The pivot_index function can be used in table calculations to extract and manipulate the value of a pivoted column based on its index (in other words, its position).
Syntax
pivot_index(expression, pivot_index)
The pivot_index function evaluates expression in the context of the pivot column at the position pivot_index (1 for first pivot, 2 second pivot, etc.) The function returns null for unpivoted results.
Examples
The following table shows a query that counts the number of orders made in each season of the year. The query includes a pivoted dimension, Products Category, which has the values Swimwear and Coats. To compute the total number of orders made in each season across all categories, create a Total Orders Count table calculation.
| Products Category | Swimwear | Coats | Total Orders Count |
|---|---|---|---|
| Season | Orders Count | Orders Count | |
| Winter | 3 | 671 | 674 |
| Spring | 278 | 120 | 398 |
| Summer | 840 | 21 | 861 |
| Fall | 30 | 432 | 462 |
The Total Orders Count table calculation adds the value of the first pivot column to the value of the second pivot column. The formula is:
pivot_index(${orders.count}, 1) + pivot_index(${orders.count}, 2)
Things to know
pivot_index cannot be used in a custom filter.