Principal Components and Classification Analysis Overview

Method
The PCCA module computes the principal components, and a wide range of the associated statistics. This module is designed to solve large-sized problems. Two types of analyses are available, depending upon whether the data needs to be standardized or centered. In the former case, the analysis is carried out through the correlation matrix, while in the latter the analysis is carried out through the covariance matrix. The basic method, however, consists of diagonalizing the symmetric matrix: correlation or covariance. The special feature of this module is the graphics that provide visual aid for the classification of variables and cases.
Active and supplementary variables and cases
Another unique feature of this module is that you can specify active and supplementary variables and cases. Active variables and cases are used in the derivation of the principal components; the supplementary variables and cases can then be projected onto the factor space computed from the active variables and cases. These facilities make the PCCA module a powerful tool for classification and data mining.
Output
The PCCA module produces results in two forms: spreadsheets and graphs. While the spreadsheets can be used for interpreting the results, the associated graphs provide visual aid for the classification of variables and cases.
Numerical results
The PCCA module produces a wide range of results, such as factor coordinates of variables and cases, contributions of variables and cases, factor scores, factor score coefficients, cosine-squares, eigenvalues, and descriptive statistics.
Graphs
Recall that the main aim in the PCA is to recover a vector space of lower dimension onto which the original points (variables or cases) can be projected, so that the underlying structure of the data could be detected. In order to facilitate this, 2D plots of the factor coordinates can be produced in this module. This option is available both for variables and cases. The module also plots the eigenvalues of the correlation or covariance matrix of the active variables, that is, the Scree plot. Various 2D and 3D graphs related to descriptive statistics are also available. Some important plotting options are available for each graph.