fix: set max time span to 1e16 for CoupledSDEs

[oameye]

For stochastic integrators, we can't have the timespan to be infinite because otherwise eps(Inf) is called in StochasticDiffEq. Instead, I have set it too 1e16. eps(Inf) is only called for non-adaptive timestep solvers.

https://github.com/SciML/StochasticDiffEq.jl/blob/9a28c57060980924d1fe040686f49c579ed9a1e2/src/integrators/integrator_utils.jl#L164-L167

At the moment, running trajectory with a non-adaptive time step solver fails:

using StatsBase: cov
diffeq_cov = (alg=EM(), abstol=1e-2, reltol=1e-2, dt=0.1)

Γ = [1.0 0.3; 0.3 1]
f(u, p, t) = [0.0, 0.0]
ds = CoupledSDEs(f, zeros(2), (); covariance=Γ, diffeq=diffeq_cov)
tr, _ = trajectory(ds, 1_000, Δt=0.1)

[1] (::Base.Math.var"#throw1#10")(x::Float64)
    @ Base.Math .\math.jl:1006
  [2] exponent
    @ .\math.jl:1009 [inlined]
  [3] eps
    @ .\float.jl:966 [inlined]
  [4] loopfooter!
    @ C:\Users\User\.julia\packages\StochasticDiffEq\NuzKA\src\integrators\integrator_utils.jl:167 [inlined]
  [5] step!
    @ C:\Users\User\.julia\packages\StochasticDiffEq\NuzKA\src\iterator_interface.jl:12 [inlined]
  [6] step!(::CoupledSDEs{false, 2, StochasticDiffEq.SDEIntegrator{…}, Tuple{}})
    @ StochasticSystemsBase C:\Users\User\.julia\packages\DynamicalSystemsBase\PJv07\ext\src\CoupledSDEs.jl:171
  [7] trajectory_continuous(ds::CoupledSDEs{…}, T::Int64; Dt::Float64, Δt::Float64, Ttr::Float64, accessor::Nothing, kw::@Kwargs{})
    @ DynamicalSystemsBase C:\Users\User\.julia\packages\DynamicalSystemsBase\PJv07\src\core\trajectory.jl:85
  [8] trajectory_continuous
    @ C:\Users\User\.julia\packages\DynamicalSystemsBase\PJv07\src\core\trajectory.jl:75 [inlined]
  [9] #trajectory#5
    @ C:\Users\User\.julia\packages\DynamicalSystemsBase\PJv07\src\core\trajectory.jl:44 [inlined]
 [10] trajectory
    @ C:\Users\User\.julia\packages\DynamicalSystemsBase\PJv07\src\core\trajectory.jl:36 [inlined]
 [11] trajectory(ds::CoupledSDEs{false, 2, StochasticDiffEq.SDEIntegrator{…}, Tuple{}}, T::Int64)
    @ DynamicalSystemsBase C:\Users\User\.julia\packages\DynamicalSystemsBase\PJv07\src\core\trajectory.jl:36
 [12] top-level scope
    @ c:\Users\User\Documents\Github\CriticalTransitions.jl\test\using CriticalTransitions.jl:9

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 86.48%. Comparing base (6c98239) to head (8a78ba1).
Report is 35 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #230      +/-   ##
==========================================
+ Coverage   82.00%   86.48%   +4.47%     
==========================================
  Files          15       17       +2     
  Lines         717      932     +215     
==========================================
+ Hits          588      806     +218     
+ Misses        129      126       -3     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[Datseris]

So the value should be 1e16 or 1e11. Does this value matter at all?

[oameye]

For some reason if I take the value to be larger then 1e11 the tests which tests trajectory fail. I am not sure why. Seems to be something internal with the integrator.

[Datseris]

okay, bump the patch version please so that i can tag a new release