I have several data points from which I'm trying to calculate a formula:
- Column 1 (
"I") is a number (inclusively) between 0 and column 2. - Column 2 (
"C") is a fixed value in each case. - Column 3 (
"R") is the ratioI / C - Column 4 (
"P") is what I'm trying to find out. - Both column 3 (
"R") and column 4 ("P") will always fall in the range0to1and can therefore be expressed as percentages.
0 | 16 | 0.000 | 1.000
1 | 16 | 0.063 | 0.518
2 | 16 | 0.125 | 0.268
3 | 16 | 0.188 | 0.139
4 | 16 | 0.250 | 0.072
5 | 16 | 0.313 | 0.037
6 | 16 | 0.375 | 0.019
7 | 16 | 0.438 | 0.010
8 | 16 | 0.500 | 0.005
9 | 16 | 0.563 | 0.003
10 | 16 | 0.625 | 0.001
11 | 16 | 0.688 | 0.001
12 | 16 | 0.750 | 0.000
13 | 16 | 0.813 | 0.000
14 | 16 | 0.875 | 0.000
15 | 16 | 0.938 | 0.000
16 | 16 | 1.000 | 0.000
The data above is limited to 3 decimal places. I can see that the relationship is exponential in nature, but I don't know exactly what the relationship is.
Is there a process that I can use to determine the formula based on these data points?
EDIT: When graphed out, I believe the above data points form a logarithmic graph.
Choosing an exponential model, the best fit would be $$P=0.98\cdot e^{-0.65348I}=0.98\cdot0.52^I$$ with a remarkable correlation of $0.98$.