" wrote:

Note, that interpolating by equation fitting will not neccesarily match
the data points in the table - there will be some error. The only time
the polynomial will exactly match the data is if the polynomial order
is equal to the number of data points minus one. Increasing the order
of the polynomial can lead to some very wild lines though.

That is why (when interpolating) you only pass TREND or FORECAST the points
necessary for the interpolation, not the entire data set.

Jerry