MATRIX Subcommand (CLUSTER command)

MATRIX reads and writes IBM® SPSS® Statistics data files.

• Either IN or OUT and a matrix file in parentheses are required. When both IN and OUT are used on the same CLUSTER procedure, they can be specified on separate MATRIX subcommands or on the same subcommand.
• The input or output matrix information is displayed in the procedure information notes.

OUT ('savfile'|'dataset'). Write a matrix data file. Specify either a quoted file specification, a previously declared dataset (DATASET DECLARE), or an asterisk in parentheses (*). If you specify an asterisk (*), the matrix data file replaces the active dataset.

IN ('savfile'|'dataset'). Read a matrix data file. Specify either a quoted file specification, a previously declared dataset (DATASET DECLARE), or an asterisk in parentheses (*). The asterisk specifies the active dataset. A matrix file read from an external file does not replace the active dataset.

When a matrix is produced using the MATRIX OUT subcommand, it corresponds to a unique dataset. All subsequent analyses performed on this matrix would match the corresponding analysis on the original data. However, if the data file is altered in any way, this would no longer be true.

For example, if the original file is edited or rearranged, it would in general no longer correspond to the initially produced matrix. You need to make sure that the data match the matrix whenever inferring the results from the matrix analysis. Specifically, when saving the cluster membership into an active dataset in the CLUSTER procedure, the proximity matrix in the MATRIX IN statement must match the current active dataset.