Files

• jupyter.ipynb summarizes very shortly what jupyter notebooks do and don't do
• networkAnalysis.ipynb is the notebook to create the graph shown above (among other things)
• networkAnalysis.html is the above notebook exported as .html file, can be downloaded and viewed in your browser (needs javascript), is still interactive!
• networkAnalysis-Google-Cloud-DataLab.ipynb is a version of the above notebook that directly works in Google DataLab; some adjustments needed to be made to make in run on Python 3.5, and it contains two lines to update packages in DataLab, since the preinstalled versions are outdated.

Setup of jupyter

The simplest way of setting up jupyter and a functioning Python installation on Windows and Mac is Anaconda: https://www.anaconda.com/distribution/

Download the Python 3.7 version. There is also a download for Linux, but I personally (as a Linux user) would recommend using your Linux distribution's built-in package manager to install all the necessary packages and versions, because Anaconda comes with it's own package manager, conda, which is again different from Python's own package manager, pip. So to avoid confusion about installed packages and versions, just stick to one.

Once everything is installed, you should be able to either open the jupyter Notebook from your applications or from the command line with $ jupyter notebook, which should open a new tab or browser window in your default browser.

Package versions

Packages like numpy are required by pandas anyway and will be installed as dependencies.

packageversion

pandas

0.23.4

plotly

3.2.1

networkx

2.2

matplotlib

3.0.2

python

>=3.6

jupyter in Google DataLab

For a quick start on how to use jupyter notebooks in the Google Cloud Platform, follow this link: https://cloud.google.com/datalab/docs/quickstart

Further reading on Python and documentations for the used packages

Python Guide(s)

The documentation of Python is extensive and well made, so that is always a good place to start: https://docs.python.org/3/tutorial/index.html

You can find a lot of different tutorials online just by searching for a combination of the words Python, data science, quick, tutorial, e.g. the first search result for me was this (even tho it's from 2016): https://www.analyticsvidhya.com/blog/2016/01/complete-tutorial-learn-data-science-python-scratch-2/

But bear in mind that Python is NOT just used for data science, but is a fully fledged programming languages, and you might miss out on many power- and helpful features if you just stick to data science tutorials.

plotly

Plotly provides a very neat interface for creating interactive plots, that base internally on JavaScript. You don't need to create an account and use their cloud functionality (there is the plotly.offline module as used in the notebooks in this repository). Have a look at their Gallery and documentation as well: https://plot.ly/python/

networkx

networkx is the backbone package for this repo, because it provides the functionality to create the network graphs we're interested in. This package provides a huge toolset of graph theoretical methods, so a read on their documentation is definitely not wasted time: https://networkx.github.io/documentation/stable/tutorial.html

pandas

pandas makes interfacing with datasets very easy, and provides many methods to transform and analyse the data. There is a package called pandas-gbq, that let's you query google BigQuery easily as well (not used in this repo though). And of course the link to the documentation: https://pandas.pydata.org/pandas-docs/stable/

Download

You can download the jupyter notebook used at the show and tell from GitHub