I have several custom sql datasets from the same datasource that I need to load into an analysis. Some of the data can be related by a key and some of it cannot be related at all - its just summaries of other data.
How best to model this with datasets and sources?
Hello Jones,
Welcome to the QS community.
Why do you need so many custom SQL dataset in the first place ? although you mentioned its related with key in some case and in some case for summaries but are they serving the purpose of many dashboards or each custom SQL is meant for single dashboard ?
I would advise rather to go with some kind of Datawarehouse tools such as AWS Redshift or snowflake etc. or data base backend where you could create or design the data product which can be reused and also serve multiple dashboards and consumers.
Else at one point it may be nightmare to manage many Custom SQL data sets. Thats my opinion.
Hope others can put some more insights on the topic.
Cheers,
Deep
Hiya!
Some of the figures i need to display are just single figures, pre-calculated elsewhere.
I’m not sure how putting these into a data warehouse would make them any more easily modelled.
Basically, I need to add several tables to the same dataset. Some of which have relationships, some of which don’t
Will you use all tables in the same visual?
You will not be able to bring the tables together without a joined column. But if you don’t need so bring data in one visual together you could just use different datasets in one analysis.
Ok, so I can’t have unrelated tables in the same dataset? I have to create a new dataset, even though its for the same database connection?
Yes. But you can use multiple datasets in one analysis