Talk: Building Interactive Visualization with Jupyter Notebook

This repository contains files for building interactive visualizations for the web using the Jupyter Notebook for scientific computing.

We utilize the Bokeh and Plotly libraries to generate the visualizations.

Install Dependencies

Install dependencies with ``pip```.

pip install -r requirements.txt

Install NBExtensions Javascript and CSS

The Jupyter Notebook Extensions add a lot of neat capabilities to Jupyter Notebook. The extensions are installed in the step above but we'll need to copy over the javascript and css libraries into your user directory and edits your configuration using the below command:

jupyter contrib nbextension install --user

You should now be able to edit your extensions at the Jupyter notebook home or by running the 'jupyter nbextension' command

Configure ipywidgets

The IPyWidget provides a common widgets interface for interactivity in a Jupyter notebook. It's used for some of our examples and is required for the qgrid data viewer. To enable it, run:

jupyter nbextension enable --py --sys-prefix widgetsnbextension

Virtual Environment Kernel

If you are using a virtualenv (highly recommended), you'll need to create a kernel specification to launch notebooks with it inside the Jupyter Notebook.

mkdir -p ~/kernelspecs/vistalk jupyter kernelspec install ~/kernelspecs/vistalk --user

Afterwards, you'll need to relaunch any running jupyter notebook servers. Your kernel containing all the packings installed within your virtualenv are now available.

Run the notebook

jupyter notebook

When the jupyter server launches, a browser window will open to the server. The URL to access the server will be displayed in the terminal along with a token (if using default config) required to discourage CSRF attacks.

Animations

To run the animations and interactive plots for bokeh, you'll want to run a bokeh server concurrently in the background. You can do so by invoking

bokeh serve