Reputation: 449
I'm trying to Average aggregated(SUM) values, but my expression keeps doing weighted averages over the whole data set.
Table Structure
REGION SITE_ID MONTH QUANTITY
A 1 01 5
A 1 02 6
A 2 01 4
B 3 01 10
B 3 02 12
Expression
Avg(
Sum([quantity]) over (All([region]))/
UniqueCount([site_id]) over (All([region]))/
UniqueCount([month]) over (All([region]))
) over (All([region]))
To Clarify, I want to average A and B's Monthly Qty per Site
But I keep getting total qty divided by total no of site_ids divided by months
Upvotes: 1
Views: 5851
Reputation: 25112
This really depends on where you are going to use it and what the REAL data looks like. This should get you started. Insert this calculated column.
SUM([QUANTITY]) OVER (Intersect([REGION],[MONTH])) / UniqueCount([REGION]) AS [AvgOverRegionByMonth]
This could be inaccurate depending on how the rest of your data looks like. You can also accomplish this in a cross table. The expressions for the Sum and Avg on the example below are as follows:
Sum([QUANTITY]) as [Sum], Sum([QUANTITY]) / Count([REGION]) as [Average]
EDIT
In order to ONLY get the average over the months, use this forumla:
AVG([QUANTITY]) OVER ([MONTH]) as [AvgOverMonth]
Here is what your data will look like:
Upvotes: 0