Replacing relationships with join functions in DAX

How to join tables in DAX when there are no relationships in the data model. The data lineage plays an essential role in this scenario.
Article and download: sql.bi/824174?aff=yt
How to learn DAX: www.sqlbi.com/guides/dax/?aff=yt
The definitive guide to DAX: www.sqlbi.com/books/the-defin...

[@teigenxayden]

Thx Marcos for your share of DAX,you are so great!

[@clerincg]

I watched your first video on DAX joining and tried to implement. I failed because I didn't evaluate the two tables and had no common fields in the variables. Then, I would have encountered the Data Lineage issue as well. Now, thanks to this companion video, I was able to get my DAX joins to work! Thank you, Marco! I think it's important that, somehow, you need to firmly instruct viewers to watch BOTH videos in order to implement a working DAX join solution with variables.

[@luisfernandomacedo1451]

Marcos, each piece of video you release is a MasterClass, and I learn great deal. Muito Obrigado!

[@mnowako]

Happy New Year 2024! I was waiting for this topic and I got it :) Thank you!

[@tibobago]

Grazie Marco !

[@brunof.s.8186]

So, what's up with the final 5 minutes? was expecting a cut scene just like in marvel movies

[@marcorusso7472]

We fixed it - thanks!

[@user-ts9kk4nd4w]

Good video! But, what is the reason to use NaturalJoin in DAX? There are generate, generateall, treatas functions. Is there a difference in performance?

[@marcorusso7472]

See my other comments - they should not be used, but the name invites people coming from SQL to use them, resulting in strange behavior I tried to explain in the video. However, the join functions could be faster than generate equivalent when the condition is a complete matching of the columns with the same lineage/name. It should be rare to use them, though.

[@totvabe1]

Fantastic! PS: The last 5 minutes of the video are superfluous, they are empty.

[@marcorusso7472]

We fixed it - thanks!

[@nikakalichava8012]

really awesome but what is the use? I don't really see any possible cases where I would use this?

[@marcorusso7472]

It depends - they should not be used often, but I used them when I had to create transformations on the fly in a measure without being able to use relationships in the model.

[@Charlie888]

Great!

[@mohsenhosseini2138]

Why the last 5 minutes are just a black screen?!

[@marcorusso7472]

We fixed it - thanks!

[@galymzhankenesbekov7242]

Thank you for the awesome video as usual. Could you please say what version of Power BI Desktop do you use? I have installed Power BI Desktop (release of December 23') but the interface of your PB and mine are different. Thanks

[@marcorusso7472]

Check the preview features, probably you didn't enable some of the latest features such as the query view.

[@JuanBretti]

Thank you for the video and article. By the way, the video after 18:10 is completely black.

[@marcorusso7472]

We fixed it - thanks!

[@douglascory]

Pardon me if this was explained somewhere else, but how to use the table presentes at the 8:47 timestamp? As in, yes you've created in the query view, but that is "just for show", how to ACTUALLY add such table to the model?

[@marcorusso7472]

You can use the code after EVALUATE and assign it to a calculated table. However, by doing that, you would have all the tables in memory, if you can apply the transformation upfront is better. This video explains how to use the join function in particular cases, for example when you cannot change the model but only query it.

[@arklur3193]

I was kinda hoping the last ~5m will contain some spoilers like in Marvel movies 😁.

[@marcorusso7472]

We fixed it - thanks!

[@RobertoStaltari]

I've stopped watching Netflix series in favor of Marco Russo's videos 😉