- #1

- 68

- 0

I'm trying to use Excel to find a cosine function, f(x)=a cos n (x-b) + c, that best fits a whole lot of data points. I'm doing this using the Solver add-on in Excel. I have it set up to minimise the sum of the squared differences between each data point and the function's value. I'm pretty confident that its all set up correctly.

The problem is, solver does NOT do what it claims to do. It says it minimises the value in the cell by changing variables a, n, b and c but I can manually input variables that give a much lower value. Solver gives a different value depending on what is initially in the cells of variables a, n, b and c. I believe it is finding some kind of local minimum which is not the overall minimum.

Any ideas? I was hoping someone might have come across this problem with solver and know a workaround