Similar Matrices Have the Same Characteristic Polynomial
What You'll Learn
You'll prove that similar matrices—those related by B = P⁻¹AP—share the same characteristic polynomial and therefore the same eigenvalues.
Why This Matters
You've learned that similar matrices have the same eigenvalues (concept 2177). Now we'll prove why by showing they have identical characteristic polynomials. This is powerful: similarity transformations preserve the fundamental eigenvalue structure of a matrix.
The Core Idea
When two matrices are similar, one is just a "change of basis" view of the other. Since the characteristic polynomial encodes eigenvalues (which are intrinsic geometric properties), it shouldn't change when we simply look at the same transformation from a different perspective.
The Proof
Let B = P⁻¹AP where P is invertible.
The characteristic polynomial of B is:
det(B - λI)
Substitute B:
det(P⁻¹AP - λI)
Notice that λI = λP⁻¹IP (since P⁻¹P = I):
det(P⁻¹AP - λP⁻¹IP)
= det(P⁻¹(A - λI)P)
Using the determinant product rule (det(XY) = det(X)det(Y)):
= det(P⁻¹) · det(A - λI) · det(P)
Since det(P⁻¹) = 1/det(P):
= (1/det(P)) · det(A - λI) · det(P)
= det(A - λI)
The det(P) terms cancel! Thus B and A have identical characteristic polynomials.
What This Means
Since eigenvalues are roots of the characteristic polynomial, similar matrices have:
- The same eigenvalues (with the same algebraic multiplicities)
- The same trace
- The same determinant
Key Takeaway: Similar matrices B = P⁻¹AP share the same characteristic polynomial because the transformation matrix P cancels out in the determinant calculation, proving that eigenvalues are invariant under similarity transformations.