Canonical Correlation with matrix input

Problem

I am attempting to use the macro provided with SPSS, 'Canonical correlation.sps', which I copied into my directory 'C:\spssjobs\', . I do not have the raw, case-level data, so I used the MATRIX DATA command to read a lower-triangular correlation matrix into an SPSS matrix data file. Here are my commands: TITLE 'CANONICAL CORRELATION'. MATRIX DATA VARIABLES = Y1 to Y6 X1 to X6 /N=300 / CONTENTS=CORR. BEGIN DATA. 1.0 .71 1.00 .12 .12 1.00 .34 .45 .43 1.00 .27 .33 .24 .33 1.00 .21 .11 .42 .46 .32 1.00 .13 .27 .21 .39 .27 .38 1.00 .18 .24 .15 .36 .33 .26 .62 1.00 .08 .14 .09 .25 .13 .23 .44 .66 1.00 .10 .16 .09 .25 .12 .28 .58 .66 .64 1.00 .13 .23 .42 .50 .41 .47 .46 .47 .37 .53 1.00 .08 .15 .36 .28 .21 .26 .30 .24 .19 .29 .58 1.00 END DATA. LIST. INCLUDE 'C:\spssjobs\Canonical correlation.sps'. cancorr SET1=y1 to y6 / SET2=x1 to x6/. The results don't look right. Something is clearly wrong with the numbers, as the output correlation matrix doesn't match what was specified as input. Is there a way to run a canonical correlation on aggregate data in SPSS?