Is there any difference if we create the calculated field in analysis or dataset level

Q&A
, , ,
1

Hi All,
Is there any difference in execution or storage if we create a calculated field at dataset level or analysis.

2

Hi
if you create calculated fields at dataset level, execution is much better.

3

Ok. What about the Storage??

4

no major difference in storage.

5

You can add calculated fields to a dataset during data preparation or from the analysis page. When you add a calculated field to a dataset during data preparation, it’s available to all analyses that use that dataset. When you add a calculated field to a dataset in an analysis, it’s available only in that analysis.

Please refer to the QuickSight documentation: Adding calculated fields - Amazon QuickSight

Here is a nice community post on [Calculated fields] Analysis vs Data Prep: [Calculated fields] Analysis vs Data Prep Level

2 Likes
6

Suppose you use a dataset calculated field that doesn’t include an aggregate function. In that case, the results are stored in SPICE instead of an analysis-calculated field, which is calculated at run-time ( and therefore slower).