I have created a table and need a formula to calculate the values in a vacant column. The relevant columns in the table are shown below. I want to calculate the Variance, which is the the difference between "Budget" and "Actuals to Date" plus "Forecast to Complete", as a percentage.
The Budget, Actuals to Date and Forecast to Complete all come from different data tables. The Budget and Actuals are sums and the Forecast is filtered to only show the latest amount.
Any help appreciated.
Solved! Go to Solution.
It will be easier to assist if you can share the PBIX file with such a scenario present.
It will save a lot of time involved with re-creating the scenario from our side to satisfy your requirement
I appreciate your help on this matter but I am not in a position to share the PBIX file.
I take that the solution is not a simple matter of a formula that uses the values in the table?
Are Budget, Actuals to Date and Forecast calculated columns/measures or original fields in source data tables? You could provide some dummy data to make the source table structure and relationships more clear.
I have extracted data from the tables to try to explain what is required. These three tables are where the base data is drawn.
Table 1 - Budget
|IdentifierId||Funding Status||Funding Type||Budget|
Table 2 Actuals
|IdentifierId||Actuals to Date||Expense Type|
Table 3 - Forecast
|Title||IdentifierId||Forecast to Complete|
|Forecast for period 1Aug2018 to 31Aug2018||5||14112|
|Forecast for period 1Aug2018 to 31Aug2018||9||289893|
The report shows the Budget as sum of CAPEX and OPEX amounts, with a Status of "Released" for each identifier, Actuals as the sum of all amounts for each of the identifier, and for Forecast, the table is filtered so only the most recent amount for each identifier is shown. For project 5 it would look like this:
I want to calculate the variance between actual costs plus what is forecast and the budget, which is calculated as "(Budget/(Actual + Forecast"))/Budget". For project 5 this calculates as 0%.
If this is too hard to do, I will fall back to Plan B and export the report to excel and add the variance to the report there.
Hope this helps
Went through your data. Can you also tell what is the expected Vaiance for your Identifier 5?
As far as I tried, the value comes as 0. But if you look deeply, it is not 0%. It is something like 0.00070581%
Going by the formula which you have given, for Identifier 5,
Budget = 151644
actuals + Forecast = 127569+14112 = 141681
The Variance Calculation would be something like this
(151644/ 141681) / 151644 = 0.0000070581
When multiplied by 100, it becomes 0.00070581%
I am afraid going by this formula, even excel is going to give the same value.
Click here for the PBIX with the same shown
Clarify if I am missing something here...?
While the solution proposed was correct, I have not been able to make it work in real life. My actual data is more complex than the example used to devise the solution and I think there is an issue with the relationships between the tables that is causing the problem.
Here is the relationship chart for the example data:
This is the chart for the actual data:
The extra table is the only real difference between the charts and it means I cannot have the same relationships between the tables.
Any help appreciated.
Try this approach. Not sure how this will suit for your data.
If possible can you remove the join between the Actual Expense and Initiative Register. That should be aloowing you to make the join beween Actual Expense and Forecast as active. My guess is it will not affect the relationship between Actual Expense and Initiative Register as these tables will continue to be related through Forecast table
Note: Keep the join between them Bi-directional, so that the Initiative Register table continues to be in the model and respond for any changes that happen with the Forecast Table while slicing and Dicing the data
Thanks, but that didn't work. My data may be more complex than the example suggests. Working on a solution, I have got to this stage where I can make the formula work used as a measure, but it throws up a large number of extra rows with incomplete data. Refer extract shown below.
|Initiative Identifier||Initiative Name||Percent Complete||Budget||Actuals to Date||Forecast to Complete||Variance|
This pattern is repeated for each project.
Interestingly, if the same formula is included as a column, the above pattern is not repeated, but the calculation is incorrect. It is -100% for each project.
My formula is:
Variance = calculate(divide(sum('Actual Expense'[Expense Amount])+sum('Forecast Expense'[Forecast Amount]),sum('Initiative Funding'[Amount]))-value(1)).
This seems to be some filtering issue.