Hey Suzzy, 

I would create a few support metrics to make the solution easier to maintain. Here are the metrics that I suggest you create:

• MET_INP_Start_of_Forecast_Calculation
  • No dimension (although if you need something version specific then I would also add version)
  • Type is the month dimension
  • Value would be Jan 25
• MET_CALC_Start_of_Forecast
  • Month dimension (although if you need something version specific then I would also add version)
  • Type is boolean
  • Formula: IF(Month.’Start Date’ >= MET_INP_Start_of_Forecast_Calculation.’Start Date’, TRUE)
  • Output: this will give you boolean values for all months that are forecast months
• I assume that your “Actual?” metric already returns a boolean value for the actual months.
• Your final formula could then look something like this:
  • IF(“Actual?”, ‘Actuals’, IF(MET_CALC_Start_of_Forecast, ‘Forecast_Values’))
  • This would return a blank value for October to December

Let me know if you need help with the average calculations. 

Also a couple things I suggest to keep in mind:

• Avoid hard coding values into formulas. In my opinion it is better to create a metric where you for example input the value for beginning of forecast months as you can use a formula to create this value and you will not need to manually update formulas in the future
• Keep in mind how different versions may behave