[Wien] About mixing again

[Stefaan Cottenier]

>       I am not very clear about mixing.
> Does the mixing influence the converged results?
> Or, does different mixing parameters can induce diffrent converged
results?
> For example,
> if we set the converging condition as :ENE=0.0001Ry
> then we use diffrent mixing parameters.
> such as 0.1 or 0.01,
> then these two setting can get the same results?

If a stable solutions exists (which is true in most cases), then mixing will
only determine how fast you will reach that solution (=how many iterations
you need). Small mixing factor = many iterations, larger mixing factor =
less iterations, too large mixing factor = divergence, you will not go to a
converged result.

If you have two or more competing solutions, then the value of the mixing
factor might be one of the factors that determine which solution you will
find. But again, this does not apply to regular cases.

Stefaan