Do you have data that you want to visually display the outliers? From all of the charts out there, which visualization is best to display anomalies?
A plain scatter plot is a good starting point. There are ways to give it that extra pop and bring your users attention to the outliers.
Reference Bands
Ryan Sleeper has a great article in the Playfair Data site on how to create a visualization using reference bands to outline the anomalies here. He is using a Distribution Band with the Standard Deviation option. He is additionally using a calculation to draw attention to the Good and Bad ranges.
I modified the calculation he used to display a standard deviation of 3, -3.
Anomaly Color
IF SUM([Profit]) < (WINDOW_AVG(SUM([Profit])) - (WINDOW_STDEV(SUM([Profit])))*3) Then "Bad"
ELSEIF SUM([Profit]) > (WINDOW_AVG(SUM([Profit])) + (WINDOW_STDEV(SUM([Profit])))*3) Then "Good"
ELSE "Expected"
END
I created a visualization using this method with States here. (In the Distribution dashboard)
Advantage: This type of chart will allow your user to quickly focus on the positive or negative points in your data.
Disadvantage: If you use too many reference bands in a chart it can diminish the value or distract the user from what is important. You need to ask yourself what point are you trying to convey?
Box and Whisker
Tableau has a great article how to read and understand box and whisker charts here. Basically anything in between your whiskers represents 50 % of your data. Anything outside of the whiskers are considered the outliers. Box and Whisker plots can be found in the Show Me tool of Tableau.
Example:
I created a visualization using this method with Regions here. (In the Distribution dashboard)
Advantage: Your users will be able to see where the majority of points are in a chart quickly.
Disadvantage: This chart is not great for providing detailed summary of the data.
Forest Plots
What is a Forest Plot? Forest plots are also known as blobbogram. They are a way to summarize multiple points to answer a single question. They also can combine multiple studies and/or multiple findings into one singular visualization.
Tableau has an article on how to create a forest plot here. It does not list how to obtain the formulas for this type of chart. (You will need to calculate hedges G etc). You would have to research how to calculate the confidence intervals for your data.
Advantage: The visual representation can make it easier to see variations between individual results.
Disadvantage: Forest plots require complex calculations and are not great for the beginner Tableau user.
There was also a Workout Wednesday session to create a twist on the forest chart here.
Displaying Anomalies Using Spatial Data
In this link here, Sarah Battersby reviews a couple of examples of how to display anomalies using spatial data in Tableau. Density maps are a great way to quickly visualize outliers in a map.
I created a visualization using this method with States here. (In the Distribution dashboard)
Advantage: Users will quickly see plots that are outliers.
Disadvantage: Spatial data is quite large and can take some time to render.