Merge pull request #511 from sathvikbhagavan/sb/opt

ChrisRackauckas · web-flow · commit 867dd7ab5db5 · 2025-03-13T00:15:40.000-01:00
refactor: add ! in surrogate_optimize as it is mutating
diff --git a/docs/src/InverseDistance.md b/docs/src/InverseDistance.md
@@ -51,10 +51,10 @@ plot!(InverseDistance, label = "Surrogate function",
 
 Having built a surrogate, we can now use it to search for minima in our original function `f`.
 
-To optimize using our surrogate we call `surrogate_optimize` method. We choose to use Stochastic RBF as the optimization technique and again Sobol sampling as the sampling technique.
+To optimize using our surrogate we call `surrogate_optimize!` method. We choose to use Stochastic RBF as the optimization technique and again Sobol sampling as the sampling technique.
 
 ```@example Inverse_Distance1D
-surrogate_optimize(
+surrogate_optimize!(
     f, SRBF(), lower_bound, upper_bound, InverseDistance, SobolSample())
 scatter(x, y, label = "Sampled points", legend = :top)
 plot!(f, label = "True function", xlims = (lower_bound, upper_bound), legend = :top)
@@ -132,7 +132,7 @@ size(xys)
 ```
 
 ```@example Inverse_DistanceND
-surrogate_optimize(schaffer, SRBF(), lower_bound, upper_bound,
+surrogate_optimize!(schaffer, SRBF(), lower_bound, upper_bound,
     InverseDistance, SobolSample(), maxiters = 10)
 ```
 
diff --git a/docs/src/LinearSurrogate.md b/docs/src/LinearSurrogate.md
@@ -50,10 +50,10 @@ plot!(my_linear_surr_1D, label = "Surrogate function", xlims = (lower_bound, upp
 
 Having built a surrogate, we can now use it to search for minima in our original function `f`.
 
-To optimize using our surrogate we call `surrogate_optimize` method. We choose to use Stochastic RBF as the optimization technique and again Sobol sampling as the sampling technique.
+To optimize using our surrogate we call `surrogate_optimize!` method. We choose to use Stochastic RBF as the optimization technique and again Sobol sampling as the sampling technique.
 
 ```@example linear_surrogate1D
-surrogate_optimize(
+surrogate_optimize!(
     f, SRBF(), lower_bound, upper_bound, my_linear_surr_1D, SobolSample())
 scatter(x, y, label = "Sampled points")
 plot!(f, label = "True function", xlims = (lower_bound, upper_bound))
@@ -130,7 +130,7 @@ size(xys)
 ```
 
 ```@example linear_surrogateND
-surrogate_optimize(
+surrogate_optimize!(
     egg, SRBF(), lower_bound, upper_bound, my_linear_ND, SobolSample(), maxiters = 10)
 ```
 
diff --git a/docs/src/abstractgps.md b/docs/src/abstractgps.md
@@ -56,7 +56,7 @@ xs = lower_bound:0.1:upper_bound
 x = sample(n_samples, lower_bound, upper_bound, SobolSample())
 y = f.(x)
 gp_surrogate = AbstractGPSurrogate(x, y)
-surrogate_optimize(f, SRBF(), lower_bound, upper_bound, gp_surrogate, SobolSample())
+surrogate_optimize!(f, SRBF(), lower_bound, upper_bound, gp_surrogate, SobolSample())
 ```
 
 Plotting the function and the sampled points:
diff --git a/docs/src/ackley.md b/docs/src/ackley.md
@@ -55,7 +55,7 @@ The fit looks good. Let's now see if we are able to find the minimum value using
 optimization methods:
 
 ```@example ackley
-surrogate_optimize(ackley, DYCORS(), lb, ub, my_rad, RandomSample())
+surrogate_optimize!(ackley, DYCORS(), lb, ub, my_rad, RandomSample())
 scatter(x, y, label = "Sampled points", xlims = (lb, ub), ylims = (0, 30), legend = :top)
 plot!(xs, ackley.(xs), label = "True function", legend = :top)
 plot!(xs, my_rad.(xs), label = "Radial basis optimized", legend = :top)
diff --git a/docs/src/gekpls.md b/docs/src/gekpls.md
@@ -78,7 +78,7 @@ x = sample(n, lb, ub, SobolSample())
 grads = gradient.(sphere_function, x)
 y = sphere_function.(x)
 g = GEKPLS(x, y, grads, n_comp, delta_x, lb, ub, extra_points, initial_theta)
-x_point, minima = surrogate_optimize(sphere_function, SRBF(), lb, ub, g,
+x_point, minima = surrogate_optimize!(sphere_function, SRBF(), lb, ub, g,
     RandomSample(); maxiters = 20,
     num_new_samples = 20, needs_gradient = true)
 minima
diff --git a/docs/src/index.md b/docs/src/index.md
@@ -114,7 +114,7 @@ my_lobachevsky = LobachevskySurrogate(x, y, lb, ub, alpha = alpha, n = n)
 value = my_lobachevsky(5.0)
 
 #Adding more data points
-surrogate_optimize(f, SRBF(), lb, ub, my_lobachevsky, RandomSample())
+surrogate_optimize!(f, SRBF(), lb, ub, my_lobachevsky, RandomSample())
 
 #New approximation
 value = my_lobachevsky(5.0)
diff --git a/docs/src/kriging.md b/docs/src/kriging.md
@@ -52,10 +52,10 @@ plot!(xs, kriging_surrogate.(xs), label = "Surrogate function",
 
 Having built a surrogate, we can now use it to search for minima in our original function `f`.
 
-To optimize using our surrogate, we call `surrogate_optimize` method. We choose to use Stochastic RBF as the optimization technique and again Sobol sampling as the sampling technique.
+To optimize using our surrogate, we call `surrogate_optimize!` method. We choose to use Stochastic RBF as the optimization technique and again Sobol sampling as the sampling technique.
 
 ```@example kriging_tutorial1d
-surrogate_optimize(
+surrogate_optimize!(
     f, SRBF(), lower_bound, upper_bound, kriging_surrogate, SobolSample())
 
 scatter(x, y, label = "Sampled points", ylims = (-7, 7), legend = :top)
@@ -139,7 +139,7 @@ size(xys)
 ```
 
 ```@example kriging_tutorialnd
-surrogate_optimize(branin, SRBF(), lower_bound, upper_bound, kriging_surrogate,
+surrogate_optimize!(branin, SRBF(), lower_bound, upper_bound, kriging_surrogate,
     SobolSample(); maxiters = 100, num_new_samples = 10)
 ```
 
diff --git a/docs/src/lobachevsky.md b/docs/src/lobachevsky.md
@@ -48,10 +48,10 @@ plot!(
 
 Having built a surrogate, we can now use it to search for minima in our original function `f`.
 
-To optimize using our surrogate we call `surrogate_optimize` method. We choose to use Stochastic RBF as the optimization technique and again Sobol sampling as the sampling technique.
+To optimize using our surrogate we call `surrogate_optimize!` method. We choose to use Stochastic RBF as the optimization technique and again Sobol sampling as the sampling technique.
 
 ```@example LobachevskySurrogate_tutorial
-surrogate_optimize(
+surrogate_optimize!(
     f, SRBF(), lower_bound, upper_bound, lobachevsky_surrogate, SobolSample())
 scatter(x, y, label = "Sampled points")
 plot!(f, label = "True function", xlims = (lower_bound, upper_bound))
@@ -132,7 +132,7 @@ size(Lobachevsky.x)
 ```
 
 ```@example LobachevskySurrogate_ND
-surrogate_optimize(schaffer, SRBF(), lower_bound, upper_bound, Lobachevsky,
+surrogate_optimize!(schaffer, SRBF(), lower_bound, upper_bound, Lobachevsky,
     SobolSample(), maxiters = 1, num_new_samples = 10)
 ```
 
diff --git a/docs/src/multi_objective_opt.md b/docs/src/multi_objective_opt.md
@@ -12,7 +12,7 @@ ub = 10.0
 x = sample(50, lb, ub, GoldenSample())
 y = f.(x)
 my_radial_basis_ego = RadialBasis(x, y, lb, ub)
-pareto_set, pareto_front = surrogate_optimize(
+pareto_set, pareto_front = surrogate_optimize!(
     f, SMB(), lb, ub, my_radial_basis_ego, SobolSample(); maxiters = 10, n_new_look = 100)
 
 m = 5
@@ -27,7 +27,7 @@ K = 2
 p_cross = 0.5
 n_c = 1.0
 sigma = 1.5
-surrogate_optimize(
+surrogate_optimize!(
     f, RTEA(Z, K, p_cross, n_c, sigma), lb, ub, my_radial_basis_rtea, SobolSample())
 ```
 
@@ -43,7 +43,7 @@ ub = [3.5, 0.5]
 x = sample(50, lb, ub, SobolSample())
 y = f.(x)
 my_radial_basis_ego = RadialBasis(x, y, lb, ub)
-#I can find my pareto set and pareto front by calling again the surrogate_optimize function:
-pareto_set, pareto_front = surrogate_optimize(
+#I can find my pareto set and pareto front by calling again the surrogate_optimize! function:
+pareto_set, pareto_front = surrogate_optimize!(
     f, SMB(), lb, ub, my_radial_basis_ego, SobolSample(); maxiters = 10, n_new_look = 100);
 ```
diff --git a/docs/src/neural.md b/docs/src/neural.md
@@ -77,6 +77,6 @@ plot(p1, p2, title = "Surrogate")
 We can now call an optimization function on the neural network:
 
 ```@example Neural_surrogate
-surrogate_optimize(schaffer, SRBF(), lower_bound, upper_bound, neural,
+surrogate_optimize!(schaffer, SRBF(), lower_bound, upper_bound, neural,
     SobolSample(), maxiters = 20, num_new_samples = 10)
 ```
diff --git a/docs/src/optimizations.md b/docs/src/optimizations.md
@@ -3,31 +3,31 @@
   - SRBF
 
 ```@docs
-surrogate_optimize(obj::Function,::SRBF,lb,ub,surr::AbstractSurrogate,sample_type::SamplingAlgorithm;maxiters=100,num_new_samples=100)
+surrogate_optimize!(obj::Function,::SRBF,lb,ub,surr::AbstractSurrogate,sample_type::SamplingAlgorithm;maxiters=100,num_new_samples=100)
 ```
 
   - LCBS
 
 ```@docs
-surrogate_optimize(obj::Function,::LCBS,lb,ub,krig,sample_type::SamplingAlgorithm;maxiters=100,num_new_samples=100)
+surrogate_optimize!(obj::Function,::LCBS,lb,ub,krig,sample_type::SamplingAlgorithm;maxiters=100,num_new_samples=100)
 ```
 
   - EI
 
 ```@docs
-surrogate_optimize(obj::Function,::EI,lb,ub,krig,sample_type::SamplingAlgorithm;maxiters=100,num_new_samples=100)
+surrogate_optimize!(obj::Function,::EI,lb,ub,krig,sample_type::SamplingAlgorithm;maxiters=100,num_new_samples=100)
 ```
 
   - DYCORS
 
 ```@docs
-surrogate_optimize(obj::Function,::DYCORS,lb,ub,surrn::AbstractSurrogate,sample_type::SamplingAlgorithm;maxiters=100,num_new_samples=100)
+surrogate_optimize!(obj::Function,::DYCORS,lb,ub,surrn::AbstractSurrogate,sample_type::SamplingAlgorithm;maxiters=100,num_new_samples=100)
 ```
 
   - SOP
 
 ```@docs
-surrogate_optimize(obj::Function,sop1::SOP,lb::Number,ub::Number,surrSOP::AbstractSurrogate,sample_type::SamplingAlgorithm;maxiters=100,num_new_samples=min(500*1,5000))
+surrogate_optimize!(obj::Function,sop1::SOP,lb::Number,ub::Number,surrSOP::AbstractSurrogate,sample_type::SamplingAlgorithm;maxiters=100,num_new_samples=min(500*1,5000))
 ```
 
 ## Adding another optimization method
@@ -36,5 +36,5 @@ To add another optimization method, you just need to define a new
 SurrogateOptimizationAlgorithm and write its corresponding algorithm, overloading the following:
 
 ```
-surrogate_optimize(obj::Function,::NewOptimizationType,lb,ub,surr::AbstractSurrogate,sample_type::SamplingAlgorithm;maxiters=100,num_new_samples=100)
+surrogate_optimize!(obj::Function,::NewOptimizationType,lb,ub,surr::AbstractSurrogate,sample_type::SamplingAlgorithm;maxiters=100,num_new_samples=100)
 ```
diff --git a/docs/src/parallel.md b/docs/src/parallel.md
@@ -4,7 +4,7 @@ There are some situations where it can be beneficial to run multiple optimizatio
 
 ## Ask-Tell Interface
 
-To enable parallel optimization, we make use of an Ask-Tell interface. The user will construct the initial surrogate model the same way as for non-parallel surrogate models, but instead of using `surrogate_optimize`, the user will use `potential_optimal_points`. This will return the coordinates of points that the optimizer has determined are most useful to evaluate next. How the user evaluates these points is up to them. The Ask-Tell interface requires more manual control than `surrogate_optimize`, but it allows for more flexibility. After the point has been evaluated, the user will *tell* the surrogate model the new points with the `update!` function.
+To enable parallel optimization, we make use of an Ask-Tell interface. The user will construct the initial surrogate model the same way as for non-parallel surrogate models, but instead of using `surrogate_optimize!`, the user will use `potential_optimal_points`. This will return the coordinates of points that the optimizer has determined are most useful to evaluate next. How the user evaluates these points is up to them. The Ask-Tell interface requires more manual control than `surrogate_optimize!`, but it allows for more flexibility. After the point has been evaluated, the user will *tell* the surrogate model the new points with the `update!` function.
 
 ## Virtual Points
 
diff --git a/docs/src/radials.md b/docs/src/radials.md
@@ -60,10 +60,10 @@ plot!(radial_surrogate, label = "Surrogate function",
 
 Having built a surrogate, we can now use it to search for minima in our original function `f`.
 
-To optimize using our surrogate, we call `surrogate_optimize` method. We choose to use Stochastic RBF as the optimization technique and again Sobol sampling as the sampling technique.
+To optimize using our surrogate, we call `surrogate_optimize!` method. We choose to use Stochastic RBF as the optimization technique and again Sobol sampling as the sampling technique.
 
 ```@example RadialBasisSurrogate
-surrogate_optimize(
+surrogate_optimize!(
     f, SRBF(), lower_bound, upper_bound, radial_surrogate, SobolSample())
 scatter(x, y, label = "Sampled points", legend = :top)
 plot!(f, label = "True function", xlims = (lower_bound, upper_bound), legend = :top)
@@ -145,7 +145,7 @@ size(xys)
 ```
 
 ```@example RadialBasisSurrogateND
-surrogate_optimize(
+surrogate_optimize!(
     booth, SRBF(), lower_bound, upper_bound, radial_basis, RandomSample(), maxiters = 50)
 ```
 
diff --git a/docs/src/randomforest.md b/docs/src/randomforest.md
@@ -52,10 +52,10 @@ plot!(randomforest_surrogate, label = "Surrogate function",
 
 Having built a surrogate, we can now use it to search for minima in our original function `f`.
 
-To optimize using our surrogate, we call `surrogate_optimize` method. We choose to use Stochastic RBF as the optimization technique and again Sobol sampling as the sampling technique.
+To optimize using our surrogate, we call `surrogate_optimize!` method. We choose to use Stochastic RBF as the optimization technique and again Sobol sampling as the sampling technique.
 
 ```@example RandomForestSurrogate_tutorial
-surrogate_optimize(
+surrogate_optimize!(
     f, SRBF(), lower_bound, upper_bound, randomforest_surrogate, SobolSample())
 scatter(x, y, label = "Sampled points")
 plot!(f, label = "True function", xlims = (lower_bound, upper_bound), legend = :top)
@@ -133,7 +133,7 @@ size(xys)
 ```
 
 ```@example RandomForestSurrogateND
-surrogate_optimize(
+surrogate_optimize!(
     bukin6, SRBF(), lower_bound, upper_bound, RandomForest, SobolSample(), maxiters = 20)
 ```
 
diff --git a/docs/src/secondorderpoly.md b/docs/src/secondorderpoly.md
@@ -37,7 +37,7 @@ plot!(sec, label = "Surrogate function", xlims = (lb, ub))
 ## Optimizing
 
 ```@example second_order_tut
-surrogate_optimize(f, SRBF(), lb, ub, sec, SobolSample())
+surrogate_optimize!(f, SRBF(), lb, ub, sec, SobolSample())
 scatter(x, y, label = "Sampled points")
 plot!(f, label = "True function", xlims = (lb, ub))
 plot!(sec, label = "Surrogate function", xlims = (lb, ub))
diff --git a/docs/src/tutorials.md b/docs/src/tutorials.md
@@ -84,7 +84,7 @@ std_err = std_error_at_point(my_krig, 5.4)
 Let's now optimize the Kriging surrogate using the lower confidence bound method. This is just a one-liner:
 
 ```@example kriging
-surrogate_optimize(
+surrogate_optimize!(
     f, LCBS(), lb, ub, my_krig, RandomSample(); maxiters = 10, num_new_samples = 10)
 ```
 
diff --git a/lib/SurrogatesFlux/test/runtests.jl b/lib/SurrogatesFlux/test/runtests.jl
@@ -72,7 +72,7 @@ using SafeTestsets
         model = Chain(Dense(2, 1), first)
         opt = Descent(0.01)
         my_neural_ND_neural = NeuralSurrogate(x, y, lb, ub, model = model, loss = Flux.mse)
-        surrogate_optimize(objective_function_ND, SRBF(), lb, ub, my_neural_ND_neural,
+        surrogate_optimize!(objective_function_ND, SRBF(), lb, ub, my_neural_ND_neural,
             SobolSample(), maxiters = 15)
     end
 
diff --git a/src/Optimization.jl b/src/Optimization.jl
diff --git a/src/Surrogates.jl b/src/Surrogates.jl
diff --git a/test/GEKPLS.jl b/test/GEKPLS.jl
diff --git a/test/SectionSampleTests.jl b/test/SectionSampleTests.jl
diff --git a/test/optimization.jl b/test/optimization.jl
