The invention relates to fitting a curve to a plurality of data points. A "seed curve" is determined from a first set of data points selected from the plurality of data points. From the remaining data points, data points are individually selected and a determination is made for each selected data point as to whether the data point is acceptable to be included with the first set of data points. When a data point is determined to be acceptable, the data point is included with the first set of data points to form another set of data points. After each of the other data points are evaluated for inclusion with the first set of data points, a best fit curve is determined from a final set of data points.