# Estimating the values of unknown subpopulation areas with only the known total population values in R

I probably should have posted it here first, but I have the full details listed on stackoverflow.

The Problem: If I have a count of 2500 County deaths and they occurred within one of 3 smaller areas (zip codes) within The County, how do I estimate the number that occurred within each of those 3 areas. Given that area 1 is 17.5%, area 2 is 55%, and area 3 is 27.5% of the total population area of the county. I'm ultimately trying to calculate the death rate of the smaller areas but only have the larger County area.

Another way to ask this: Is if I have 2500 balls and drop them, and it will land in 1 of 3 bins, how can I estimate how many will fall within each of the bins being that each bin is different sizes. Would RandomForest be appropriate to use in R to determine this? Thank you for your help.

• This cannot be solved without additional information about the smaller areas. Right now, all you could say is that 17.5 % of people died in area 1, 55 % in area 2 and 27.5 % in area 3, if you are willing to assume that deaths are uniformly distributed. – user2974951 Jan 30 at 8:55