hierarchy of errors #247
hierarchy of errors #247
Conversation
mostly doing this to test out handling errors this way
There was a problem hiding this comment.
This looks pretty nice. The error contains a lot of information (maybe too much? I can't estimate if this level of detail is needed) and the source annotation from thiserror should create better dev feedback for little effort.
There's one thing I don't really understand but I like this in general
| pub struct Parameters { | ||
| shape: f64, | ||
| rate: f64, | ||
| } |
There was a problem hiding this comment.
Can you explain why you added this abstraction? You could also check the parameters inside Gamma::new and return a suitable error from there, right?
There was a problem hiding this comment.
Mostly just playing around with having input checking in another type to support the use of infallible new methods.
But more interestingly perhaps it could be a way to get alternative parametrizations #198 as well as new_from_moments to do something like,
let m = Gamma::new(shape, rate).into_moments();
let p = Poisson::new_from_moments(m);But I've lots of thoughts floating around about that none motivated by something I've actually needed thus far, so happy to drop it until specific need arises. As a solution to #198 could be adding just one more method.
|
#265 will do this without unused complexity. Thanks @FreezyLemon |
looking for some feedback on possible error hierarchy.
I don't think I did the source error pattern nicely, but I think there's something here worth looking at.