add inverse hyperbolic sin transform

[pearsonca]

Saw #297, and seemed easy enough to fix while I was looking at that area anyway.

[thomasp85]

Would you like to update the PR to remove merge conflicts

[pearsonca]

Happy to - anything else I should address while in there?

[thomasp85]

No, thanks. Better to keep the PRs lean and focused

[thomasp85]

However, can I get you to add a bullet in NEWS about the new feature

[pearsonca]

Can do re NEWS.

Question - I've written merge conflict resolution on probability_trans as moving the domain default setting to the signature. The existing resolution is setting it internally - what's the style preference here? I'd lean towards making it visible to the user via signature (and also allowing for some clever hacks with user defined q/p functions), but am also sympathetic to the-mathematical-definition-of-a-probability-is-...

[thomasp85]

I think the domain should be fixed and set internally rather than as part of the construction

[pearsonca]

I think the domain should be fixed and set internally rather than as part of the construction

Alright - I will adjust that (plus tweak test @olivroy mentions in comments).

I will also add the derivatives per #341

[thomasp85]

Also, can you remove the derivative part - I rather we add that all together in one PR

[thomasp85]

Thank you 🙏