PT1_week08

@apietak @gfederic @jbachmann @lbrodbeck @mkryukov Hi! In the Simpson Benchmark I didn't understand if we were supposed to test the different simpson implementations on all integrand functions simultaneously.

In the Penna simulation I decided to create a derived class for the population subject to fishing, whereas in the slides we were suggested to derived a new simulation class; which way is the better one?

Please be picky with your corrections!