Tableau Visual Analytics
Tableau visual analytics pane is sometimes overlooked because its usually hidden from the data pane but also because most users don’t request visual analytics outside of a trend line or forecast usually. I honestly don’t use it as much as I should but there is some hand visual analytics you can add to most of your charts. Normally most of them are relevant to line charts but as you will see in the video they can also be applied to other charts like bar charts. After learning about all the different visual analytics try to use them in your next dashboard and see if you can add any additional value to it for the end user.
Enter your email to download the data
Constant Line
The constant line is probably the most simple of all the visual analytics. You can use this to out a target line on your chart so you can visually see if the marks are above or below the line. Just remember this is constant so wont change it unless you change it. I usually do a parameter and a reference line if I am doing a line so that the end user can change it in the future if things change.
Average Line
The average line will give you the average of the data and display it across the chart. This one is dynamic so will update with the data and also when using filters.
Median Quartiles
This one I should use more often to find outliers cause it will show you the median line and band the upper and lower quartiles. This will help a lot with showing volatility in your line charts.
Box Plot
Box plots gets some love from me, it usually used for ad hoc analysis because it does a really good job at finding outliers that you generally would see. Its a clear way of seeing outlier marks that you can then drill down into.
Totals
I generally use totals on text tables by going to the top to analysis–> Totals but this is also another way to do it. You can add totals to any charts not just text tables. Sometimes it looks funky and throughs the chart off but that is something you will have to decide if its worth it or not.
Average with 95% CI
The average line with the 95% Confidence Interval is also a powerful tool, especially for people that appreciate statistics. Its telling you that with 95% confidence the marks will be within the grey area. So this is another way to find those old marks that are messing up your averages.
Median with 95% CI
This is sometimes coupled with the Average so you can see how bad the outliers are skewing your data. In my example I overlapped them which is busy but also shows you what those outliers are doing to the mean.
Trend Line
The most popular of them all! It really never hurts to put a trend line on a line chart, I use it all the time. Plus for the statistic people it gives them some good information if they want it.
Forecast
Forecast scares me because it people tend to just believe what they see especially if they are not data people and then make decision based on it and come back to you if it is wrong. I tend to stay away from this unless it is requestion and then I still try to talk them out of it. Use the R integration and find a statistician if you want to do some forecasting.
Cluster
Cluster are actually very easy to use. They group marks together based on logic instead of doing the standard X Y axis break up. This is mostly used on scatter plots.
Reference Line
Reference lines are very useful because you can show a line for another calculation that is not even on the view. This is handy when you have targets or goal type data where you want to show your current situation compared to where you want to be.
Reference Band
The reference band allows you to pick a band size base on the bottom point and the top point. This can be any aggregation based on any measure that is in the view. This will create a shading in the background that you can use to see if your marks are in that band or out that band.
Distribution Band
The distribution band looks just like the reference band but it uses logic to display the band instead of you pick the top and bottom aggregation. You can use percentages to total or do things like percentiles.
Conclusion
As you see there is a lot of different options when it comes to visual analytics in Tableau. The most common ones I use are the trend line, reference line and totals but after writing this I am going to try and incorporate more into work. See which ones you can use in you next dashboard!
