About the author
Horatio Bota is a freelance Data Scientist with over seven years of experience in data analytics
and data science. He has previously worked at Microsoft Research, J.P. Morgan Chase, and several
startups in the U.K. He holds a B.Sc. and a Ph.D. in Computing Science from the University of
Glasgow.
About the content reviewer
Adrian Gosa is a Senior Accountant at Nike Europe, with over seven years of business and finance
experience. He has previously worked at PricewaterhouseCoopers and Deloitte, covering a wide
variety of industries. Adrian holds an M.A. in Business Economics and an M.Sc. in Quantitative
Finance from the University of Glasgow.
About the technical reviewer
Alexandra Vtyurina is a Research Scientist at Kira Systems in Toronto, where she uses Python to
analyze user behavior. She holds a B.Sc. and M.Sc. in Computer Science from the Southern Federal
University, and is about to get her Ph.D. in Computer Science from the University of Waterloo

Introduction
If you work in accounting, you’ve probably heard the rumors: artificial intelligence and automation are set to “reshape the accounting function” over the next few years. When exactly the robots take over1 is still uncertain, but how they’ll do that is already clear: someone will program them. More likely than not, that someone is you.
You’re best suited to automate the repetitive tasks in your job because you already know how: you have lots of domain knowledge and are used to working with computers and data.2 You even have programming experience: most accounting work involves creating custom computer programs through point-and-click operations instead of code — these programs are called spreadsheets. You probably know what “functions” or other programming concepts are, and more importantly, you know how to think about data manipulation in powerful ways.
Unfortunately, the tools used in accounting today limit how expressive you can be when working with data. How many times have you found yourself digging through Excel’s menus for a command that almost does what you want it to, but not entirely? How painless is it to work with large files in Excel? How simple is it to automate the boring parts of your workflow?
Unlike Excel, Python is a tool designed for expressing any data manipulation you can think of. On top of that, Python is fast, handles spreadsheets that Excel can’t even open, and working with it always leaves a trace (i.e., code) that you can check and run whenever you need to. Even better, because you’re already familiar with programming ideas and have the right mental models for working with data, all you need to start using Python is a guide on how to add it to your accounting toolkit. This book is your guide. It starts with Python programming basics and goes all the way to making interactive data visualization using Python. On the way, it introduces many of the tools that have become a foundation for data science — all the while keeping its focus on accounting tasks and data (i.e., on using Python with spreadsheets, not instead of them). The rest of this introduction explains how Python enables powerful data analysis practices that can strip away much of the boring parts of your work. Hopefully, by the end of this chapter, it will be clear why reading the rest of this book is worth your time.

What this book is about
This book is about computer programming more than it is about accounting practices: it teaches you how to use the Python programming language to work with tables. You’ll use general ledgers, cash flow statements, and other accounting datasets throughout the book, but the goal is to teach you how to work with those datasets using Python code rather than introduce new accounting ideas on the way.
There’s more than one kind of code: websites, smartphone apps, or the operating system on your computer are all built with code, but each is made with different programming languages and tools. Even when handling tables with Python, you can quickly glue spreadsheets together in a few lines of code, or you can code up an entire data analysis project. Both approaches use code, but larger projects employ more of Python’s gears and tools.
As it happens, accounting tasks are a mix of quick data wrangling (e.g., moving rows between spreadsheets, cleaning text entries, reconciling values across files) and extensive data analysis projects (e.g., margin or costing analysis, inventory forecasts). You can easily use Python for both, but you need to use more of Python’s machinery for larger projects. As we progress with the book, I’ll introduce different parts of that machinery and go through several data wrangling projects using accounting datasets before attempting a Python-based sales analysis project. This way, you’ll see how to use Python for the boring parts in your work first, and how to set up an entire data analysis project in Python towards the end of the book.


Whom this book is for
If you use a computer at work in any way, knowing how to write even a few lines of Python code will make you more productive (and you might even find coding fun). This book is for anyone who works with accounting data in Excel and wants to use Python to improve their workflow:  

▲ Accounting, finance, or economics students
▲ ccountants working with anything from payroll to costing
▲ Business controllers
▲ Auditors
▲ Financial planners