Data Analysis for Corporate Finance

Building financial models using SQL, Python, and MS PowerBI

Mariano F. Scandizzo CFA CQF

• Publisher: Fulton Books, Inc.
• Year: 2021
• ISBN:
  9781649527226
• Topic: Business
• Subtopic: Corporate Finance
• Index: Business

Have you ever tried to learn to code or to use advanced visualization tools? If so, I am sure you know how daunting it is to learn by yourself.

Generally, tools and books follow an encyclopedism approach, i.e., books attempt to teach every feature about a coding language or tool. This implies hundreds, if not thousands of pages simply to tackle a single topic, whether SQL, Python, MS Excel, MS PowerBI, you name it. The journey from zero to hero to become proficient using numerical and visualization tools to take your career to the next level becomes an ordeal that requires years and thousands of pages just to begin putting the pieces of the puzzle together. However, the reality is that you do not need to learn absolutely every available feature to use those tools and deliver a superior project.

Rather than teaching you about the forest, I will discuss specific trees. Why? Because once you become familiar and confident nurturing a few trees, growing a forest becomes a simple process of planting new trees. This book provides the fundamental blocks so that you can learn about financial data science and take these tools and start using them tomorrow. The scope of the selected tools will empower you to see a considerable improvement in your financial modeling skills.

The book is designed to provide corporate finance professionals the ability to start immediately using advance tools for concrete real-world tasks.

Therefore, this book is all about functionalism. It is about providing you with tools that will put you to work and dramatically change the way you analyze data. Once you see the benefits, it will become natural to keep expanding your domain knowledge, leveraging today's endless available educational resources.

Chapter 5

Pandas

“The goal is to turn data into information, and information in insight.”

Carly Fiorina

Introduction

We have reached the cornerstone of our data science applied to finance journey. It is as certain as death and taxes that anything you do regarding numerical analysis, visualizations, machine learning, simulations, etc. will be related in one way or the other to a Pandas dataframe. In line with that fundamental aspect of the Python’s state of the art, this chapter will be fully dedicated to exploring Pandas. There are three clear aspects which summarize dataframe operations:

1. Importing and exporting data into/ from a dataframe
2. Manipulating data
3. Table manipulations such as merge, join or concatenate

We will touch each of the above topics in great detail to ensure a good understanding of this de facto library.

I will begin by formally introducing pandas. Pandas is named after the panel data econometric term. In a sense, Pandas is built on top of NumPy. For example, NumPy universal functions will generally work on Pandas objects as well. On a rather fundamental level, the dataframe class is designed to manage indexed and labeled data.

Dataframes can be created in the following ways...