Advance your Data & AI career with 50 days of live learning, dataviz contests, hands-on challenges, study groups & certifications and more!
Get registeredGet Fabric Certified for FREE during Fabric Data Days. Don't miss your chance! Learn more
Hello!
I'm having an issue where several rows of my data are duplicating causing many different measures ($, Price, etc) to be overstated.
I'm merging two tables by a singular column (UPC) - I know why duplicates exist due to one of the tables having several of the duplicate values within it. As the UPC's are sold at several different markets.
What I think I need to do is create a unique index column based off of several different columns (Market, UPC, Timing) - this column would need to be dynamic as the data is always changing. I think once this column is done I could remove duplicates?
I'm not thtat well versed in this but any help provided would be incredbily beneifical!
Hi @Anonymous
You can Merge Queries on Multiple Columns, you just need to select them in the right order
Hey @Mariusz thanks for the follow up. Unfortunately one of the tables I'm merging only has 3 columns. (As I'm only adding essential segmentation to another table). So one of the tables only contains 1 like column (UPC).
Hi @Anonymous
Could you show some example so we can make a test?
It is appreciated to upload some screenshots or paste some data.
Hi @Anonymous
You can use Group By
Advance your Data & AI career with 50 days of live learning, contests, hands-on challenges, study groups & certifications and more!
Check out the October 2025 Power BI update to learn about new features.