Rolling 4 week average conversion rate

[Jacob1832]

I am trying to create a rolling 4 week average conversion rate. The column LTA is the conversion rate and equals (Appts/Leads). RIght now, LTA is week by week. I need to create a new column that is a 4 rolling conversion rate. 
Here is the data.

So if we started at the bottom, the RollingAvg for May 29 would be (19+19+32+38)/(25+45+60+82) = 50.943 %

For the week may 22, the numbers would roll back one week, so itd be (19+32+38+0)/(45+60+82+117) = 29.276 %

Help would be appricated.

[amitchandak · Accepted solution]

@Jacob1832 , Try a new column like

new column =
var _date = [Week]
return
divide(sumx(filter(Table, [Week] > _date -28 && [Week]<= _date ),[Appts]),sumx(filter(Table, [Week] > _date -28 && [Week]<= _date ),[leads]))

[Jacob1832]

How would i adjust the calculation if i had a single number (Actuals) and a grouping variable (state)?
The rolling average would still have the same numerator, the denominator would now be 4 since its a 4 week rolling average and not a conversion rate. How would I incorporate the group of state?