What is the difference between discrete and continuous data and how does it impact the way you work in Tableau? Today I will be covering the difference between discrete vs continuous data.
This blog is written for those who are getting started with Tableau as well. I will be using Tableau Worksheet anatomy terms such as “shelf”, “canvas” and “side bar” – if you are unfamiliar with these terms in a Tableau context, read more about them here.
Blue vs Green in Tableau
When you first start using Tableau, you may notice that the values in the side bar will have small blue or green icons next to them. You will also notice that when you drag these values into a shelf that the colour of their pill (Tableau speak for the coloured rectangle with rounded ends that looks kind of like a “pill”) will be the respective blue or green.
I didn’t know what this meant at first, but it turns out that blue and green are Tableau’s way of colour-coding discrete and continuous data.
What is discrete data? (Blue pills in Tableau)
Discrete data is shown with blue pills in Tableau that are also known as dimensions. Discrete data has the following characteristics: finite, countable and whole (non-divisible) numbers. Though discrete data is non-divisible, averages of this data may result in decimal numbers. For example, a family has 1, 2, 3… children. You cannot have half a child, however, the average number of children in all of the families in your data set may be 2.5.
Categorical data
Discrete data can also be made up of categorical data, and categorical data can be broken down into nominal or ordinal data.
Nominal Data
Nominal data is used for qualitative data that cannot be put into a logical sequence. It includes categorical names rather than numbers. For example, genres of film and TV can be described with words such as “Comedy”, “Drama”, “Action” etc. Though these are not numbers, this is still discrete data. Nominal data should be made up of mutually exclusive categories; for example Blood type is A, B, AB or O, you cannot have a bit A and a bit O type blood.
Ordinal Data
In contrast, ordinal data may also use categories, however these can be logically sequenced. For example, temperature may be described as “freezing”, “cold”, “mild”, “warm”, “hot”. Or a survey response may be ordered from “strongly disagree”, “disagree”, “neither agree nor disagree”, “agree” to “strongly agree”.
It is important to note that though nominal and ordinal data is usually representative of words, it may be encoded and represented with numbers in a database.
Recap:
Discrete data is either whole, countable, finite numbers or categorical data. Think of real-world examples, you cannot have half a dog at the animal shelter, but the average number of dogs per year may result in a decimal number.
The dog breed is a good example of categorical data, even breeds that are mixed together result in their own breed – think of all the -oodle breeds you’ve heard of. If a dog has so many breeds in it, that it is completely unknown, then the category for the nominal data of “Dog Breed” may be “mut”.
Discrete Data examples:
|
Numerical/Quantitative |
Nominal |
Ordinal |
|
Number of students in a school |
Subjects offered by a school |
Grades offered by a school |
|
Population of a country |
Names of states in a country |
Age groups of a population |
|
Number of animals in a shelter |
Types of animals in a shelter |
Date each animal leaves the shelter |
|
Number of bags of blood in a blood bank |
Blood types |
|
|
Result of rolling a die |
Names of winners of a game |
Game number |
|
Longitude and latitude |
Name of a street in that long/lat |
House/building numbers on the street |
What is continuous data? (Green pills in Tableau)
Continuous data is shown with green pills in Tableau that are also known as measures. Continuous data is representative of continuous measures – meaning it can be broken up into fractions and decimal values.
Examples of continuous data include height, weight, mass, speed, temperature, length and distance. Time is also commonly measured in a continuous manner.
In some cases, you might need to convert a discrete value to continuous or vice versa – this is most likely required with date values.
You can create discrete data from continuous data by grouping it and converting it into ordinal data (discussed above). For example, weight is a continuous measure, however it is often divided into certain groups for certain sports – e.g. the eight boxing weight classes ranging from flyweight to heavyweight boxers.
Line charts are more common for continuous data because a line suggests that the data continues from one point to another.
Rounding up
To see examples of how the nature of your data can influence your visualisations in Tableau, read my blog on dealing with discrete and continuous data in Tableau. For now, I hope this blog has helped you to understand the difference between measures and dimensions in Tableau. Happy data analyzing!