As mentioned in issue 14, we need to add more search algorithms, and each one will have its own issue to track progress. This issue will track progress implementing Variational Bayes with snowline.
Snowline is designed to provide a nearly drop-in replacement for emcee or dynesty, which should make this easy. However, snowline is optimised for low-dimensional (N < 10) problems, so may struggle when users want to fit multiple spectra simultaneously. In these low-dimensional problems, it will outperform emcee for monomodal, approximately-gaussian posteriors (where emcee performs the best).
As mentioned in issue 14, we need to add more search algorithms, and each one will have its own issue to track progress. This issue will track progress implementing Variational Bayes with snowline.
Snowline is designed to provide a nearly drop-in replacement for emcee or dynesty, which should make this easy. However, snowline is optimised for low-dimensional (N < 10) problems, so may struggle when users want to fit multiple spectra simultaneously. In these low-dimensional problems, it will outperform emcee for monomodal, approximately-gaussian posteriors (where emcee performs the best).