Tableau Prep: Up & Running

Chapter 4. Shaping Data

As discussed in the previous chapter, the first step of data preparation is understanding how the original data set is structured, followed by determining what structure the data set needs to be in—that is, the desired state—for analysis. This chapter looks at these factors and Prep Builder’s functionality for reshaping data to help you quickly identify how to shape future data sets for analysis.

What to Look for in Incoming Data Sets

Let’s look at a typical input data set from Excel by building a pivot table (Figure 4-1). This example uses sales of ice cream-scented soap.

Figure 4-1. Ice Cream Scents sales and profit data

When assessing an incoming data set, it’s important to identify both the dimensions (or categorical values) of the data and the measures. Dimensions is the term Tableau Desktop uses to refer to the columns of data that describe the records example, the regions a product is sold, in or the category that product belongs to). Measures refers to the numeric values of the data set that are being analyzed (such as the number of students in a college class or the tuition they are paying). If the dimensions are all in individual columns, you can move on to measures without having to think about any structural changes. Likewise, you need to assess the measures to ensure each has a separate column in the data set.

The data set from Figure 4-1 ...