Drift and gene flow

You will model a small population in which the next generation is a random sample of gene copies. Then you will add gene flow and count how often an allele is lost or fixed.

The idea. A population of size individuals has 2 * size gene copies. Each copy of the next generation is the allele $A$ with the chance $p$, and the new frequency is the share of copies that are $A$.

Your tasks.

1. Finish next_p. It gives the allele frequency of the next generation.
2. Finish with_migrants. It gives the frequency after a share of the breeders are replaced by migrants.
3. Finish lost_or_fixed. It starts runs populations at $p = 0.5$, runs each for generations generations, and counts those that end at 0 or 1.