In Tableau - Fixed

Change the aggregation of the new field to AVG in the view (or wrap it):

| Expression | Respects view dimensions? | Respects view filters? | |------------|---------------------------|------------------------| | | No (uses specified dimensions only) | No (unless context filter) | | INCLUDE | Yes (adds specified dimensions) | Yes | | EXCLUDE | Yes (removes specified dimensions) | Yes |

The primary advantage of using FIXED is that it operates independently of the "Viz LOD" (the dimensions in your chart). This makes it essential for several high-level analytical tasks: Intro to Tableau LODs (fixed) - The Data School fixed in tableau

You want to find the maximum order value for each Customer, and then find the average of those maximums across all customers.

It is vital to understand when to use which LOD. Change the aggregation of the new field to

| Scenario | FIXED Expression | |----------|------------------| | Highest sales per customer, regardless of product or date | FIXED [Customer ID] : MAX([Sales]) | | Average regional target, no matter what city or store is in the view | FIXED [Region] : AVG([Target]) | | Percentage of total sales per category, not affected by year filters | FIXED [Category] : SUM([Sales]) / FIXED : SUM([Sales]) |

| LOD Type | Behavior | Use Case | | :--- | :--- | :--- | | | Computes at only the dimensions listed in the brackets. Ignores view dimensions. | Computing values at a specific hierarchy level (e.g., State totals while viewing City data). | | INCLUDE | Computes at the dimensions in the brackets plus whatever dimensions are in the view. | Useful when you want to add detail. (e.g., "Calculate the max sale per customer, but then average that across the view"). | | EXCLUDE | Computes at the dimensions in the view minus the dimensions in the brackets. | Useful for "total" or "grand total" comparisons. (e.g., "Exclude Region so I can see the Country total"). | This makes it essential for several high-level analytical

Sometimes you will add a FIXED LOD to a view and get a red exclamation mark saying the fields cannot be compared. This happens when you try to mix granularities that don't align.