Inferring the Gravitational Potential of the Milky Way

We present a simple numerical algorithm to infer the gravitational potential of a galaxy from the observed positions and velocities of stars. The method is novel in that it works directly in Cartesian phase space and thus does not require any assumptions about the integrability of the Hamiltonian, as some other methods do. We have tested the algorithm on a two-dimensional logarithmic potential with good results and hope to be able to extend the algorithm to infer the Milky Way's potential from the full Gaia dataset.