Hi

I have to generate a covariance matrix from a data set for input to
another statistical program (LISREL). How does arrange for SPSS to
output this covariance matrix to a file?

Thanks

Max

Try:

CORRELATIONS varlist
 /MATRIX=OUT(*).
MCONVERT.
SELECT IF ROWTYPE_='COV'.
SAVE OUTFILE=filename /DROP=ROWTYPE_ VARNAME.                                   

Where, 'varlist' is the names of variables and 'filename' is the name
of the file which will hold the covariance matrix.


How to Create a File with a Covariance Matrix
http://www.asu.edu/it/fyi/dst/helpdocs/statistics/spss/tips/cov_matrix.html


The most current version of LISREL (v.8.54--05/2003) uses the Import
Data in Free Format option in order to:

Read comma separated files (.csv) 
Read Tab-delimited files (.txt) 
Read SPSS for Windows (.sav) files 

LISREL Upgrade Patches are available at:
http://www.ssicentral.com/other/lispatch.htm




Alternative Methods
-------------------

Export Co-Variance Matrix to ASCII
http://pages.infinit.net/rlevesqu/Syntax/Matrix/ExportVarianceCovarMatrixToASCIIfile.txt

Export CoVariance Matrix to Save File
http://pages.infinit.net/rlevesqu/Syntax/Matrix/ExportVarianceCovarMatrixToSavFile.txt

MACRO To Export Covariance Matrix of Variables to Text File
http://www2.jura.uni-hamburg.de/instkrim/kriminologie/Mitarbeiter/Enzmann/Software/covmat.sps


I hope this helps. 

---l

---

Request for Answer Clarification by maxb_max-ga on 29 Sep 2004 01:10 PDT

As a matter of interest, would there have been a way of achieving this
without using SPSS syntax e.g. from the menu, or using an SPSS output
file?

Kind regards

Max

---

Clarification of Answer by larre-ga on 29 Sep 2004 08:41 PDT

If so, I'm not familiar it. I'm pretty sure a custom application could
be written to convert between the formats, but several additional
searches didn't manage to locate a universal independent conversion
utility.

----------------------------------------------------------------------


You're quite welcome. My thanks to you for the kind words, rating and
tip. Glad to be of service.  ---l

What can I say - this is the perfect answer, complete with macro
example. Truely, I searched Google for days to find this - worth every
penny (or cent depending where you live!)

Many, many thanks.

Max