Forum Discussion

Ffitzpatrick47's avatar
Ffitzpatrick47
Helper II
8 years ago

table qualified columns

Much as been written about qualifying all columns by the table name in dimensions, but not in measures when making these dax formulas, yet every single dax formula I make requires every single column...
Need Help
MattAllington's avatar
MattAllington
Community Champion
8 years ago

It is well documented best practice to ALWAYS qualify all column names in dax with a table name.  I have covered it here,  https://exceleratorbi.com.au/best-practices-power-pivot-power-query-power-bi/ and in my book, as have countless others

 

your formula posted is not valid. If you are going to use MIN, you must also use FILTER. 

 

=calculate([measure],filter(table,tablename[column] =min(table[othercolumn])))

Ffitzpatrick47's avatar
Ffitzpatrick47
Helper II
8 years ago

Yes, I read your book first of all the power bi books I bought.  Since I have your attention, I want to discuss with you the quality of your book.  There are three major types of paper.  There's just the normal paper you see every day, there's the pressed paper that gets a very smooth finish and there's the calendered paper which has a glossy sheen to it.  Calendered paper is made by coating paper with a clay and pressing it to that glossy coffee resistant finish that really all computer books should have.  Your book is very nice, but unlike the power pivot and power bi book by collie, it doesn't have that same coffee resistance.  It seems like it would with the finish and all, but it doesn't.  I don't know why but you should take this matter up with your publishers.

 

Second, if you read page 56 of the collie book, it says right there: to reference a measure we omit the table name.  Also, if you watch the Ferrari video https://www.sqlbi.com/tv/advanced-dax/, he explicitly says he refuses to touch a model where the measure convention with the table names are not followed.  I thought about this for awhile now.  You may have misinterpreted his sentance and thought he meant the return values of an aggregate function are measures, and not the measures in a fact table, but of course this can not be the interpretation.  After all, what would be the proper table for an aggregate function that spans two tables.  In sql I'd write it as: select sum(if(a, t.a, t.b)) from a join b on something.  I don't know how to write it in dax, but I"m sure it must exist when you do FvA analysis (aka pacing, variance, mda analysis)