This folder contains the presentation about decision making with Bayesian methods for PyData Global 2020. The presentation is created from a Jupyter notebook by using RISE and is then served with Binder.
The libraries required to build the slides locally can be installed from the requirements file:
pip install -r requirements.txt
RISE allows to generate the slides from the Jupyter notebook editor itself with a single click:
To render slides use jupyter nbconvert foo.ipynb --to slides --post serve
The slides are served using the environment in arviz_sandbox for convenience and speed, see this post in Jupyter Discourse for a detailed description.
To generate the binder shield with the link to the presentations the following steps should be followed:
- Make sure the notebook metadata has
"livereveal": {"autolaunch": true}
. If you create you presentation as a copy of the English one it will already be done.- This (as you'll see when editing) generates the presentation automatically whenever the notebook is opened. To edit the notebook close the presentation and modify the cells contents. More details in RISE documentation
- Generate the binder link to run the notebook in the sandbox environment.
There is a helper page nbgitpuller link generator.
The
Binder
tab allows to specifyGit Environment Repository URL
:https://github.com/arviz-devs/arviz_sandbox
Git Content Repository URL
:https://github.com/arviz-devs/arviz_misc
- and the file to be opened
- Create a custom shield from Binder docs