APPLIED MULTIVARIATE STATISTICS: WITH SAS® SOFTWARE

2.3. Profile Plots

One of the simplest ways of representing p-dimensional measurements is by using profile plots. These plots are the polygonal representations of p-dimensional observation vectors. Each p-dimensional observation vector is represented by p points with the vertical coordinate of each point proportional to the value of the corresponding variable. The successive points are joined using straight line segments. The resulting curve is called the profile of that observation. These plots can be very helpful in identifying clusters of the observations and outliers. Many times it may be more meaningful to plot the standardized variables in order to have a uniform scale for each variable. The standardization of variables can easily be achieved ...