If you use a correlation (or covariance) matrix as input to a procedure, the procedure might give the following error:

If you used a procedure or SAS-supported macro to create the correlation (or covariance) data set, the most common cause of the error is a missing value in the correlation (or covariance) matrix. Run PROC PRINT to look at the matrix to see if any of the values are missing. Sometimes a correlation (or covariance) is missing because there is not enough variability in the raw data, or because there are too many missing values in the raw data.

If you used a DATA step to create your own matrix, the most common cause of the error is a spelling mistake. Run a PROC PRINT of your matrix to carefully verify that the names of the variables exactly match the values of the _NAME_ variable. Make sure that the length of the _NAME_ variable is long enough for your variable names. If the _NAME_ value truncated your variable name, then use a LENGTH statement to increase the length of the _NAME_ value.

For more information, see the chapter Special SAS Data Sets in SAS/STAT User's Guide.

Operating System and Release Information