diagonalization II
Summary
Recognising eigenvalues
| Situation | Conclusion |
|---|---|
| Triangular matrix | Eigenvalues are the diagonal entries |
| Diagonal matrix | Already diagonalized |
| Diagonalizable | |
| Nonreal eigenvalues | Cannot diagonalize over |
| Only one eigenvalue | Diagonalizable iff |
| Matrix is not invertible | |
| Invertible matrix | May still fail to be diagonalizable |
Concept
Eigenvector and eigenvalue
- eigenvector - vector that is only scaled by the linear operator
- eigenvalue - scalar that the vector is scaled by
Eigenspace
Characteristic polynomial
Diagonalizability
Formally
diagonalizing is just finding the eigenbasis
Diagonalizability test
- (3.2.18)
- algebraic multiplicity - degree of the root
- geometric multiplicity - dimension of the eigenspace
Powers of matrix
Similarity
must be invertible
Revering diagonalizability
Similarity is just change of coordinates
essentially, two matricies are similar if they represent the same transformation
Similar matricies share properties
- characteristic polynomial
- diagonalizability
Application
Diagonalizing a linear map
Eigenvector in
Characteristic polynomial of linear operator
- no roots = no eigenvectors
- two roots
Interesting linear operator
Fixed point
- special case where eigenvalue is 1
Properties from characteristic polynomial