auxiliary walkers

[Lucas Wagner]

I'm currently removing the concept of auxiliary walkers from the code.
The problem with the concept is that if you sample a primary walker
and then try to evaluate the energy difference, then you end up with a
nonsystematic sampling error. Rmc_corr_method.cpp fixes this problem
by using multiple trialfunc sections.

So I would suggest that you not do anything with auxiliary walkers.

Cheers,

Lucas