Data Science Desktop Survival Guide
by Graham Williams
Chapter: Data Wrangling
20200607 Data will need to be transformed in many different ways to get it ready for visualisation and analysis. This chapter provides a comprehensive catalogue of different tasks useful for wrangling our data. The goal is to transform the acquired raw and untidy data into a tidy dataset.
Chapter 7 presented an approach to data wrangling based on a template that utilises many of the transformations presented here. The template provides the basic steps which in any project will be combined with transformations found in this current chapter. What is to be transformed is usually determined through our data exploration as presented in Chapter 8.
We will also find ourselves coming back to the data wrangling after visualising (Chapter 10) and building machine learning models from the data.