You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Given the new Fit object API in 0.19, I was wondering how other fitting backends (e.g. emcee or Ultranest) could be integrated and what the structure of the input/output would need to be. What are the current road blockers that you see with current API ?
The current example of MCMC is in the recipe here.
One of the difficulties raised by Bayesian frameworks is that you need to define priors that are more complex than just init parameters. At minimum a flat [min, max] prior is needed which is currently doable with the existing parameter interface. A gaussian prior will require an extra parameter.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
Hi,
Given the new
Fit
object API in 0.19, I was wondering how other fitting backends (e.g. emcee or Ultranest) could be integrated and what the structure of the input/output would need to be. What are the current road blockers that you see with current API ?The current example of MCMC is in the recipe here.
One of the difficulties raised by Bayesian frameworks is that you need to define priors that are more complex than just init parameters. At minimum a flat [min, max] prior is needed which is currently doable with the existing parameter interface. A gaussian prior will require an extra parameter.
Thanks
Beta Was this translation helpful? Give feedback.
All reactions