Leveraging Analytics Capabilities

In the previous three chapters, we learned how to create increasingly more complex calculations in Tableau. Now that we have the tools to develop powerful workbooks, let's see if we can leverage Tableau's analytics capabilities to bring some color to our data, either with graphical elements, such as reference lines or bands, or using forecasting to predict future data.

The following topics will be covered in this chapter:

• Basic tools in the Tableau Analytics pane
• Additional analytical options
• Using forecasting
• A practical example

Technical requirements

This chapter uses the Global Superstore dataset, which can be found at http://www.tableau.com/sites/default/files/training/global_superstore.zip.

Basic tools in the Tableau Analytics pane

Tableau holds a set of ready-made analytical tools that can be simply dragged and dropped onto a view. Those functions can be found in the Analytics pane, usually by clicking on the tab next to Data:

If you can't see this on the left-hand side of your screen, you can go to Window > Show Side Bar in the menu bar to display it. The pane holds tools that have been separated into three sections: Summarize, Model, and Custom.

Using the options

Each of the analytics tools in this toolbar functions in the same manner. You can either double-click the option or drag one of them onto the view, which will make a panel containing a few options appear:

In general, it will display either...

Additional analytical options

There are four options remaining in the Analytics pane: Totals, Trend Lines, Clustering, and Forecasting. The last of these will be detailed in the next section, so let's focus on the first three.

Totals

The Totals option can be dragged onto the view to turn on totals either for all the rows, all the columns, or to add subtotals to all the rows or and columns:

It is not possible to turn on subtotals for selected dimensions; it has to be done for all or none.

Totals can also be turned on using the Analysis > Totals options, which can be found in the menu.

Trend lines

...

Using forecasting

While the previous tools were only performing analytics on existing data, the Forecast option allows us to extrapolate the data to get a glimpse into the future.

By default, Tableau includes a package that takes the best of a few options of models to create a simple forecast. Tableau uses an exponential smoothing model, which gives more weight to the recent data points over the older ones. Those forecasts can then be fine-tuned with the use of the options that can be found in the Edit Forecast toolbox.

To start using forecasts, simply drag and drop the Forecast option from the Analytics bar onto a chart or table that includes one and only one date field (otherwise, you will see a message indicating that no forecast can be created for views with multiple date fields). You can also use the Analysis > Forecast > Show Forecast menu to achieve the same result...

A practical example

Now that we know all about Tableau's analytical tools, let's see whether we can solve this sample exam question:

Which subcategories in the LATAM market are more than a standard deviation away from the mean for both sales and profit?

We can answer this question in two ways.

First, we can go with the previous chapter and use INCLUDE statements to show the average and standard deviation of the sum of the two measures. Here, we will define {INCLUDE [Sub-Category]: SUM([Profit])} and {INCLUDE [Sub-Category]: SUM([Sales])} and look at the average and standard deviation of each, as follows:

By performing some simple additions, we can get the lower and upper bounds (for example, $13,038 + $14,161 = $27,199 as the upper bound for profit and $127,330-$111,886 = $15,444 as the lower bound for sales). Subsequently, we can plot Sales and Profit by Sub-category...

Summary

In this chapter, we covered some of Tableau's analytic tools, which allow us to create reference lines or bands, cluster data in similar buckets, identify trends, and forecast what our data will look like in the future. This allows us to highlight some of the useful information that might have been missing from the views (such as an average line or maybe a constant band to tell us when a project started and finished), and this helps bring meaningful insight to our data.

This chapter concludes our exploration of worksheets, for which we now have all the tools required for the exam. We will study how to link and present different worksheets in a single consolidated view using dashboards in the next and last chapter of this book.

Questions

Answer the following questions to test your knowledge of the information in this chapter.

Q: Using the Global Superstore dataset, if you create a forecast for Sales by Year, how many periods forward does Tableau forecast by default?

A: Two periods (with Year(Order Date) as columns and SUM(Sales) as rows, and dragging the Forecast option).

Q: How can you create a single line for the 90th percentile on a view?

A: By adding a Distribution Band, with Percentile, 90 as options, not by adding a Reference Line, as there are no options for percentiles.

Q: Using the Global Superstore dataset, which country in Central Asia has the widest distribution of profits, using interquartile ranges?

A: Creating a box plot, as described in this chapter, India is the country with the widest IQR (represented by the size of the box).

Further reading

You can check out the following links for more information about the topics that were covered in this chapter:

• Forecasting options: https://help.tableau.com/current/pro/desktop/en-us/forecast_options.htm.
• Forecast field results: https://help.tableau.com/current/pro/desktop/en-us/forecast_field_results.htm.
• How clustering works: https://help.tableau.com/current/pro/desktop/en-us/clustering.htm#how-clustering-works.