Time Intelligence Puzzle
Looking for a DAX formula that has some sort of dynamic PARRALLELPERIOD functionality. I have 2 periods as follows: "Period A" (a baseline if you will) and "Period B". Period A spans 3.5 months. Period B is all dates after Period A ends. When a user selects a date range in Period B for a measure (e.g., Sales) I want the DAX formula to compute the average Sales for all exactly matching periods within Period A. For example, lets say the user randomly picks 11 days is Period B (like Sunday June 5 thru Wednesday June 15). And let's say Sales for that period equal 10,050. And lets say Period A begins on March 1 and ends on May 15. I would want to compare the 10,050 to the average of each matching 11 day period within Period A. A matching period (within Period A) would have to start on Sunday and end on the Wednesday 11 days later. Specifically, based on my logic, the periods in this particular example are as follows: 3/6-3/16, 3/20-3/30, 4/3-4/13, 4/17-4/27, and 5/1-511. Each of those 5 periods have different total Sales. I then want the average of those 5 sums. That average would then be comparable to the 10,050 for 6/5-6/15. Again, this formula would have to dynamically support any range of dates a user might select within Period B whether that be 1 day, 5 days, 10 days, 13 days or whatever. Lastly the model is a simple transactional sales fact table attached to a calendar table.