Forum Discussion
Running total measures performance issues
Hello all,
Has anyone had issues trying to use a running total measure for large sets of data?
Some context on the data I'm using:
- 76,000+ rows of data
- 100+ columns
When I try to use a running total measure, power bi crashes ("not enough memory to complete this operation").
I've heard that large amounts of columns affects the performance of power bi. If I remove columns at the 'edit queries' level will that solve my issue? Or does the actual dataset need to have the number of columns reduced?
Are there any other issues I'm missing or any tips that I can use to resolve this issue? A running total seems like a pretty simple operation, so I'm not sure why Power Bi can't handle it in my case.
Appreciate any help that I can get..
Best,
Eliot
1 Reply
Hey,
I'm wondering if you are using a dedicated calendar table and if you're following the patterns for time calculations that are described here: https://www.daxpatterns.com/time-patterns/
Can you please describe your data model, it's always a good idea to create a data model based on the design principles of a star schema. Here, long (many rows) but narrow (few columns) fact tables are joined with short (few rows) but broad (more columns) dimension tables are joined.
76k are very very few rows, but 100+ columns is broad.
Regards,
Tom
