Product of a Rotor with a Vector of Rotor(s)

[empet]

Thank you for Quaternionic.jl. It is almost perfect for my needs in developing HopfFibrations.jl.
There is a small issue: the product between a Rotor{T} and a Vector{Rotor{T}} is a Vector{Quaternion{T}}, instead of a Vector{Rotor{T}}:

julia> p = Rotor(3,-1,2, 1.2)
rotor(0.7634798978389314 - 0.2544932992796438𝐢 + 0.5089865985592876𝐣 + 0.30539195913557254𝐤)

julia> vr = [Rotor(-2,1,1,3), Rotor(2,0,-1,3)]
2-element Vector{RotorF64}:
 rotor(-0.5163977794943222 + 0.2581988897471611𝐢 + 0.2581988897471611𝐣 + 0.7745966692414834𝐤)
                 rotor(0.5345224838248488 + 0.0𝐢 - 0.2672612419124244𝐣 + 0.8017837257372732𝐤)

julia> p*vr
2-element Vector{QuaternionF64}:
 -0.6965248056142178 + 0.6439568957565411𝐢 + 0.2102716394307073𝐣 + 0.23655559435954573𝐤
  0.29927125898501883 + 0.3536842151641131𝐢 + 0.27206478089547165𝐣 + 0.843400820775962𝐤

julia> vr*p
2-element Vector{QuaternionF64}:
 -0.6965248056142179 + 0.013141977464419141𝐢 - 0.34169141407489934𝐣 + 0.630814918292122𝐤
   0.29927125898501883 - 0.6257489960595848𝐢 - 0.1360323904477358𝐣 + 0.7073684303282262𝐤

Taking into account that operations involved in computing Hopf fibres and Hopf tori are such multiplications it would be nice to get the respective products as Vector(s) of Rotor(s), according to the theory.

[moble]

Thanks for reporting this. The problem was buried as deep in the guts of Julia as I've ever gone, but turned out to be fairly straightforward to fix. It will be a breaking change, so if you happened to be using Base.eltype on any quaternionic type, you'll need to replace it with Quaternionic.basetype.