Arbitrarily sample a Gaussian distribution

Is there an idiomatic way to create a Gaussian distribution whose probability density can then be queried for a known, arbitrary value?

Hoping to take advantage of rand::distributions::Normal, I looked into rand::distributions::Distribution::sample and ::sample_iter, and it looks like both of those methods take an rng argument. However, for my application it would be convenient to be able to pass a float instead.

Yes: Normal distribution - Wikipedia

use statrs::distribution::{Normal, Continuous};

let n = Normal::new(0.0, 1.0).unwrap();
assert_eq!(n.pdf(1.0), 0.2419707245191433497978);

I think the rand family of crates is geared towards sampling and doesn't support evaluating PDFs or CDFs of distributions.