Alright! That help was greatly appreciated! Thank you! :)

Alright! I think that Robert Israel described exactly what i was looking for, thou I thought that it would be a lot easier, as my book (A First Course In Mathematical Modeling) describes this as the chebyshev approximation, and make it sound like its something common, in the same way as Least Square and Polynomial Interpolation, witch made me think there would be a function already in place for this... Now i just gotta spend some days with the code to figure out how it works.. ;) Thx a lot for all the help guys, I really appreciate it!

Thanks for the replys. To explain why im doing this it is for a homework where the mission was to make a curve that fits with some data, and we had to compare the chebyshev approximation, Least Squares Crierion and the Polynomial Interpolation, and then say witch one we prefered and why. Least Squres and Polynomial is easy to do with maple and compare, but the Chebyshev got me stranded... It seems like theres a lot of different stuff involving Chebyshev, it feels like im after something like this: > with(stats); > fit[chebyshev[[x, y]]]([[1, 2, 3, 4], [2, 8, 6, 15]]); except, that function does not exist.. :/ The book im working with states that the chebyshev approximation determins the parameters of the function type y = f(x_i) that minimizes the number Max |y_i-f(x_i)| where i = 1,2,...,m.. ( the _i means that the i is in subscript/lowered or what do you call it?) I also found a this saying that: Minimax Polynomial: It is closely approximated by the Chebyshev polynomials of the first kind. So maybe what im after is called something with Minimax Polynomial?