Colt Allen

Ok I've forked Aesara and started working on this in a new branch, but I want to ensure all the tests are passing before creating a PR. Here's what I...

I've done some digging and found the derivations for this function: https://functions.wolfram.com/HypergeometricFunctions/Hypergeometric2F1/20/01/ I've modified my previous post with some pseudocode for the `grad` method. However, these derivations seem to vary...

Thanks. After looking into the likelihood function for my specific use case, gradients only need to be implemented for `|z|< 1`: (See expressions 19 and 20) https://www.brucehardie.com/notes/009/pareto_nbd_derivations_2005-11-05.pdf Implementing gradients for...

This is almost ready for a PR. The only test failing now for `TestHyp2F1Broadcast` and `TestHyp2F1InplaceBroadcast` is `test_grad`. I'm using the `mpmath` library (a transitive dependency of Aesara) for the...

Sorry for not providing a more complete code example. These methods are already being wrapped in an Op: ``` class Hyp2F1Der(ScalarOp): """ Derivatives of the Gaussian hypergeometric function ``2F1(a, b;...

Awesome; if I can approximate an infinite sum via `Scan`, then I could write this entirely in terms of existing `Op`s. What would that approximation look like? Here's my first...

@brandonwillard [Draft PR has been created.](https://github.com/aesara-devs/aesara/pull/1288)

Would it be possible to proceed with publishing the next version release seeing as this issue is now fixed?

Would it be possible to proceed with publishing the next version release seeing as this issue is now fixed?

Third beta release of `btyd` is now available for pip installation! This one includes a Bayesian variant of the Modified BG/NBD model, a few bug fixes, and some requested additions...