Forum Discussion
MAX value separately by one column's different values
Hi
I would like to have a DAX to solve this (I could create a helper table quite easily in QE). Example table:
| Country | Cost per item | Date |
US
| 100 | 1.1.2021 |
| US | 100 | 1.2.2021 |
| US | 120 | 1.3.2021 |
| GB | 1000 | 1.4.2021 |
What I would like to achieve?
- DAX would be called "Cost per item by country"
- Logic: the MAX Value by country regardless of the date
-Use case: this would be used as helper measure to calculate total cost ([Cost per item by country] * [Volume] -> I need to take into consideration country differences in cost price, but no timeline/date differences (just use the max of the country).
- Anonymous4 years ago
I think your suggestion only does the removal of time/date in the formula. In my model the main problem is that it that current formula suggestions do not do the calculations by row/country if my table does not have a country in it. So if I dont have country the MAX chooses the largest number of the whole table (without considering the country differences) and thus the total result is wrong. If country is a dimension in the table, then it calculates it correctly.
I did a workaround and created just a helper table for this and it of course did the trick. Maybe some day I know how to work these around directly with DAX (or understand when it does not make sense to try to tackle with DAX).
2 Replies
- Fowmy
Super User
Anonymous
Use this measure:Cost per item by country = CALCULATE( MAX(Table[Cost per item]), REMOVEFILTERS(Table[Date]) )- AnonymousNot applicable
I think your suggestion only does the removal of time/date in the formula. In my model the main problem is that it that current formula suggestions do not do the calculations by row/country if my table does not have a country in it. So if I dont have country the MAX chooses the largest number of the whole table (without considering the country differences) and thus the total result is wrong. If country is a dimension in the table, then it calculates it correctly.
I did a workaround and created just a helper table for this and it of course did the trick. Maybe some day I know how to work these around directly with DAX (or understand when it does not make sense to try to tackle with DAX).