Sean Pinkney

Results 151 comments of Sean Pinkney

Allow multivariate distributions to take in matrices

We just need to be super clear on what index the observations are on. Do we choose rows or do we allow either rows or columns? If we allow both...

Add student_t_qf

I believe I can wrap the qf from boost. They say that the method is faster than using the `inv_inc_beta` implementation above. You can see the comment in boost at...

BesselK continuous function

With the introduction of hypergeometric pFq I believe we can use that to calculate the derivative of the modified bessel function of the second kind when $\nu \notin \mathbb{Z}$. From...

allow `array[] int` to add and subtract elementwise an `int` or `array[] int` of the same size

Yes, I'd like those 3 cases for addition and subtraction. For example, I had some code like below ```stan int length_x; array[length_x] int x = linspaced_int_array(length_x, L, U); int k;...

allow `array[] int` to add and subtract elementwise an `int` or `array[] int` of the same size

I realize there's no efficiency gain. As a user, I just expected the vectorized indexing to work with addition/subtraction and array integers.

Hypergeometric function naming

Do you want to expose the `grad_2F1` and `grad_3F2` functions as well as `hyper_2F1` and `hyper_3F2`?

gp_dot_prod_cov allow vector sigma (ard)

@rok-cesnovar can we close this?

Discrete 1D convolutions [FR]

Would the row/col-wise framework work for this? https://github.com/stan-dev/design-docs/pull/35#issuecomment-821812405

add `cospi()`, `sinpi()`, `tanpi()`

I'm doing a bunch of stuff with trig functions and I'm finding this issue. It's problematic as the 0s don't propagate through all the calculations leaving that small remainder.

Update `multi_normal` to call `multi_normal_cholesky` after the cholesky

The precision parameterization should have derivatives as well. A few ad hoc tests of mine show that the cholesky version is faster than precision, do others see the same?