search

CamDavidsonPilon / Probabilistic-Programming-and-Bayesian-Methods-for-Hackers

aka "Bayesian Methods for Hackers": An introduction to Bayesian methods + probabilistic programming with a computation/understanding-first, mathematics-second point of view. All in pure Python ;)
http://camdavidsonpilon.github.io/Probabilistic-Programming-and-Bayesian-Methods-for-Hackers/
MIT License
26.51k stars 7.84k forks source link

Edward #365

Open chananshgong opened 6 years ago

chananshgong commented 6 years ago

I would like to repeat all the examples with Edward. What do you think?

CamDavidsonPilon commented 6 years ago

Neat! I’d love to see this!

chananshgong commented 6 years ago

I would probably need help

On Nov 29, 2017 3:09 PM, "Cameron Davidson-Pilon" notifications@github.com wrote:

Neat! I’d love to see this!

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/CamDavidsonPilon/Probabilistic-Programming-and-Bayesian-Methods-for-Hackers/issues/365#issuecomment-347855129, or mute the thread https://github.com/notifications/unsubscribe-auth/AXPZfhIFj3luFuhzkSC99YzRN4bfBE9Jks5s7VdZgaJpZM4QukoZ .

petecog commented 6 years ago

i was thinking the same. Happy to help - in my limited way.

petecog commented 6 years ago

349 looks like @cavities has tried, but got stuck at the first hurdle.

petecog commented 6 years ago

Looking at the possibility of porting the examples to pyro also. This excellent book could become a Rosetta stone for different PP approaches.

tigershen23 commented 6 years ago

@petecog any progress on a Pyro port? @jvans1 and I are thinking of doing the same. TensorFlow also recently released a PP library: https://medium.com/tensorflow/introducing-tensorflow-probability-dca4c304e245

nathanin commented 5 years ago

Hi everyone, I'm working through this book by re-implementing examples in tensorflow_probability -- killing two birds with one stone.

Here's my version of the introduction chapter main example: https://github.com/nathanin/Probabilistic-Programming-and-Bayesian-Methods-for-Hackers/blob/master/Chapter1_Introduction/Chapter_1_Introduction_tensorflow_probability.ipynb

It's good to go with the seed, and MCMC settings as-is. On other runs, I've seen some weird results for tau, and bimodal distributions for the lambdas that I can't quite explain. I'm new to this whole thing, so critiques welcome!