My excellent @LancUniPhysics summer intern Mo Ghani has produced a great webpage showing the use of a variety of Python-based #Bayesian inference packages on real physics/astronomy datasets https://t.co/8Mznc7oyUR #mcmc#nestedsampling
@exoplaneteer@pymc_devs @kylebarbary If anyone wants to try and it, please do, and send me any feedback/comments/suggestions via the code's github repo here https://t.co/l32Fcrgpg2
It's been a while coming, but I've finally got round to updating https://t.co/ZzwDO9Aggi (thanks to the effort of a great undergrad project student) - it now uses the bilby package (https://t.co/8cJqRO6pzj) as the back-end... (1/n)
and thanks to bilby's interface the website now allows you to choose between the emcee (@exoplaneteer) and PyMC3 (@pymc_devs) MCMC-based algorithms, and the dynesty and nestle (@kylebarbary) nested sampling based algorithms (2/n)
I created a webpage with a simple demonstration example (fitting parameters of a straight line) of using of four Python-based #MCMC packages, and five Python based #NestedSampling packages #bayes https://t.co/kMBwtDrCQo. Details on which ones to follow...
@pymc_devs@mcmc_stan@colindcarroll @AustinRochford Thanks. Our site is fairly basic at the moment, but a proper API would very be nice to have in the future. Please do ask me (aka @matt_pitkin) any questions, and I'm more than happy to exchange ideas.
Now I've also got the example using #CPNest (https://t.co/f1aQbyVz1y) @JohnVeitch https://t.co/eHU4Hy8bUm https://t.co/phQAyQAhZm to add some #NestedSampling into the mix!
We'll hopefully be adding samplers other than #emcee soon-ish. As such, to figure out how to code them up, I've attempted a toy example using #PyMC3@pymc_devs ()https://t.co/wfFiuwTwU9 and #PyStan@mcmc_stan (https://t.co/fRTUsN74L2)
The site (https://t.co/ZzwDO9Aggi) has a new, and hopefully improved, look. You can also find a video demonstration of how to use it here https://t.co/ZdlJftNYVI. Please let me know what you think of the new layout. #mcmc#bayes#inference
Researchers at our @UofGravity and @UofGPhysAstro will benefit from a £500k bequest from pioneering physicist Prof Ronald Drever: https://t.co/aZNQTLBe6s