How to ignore nonexistent rows when using multi-row formula tool

[eodhiambo]

I am using a multi-row formula to calculate averages over a range of rows, grouped by values in another row. I am averaging values backward - i.e., from the active row (selected by a certain criteria and usually not the first row), I am averaging the previous 10 rows. For some groups, there are less than 10 rows. Alteryx is treating those nonexistent rows as zeros and therefor "undercalculating" my averages. Is there a way to instruct Alteryx to ignore non-existent rows? Thank you!

If you do open the data attached, I am trying to calculate the average score 10 days before a score of zero (or all days if days before zero are less than 10) 

[Tyro_abc]

Hi 

Can you check if the attached workflow is as per your expectation?  I added one constant field with a value of 1 and added them to get the count of records till we hit the last 10 rows limit and then divided the total score with that count. 

Regards

Arundhuti