Hello Power BI Community,
I have an model where we are trying to calculate contract optimization where we take all transactions and compare them against alternatives. When analyzing these alternatives, we look at volume (A) on a MTD and YTD basis by supplier, where the MTD/YTD dates are all dates before the date of the transaction or, alternatively, the RowIndex before the transaction's RowIndex. Then, after analyzing A, if conditions are met (B) the report will say whether the original transaction supplier was the best option or if we should have selected a different supplier (C). Because of the MTD and YTD criteria, if we suggest the supplier is different, we want to track how much volume we are changing from the original supplier to the new supplier (if C is not the original supplier, flag as changed volume (D)). To properly calculate MTD and YTD on the conditions met column (B), we then have to reference A (original volume by supplier) and D (changed volume by supplier) but we only want the totals for all rows before the current row we are trying to calculate. However, we get a circular dependency on a columnar level, because we have a calculation step that looks like A > B> C > D > B. We tried things like CALCULATE(CALCULATE(sum(Volume),Supplier=earlier(Supplier)),RowIndex<earlier(RowIndex) as well as trying to emulate Dany Hoter's LookUpValue solution, but to no avail so far. All described steps are calculated columns as analysis is done at a row (transactional) level. Any insight would be very helpful and appreciate your time,
Shawn