Book Image

Python Data Cleaning Cookbook

By : Michael Walker
Book Image

Python Data Cleaning Cookbook

By: Michael Walker

Overview of this book

Getting clean data to reveal insights is essential, as directly jumping into data analysis without proper data cleaning may lead to incorrect results. This book shows you tools and techniques that you can apply to clean and handle data with Python. You'll begin by getting familiar with the shape of data by using practices that can be deployed routinely with most data sources. Then, the book teaches you how to manipulate data to get it into a useful form. You'll also learn how to filter and summarize data to gain insights and better understand what makes sense and what does not, along with discovering how to operate on data to address the issues you've identified. Moving on, you'll perform key tasks, such as handling missing values, validating errors, removing duplicate data, monitoring high volumes of data, and handling outliers and invalid dates. Next, you'll cover recipes on using supervised learning and Naive Bayes analysis to identify unexpected values and classification errors, and generate visualizations for exploratory data analysis (EDA) to visualize unexpected values. Finally, you'll build functions and classes that you can reuse without modification when you have new data. By the end of this Python book, you'll be equipped with all the key skills that you need to clean data and diagnose problems within it.
Table of Contents (12 chapters)

Chapter 9: Tidying and Reshaping Data

As Leo Tolstoy and Hadley Wickham tell us, all tidy data is fundamentally alike, but all untidy data is messy in its own special way. How many times have we all stared at some rows of data and thought, "what..... how...... why did they do that?" This overstates the case somewhat. Although there are many ways that data can be poorly structured, there are limits to human creativity in this regard. It is possible to categorize the most frequent ways in which datasets deviate from normalized or tidy forms.

This was Hadley Wickham's observation in his seminal work on tidy data. We can lean on that work, and our own experiences with oddly structured data, to prepare for the reshaping we have to do. Untidy data often has one or more of the following characteristics: a lack of clarity about merge-by column relationships; data redundancy on the one side of one-to-many relationships; data redundancy due to many-to-many relationships; values...