Hi Paige, Sorry for the confusion. Still new in posting questions here. Your answer is kind of what I need, but I also need to use the forecasted moving average data for the future month forecast, let me show your an example: for proportion data, I only have actuals till 1/8/2019, then I want to have the three month moving average for Sep'19 to Dec'19, using forecasted data as an input too. date proportion moving_average
1/4/2019 0.1
1/5/2019 0.2
1/6/2019 0.3
1/7/2019 0.25 0.2 (average of 0.1, 0.2, 0.3)
1/8/2019 0.5 0.75 (average of 0.2, 0.3, 0.25) 1/9/2019 0.35 (average of 0.3, 0.25, 0.5) 1/10/2019 0.37 (average of 0.25, 0.5, 0.35(forecasted)) 1/11/2019 0.41 (average of 0.5, 0.35(forecasted), 0.37(forecasted)) 1/12/2019 0.38 (average of 0.35, 0.37, 0.41) I tried proc expand, which gives exactly the result you offered. I also tried proc arima to compute the moving average. The interesting part by using proc arima is that, for the computed moving average, there will be a slightly difference comparing the result that I manually computed in excel (in 10e-4 level). Since this ratio is needed to multiply a big number (10 billion level), a slight difference will produce a big number after that. I am wondering if any other simple way we can do the moving average with the forecasted data?
... View more