonconvex.add_ineq_constraint!(m, constr) # setup volume inequality constraint
+    options = TOBSOptions() # optimization options with default values
+    TopOpt.setpenalty!(solver, p)
+
+    # Perform TOBS optimization
+    @time r = Nonconvex.optimize(m, TOBSAlg(), x0; options = options)
+
+    # Results
+    @show obj(r.minimizer)
+    @show constr(r.minimizer)
+    topology = r.minimizer
+
 end