Tutorials Roadmap about turingtutorials HOT 10 CLOSED

@mkborregaard Happy we could showcase some of our cool stuff. Let us know if there's anything specific you think might be interesting to include in the tutorials.

from turingtutorials.

I will! By the way, StatPlots is better at plotting distributions than just Plots - if you use StatPlots you can replace the line

   plot(x, pdf.(Ref(updated_belief), x), 

with

   plot(updated_belief, x,

from turingtutorials.

@mkborregaard The introduction notebook has been updated to use StatPlots and also handles the changes to MCMCChain's syntax.

from turingtutorials.

Regarding 3., I think it should be possible to find a dataset in the mentioned pkgs or via other sources for every show-case application. Synthetic data, however, might sometimes simplify the documentation, e.g. for a GMM example.

from turingtutorials.

That's a great initiative @cpfiffer !
Regarding 3: Few months ago, @xukai92 suggested using examples from OpenBugs for Turing:
http://www.openbugs.net/w/Examples

from turingtutorials.

@revendrat My, there's quite a lot of good tutorials on there. Thanks for sending that over!

from turingtutorials.

Hey, I just found this repo. The first notebook "introduction" and the planned expansions look really sweet. Looking forward to following this!

from turingtutorials.

In fact you shouldn't need the x at all, plot(updated_belief, should be sufficient. I could make a PR, but a little uncertain how to do that on a Jupyter notebook, where so much of the file content is auto-generatd.

from turingtutorials.

Typically we'd make the change and then just rerun all the cells before committing the file, if you'd like to do so. Otherwise I can probably get to it this evening when I'm back at my machine.

from turingtutorials.

I don't mind but I'm working on a plotting pr on mcmcchains first

from turingtutorials.