An Example of the Use of Andrews' Plots to Detect Time Variations in Model Parameters and Outlying Observations

摘要

Andrews (1972) introduced a method of plotting high-dimensional data in two dimensions. In his method, Andrews represents each multidimensional point by a Fourier function. The clustering of plots of these functions is equivalaent to the clustering of the multidimensional points. Andrews method is exploited as a graphical tool for exploratory data analysis for the examination of changes over time in the parameters of a time series model. An example using the total Canadian unemployment figures from 1956-1975 is used to illustrate the method. These data have four spurious (outlying) observations and it is shown how these may be detected by the use of Andrews' plots.