Remove versions taking various argument orderings?

[emstoudenmire]

Currently there are some interfaces like the tdvp function which have wrappers that allow passing the arguments in a different order, for example:

tdvp(H,psi,t, ...)

instead of

tdvp(H,t,psi,...)

Before finalizing this package, I would suggest we decide on a standard ordering for interfaces like this and remove the other orderings.

[mtfishman]

One reason I like the second one is that it is more similar to the equation exp(Ht)|psi> (in fact we could consider calling tdvp something like exp or exponentiate, in a future where there are multiple MPO-based time evolution backends).

[emstoudenmire]

Agreed that the second interface is nicer. I think we should just have one of them (since this package hasn't been released yet) so I'm happy with that one being the interface.