Restore the docs to their full potential

Author: ptiede

docs/make.jl

@@ -1,6 +1,6 @@
 # THis is because we are on a headless system
 # see https://github.com/jheinen/GR.jl/issues/510
-ENV["GKS_WSTYPE"]="nul"
+# ENV["GKS_WSTYPE"]="nul"
 
 
 using Documenter

examples/geometric_modeling.jl

@@ -134,7 +134,7 @@ prob = Optimization.OptimizationProblem(f, randn(rng, ndim), nothing, lb=fill(-5
 
 # Now we solve for our optimial image.
 
-sol = solve(prob, BBO_adaptive_de_rand_1_bin_radiuslimited(); maxiters=50_000)
+sol = solve(prob, BBO_adaptive_de_rand_1_bin_radiuslimited(); maxiters=50_000);
 
 # The sol vector is in the transformed space, so first we need to transform back to parameter space
 # to that we can interpret the solution.

examples/hybrid_imaging.jl

@@ -161,7 +161,7 @@ using OptimizationBBO
 using Zygote
 f = OptimizationFunction(tpost, Optimization.AutoZygote())
 prob = Optimization.OptimizationProblem(f, prior_sample(rng, tpost), nothing, lb=fill(-5.0, ndim), ub=fill(5.0,ndim))
-sol = solve(prob, BBO_adaptive_de_rand_1_bin_radiuslimited(); maxiters=100_000)
+sol = solve(prob, BBO_adaptive_de_rand_1_bin_radiuslimited(); maxiters=100_000);
 
 # Alright now we can zoom to the peak!
 using OptimizationOptimJL

examples/imaging_closures.jl

@@ -107,7 +107,7 @@ using ComradeOptimization
 using OptimizationBBO
 f = OptimizationFunction(tpost, Optimization.AutoForwardDiff())
 prob = Optimization.OptimizationProblem(f, prior_sample(tpost), nothing, lb=fill(-5.0, ndim), ub=fill(5.0,ndim))
-sol = solve(prob, BBO_adaptive_de_rand_1_bin_radiuslimited(); maxiters=100_000)
+sol = solve(prob, BBO_adaptive_de_rand_1_bin_radiuslimited(); maxiters=100_000);
 
 
 # Before we analyze our solution we first need to transform back to parameter space.

examples/imaging_pol.jl

@@ -306,7 +306,7 @@ using Zygote
 f = OptimizationFunction(tpost, Optimization.AutoZygote())
 ℓ = logdensityof(tpost)
 prob = Optimization.OptimizationProblem(f, prior_sample(rng, tpost), nothing)
-sol = solve(prob, LBFGS(), maxiters=15_000, callback=((x,p)->(@info ℓ(x);false)), g_tol=1e-1)
+sol = solve(prob, LBFGS(), maxiters=15_000, callback=((x,p)->(@info ℓ(x);false)), g_tol=1e-1);
 
 # !!! warning
 #     Fitting polarized images is generally much harder than Stokes I imaging. This difficulty means that

examples/imaging_vis.jl

@@ -203,7 +203,7 @@ using OptimizationOptimJL
 f = OptimizationFunction(tpost, Optimization.AutoZygote())
 prob = Optimization.OptimizationProblem(f, 2*rand(rng, ndim) .- 1.0, nothing)
 ℓ = logdensityof(tpost)
-sol = solve(prob, LBFGS(), maxiters=5_000, g_tol=1e-1, callback=((x,p)->(@info f(x,p); false)))
+sol = solve(prob, LBFGS(), maxiters=5_000, g_tol=1e-1, callback=((x,p)->(@info f(x,p); false)));
 
 # Now transform back to parameter space
 xopt = transform(tpost, sol.u)