1. Add a function to create two versions of default 'cnModels'. First
version is all possibilities from 0 to 5 copies. Second version
imposes the strong prior that you can't both gain and lose the same
segment in different clones. 

2. Make the second version of 'cnModels' the default input to the
'seekClones' function. 

3. Add a default set of psi vectors as the input to 'seekClones'.

4. Make the 'pars' argument to 'seekClones' a proper object, and
supply a default value.

5. Make a function to display psi vectors (projected onto the
2-simplex, and flattened to the Euclidean plane). Think about allowing
the 3-simplex for visualization.

6. Make all the arguments to the 'generateTumorData' function into a
proper object with a friendly constructor that assigns sensible
default values.

7. Change the simulated copy number data to carry along the segment
locations and not just the "seg" indirect index.

8. Think about changing "getClones" to a bracket operator (like [i] or
[[i]] depending on what makes the most sense).

