Implement Distribution<u64> support for Zeta, Zipf

[dhardy]

• Added a CHANGELOG.md entry

Closes #1323. I would have added usize and possibly u32 support, but @vks preferred not to, and this is reasonable.

These impls use @JamboChen's tests.

[benjamin-lieser]

Looks good.

Having worked a bit more with the two implementation of Poisson and seeing the troubles it caused, I feel like we should not add them and remove them for Possion. For example having an impl Distribution<T> as arg in rust 1.61 where T cannot be specified or the arguable lengthy <Poisson<f64> as Distribution<f64>>::sample(self, rng)

We can just have a doc, where we explicitly say that is is safe to do as u64 or even smaller int types when the user does not care about saturation or is sure it cannot happen.

[dhardy]

It is unfortunate that there is no simple notation for type-ascribing in this case, aside from let x: f64 = distr.sample(rng);. (There were some unsuccessful RFCs.) Distribution::<f64>::sample(distr, rng) should be enough, but it's still unwieldy and not very intuitive.

To some extent this was baked into the original design: the result is a type parameter of the trait Distribution<T> not an associated type. Yet, other than Standard, this ability to support multiple output types hasn't been used much.

I feel that supporting only floats + u64 is the worst approach: it's enough target types to prevent inference in many cases, yet still doesn't support all potentially desired types (e.g. u32, usize).

[benjamin-lieser]

I feel that supporting only floats + u64 is the worst approach: it's enough target types to prevent inference in many cases, yet still doesn't support all potentially desired types (e.g. u32, usize).

I feel a silent saturation is often a serious bug and I don't want to make it too easy. For Zipf I would be ok, because the upper bound is part of the parameters, but for Zeta and Poisson I would not implement the smaller intergers. And I agree, the current approach seems to be the worst of both worlds, because getting the u64 impl does not do too much.

[dhardy]

Replaced by #1517