Merge pull request #653 from ArnoStrouwen/links

strict cross references and links

Author: ChrisRackauckas

docs/make.jl

@@ -27,14 +27,18 @@ makedocs(modules = [
              "http://www.radford.edu/~thompson/vodef90web/problems/demosnodislin/Demos_Pitagora/DemoHires/demohires.pdf",
              "https://www.radford.edu/%7Ethompson/RP/nonnegative.pdf",
              "http://www.radford.edu/~thompson/vodef90web/problems/demosnodislin/Demos_Pitagora/DemoOrego/demoorego.pdf",
+             "https://zenodo.org/record/5883749#.Yg-d698xmu4",
+             "https://www.sciencedirect.com/science/article/abs/pii/S0375960109009591",
+             "https://www.sciencedirect.com/science/article/abs/pii/0375960176901018",
          ],
          strict = [
              :doctest,
              :linkcheck,
              :parse_error,
              :example_block,
+             :cross_references,
              # Other available options are
-             # :autodocs_block, :cross_references, :docs_block, :eval_block, :example_block, :footnote, :meta_block, :missing_docs, :setup_block
+             # :autodocs_block, :docs_block, :eval_block, :example_block, :footnote, :meta_block, :missing_docs, :setup_block
          ],
          doctest = false, clean = true,
          format = Documenter.HTML(analytics = "UA-90474609-3",

docs/src/basics/faq.md

@@ -443,7 +443,7 @@ like BigFloats or [ArbFloats.jl](https://github.com/JuliaArbTypes/ArbFloats.jl).
 #### Native Julia solvers compatibility with autodifferentiation
 
 Yes, they are compatible with automatic differentiation! Take a look at the
-[sensitivity analysis](@ref sensitivity) page for more details.
+[sensitivity analysis](https://docs.sciml.ai/SciMLSensitivity/stable/) page for more details.
 
 If the algorithm does not have differentiation of parameter-dependent events,
 then you simply need to make the initial condition have elements of Dual numbers.

docs/src/solvers/sde_solve.md

@@ -43,7 +43,7 @@ a given SDE, the use of high weak order solvers is recommended. Specifically,
 Runge-Kutta solvers with weak order 2 due to Roessler are adaptive. All other
 high weak order solvers currently require a fixed step size.
 
-## Special Noise Forms
+## [Special Noise Forms](@id special_noise_forms)
 
 Some solvers are for specialized forms of noise. Diagonal noise is the default
 setup. Non-diagonal noise is specified via setting `noise_rate_prototype` to
@@ -69,7 +69,7 @@ I(h) = \int_0^h\int_0^sdW^1_tdW^2_s
 in these schemes.
 
 The approximation of these iterated integrals can be avoided, if the diffusion matrix
-satisfies the special commutativity condition given [above](@ref special noise forms).
+satisfies the special commutativity condition given [above](@ref special_noise_forms).
 Because of this, many methods are only applicable to problems that satisfy the commutativity
 condition. In other words, many methods can only handle specific noise cases, like
 diagonal noise or commutative noise, because of how this iterated integral approximation

docs/src/tutorials/sde_example.md

@@ -350,4 +350,4 @@ SDEProblem(f, g, ones(2), tspan, noise = heston_noise)
 Of course, to fully define this problem, we need to define our constants. Constructors
 for making common models like this easier to define can be found in the modeling
 toolkits. For example, the `HestonProblem` is pre-defined as part of the
-[financial modeling tools](@ref financial_models).
+[financial modeling tools](https://github.com/SciML/DiffEqFinancial.jl).