 Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.

# st: Problem computing eigenvalues, STATA vs. MATA

 From Zachary Neal To statalist@hsphsun2.harvard.edu Subject st: Problem computing eigenvalues, STATA vs. MATA Date Sun, 9 Jan 2011 08:08:45 -0500

```I am trying to obtain the eigenvalues of a symmetric matrix.  However,
I get different results depending on whether I use STATA or MATA.  For
example:

matrix define A = 0,1,1,0,0 \ 1,0,0,1,0 \ 1,0,0,0,1 \ 0,1,0,0,0 \ 0,0,1,0,0
matrix symeigen eigenvalues1 eigenvectors1 = A
matrix list eigenvalues1

mata: A = 0,1,1,0,0 \ 1,0,0,1,0 \ 1,0,0,0,1 \ 0,1,0,0,0 \ 0,0,1,0,0
mata: eigenvalues2 = 0
mata: eigenvectors2 = 0
mata: symeigensystem(A,eigenvalues2,eigenvectors2)
mata: eigenvalues2

In this case, eigenvalues1 does not equal eigenvalues2.  I believe the
results yielded by STATA (i.e. eigenvalues1) are what I'm looking for.

Why do these two sets of commands yield different results?  What
commands are necessary in MATA to yield the same eigenvalues that are
given by STATA?

Thank you
Zachary Neal

--
Zachary Neal, PhD
Department of Sociology
Michigan State University
zpneal@msu.edu
http://www.msu.edu/~zpneal
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
```