This site uses different types of cookies, including analytics and functional cookies (its own and from other sites). To change your cookie settings or find out more, click here. If you continue browsing our website, you accept these cookies.
Sean,
I'm not a statistician, but I don't think calculating a moving average with too few data points is correct either. The "closest value" approach and your approach could both yield significantly skewed results. Using a stock chart as a familiar use case, the 50-day and 200-day moving average charts do not begin until there are a sufficient number of data points to calculate the average. With your solution, in Month 2 the 3-month and 6-month averages are the same calculation because only two data points are available. This does not seem right. Within Alteryx, I think the correct approach is to select 'NULL' for the 'Values for Rows that don't Exist' setting. This matches the behavior seen in the stock chart example.
Ken
My solution for Challenge 3.
As some of you already pointed out, the result set does not necessarily match those on the problem. The root cause seems to be that the result set seems to have been compiled without grouping the HP Category on the 6 month multi-row function.
Thus, I had difficulty matching the results.
I went through multiple iterations of Multi-Row Tools using IfElse statements with modifying each field for 3mo or 6mo. While this was the right idea, it was lengthy and inefficient. I took a look at what others were doing, and, unfortunately, got so stuck that I referenced the solution file. I have modified for a more efficient process and more like the solution.
See attached.