Hi guys,

how can I calculate the rank of a symmetric 3x3 matrix with positive values on the diagonal? Also, can I get a rank as a byproduct of applying SelfAdjointEigenSolver?

Regards,
Dženan

I guess one method is described here
Documentation Chapter 6
matse

Chapter 6 docs say: "Rank-revealing decompositions offer at least a rank() method."
Also, this table: http://eigen.tuxfamily.org/dox/TopicLin ... tions.html
suggest that SelfAdjointEigenSolver is rank-revealing. But the following code:

Eigen::SelfAdjointEigenSolver<EigenMatrix3> es(myMatrix);
std::cout<<"Rank: "<<es.rank();

gives error C2039: 'rank' : is not a member of 'Eigen::SelfAdjointEigenSolver'

So, I guess I would have to use FullPivLU to calculate rank.

For anyone stumbling onto this thread later:
For hermitian matrices, number of non-zero eigenvalues=rank. So no need for extra FullPivLU decomposition