Hi GD team, Running into an issue with a visualization. I can get it to work if I use an ID column to join the two tables but I assumed I could use a column from a table that could be joined implicitly. Trace ID: f76351297f54803e27b987275dfb2a17

Hi Tim, This is pretty standard based on your LDM. You can connect datasets by using a BY clause in your metric to make those connections within your LDM

So I am using a BY clause in one of the metrics but it is a composite metric (metric in the table is made up of metric1 / metric2 where metric2 has the BY clause). If my visualization, if the rows are

Company, Department

the visualization doesn't work. If the rows are

Company, Department ID, Department

it does work. Do I need the extra BY clause on Department? Wouldn't this reduce the reusability of the metrics?

This would all depend on how you intend to have those objects interact with the other objects in your LDM, and how plan to evolve your LDM over time. The BY clause will continue to make connections from right to left in the model.