Forum Discussion
Variance vs. Budget and Filters
I have a table of projects, expenses and budgets. I wanted to calculated variance from budget then do some filtering on those that are say 10,000 under budget. I have run into a few problems
I used a new column of Variance = table1[actuals ytd] - table1[budget ytd]
However if there are multiple lines for a project I get the variance per line when I need it aggregated at the project level.
I get:
Proj1 Amount1 variance 1
Proj1 Amount2 variance 2
Proj1 Amount3 variance 3
Proj2 Amount4 variance 4
Proj2 Amount5 variance 5
Proj2 Amount6 variance 6
Proj2 Amount7 variance 7
I really just need Project 1 and Project 2 the variances aggregated.
Seems like it should be simple, but looks to be hard.
then after i would want to be able to filter on variance totals.
Any thoughts?
Alan
No that should work. If you put a chart/table in your report with Project Number as axis and the measure as the value, it will calculate that measure for each Project Number as if that were a filter applied to the underlying data.
In more complex examples, there may be performance differences between calculated DAX columns and columns created in the query editor. The query editor generally will apply query folding and push the processing workload to the data source (where applicable), whereas the calculated DAX column happens within Power BI. DAX is way more flexible than the query editor, however, so my best advice is to do the ETL and mashing phase as much as possible in the query editor, but save the advanced pivoting and calculating for DAX.
5 Replies
I would create the calculated variance column in the query editor, then use the "Group By" function to group rows by Project and sum the variance column.
thanks for the fast reply....
I also just changed the calculated column to a measure and it seemed to work.
YTD Variance = sum(CurYear[Actuals YTD]) - sum( CurYear[Budget YTD])
am i missing something as to why this might be bad?
thanks again
Alan
No that should work. If you put a chart/table in your report with Project Number as axis and the measure as the value, it will calculate that measure for each Project Number as if that were a filter applied to the underlying data.
In more complex examples, there may be performance differences between calculated DAX columns and columns created in the query editor. The query editor generally will apply query folding and push the processing workload to the data source (where applicable), whereas the calculated DAX column happens within Power BI. DAX is way more flexible than the query editor, however, so my best advice is to do the ETL and mashing phase as much as possible in the query editor, but save the advanced pivoting and calculating for DAX.
If you just want a measure, you could use something like the following:
Proj1Variance = Calculate(Sum(TableName[Variance]),Filter(TableName, [Project] = "Proj1"))
