Skip navigation
Mark Wu

Tableau Filters

Posted by Mark Wu Jan 10, 2016

Tableau filters change the content of the data that may enter a Tableau workbook, dashboard, or view. Tableau has multiple filter types and each type is created with different purposes. It is important to understand who can change them and the order of each type of filter is executed. The following filters are numbered based on the order of execution.

A. Secure Filters: Filters that can be locked down to prevent unauthorized data access in all interfaces (i.e., Tableau Desktop, Web Edit mode, or standard dashboard mode in a web browser).

1. Data source filters: To be “secure” they must be defined on a data source when it is published. If they are defined in the workbook with live connection, Tableau Desktop users can still edit them. Think of these as a “global” filter that applies to all data that comes out of the data source. There is no way to bypass a data source filter.
2. Extract filters: These filters are only effective at the time the extract is generated. They will not automatically change the dashboard contents until the extract is regenerated/refreshed.

B. Accessible Filters: Can be changed by anyone that opens the dashboard in Tableau Desktop or in Web Edit mode, but not in regular dashboard mode in a web browser.

3. Context filters: You can think of a context filter as being an independent filter. Any other filters that you set are defined as dependent filters because they process only the data that passes through the context filter. Context filters are often used to improve performance. However if the context filter won’t reduce the number of records by 10% or more, it may actually slow the dashboard down.
4. Dimension filters: Filters on dimensions, you can think of as SQL WHERE clause.
5. Measure filters: Filters on measures, you can think of as SQL HAVING clause.

C. User Filters: Can be changed by anyone in Tableau Desktop, in Web Edit mode, or in regular dashboard mode in a web browser.

6. Quick filters: Commonly used end user filters.
7. Dependent quick filters: There are quick filters depends on another quick filter. Dependent quick filters can quickly multiply and slow down dashboard performance.
8. Filter actions: To show related information between a source sheet and one or more target sheets. This type of action works well when you are building guided analytical paths through a workbook or in dashboards that filter from a master sheet to show more details. These will seem the most “responsive” to end users in terms of user experience, as they don’t incur any processing time unless they are clicked on by the user.
9. Table calculation filters: Filters on the calculated fields.