This chapter is from the book
This chapter is from the book
This chapter is from the book
Understanding Dimensions and Measures
When you bring a data source into Tableau, the software automatically classifies each field as either a dimension or a measure. The differences between these two are important, though they can be tricky to understand for those who are new at analytics. Perhaps the best way to differentiate these two classifications is to think about them this way: Dimensions are categories, whereas measures are fields you can do math with.
Dimensions
Dimensions are things that you can use to group data by or drill down by. They are usually, but not always, categories (e.g., City, Product Name, or Color), and they can be logically grouped into strings, dates, or geographic fields. Dimensions can also be organized into Tableau groups and hierarchies, which we’ll discuss shortly.
Measures
Measures are generally numerical data on which you want to perform calculations— summing, averaging, and so on. The definition of a field as a measure or dimension can be adjusted in the Data Source screen by clicking the data type icon. You can also change this directly in the sheet by either dragging and dropping a dimension to measure, and vice versa, or by clicking the drop-down menu by any field and selecting the Convert to Measure (or Dimension) option.
Continuous and Discrete
Generally, dimensions are discrete, whereas measures are continuous. We could break this down a little more into four types or levels of measurement:
Nominal measures are discrete and categorical (e.g., for/against, true/false, yes/no).
Ordinal measures have order but there are not distinct, equal values (e.g., rankings).
Interval measures have order and distinct, equal values—or at least we assume they are equal (e.g., Likert scales).
Ratio measures have order, distinct/equal values, and a true zero point (e.g., length, weight).
In Tableau, continuous fields produce axes, whereas discrete fields create headers. Continuous means “forming an unbroken whole, without interruption.” Discrete means “individually separate and distinct.” Be sure you understand the difference between these mathematical terms. Text and categories (dimensions) are inherently discrete. Numbers can be discrete if they can take only one of a limited set of distinct, separate values (e.g., a rating). Numbers, including dates, can be continuous if they can take on any value in a range.
COLORFUL PILLS
When a field is brought from the Data pane and dropped into the Rows and Columns shelves, Tableau creates a “pill.” These pills are color coded: Blue pills represent discrete variables, whereas green pills are continuous. The data type icons also reflect these color codes (Figure 3.15).
)
Figure 3.15 Color-coded pills reflect continuous (green) measures and discrete (blue) dimensions.