Implement triangular distribution #575
Conversation
dhardy
left a comment
dhardy
left a comment
There was a problem hiding this comment.
Also needs some kind of value-stability test.
|
|
||
| #[test] | ||
| fn test_new() { | ||
| for &(min, max, mode) in &[ |
There was a problem hiding this comment.
It just tests that the asserts are implemented correctly. I ported the test from statrs.
There was a problem hiding this comment.
So you're just copying from statrs. I wondered. Why?
There was a problem hiding this comment.
You mean this test specifically or the distribution in general?
| let norm = Triangular::new(0., 1., 0.5); | ||
| let mut rng = ::test::rng(1); | ||
| for _ in 0..1000 { | ||
| norm.sample(&mut rng); |
There was a problem hiding this comment.
This also doesn't do much. It could at least test the mean.
There was a problem hiding this comment.
Fair enough. This is the status quo of most distribution tests in Rand though.
In principle we should use histograms and test against the PDF. (I plan to work on that in the midterm, probably after implementing PDFs for #290 and by generalizing tests/uniformity.rs.)
|
Yes, I think it can be merged. The tests can be added later and are needed for the existing distributions as well. I think only concern is that we might not want to expose new distributions for 0.6 and rather move them to a |
2 participants
No description provided.