Fix incorrect docs for Bernoulli::from_ratio and add a doc-test

[teddytennant]

What

Corrects the doc comment on Bernoulli::from_ratio and adds a runnable example.

The previous docs had three inaccuracies:

• The example referenced new_ratio(2, 3), but the method is named from_ratio; new_ratio no longer exists, so following the example gives a compile error.
• A prior edit left an orphaned, subjectless fragment: "return true. If numerator == 0 it will always return false."
• The error condition was described as "For numerator > denominator and denominator == 0", but the implementation is if numerator > denominator || denominator == 0 — two independent conditions joined by OR, not AND.

from_ratio also lacked a runnable example, unlike its siblings (new, etc.).

Why

The rendered documentation was misleading: it named a nonexistent method, contained a garbled sentence, and mis-stated when an error is returned. A reader following the old example would hit a compile error, and the AND wording contradicts the actual OR behavior.

Testing

• cargo test --doc from_ratio — the new doc-test compiles and passes. It checks d.p() is within 1e-9 of 2/3, that from_ratio(3, 3).p() == 1.0 and from_ratio(0, 3).p() == 0.0, and that from_ratio(4, 3) and from_ratio(1, 0) both return Err.
• cargo fmt --check — clean.
• cargo clippy --lib — clean.

Doc-only change; the function body is untouched.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[dhardy]

• The error condition was described as "For numerator > denominator and denominator == 0", but the implementation is if numerator > denominator || denominator == 0 — two independent conditions joined by OR, not AND.

This is a case of ambiguous language ("for the cases where ..."), not technically an error.

[dhardy]

(Sorry, mis-click.)