Search
Home

Introduction

Welcome to the start of your journey into analysing time-series data within Python.

The focus of this book is to enable you to get started as soon as possible with analysing, modelling and forecasting time-series data.

The Platform: Jupyter Book

Jupyter Book lets you build an online book using a collection of Jupyter Notebooks and Markdown files. Its output is similar to the excellent Bookdown tool, and adds extra functionality for people running a Jupyter stack.

For an example of a book built with Jupyter Book, see the textbook for Data 100 at UC Berkeley.

Here are a few features of Jupyter Book

  • All course content is written in markdown and Jupyter Notebooks, stored in notebooks/
  • The Jupyter Book repo comes packaged with helper scripts to convert these into Jekyll pages (in scripts/) that can be hosted for free on GitHub
  • Pages can have Binder or JupyterHub links automatically added for interactivity.
  • The website itself is based on Jekyll, and is highly extensible and can be freely-hosted on GitHub.
  • There are lots of nifty HTML features under-the-hood, such as Turbolinks fast-navigation and click-to-copy in code cells.

Getting started

To get started, you may be interested in the following links. Here are a few links of interest:

  • Quickstart is a quick demo and overview of Jupyter Book.

  • The Jupyter Book Guide will step you through the process of configuring and building your own Jupyter Book.

  • The Jupyter Book template repo is the template repository you'll use as a start for your Jupyter Book.

  • A demo of the Jupyter Book can be browsed via the sidebar to the left.

Installation

Here's a brief rundown of how to create your own Jupyter Book using this site. For a more complete guide, see the Jupyter Book guide.

  • Fork the Jupyter Book template repo
  • Replace the demo notebooks in content/ with your own notebooks and markdown files.
  • Create a Table of Contents yaml file by editing _data/toc.yaml.
  • Generate the Jekyll markdown for your notebooks by running scripts/generate_book.py
  • Push your changes to GitHub (or wherever you host your site)!

Acknowledgements

Before we proceed further, enormous debt is owed to Rob Hyndman's and George Athanasopoulos' excellent ebook Forecasting: Principles and Practice for which this book owes plenty to.

Also, a huge thanks to Alessia Tosi for her expert judgement and guidance throughout.

Jupyter Book was originally created by Sam Lau and Chris Holdgraf with support of the UC Berkeley Data Science Education Program and the Berkeley Institute for Data Science.