Optimization interface added.

[efposadac]

Added the file to implement the optimization interface and the test for optimization interface.

[luca-heltai]

You and @gpitton have just won issue #12. Please, make a feasibility study, together with @gpitton, in which you design an interface for this class, compatible with the optimisation utilities in numpy and scipy, that should be able to solve the following problem:

• Given a time varying shape shape in time (represented by a n x m matrix shape), assume that there are functions which return:
  • a total displacement pos = position(shape) (a vector of dimension maximum 6)
  • the energy consumption ene = energy(shape) (for the moment one number, but consider the possibility of having more than one)
  • all admissible shapes, in the form (S_min, S_max) = admissible_shapes()

Design your interface in a way that would allow the solution of the following problem:

shape = min_{S in all admissible shapes} ( E(S) ) such that position(S)=g_T

where g_T is a given target displacement

[gpitton]

Sorry, I chose Optimization randomly, just to learn git.
Can I switch to VectorSpace? It would be closer to my PhD work interests I think (n-widths, RB and so on).

[luca-heltai]

Sure, no problems! ;)