Implement tau leaping

[jmason42]

The exact SSA from Gillespie isn't practical at scale. An approximate (i.e. tau leaping) approach could use the same interface (plus some additional simulation precision parameters).

[prismofeverything]

I would love to see this in action.... I had some concerns about tau leaping before reverting to the degenerate case (getting stuck in a state where it is just advancing at the smallest time step) but it would be great to see a more convincing demonstration.

[jmason42]

Tau leaping will really shine in cases where a system is well modeled as ODEs, but for one reason or another you still want discretization or stochasticity.

[eagmon]

@jmason42 -- you had a working example of tau leaping. Could that be brought in here?

[jmason42]

Unfortunately not; that implementation was very system-specific. There may be some adaptable elements but it will largely be a ground-up implementation.