Statistical Estimation Algorithm

Question

I'm not sure what this question is suitable for, but I will try to try it anyway. I have some data:

I also have a different set of data that seems to correspond to a similar distribution, but I only know the total percentage (for example, 30%, not 17%.) Can anyone suggest an algorithm for estimating% s for each individual level based on the new total% and initial distribution?

Answer 1

You doubt it. If you want to estimate the new total percentage by including the additional data that you receive, you must have the quantity associated with your percentage so that you can create a weighted average.

, , , , , . .

Wilcoxon-Rank , . , .

Answer 2

-1: 17% → 30%, () 105 → ~ 189.

-2:

, - Actual POssible. .

i.e., function (, total ()) = .

, ;)

Answer 3

If your new value is x, then put (22/627) * x as much as possible for level 1, and (21/627) * x as relevant for level 1, which will give you the same percentage as before for level 1. Then do the same for the other levels (possibly for level 2 (45/627) * x, etc.).