Analysing surveys of our Galaxy I: basic astrometric data

We consider what is the best way to extract science from large surveys of the Milky Way galaxy. The diversity of data gathered in these surveys, together with our position within the Galaxy, imply that science must be extracted by fitting dynamical models to the data in the space of the observables. Models based on orbital tori promise to be superior for this task than traditional types of models, such as N-body models and Schwarzschild models. A formalism that allows such models to be fitted to data is developed and tested on pseudodata of varying richness.