enable strict doctesting

docs/Project.toml

@@ -14,6 +14,7 @@ OptimizationFlux = "253f991c-a7b2-45f8-8852-8b9a9df78a86"
 OptimizationOptimJL = "36348300-93cb-4f02-beb5-3c3902f8871e"
 OptimizationPolyalgorithms = "500b13db-7e66-49ce-bda4-eed966be6282"
 QuadGK = "1fd47b50-473d-5c70-9696-f719f8f3bcdc"
+ReverseDiff = "37e2e3b7-166d-5795-8a7a-e32c996b4267"
 Tracker = "9f7883ad-71c0-57eb-9f7f-b5c9e6d3789c"
 Zygote = "e88e6eb3-aa80-5325-afca-941959d7151f"
 

docs/make.jl

@@ -1,5 +1,9 @@
 using Documenter, DiffEqSensitivity
 
+# Make sure that plots don't throw a bunch of warnings / errors!
+ENV["GKSwstype"] = "100"
+using Plots
+
 include("pages.jl")
 
 makedocs(
@@ -9,6 +13,15 @@ makedocs(
     doctest = false,
     modules = [DiffEqSensitivity],
 
+    strict = [
+             :doctest,
+             :linkcheck,
+             :parse_error,
+             :example_block,
+             # Other available options are
+             # :autodocs_block, :cross_references, :docs_block, :eval_block, :example_block, :footnote, :meta_block, :missing_docs, :setup_block
+         ],
+
     format = Documenter.HTML(#analytics = "",
                              assets = ["assets/favicon.ico"],
                              canonical="https://sensitivity.sciml.ai/stable/"),

docs/src/ad_examples/adjoint_continuous_functional.md

@@ -81,9 +81,10 @@ end
 
 To get the adjoint sensitivities, we call:
 
-```julia
-res = adjoint_sensitivities(sol,Vern9(),g,nothing,dg,abstol=1e-8,
-                                 reltol=1e-8,iabstol=1e-8,ireltol=1e-8)
+```@example continuousadjoint
+prob = ODEProblem(f,[1.0;1.0],(0.0,10.0),p)
+sol = solve(prob,DP8())
+res = adjoint_sensitivities(sol,Vern9(),g,nothing,dg,abstol=1e-8,reltol=1e-8)
 ```
 
 Notice that we can check this against autodifferentiation and numerical

docs/src/ad_examples/direct_sensitivity.md

@@ -121,7 +121,7 @@ and computation time. We can check this almost exactly matches the
 autodifferentiation and numerical differentiation results:
 
 ```@example directsense
-using ForwardDiff,Calculus,Tracker
+using ForwardDiff,Calculus,ReverseDiff,Tracker
 function G(p)
   tmp_prob = remake(prob,u0=convert.(eltype(p),prob.u0),p=p)
   sol = solve(tmp_prob,Vern9(),abstol=1e-14,reltol=1e-14,saveat=ts,

docs/src/neural_ode/GPUs.md

@@ -126,6 +126,6 @@ optprob = Optimization.OptimizationProblem(optf, Lux.ComponentArray(p))
 
 result_neuralode = Optimization.solve(optfunc,
                                        ADAM(0.05), 
-                                       cb = callback, 
+                                       callback = callback, 
                                        maxiters = 300)
 ```

docs/src/neural_ode/neural_ode_flux.md

@@ -136,14 +136,14 @@ optprob = GalacticOptim.OptimizationProblem(optfunc, prob_neuralode.p)
 
 result_neuralode = GalacticOptim.solve(optprob,
                                        ADAM(0.05),
-                                       cb = callback,
+                                       callback = callback,
                                        maxiters = 300)
 
 optprob2 = remake(optprob,u0 = result_neuralode.u)
 
 result_neuralode2 = GalacticOptim.solve(optprob2,
                                         LBFGS(),
-                                        cb = callback,
+                                        callback = callback,
                                         allow_f_increases = false)
 ```
 

docs/src/neural_ode/neural_ode_galacticoptim.md

@@ -68,14 +68,14 @@ optprob = GalacticOptim.OptimizationProblem(optfunc, prob_neuralode.p)
 
 result_neuralode = GalacticOptim.solve(optprob,
                                        ADAM(0.05),
-                                       cb = callback,
+                                       callback = callback,
                                        maxiters = 300)
 
 optprob2 = remake(optprob,u0 = result_neuralode.u)
 
 result_neuralode2 = GalacticOptim.solve(optprob2,
                                         LBFGS(),
-                                        cb = callback,
+                                        callback = callback,
                                         allow_f_increases = false)
 ```
 
@@ -181,7 +181,7 @@ optprob = GalacticOptim.OptimizationProblem(optfunc, prob_neuralode.p)
 
 result_neuralode = GalacticOptim.solve(optprob,
                                        ADAM(0.05),
-                                       cb = callback,
+                                       callback = callback,
                                        maxiters = 300)
 # output
 * Status: success
@@ -205,7 +205,7 @@ optprob2 = remake(optprob,u0 = result_neuralode.u)
 
 result_neuralode2 = GalacticOptim.solve(optprob2,
                                         LBFGS(),
-                                        cb = callback,
+                                        callback = callback,
                                         allow_f_increases = false)
 # output
 * Status: success

docs/src/ode_fitting/optimization_ode.md

@@ -5,7 +5,7 @@
 If you want to just get things running, try the following! Explanation will
 follow.
 
-```@example optode
+```@example optode_cp
 using DifferentialEquations, Optimization, OptimizationPolyalgorithms, OptimizationOptimJL, Plots
 
 function lotka_volterra!(du, u, p, t)
@@ -54,9 +54,8 @@ optf = Optimization.OptimizationFunction((x,p)->loss(x), adtype)
 optprob = Optimization.OptimizationProblem(optf, p)
 
 result_ode = Optimization.solve(optprob, PolyOpt(),
-                                    cb = callback,
-                                    maxiters = 100)
-# result_ode = Optimization.solve(optprob, ADAM(0.1), cb = callback)
+                                callback = callback,
+                                maxiters = 100)
 ```
 
 ## Explanation
@@ -145,8 +144,8 @@ optf = Optimization.OptimizationFunction((x,p)->loss(x), adtype)
 optprob = Optimization.OptimizationProblem(optf, p)
 
 result_ode = Optimization.solve(optprob, PolyOpt(),
-                                    cb = callback,
-                                    maxiters = 100)
+                                callback = callback,
+                                maxiters = 100)
 ```
 
 In just seconds we found parameters which give a relative loss of `1e-16`! We can
@@ -162,13 +161,3 @@ plot(remade_solution, ylim = (0, 6))
 ```
 
 ![Final plot](https://user-images.githubusercontent.com/1814174/51399500-1f4dd080-1b14-11e9-8c9d-144f93b6eac2.gif)
-
-Note that this was done with the default optimizer. One can also pass an
-optimization method, like `ADAM(0.1)`, and tweak settings like set `maxiters=100`
-to force at most 100 iterations of the optimization. This looks like:
-
-```@example optode
-result_ode = Optimization.solve(optprob, ADAM(0.1),
-                                    cb = callback,
-                                    maxiters = 100)
-```

docs/src/optimal_control/feedback_control.md

@@ -89,5 +89,5 @@ adtype = Optimization.AutoZygote()
 optf = Optimization.OptimizationFunction((x,p)->loss_univ(x), adtype)
 optprob = Optimization.OptimizationProblem(optf, θ)
 result_univ = Optimization.solve(optprob, PolyOpt(),
-                                     cb = callback)
+                                     callback = callback)
 ```

docs/src/sde_fitting/neural_sde.md

@@ -175,7 +175,7 @@ adtype = Optimization.AutoZygote()
 optf = Optimization.OptimizationFunction((x,p) -> loss_neuralsde(x, n=10), adtype)
 optprob = Optimization.OptimizationProblem(optf, p)
 result1 = Optimization.solve(optprob, opt,
-                                 cb = callback, maxiters = 100)
+                                 callback = callback, maxiters = 100)
 ```
 
 We resume the training with a larger `n`. (WARNING - this step is a couple of
@@ -185,7 +185,7 @@ orders of magnitude longer than the previous one).
 optf2 = Optimization.OptimizationFunction((x,p) -> loss_neuralsde(x, n=100), adtype)
 optprob2 = Optimization.OptimizationProblem(optf2, result1.u)
 result2 = Optimization.solve(optprob2, opt,
-                                 cb = callback, maxiters = 100)
+                                 callback = callback, maxiters = 100)
 ```
 
 And now we plot the solution to an ensemble of the trained neural SDE:

docs/src/sde_fitting/optimization_sde.md

@@ -176,7 +176,7 @@ optf = Optimization.OptimizationFunction((x,p) -> loss_sde(x), adtype)
 
 optprob = Optimization.OptimizationProblem(optf, p)
 result_sde = Optimization.solve(optprob, ADAM(0.1),
-                                    cb = callback, maxiters = 100)
+                                    callback = callback, maxiters = 100)
 ```
 
 ![](https://user-images.githubusercontent.com/1814174/51399524-2c6abf80-1b14-11e9-96ae-0192f7debd03.gif)

docs/src/training_tips/local_minima.md

@@ -84,7 +84,7 @@ optf = Optimization.OptimizationFunction((x,p) -> loss_neuralode(x), adtype)
 
 optprob = Optimization.OptimizationProblem(optf, Lux.ComponentArray(pinit))
 result_neuralode = Optimization.solve(optprob,
-                                      ADAM(0.05), cb = callback,
+                                      ADAM(0.05), callback = callback,
                                       maxiters = 300)
 
 callback(result_neuralode.u,loss_neuralode(result_neuralode.u)...;doplot=true)
@@ -110,7 +110,7 @@ optf = Optimization.OptimizationFunction((x,p) -> loss_neuralode(x), adtype)
 
 optprob = Optimization.OptimizationProblem(optf, ComponentArray(pinit))
 result_neuralode2 = Optimization.solve(optprob,
-                                      ADAM(0.05), cb = callback,
+                                      ADAM(0.05), callback = callback,
                                       maxiters = 300)
 
 callback(result_neuralode2.u,loss_neuralode(result_neuralode2.u)...;doplot=true)
@@ -128,7 +128,7 @@ prob_neuralode = NeuralODE(dudt2, (0.0,3.0), Tsit5(), saveat = tsteps[tsteps .<=
 optprob = Optimization.OptimizationProblem(optf, result_neuralode.u)
 result_neuralode3 = Optimization.solve(optprob,
                                         ADAM(0.05), maxiters = 300,
-                                        cb = callback)
+                                        callback = callback)
 callback(result_neuralode3.u,loss_neuralode(result_neuralode3.u)...;doplot=true)
 savefig("shortplot2.png")
 ```
@@ -143,7 +143,7 @@ prob_neuralode = NeuralODE(dudt2, (0.0,5.0), Tsit5(), saveat = tsteps)
 optprob = Optimization.OptimizationProblem(optf, result_neuralode3.u)
 result_neuralode4 = Optimization.solve(optprob,
                                       ADAM(0.01), maxiters = 300,
-                                      cb = callback)
+                                      callback = callback)
 callback(result_neuralode4.u,loss_neuralode(result_neuralode4.u)...;doplot=true)
 savefig("fullplot.png")
 ```