Replace the data of the main window by a new data set. Each point of the new data set is the average of the original value of the point and the y-values of the two neighoured points. Applying this action several times leads to a considerable smoothing of the data.