Help Diagnosing Many to Many Join Error

Legacy Contributor

Hello, 

I keep getting a Many to Many join error message. I exported my list of 76 items from Visualizer window into excel, and checked if there was any duplication in excel. No duplicate records were removed, meaning my list of items is unique in my dimension table. I also put the item field in ASC order and scanned over and saw no duplicates within Birst Visualizer Table viz. Why  is this message coming up and not letting me join in this context? What else can I check to troubleshoot?

Legacy Contributor

Kevin, if you're still using the Pronto documentation you are barking up the wrong tree.  In Birst you would normally never do joins like this.  If you describe the data in the two tables I could advise you more directly, but assuming that one has attributes describing something (a dimension) and the other has facts/measures that you would use to sum / count, here are the basics:

In Relate, click on the icon in the top right for Hierarchies.  Add a Hierarchy (I'll call DimA) and tell it what the primary key of that hierarchy is (presumably the column that also exists in your fact table that you are trying to join on).  Close that.  Still in Relate, click on the table that has the attributes, click Model Columns, and designate the appropriate columns as being an attribute of your hierarchy (DimA).  Then go to Source Properties for the same table and select DimA in the Grain Configuration section.

Next, do the same for the other table, making sure you mark any column that you wish to sum as a Measure, any date column that should be used in date analysis as Analyze by Date, and the level-key (primary key) from DimA as an attribute of DimA.  Again, in Source Properties select DimA in the Grain Configuration section.

Then click the icon in the top right of Birst to publish your data.  Afterwards it should be available to report on in Visualizer.