Hello, I can’t believe how time flies, it is already half-time. This thought makes me imminently nervous because the exam day is coming closer. Good to have such a nice and practical topic this week: Create basic charts in Tableau. I’m sure a lot of you already know how to create the most basic chart types, but I will also mention characteristics of the different types, so this might be interesting for everyone.
Check out the further posts of this series to learn more about it:
- Create and Save Data Connections
- Connecting and Preparing Data
- Dimensions and Measures
- Discrete vs. Continuous & Aggregated vs. Dis-aggregated
- Create basic charts
General thoughts about chart types
Not every chart type fits every topic. The right chart type depends on the message you want to communicate. To share your insights the best way possible it is good to know the advantages and disadvantages of the different types. The exam guide focuses on basic chart types. However, basic doesn’t mean that these are less important, on the contrary, these types have almost unlimited use cases.
I don’t know if the following phenomenon only applies to german students or if you all know this feeling. When my class started to create charts we all wanted to show what we had learned and used all the ‘new’ possibilities to create a viz. So, creating a bar chart seemed too boring for me and I also think to most of the other students because everyone is familiar with bar charts since the fifth grade. But Klaus often told us when he saw our approaches that we could better use a bar chart for this viz.
So what I learned is: When things are simple, keep them simple!
The biggest advantage of the bar is that you can compare values that are close together very well, they are very accurate. Bars can be horizontal or vertical (column bar chart). A horizontal bar chart is favorable when you have long labels. It is easier to read the labels when they displayed horizontally. Using a column bar chart will often cause the problem that you have to rotate your labels.
There are many variations of bar charts. Check out, for example, Jeffrey Shaffer’s bar chart gallery here for inspiration.
The favorite use case for line charts is probably to show development over time. In comparison to bar charts (which can be also used to show development) line charts make it easier to bring additional dimensions into the view or rather to compare the members of a dimension.
Perfect timing: Yesterday Jeffery Shaffer published a new example for variations of Line charts. You can find his blog post here.
The scatter plot is another very powerful chart type that visualizes the intersection of two measures. In his book, ‘Practical Tableau’ Ryan Sleeper gives the advice to put the most dependent metric on the y-axis and the explanatory metric on the x-axis. A scatter plot can help you to find patterns.
To create a scatter plot you need at least two measures and you can use dimensions, but you don’t have to. When you put your first measure on rows and the other one on columns, Tableau will aggregate the measures and show only one mark. This is because of the aggregation (remember last week!), to see a mark for every row of your data, go to ‘Analysis’ and click ‘Aggregate Measures’ to remove the checkmark.
The second option is to keep the measures aggregated but to bring dimensions onto the marks shelf. Besides ‘Detail’ you can also use shapes, color, size or labels to increase the level of detail.
Mapping geographic data
The variation of maps that are created by the Tableau Community is always exciting to me. It seems like there are no limits to creativity.
Maps make it very easy to process geographic data. In this blog, I want to concentrate on the easiest ones: symbol maps and filled maps.
When your data represents a geographical role you only have to double click this dimension and Tableau will create a map.
While in a symbol map Tableau draws a circle for every combination of longitude and latitude in this dimension, it creates polygons for a filled map. Once we created the map we can put more dimensions or measures to the mark shelf. In this way, we can add context to the view or increase the level of detail. For the symbol map you typically use size and for the filled map color is used in most cases.
Dual-axis vs. combined axis
I still cannot believe that I’ve just learned about the concept of combined axes only two weeks ago, although I’ve learned the concept of a dual-axis in one of my first lessons. Both ways are very useful, so let’s look at the differences.
Most chart types can display only one measure, dual-axis and combined axis give you the power to bring more measures into the view. So, how does it work?
When you create a dual-axis your measures share the same x-axis but every measure has its own y-axis. You can synchronize the y-axes, but you don’t have to (but it is recommended to synchronize axes because different axes can cause confusion or misinterpretations). The big advantage of a dual-axis chart is that you can format your graphs individually. When you look at the mark shelf you see a card for each measure. This allows you, for instance, to show one measure as a bar and the other one as a line.
A combined axis gives you the power to bring as many measures in your view as you want. To create a viz like this you use the generated field ‘measure values’ and put it on rows or columns. Now a new shelf with the header ‘Measure Values’ appears under the mark shelf. Tip: Tableau will automatically put all measures into this new shelf, to avoid this put measure names on the filter shelf and select the measures you want to include in your view.
A second option to create a combined axis is to drag the second measure onto the first measure’s axis until two green parallel bars appear. When you then drop the measure, this will create the combined axis.
As you can see in your view, you still have one axis and one mark card, which means that you can’t format the measures independently. This disadvantage leads to a big advantage because it enables you to add an additional dual axis.
Stacked bar charts
Stacked bar charts slice the measure in your view into different parts. By doing this, you can, for example, break down a measure into different dimension members.
Something you should keep in mind when you compare more than two categories in a stacked bar chart: the categories in the middle and at the top (in the first example) don’t have the same start- or endpoint. Therefore, it is much harder to compare their lengths.
You can also unstack bar charts (or any other ‘stackable’ chart type) by turning off stacking.
Tables provide a lot of information but it’s hard to process all the information and to find the interesting points. A highlight table makes it easier, for instance, to identify high or low measures.
What do you need? One or more dimensions and one measure. The first step is to create a basic table. Then put your measure on color (now you have the same measure twice in the mark shelf. One pill is on Label and the other one on Color). After that, change the mark type from ‘Text’ to ‘Square’.
I really enjoyed writing this blog because I didn’t have to research everything. Most of the time I could just write down what I already knew. Even it is a simple topic, I hope you have received some valuable information.
I am happy to get your feedback, you can reach out to me on Twitter or in the comments below.