Sum ( Distinct Total <Region,Year> Total_Pop) resulting out of obejct memory

[#1]

Dear Experts:

I got an out of object memory when trying to use the set analysis code below.

Each Region have multiple cities. The Population is at city level per monthly basis. So in order to get the Percentage Population

I need to divide Population for that city / Overall Population in that Region (where that city existed)

Then I had to * the result with the GRP to get the GRP for that city alone.

SUM(P_18_49_GRP*DMA_A18_49_POP)/SUM(DISTINCT TOTAL <REGION,[Brcst Year]>DMA_A18_49_POP)

If I only do 1 year - everything will be fine - The problem existed when I do more than 1 year - It will give me out of object memory issues.

Thanks

Senjaya

[#2]

The week field is '2010-05-12' in varchar format -

I had the problem with the table 2 - that''s why I change it to : Week,State, Population - which force to to do the Sum(Distinct Total <Region, Year>) which cause me to have the out of object memory

Currently I try to add a hashkey128 ( Year, Region) on table 2

so it will become : Week, Haskey128(Year,Region), State, Pop

and then add the same haskey128) on table 1 as well - but I still have the same issue

[johnw]

Senjaya wrote:Table 1 : Week, State, GRP (at State level), Spending $
Table 2 : Year , State,, Population
Table 3 : State, Region
Table 4 : Week, Month, Year

Well, that's ONE of the three things I asked for, but it has a loop in it:

T1:State -> T2 State T2:Year -> T4:Year T4:Week -> T1:Week

So it's incorrect and therefore not helpful. I'm done wasting my and your time. I'm unsubscribing from the thread.