PBI Dataflow - Relationships between columns of decimal type and impact on performance

[asuteu10]

I have the following situations:

Dataflows created on SQL views. Original SQL columns exceed the INT SQL data type limitations and therefore are created as numeric(10,0).

If I select the column values as they are in View, the numbers returned don't have any decimals. Hovewer Power BI Dataflow reads the original data type from SQL (numeric) and then determines data type = decimal. These columns are keys used in Power BI relationships!

I'm trying to figure out if I should change data type to get whole number in Power BI or just leave it as decimal. Is there any implication in performance if decimal data type is used, although all the numbers are actually whole numbers?

[v-kkf-msft · Accepted solution]

Hi @asuteu10 ,

If your data is integer, then I think it is best to use Whole number type in Power Query. 

For more information about the impact of different data types on performance, please refer to:

https://www.sqlbi.com/articles/choosing-numeric-data-types-in-dax/ 

https://www.excelguru.ca/blog/2019/08/27/do-data-types-matter-in-power-query/ 

Best Regards,
Winniz
If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.