improve docs (#146)

* make output of build_function tutorial more readable

* bump version

* mention AverageVectorFieldMethod in docs

* fix printing of benchmarks

* update commend on Requires.jl

* fix docstring of energy_preserving_order

Project.toml

@@ -1,7 +1,7 @@
 name = "BSeries"
 uuid = "ebb8d67c-85b4-416c-b05f-5f409e808f32"
 authors = ["Hendrik Ranocha <mail@ranocha.de> and contributors"]
-version = "0.1.54"
+version = "0.1.55"
 
 [deps]
 Combinatorics = "861a8166-3701-5b0c-9a16-15d98fcdc6aa"

docs/src/benchmark_python_orderconditions.py

@@ -40,7 +40,8 @@
   print("", end_time - start_time, "seconds", file=io)
 
 
-print("\nModifying integrator")
+with open(io_file, 'a') as io:
+  print("\nModifying integrator", file=io)
 
 start_time = time.time()
 BSeries.set_order(up_to_order)

docs/src/tutorials/bseries_creation.md

@@ -204,6 +204,14 @@ series = bseries(5) do t, series
 end
 ```
 
+[BSeries.jl](https://github.com/ranocha/BSeries.jl) also offers a
+convenience constructor using the type [`AverageVectorFieldMethod`](@ref)
+as follows.
+
+```@example ex:AVF
+series == bseries(AverageVectorFieldMethod(), 5)
+```
+
 We can check that this method is second-order accurate by comparing it to
 the B-series of the exact solution, truncated at the same order.
 

docs/src/tutorials/code_generation.md

@@ -39,7 +39,7 @@ Finally, we generate a C function that evaluates the expressions above.
 
 
 ```@example code-generation
-build_function(du, α, p, q, h, target = Symbolics.CTarget())
+println(build_function(du, α, p, q, h, target = Symbolics.CTarget()))
 ```
 
 The [Symbolics.jl](https://github.com/JuliaSymbolics/Symbolics.jl) function `build_function`

src/BSeries.jl

@@ -1394,7 +1394,8 @@ Internal function specialized on symbolic variables and expressions from
 - [SymPy.jl](https://github.com/JuliaPy/SymPy.jl), and
 - [Symbolics.jl](https://github.com/JuliaSymbolics/Symbolics.jl)
 
-if these packages are loaded (via Requires.jl).
+if these packages are loaded (via Requires.jl or weak dependencies on
+Julia v1.9 and newer).
 """
 function compute_derivative end
 
@@ -1546,12 +1547,12 @@ function renormalize!(series)
 end
 
 """
-    energy_preserving(rk::RungeKuttaMethod, max_order)
+    energy_preserving_order(rk::RungeKuttaMethod, max_order)
 
-This function checks up to which order a Runge-Kitta method 'rk' is
+This function checks up to which order a Runge-Kutta method `rk` is
 energy-preserving for Hamiltonian problems.
-It requires a 'max_order' so that it does not run forever if the order up to
-which the method is energy_preserving is too big or infinite.
+It requires a `max_order` so that it does not run forever if the order up to
+which the method is energy-preserving is too big or infinite.
 
 See also [`is_energy_preserving`](@ref)
 """