TOPICS
Search

Geometric Multiplicity


The geometric multiplicity of an eigenvalue lambda of an n×n matrix A is the dimension of its eigenspace,

 g_lambda=dimker(A-lambdaI)=n-rank(A-lambdaI),

where ker denotes the null space, I is the identity matrix, and rank denotes matrix rank. It is the maximum number of linearly independent eigenvectors belonging to lambda.

If m_lambda is the algebraic multiplicity of lambda, then

 1<=g_lambda<=m_lambda.

A complex matrix can be diagonalized iff these multiplicities agree for every eigenvalue. For a real matrix to be diagonalizable over R, all its eigenvalues must additionally be real numbers.


See also

Algebraic Multiplicity, Eigenspace, Eigenvector, Matrix Diagonalization, Nullity

Explore with Wolfram|Alpha

References

Margalit, D. and Rabinoff, J. "Diagonalization." §5.4 in Interactive Linear Algebra. https://textbooks.math.gatech.edu/ila/diagonalization.html.

Cite this as:

Weisstein, Eric W. "Geometric Multiplicity." From MathWorld--A Wolfram Resource. https://mathworld.wolfram.com/GeometricMultiplicity.html

Subject classifications