Impact of data types in DAX arithmetical calculations

[@captivatedCoder]

"Level of Wrongness" is going to be my band name.

[@lucasdenker3734]

Love your channel so insightful, can you guys make a topic about “exceeded the available resources” error? This is a constant issue i run into in my reports and I don’t know what causes them or how to solve it.

[@babkenhakobyan6847]

Very interesting and insightful, as always.

[@SteiniJonsson]

The conversion in this video is to demonstrate rounding (errors) and how they affect the result. In a real-world scenario, I would store the price in the original currency to avoid the conversion all together. That way I would always get the same amount, both in original currency and USD.

[@shashipaul6279]

love it Alberto.. as always

[@xyclos]

Very very interesting...granular analysis. Thanks!!!

[@richardmartin7904]

Very interesting as always. So what if you have banking deposit data for hundreds of millions of customers (with daily transactions) and you are required to be accurate to the penny for all of them. How do you address that within Power BI?

[@SQLBI]

You should use the Currency (Fixed Decimal Number) data type. However, it represents numbers up to 922,337,203,685,477.5807 - it's definitely an issue for certain currencies.

[@jensvonkracht6770]

Nice Video. What happens performancewise, when you store the calculated USD value as Decimal instead of Fixed Decimal number?

[@SQLBI]

Not much, you use a floating point instead of an integer, but the performance difference is negligible for the data type itself - you should look at data type conversions involved, which could be more expensive.

[@matiasoleksikiw4626]

Great video! I am curious about the software you are using on your mouse to mark the screen while presenting. Which one is it?

[@SQLBI]

ZoomIt: learn.microsoft.com/en-us/sysinternals/downloads/zoomit

[@teach127]

Thank you ​@@SQLBI

[@SandipanSarkar-c8v]

Finish watching

[@mienzillaz]

Who stores rates in row context. Even if it changes over time i prefer to have own table just for that and compute figures when there's a need.

[@SQLBI]

It's an example to show the issue in arithmetical calculations, it's not meant to be a best practice for data modeling in this specific case.

[@mienzillaz]

@@SQLBI I know, but had to point that out as your audience represent many skill levels and I bet some people will take this as saint practice. :)