Relationship Needed?

[Anonymous]

Hi,

I currently have a table in my dataset from SQL which has a column named 'RegistrationLog_id', which is for when different companies submit registration details, which can be more than once (i.e. multiple rows of the same company_id but a new log_id).

I am adding a View into the dataset to use for another report, which uses the same table but filters out some companies. The 'RegistrationLog_id' will have the same numbers and rows per company as the current table in the dataset, but would I need a relationship between these tables? Would it be one-to-one? Or possibly bad data modelling? Thank you.

[ToddChitt]

I would not create a relationship. I would not even add this second table if it is strictly a subset of the first. If you need to create a visual off of it, simply filter the visual (at the visual or page level).

Why have two copies of the data? That's "Department of Redundancy Department" 🙂

[ToddChitt]

Can you show us some sample data? Sounds like Table 2 is simply a subset of Table 1?

[Anonymous]

Hi @ToddChitt , sorry I cannot show sample data, but yes the new table is a subset of the original table (almost like a calculated table I believe?) Thanks

[ToddChitt]

Can you MAKE UP something for us to see? Objuscate the crap out of it:

Table 1:

Company           Log ID

======            =-===

1                          101

1                          102

??

[Anonymous]

Hi @ToddChitt, no problem, please see the below.

It is near enough identical, whereby companies are filtered out for their status, and belonging to a region from another database (right join). The latestreg is just a partition by company id for numbering the latest registrations. I hope this helps, thank you