The HPPRINCOMP Procedure

PROC HPPRINCOMP Features

The main features of the HPPRINCOMP procedure are as follows:

  • supports a PARTIAL statement for analyzing a partial correlation or covariance matrix

  • supports a FREQ statement for grouped analysis

  • supports a WEIGHT statement for weighted analysis

  • produces an output data set that contains principal component scores and other observationwise statistics

  • produces an output data set that contains means, standard deviations, number of observations, correlations or covariances, eigenvalues, and eigenvectors

The HPPRINCOMP procedure implements the following algorithms:

  • eigenvalue decomposition, which uses the correlation or covariance of the data matrix and calculates all the principal components simultaneously

  • nonlinear iterative partial least squares (NIPALS), which uses the data matrix and extracts the principal components successively

  • the iterative method based on Gram-Schmidt orthogonalization (ITERGS) of Andrecut (2009), which uses the data matrix and extracts the principal components successively. The algorithm applies reorthogonalization correction to both the scores and the loadings at each iteration step.

Because the HPPRINCOMP procedure is a high-performance analytical procedure, it also does the following:

  • enables you to run in distributed mode on a cluster of machines that distribute the data and the computations when you license SAS High-Performance Statistics

  • enables you to run in single-machine mode on the server where SAS is installed

  • exploits all the available cores and concurrent threads, regardless of execution mode

For more information, see the section Processing Modes.