Dear all,
Good day.
Is there an example for Euler Maruyama application in numpyro?
I am basically PyMC user and wish to try out in numpyro too.
There are alot of tutorials in numpyro.
May I know is there an example that show systematically model define and infer in numpyro?
Thanks
Currently, we don’t have examples for Euler Maruyama (you can make a feature request for it). I think bayesian regression would be a starting point. For a problem like Euler Maruyama, you can either define a custom distribution or manually compute log-likelihood and use numpyro.factor
to add that log-likelihood to the joint density.