numpyro
numpyro copied to clipboard
Probabilistic programming with NumPy powered by JAX for autograd and JIT compilation to GPU/TPU/CPU.
```python >>> import numpyro.distributions as dist >>> d = dist.Beta(1, 5) >>> d.icdf(0.975) Traceback (most recent call last): File "", line 1, in File "/home/ayaka/venv/lib/python3.10/site-packages/numpyro/distributions/distribution.py", line 463, in icdf raise...
Following the discussion in #1403, this PR introduces a more general mixture distribution combining arbitrary child distributions (so long as they have similar shapes and supports) with fixed weights.
Hi, I understand that I can use `funsor.log_density()` to compute the log-joint probability of a sample from a model whose discrete latent sites can be enumerated. However, I'm not sure...
Currently, enumeration is supported in MCMC but is not yet tested in SVI. It would be nice to support this feature for SVI too.
I have implemented versions of both of these: ```py from __future__ import annotations from typing import Union, cast import jax from jax import lax from jax import random import jax.numpy...
Hi, I'm having issues using NumPyro's latest version of `initialize_model()` on a model with discrete variables. The model samples random variables within a plate context with a data-dependent size. When...
Hi, I'm wondering if there is a need to contribute with a tutorial for integrating ODEs multiple times based on different initial conditions. It's also possible to extend the current...
Hi, I am trying to use random_flax_module on a class that uses flax.linen.BatchNorm that uses mutable parameters. Is there any example on how to use that? Here is my code:...
Hi all, since it's been a while I thought I should maybe give a sign of life and continue from here. This tries to implement https://github.com/pyro-ppl/numpyro/issues/559. There are still some...