celerite2
celerite2 copied to clipboard
Fast & scalable Gaussian Processes in one dimension
Hi, I was trying to install celerite2 from source code following the instructions in the documentation. But while testing the code I got an error: ``` python -m pytest -v...
I am happily using `celerite2` with no issues on my local computer, but encountered a problem when importing `celerite2` on readthedocs. I found a workaround but am posting here in...
Hi folks, I just discovered this updated version of celerite and am excited to add `celerite2`'s `PyMC3` GP into the [`Eureka!` ](https://github.com/kevin218/Eureka) package (as I'm finding more and more cases...
Hi @dfm, I'm trying to use numpyro to sample a GP with the SHO kernel as follows: ```python from jax.config import config config.update('jax_enable_x64', True) import jax import jax.numpy as jnp...
Hi @dfm, Today I've built celerite2 from source following the recommendations on the install docs. I'm trying to do something simple, like this ```python from celerite2.jax import terms sho =...
Hi, I am using emcee to maximize a likelihood function to which I pass a celerite2.GaussianProcess object. I store the chain in a h5 file. When I start the emcee...
Please, can you add a C++ example to the documentation? `celerite` C++ doc's example looks helpful but I'm not sure if it is applicable here. Thank you!
I've derived the algorithms for matrix multiplication and solves, but I haven't been able to work out the factorization algorithm yet. There don't seem to be numerical issues for the...
This will require updating the backend to iterate over the batch dimension, but that shouldn't be too terribly hard. Then, we'd need to add a simple batching function. One question...
This should be possible using semi-separable matrices, but it'll take a little bit of reimplementation in the backend. Hopefully it's not too terrible!