Forum Discussion
Reference different tables when creating a measure
Hi,
Create a single column table called Records with only unique Record ID in that table. Create a relationship (Mnay to One and Single) from the RecordID column of the 2 data tables to the RecordID column of the new table. Write these measures and drag the last one to a card visual
Measure1 = calculate(COUNTROWS('sales'),'Sales'['Sales Numbers'], == "1")
Measure2 = COUNTROWS('All Sales')
measure3 = divide([Measure1],[Measure2])
Hope this helps.
Read about CALCULATE()
OK I think with the above suggestions and correcting the syntax I've worked it out.
As Ahsish Mathur says you can create two seperate Measures, then divide those. This makes it easier to follow. Originally I wanted to avoid this to minimise processing time of 3 measures when loading in millions of row data situations, and creating more items in the Fields section.
So combining this into one equation reads:
Measure = DIVIDE(COUNTROWS('All Sales'), CALCULATE(COUNTROWS('Sales'),'Sales'[Sales Numbers] == "1"))So the key point for others if I am correct:COUNTROWS only counts the rows in a Table not a column, so you can only reference tables.To find the quantity of specific values in a column, you filter the table using CALCULATE then count the rows remaining.So:CALCULATE(COUNTROWS('Sales'),'Sales'[Sales Numbers] == "1"reads as:Count the rows left in the Sales table after the filter Sales Numbers = 1. Where the term CALCULATE creates the filter , not COUNTROWS.Interesting thought though. Would using the 3 measures method be just as quick to load as the 1 measure method?Use DAX Studio to examine the query plans.
