improve docs re: set.seed and reproducibility

[simonpcouch]

This PR adds a section to the docs for generate and functions used after generate()ing about reproducibility when resampling! Notes that infer respects the state of the RNG and provides an example of how to reproduce exact results.

Worth mentioning that this PR doesn't introduce a more integrated solution to setting the RNG seed, as @kellieotto suggested in #416. I think this avoids introducing an additional layer of abstraction for RNGs (which I find tricky, personally?), though am open to discuss further if folks feel this might be worthwhile.

Closes #416.

[echasnovski]

I don't see as reasonable integrating any sort of randomness management into function arguments. The most adopted way of R community to manage randomness is with set.seed(). Its extra documentation should be enough.

[andrewpbray]

I agree. Thanks for pulling this together Simon!

[mine-cetinkaya-rundel]

This looks good, just a minor wording suggestion.

An additional place this sort of information could go to is the "Getting to know you" vignette, right around the first appearance of the example including generate(). Much of the same language can be reused there and it can be a good place to draw the users' attention to the importance of setting a seed for reproducibility, without making it a required argument in generate().

[github-actions]

This pull request has been automatically locked. If you believe you have found a related problem, please file a new issue (with a reprex: https://reprex.tidyverse.org) and link to this issue.