Book Image

Data Wrangling with R

By : Gustavo R Santos
Book Image

Data Wrangling with R

By: Gustavo R Santos

Overview of this book

In this information era, where large volumes of data are being generated every day, companies want to get a better grip on it to perform more efficiently than before. This is where skillful data analysts and data scientists come into play, wrangling and exploring data to generate valuable business insights. In order to do that, you’ll need plenty of tools that enable you to extract the most useful knowledge from data. Data Wrangling with R will help you to gain a deep understanding of ways to wrangle and prepare datasets for exploration, analysis, and modeling. This data book enables you to get your data ready for more optimized analyses, develop your first data model, and perform effective data visualization. The book begins by teaching you how to load and explore datasets. Then, you’ll get to grips with the modern concepts and tools of data wrangling. As data wrangling and visualization are intrinsically connected, you’ll go over best practices to plot data and extract insights from it. The chapters are designed in a way to help you learn all about modeling, as you will go through the construction of a data science project from end to end, and become familiar with the built-in RStudio, including an application built with Shiny dashboards. By the end of this book, you’ll have learned how to create your first data model and build an application with Shiny in R.
Table of Contents (21 chapters)
1
Part 1: Load and Explore Data
5
Part 2: Data Wrangling
12
Part 3: Data Visualization
16
Part 4: Modeling

Summary

In this chapter, we studied one of the main graphic packages in marketing. The ggplot2 library is capable of so much that it was even translated into other languages, such as Python.

We began the chapter discussing the interesting theory of the grammar of graphics, using an analogy of textual grammatical elements and looking at the elements needed to construct and plot a good visualization. ggplot2 was built on top of that concept, enabling analysts to code a graphic by layers, adding one piece at a time. We then introduced a template of questions to help organize our thinking when creating code: (1) start with a dataset, (2) choose a geometry, (3) provide axes and aesthetics, and (4) add a title, labels, statistics, and themes.

After familiarizing ourselves with the syntax, we studied the code for the most commonly used types of graphics, such as histograms, boxplots, scatterplots, bar plots, and line plots. Then, we introduced smooth geometry, which helps us to create...