Update to CellCollection.select

[quaquel]

This PR updates CellCollect.select to reflect the updated API of AgentSet.select. So, we switch to at_most instead of using n. Since n was not used anyway before and this is still experimental, I have just removed n altogether.

[github-actions]

Performance benchmarks:

Model	Size	Init time [95% CI]	Run time [95% CI]
BoltzmannWealth	small	🟢 -4.3% [-5.1%, -3.5%]	🔴 +8.5% [+8.3%, +8.7%]
BoltzmannWealth	large	🔵 -2.0% [-2.5%, -1.4%]	🔴 +16.0% [+11.7%, +21.6%]
Schelling	small	🔵 -0.0% [-0.5%, +0.4%]	🔵 +0.6% [+0.4%, +0.9%]
Schelling	large	🔵 -0.1% [-0.9%, +0.5%]	🔵 +1.5% [-0.6%, +3.8%]
WolfSheep	small	🔵 +0.0% [-0.4%, +0.4%]	🔵 +2.9% [-0.3%, +6.2%]
WolfSheep	large	🔵 +1.0% [-0.2%, +2.2%]	🔴 +6.6% [+4.8%, +8.4%]
BoidFlockers	small	🔵 +0.2% [-0.2%, +0.6%]	🔵 -0.3% [-1.0%, +0.4%]
BoidFlockers	large	🔵 -0.0% [-0.5%, +0.4%]	🔵 -0.6% [-1.1%, -0.1%]

[quaquel]

these performance benchmarks are strange. Neither one uses CellCollection.select.

[github-actions]

Performance benchmarks:

Model	Size	Init time [95% CI]	Run time [95% CI]
BoltzmannWealth	small	🔵 -3.5% [-4.6%, -2.5%]	🔵 +0.4% [+0.2%, +0.6%]
BoltzmannWealth	large	🔵 -1.3% [-1.6%, -0.9%]	🔵 -0.2% [-1.3%, +1.0%]
Schelling	small	🔵 -3.3% [-3.7%, -2.9%]	🔵 -0.1% [-0.4%, +0.2%]
Schelling	large	🟢 -5.6% [-6.0%, -5.2%]	🟢 -12.5% [-13.4%, -11.8%]
WolfSheep	small	🟢 -3.4% [-3.7%, -3.2%]	🔵 -1.0% [-1.2%, -0.8%]
WolfSheep	large	🟢 -6.3% [-7.5%, -5.3%]	🟢 -11.5% [-12.4%, -10.7%]
BoidFlockers	small	🟢 -4.2% [-4.6%, -3.7%]	🔵 -1.4% [-1.9%, -0.8%]
BoidFlockers	large	🔵 -3.1% [-3.7%, -2.6%]	🔵 -1.3% [-1.9%, -0.6%]

[EwoutH]

Nice to align them. Reminds me I wanted to add a sample method.

We merged some stuff, I think that's causing this PR branch to have more differences with main. Maybe rebase it.