I have two use-cases for this question.
Can calculated columns quickly be added to new datasets without having to re-build from scratch?
I have a dataset with multiple calculated fields which I would like to easily extract for a data catalog, to include the formula. Is this possible?
You can connect an existing dataset to a new dataset and reuse any calculated fields that were used in the parent dataset. This is a great way to enable self-service BI scenarios. The “certified” calculations will be managed from the centralized dataset. Different teams can build new datasets on top of it and reuse the parent’s calculations. Read about it here.
Can you provide more details regarding your second question - Are you looking for an API to read all calculated fields? If your calculated fields are implemented on the dataset level, you can use QuickSight API to read them. Here is the relevant reference - with CalculatedColumn object that is explained here, and here is the CLI version.
With this python script (wrap in Lambda function), amazon-quicksight-sdk-proserve/Dataset_info.py at master · aws-samples/amazon-quicksight-sdk-proserve · GitHub, you can get the dataset, field, dashboard, and analysis information and relationships. FYI.