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)

Working with dates

Working with dates is rarely straightforward. Data analysts need to successfully parse date values, identify invalid or out-of-range dates, impute dates when they're missing, and calculate time intervals. There are surprising hurdles at each of these steps, but we are halfway there once we've parsed the date value and have a datetime value in pandas. We will start by parsing date values in this recipe before working our way through the other challenges.

Getting ready

We will work with the National Longitudinal Survey and COVID case daily data in this recipe. The COVID daily data contains one row for each reporting day for each country. (The NLS data was actually a little too clean for this purpose. To illustrate working with missing date values, I set one of the values for birth month to missing.)

Data note

Our World in Data provides COVID-19 public use data at https://ourworldindata.org/coronavirus-source-data. The data that will be used in...