Join us at FabCon Atlanta from March 16 - 20, 2026, for the ultimate Fabric, Power BI, AI and SQL community-led event. Save $200 with code FABCOMM.
Register now!To celebrate FabCon Vienna, we are offering 50% off select exams. Ends October 3rd. Request your discount now.
Question:
Is it better to use a custom column than a constantly performed row-wise measure? For instance, in a particular table, I have a date column and a time column. Throughout multiple dozens of measures, I have the date and time being added together row-wise in the measure. In this case is it better practice to use a custom column to simplify the measure? Is this a performance concern? Has anyone measured the performance difference with this particular case?
Background:
I was reading best practices from a number of various sources (which I have unfortunately lost the references to due to receiving a new work computer). I remember that a common theme was that using measures was a better practice than using custom columns in almost all situations. I have adopted this methodology and I have to say that I agree to a large extent that it is generally a good practice.
That said, I know of 1 situation already where custom columns are better. In terms of performance, it is occasionally better to use a custom column for extraordinarily large data sets. Even then, custom columns are deep down in my steps of optimization, none-the-less they are there.
Solved! Go to Solution.
Here is a good article for your reference. By the way, calculated tables sometimes make things easy.
Thank you for the article, this is a better description of not simply which is preferred in which situation, but also why. Applying that knowledge to my situation, I should stick with a measure performing identical calculations within the measure to save on RAM. That said, since it is performed so frequently, I will simply need to look at the environment it is running in, and determine if the longer CPU time to perform the measure calculation repeatedly is a higher "cost" than the extra RAM by adding a column.
It is unfortunate that I have to sacrifice readability in favor of practicality here. But at least now I havea strong udnerstanding as to why.
User | Count |
---|---|
98 | |
76 | |
69 | |
53 | |
27 |