Bugfix for non deterministic rng behavior

[quaquel]

This fixes a bug in how the numpy rng is seeded. If instantiating a model with Model(seed=42), rng should default to 42, not None. The net result of this bug is that at the moment, `Model(seed=42) will leave the numpy rgn undeterministic. This adds an explicit test to ensure that both model.random and model.rng behave in a deterministic manner when seeded with the same initial value and fixes the bug.

[github-actions]

Performance benchmarks:

Model	Size	Init time [95% CI]	Run time [95% CI]
BoltzmannWealth	small	🔵 -3.0% [-5.6%, -0.2%]	🔵 +1.0% [+0.8%, +1.1%]
BoltzmannWealth	large	🔵 +1.2% [+0.6%, +1.7%]	🔵 +3.1% [+1.9%, +4.3%]
Schelling	small	🔵 -0.2% [-0.5%, +0.2%]	🔵 +1.0% [+0.9%, +1.1%]
Schelling	large	🔵 +0.7% [+0.3%, +1.0%]	🔵 -0.6% [-1.6%, +0.2%]
WolfSheep	small	🔵 -0.5% [-1.1%, +0.2%]	🔴 +22.0% [+16.4%, +27.7%]
WolfSheep	large	🔵 -2.0% [-2.7%, -1.5%]	🔵 -4.0% [-5.4%, -2.5%]
BoidFlockers	small	🔵 +1.8% [+1.3%, +2.3%]	🔵 +0.2% [-0.5%, +0.9%]
BoidFlockers	large	🔵 +3.3% [+2.3%, +4.3%]	🔵 +0.5% [+0.1%, +1.0%]

[EwoutH]

Good catch!