Forum Discussion
Merging Tables Efficiency Question
Hi boldwake,
According to your description, I would suggest you to keeping doing it in DAX for performance purpose. By this way, you can keep making use of the physical relationships to get best performance and the best use of the VertiPaq engine.:smileyhappy:
In addtion, another possible solution is to use the LOOKUPVALUE function. Using LOOKUPVALUE you can denormalize the Brand & Category directly in the Sales table, by defining a new calculated column in Sales table. This solution can make the code simple to author.
Regards
Hi boldwake,
According to your description, I would suggest you to keeping doing it in DAX for performance purpose. By this way, you can keep making use of the physical relationships to get best performance and the best use of the VertiPaq engine.:smileyhappy:
In addtion, another possible solution is to use the LOOKUPVALUE function. Using LOOKUPVALUE you can denormalize the Brand & Category directly in the Sales table, by defining a new calculated column in Sales table. This solution can make the code simple to author.
Regards
Late to the party here, but it sure looks like every time I do a merge query, it has to load up all the data into memory again for each merge. I am hitting the magic combination of suck because one of the tables that I am merging with loads up multiple data files from a Folder data source.
I am going to switch to a DAX approach but thats a bummer because it sure is easier to use the merge query in the editor.
