누구나 사용할 수 있는 분석 도구
누구나 사용할 수 있는 데이터 준비 도구
조직을 위한 분석 도구
조직을 위한 클라우드 분석 도구
Interactivity is a core component of rich analytical systems. It helps answer deeper questions, solve more end user tasks, and create improved user experiences. More expressive interactivity means you can provide richer content for audiences such as clients and stakeholders. Playful experiences drive curiosity and engagement, two crucial ingredients to improving data literacy throughout the world.
Until now, Tableau’s interactivity has centered on specific action types, such as filtering and highlighting. While authors can control the source and targets of these actions, their behaviour is fixed. Highlight actions always highlight, but what if you wanted to colour your selection? Filters always filter all fields, but what if you wanted to filter a single axis to compare it to the total?
Sets are powerful. They allow user selections to be used anywhere in a visualization and in calculations. Sets can colour, group, and filter an axis in a viz or a term in a calculation. They can also conditionally aggregate, combine fields, hide data that is used in a table calculation, and filter on a related field—and that’s just the start.
The new set action feature in Tableau 2018.3 means that a selection of data points in a visualization can be stored in a set. Set actions are much more than a new type of action; they extend Tableau's interactivity to support custom user-defined behaviour.
A set can apply different behaviours to various target sheets. For example, the same set could colour viz A, hide data in viz B, and filter an axis in viz C. Now with set actions, a user selection in any of the sheets can update the set, thereby modifying all target sheets in a single coordinated selection. Coordinating multiple actions through a single selection dramatically increases the breadth and depth of scenarios that can be addressed for end users though interactive analytic applications.
While this post focuses specifically on comparative analytics, the second post introduces many new interactive analytics techniques expressed with set actions.
Set actions connect two existing features, offering a wealth of opportunity to create new compositions from existing concepts. In this post, we'll explore examples of part-to-whole analysis, part-to-part analysis, and difference from a selection. While set actions have made all of the scenarios presented here possible, the goal of the examples is to demonstrate end-to-end user flows. Some of the things you can expect to learn from the following examples:
Have you ever wanted to analyse how a selection contributes to the total? Sometimes you may want to calculate its contribution as a percent of total. Alternatively, you may want to see its magnitude against the total. In other cases, you may wish to calculate differences between your selection and the total. This section will demonstrate all of these concepts. The first 3 examples incrementally build on one another, so you may want to watch the video demos in order!
In relational data structures, it is easy to perform calculations across columns, such as sales and profit. It is much less trivial to calculate across rows within a single column, such as sales in Europe versus sales in America, or sales in 2017 vs sales in 2018. Set actions have made comparisons across rows dynamic, interactive, and simple!
Concept: Change over time is one of the most common types of comparative analysis, where flexible selection of the start and end periods is available to viz consumers.
Comparative analytics are more powerful and flexible now that sets are accessible to end users. Instead of using parameters for many types of interactive comparisons, you can now use sets to enable multi-selection and direct interaction with a viz instead of a list. Moreover, sets are live linked to your data, and you can express some comparisons simply through dragging and dropping!
The next post introduces 8 new techniques expressed with set actions. In particular, filtering on relationship, sorting and aligning on a selection, applying OR conditions across filters, dynamic grouping, and controlling the order of operations. Check it out!