Abstract:
|
The scatter-plot matrix is a useful instrument for assessing visually the strength of the linear association that may exist between the columns of a given data matrix. It seems, however, that its use for representing graphically the process of data reduction performed by Principal Component Analysis (PCA) has not been yet fully explored in the literature. This communication presents a possible method for constructing a scatter-plot matrix that can be a complement the usual graphical methods in PCA, as for instance the scree plot, or the standard 2D component plots. The analytical tools required are some functions included in the package gclus of R.
|