Forum Discussion
Summarizing & reworking Data
- 5 years ago
Hi Anonymous ,
You can do this without the need for the Ageband table.
On your Database unpivot your two Age columns this will give you two columns one witht the Attribute and value.
Create a new table with the age band you need:
AgeBandMinimumMaximum
0-10 0 9,99 10-20 10 19,99 20-30 20 29,99 30-40 30 39,99 +40 40 9999 Now create the following measure:
Count by age = VAR temnp_table = FILTER ( DataBase_Unpivot, DataBase_Unpivot[Value] <= MAX ( Ageband[Maximum] ) && DataBase_Unpivot[Value] >= MIN ( Ageband[Minimum] ) ) RETURN COUNTROWS ( temnp_table )If the Unpivot of the columns is not an option beside the age band table create one for the ages:
Ages
Age1 Age2
Now create the followin measure:
Count by age no unpivot = SWITCH ( SELECTEDVALUE ( Ages[Ages] ), "Age1", COUNTROWS ( FILTER ( DataBase, Database[Age] <= MAX ( Ageband[Maximum] ) && Database[Age] >= MIN ( Ageband[Minimum] ) ) ), "Age2", COUNTROWS ( FILTER ( DataBase, Database[Age 2] <= MAX ( Ageband[Maximum] ) && Database[Age 2] >= MIN ( Ageband[Minimum] ) ) ) )Check PBIX file with both options
I have kept your model for comparision but has refered no need to keep the age bands for both ages, also be aware that the two tables are not related with the other tables.
Thank you both MFelix & v-easonf-msft .
I had to use the second solution from MFelix as the actual data are coming from a DAX code. The PQ code is really interesting too, thanks to v-easonf-msft