1. Building first visualizations in Tableau
In this demo, we will be working with a similar dataset to the one which you will see in the exercises. Let’s have a look at the data.
We have two tables, titles, and credits joined on film title ID.
In the sheet, we find the same tables and their fields.
Often it is not important which table the data comes from, so it may be clearer to group these measures and dimensions differently. By clicking on the little arrow of the Data Pane, we can change it to Group by Folder to have one handy list of dimensions and measures.
This is a good moment to check our fields. We already see that Seasons is incorrectly marked as a measure (which often happens when Tableau sees a numeric field), so we will drag it over to the dimensions. Release Year, on the other hand, is considered a number, while we would like to use it as a Date. Let’s change that.
How to build any visualization? Notice that when we drag a field to the canvas, some UI elements highlight in orange or blue. This is where we can drop our dimensions or measures.
Let’s drop Release Year onto the columns and add the average IMDb Score to the rows.
Tableau immediately visualizes our data.
But there are even easier ways to achieve the same, the famous Show Me.
Let’s clear the Canvas by clicking here.
Control selecting (or command selecting on Macs) these fields, we open the Show Me tab. Tableau will highlight all possible basic charts. Some options are grayed out and hovering over, for example. Dual combination chart, we will see it’s because this visualization is not possible with just one measure.
Let’s choose a line chart. We achieve almost the same; the only difference is that Tableau will use the default aggregation, SUM, and, in the case of dates, will auto-convert the dimension to a continuous one. So our chart needs a bit of tweaking, but it’s always a good start.
One often overlooked piece of information is in the bottom left corner of Tableau.
It tells us how many marks, so the combination of all used dimensions we currently have in our worksheet. Currently, we present the data for 66 of 67 possible years (there is one null year with no data). If we now add Type to the Color Marks Card, the marks count increases to 106.
Dropping any dimension onto any Marks may increase the count of the data points in our visualization, and it is important to observe this number as this often impacts how legible the chart will be in the end.
Let’s remove the Type from Color and add another measure to the rows: the count of IDs.
If we open Show me and select the dual combination chart, we get an overlaid chart.
That’s a start and a good inspiration, but let’s try to build it ourselves and even improve it! We will take a few steps back.
To overlay two charts, we click on the arrow on the second measure and select “Dual Axis”.
This merges the visualization into one graph.
Using the Marks on the left, we can adapt the individual chart types used. We will choose to display the count of IDs as bars and keep the score as a line.
Lastly, we can change the order of the charts by swapping the order of the measures in the Rows shelf.
We now clearly see fast growth in the number of title IDs since 2010, which slowed down in 2020 due to the Covid pandemic.
The IMDb scores display a slightly dropping trend over the past few years.
That’s it for now; let’s explore some basic Show me charts together in the exercises.
2. Let's practice!