Fix docs build

[JonasKoziorek]

One of the algorithms stopped converging for examples in the docs. I changed the initial guesses and precision. This must have resulted from some changes in other libraries, probably NonlinearSolve.

[JonasKoziorek]

Strange, I pass the tests locally.

[Datseris]

Any way you can fix the tests? change tolerances or something? Docs wont build if tests fail.

Thanks for the PR!

[JonasKoziorek]

I will play around with parameters and see if it passes the tests.

[JonasKoziorek]

Any tips on how to debug why the tests run locally and not here?

[Datseris]

Can you tell me what is going wrong with the test? What is it supposed to be testing and what it finds instead?

[JonasKoziorek]

PeriodicOrbits.jl/test/algorithms/continuous_time/optimized_shooting.jl

Lines 29 to 30 in 11063cb

	res = periodic_orbits(ds, alg, igs)
	@test length(res) == 2

Given two initial guesses, the algorithm finds an orbit only for one of the guesses.
Maybe after the update of NonlinearSolve the SciML solver doesn't converge to the same result for the given initial guess. (11063cb#diff-72ed386c2a0cd1d23c0968297e70023ed98c22490d146dd89fc91f48369bad4dL18)

[Datseris]

Can you just check out the branch, run the initial conditions and see if you get the orbit you should, otherwise change the guess?

[Datseris]

Or, significantly increase the starting guess period for the second orbit.