solve_exp_ode does not ever update position

[r0uv3n]

solve_exp_ode seems to always return the starting point p for any input. See e.g.

using Manifolds
using OrdinaryDiffEq
E = Euclidean(3)
p = [0, 0, 0] # point in E
A = get_default_atlas(E)
i = Manifolds.get_chart_index(E, A, p)
B = Manifolds.induced_basis(E, A, i, TangentSpaceType())
X = [0, 0, 1] # tangent vector
time = 1

Manifolds.solve_exp_ode(E, p, X, time; basis=B)

which returns [0, 0, 0] (i.e. just the starting point p).

[r0uv3n]

This seems to be due to a small typo in

Manifolds.jl/ext/ManifoldsOrdinaryDiffEqExt.jl

Line 64 in 4a6ca7f

q = sol.u[1][(n + 1):(2 * n)]

where the first element instead of the last element of the solution u is accessed.

[kellertuer]

If you see how this can be fixed, feel free to also open a PR for this. Otherwise, probably Mateusz will step in, he I´might just be a bit busy currently, maybe.

[mateuszbaran]

Right, sorry for the mistake, I will fix that this weekend.