The line is preserved, but the arrow can reverse
An eigenvector is explicitly nonzero. Zero is nevertheless a valid eigenvalue.[1]
A matrix changes an eigenvector only by a scalar factor. Positive λ preserves its orientation, negative λ reverses it, and λ=0 maps it to zero. Thus “always points in the same direction” is too restrictive. For nonzero λ, v and Av lie on the same line. Every nonzero scalar multiple of an eigenvector belongs to the same eigenvalue.[1]
Eigenvalues describe special directions of a linear transformation and help with repeated applications, differential equations and oscillations. They are not simply the diagonal entries of every matrix. That shortcut works for triangular matrices; in general one needs the characteristic equation.[1]
Find λ, then solve for the eigenspace
A nontrivial nullspace requires A−λI to be singular.[1]
For a real 2×2 matrix, Δ≥0 gives real eigenvalues. For Δ<0, the pair is complex conjugate.[1]
After finding λ, solve (A−λI)v=0 and choose a nonzero vector from its solution space. Distinct eigenvalues have independent eigenvectors. A repeated eigenvalue, however, may have only one eigenvector direction. Algebraic multiplicity as a root of the characteristic polynomial does not automatically equal eigenspace dimension. The determinant first supplies the eigenvalue condition; nullspace and rank then determine the possible eigenvectors.[1]
Two eigenvector directions and a defective case
Symmetric matrix with two directions
- Given
- A has rows (2,1) and (1,2).
- Find
- Find the eigenvalues and one eigenvector for each.
Form the characteristic equation.
For λ=3 the system gives v2=v1; for λ=1, v2=−v1.
Check the vector equation directly.
Result: Eigenvalues are 3 and 1; their eigenspaces are the lines through the displayed vectors.
Check: Trace 4 equals their sum and determinant 3 their product.
Repeated value, only one eigenvector direction
- Given
- J has rows (2,1) and (0,2).
- Find
- Check whether two independent eigenvectors exist.
The characteristic polynomial has a repeated root.
The nullspace of J−2I requires v2=0.
Result: λ=2 has algebraic multiplicity 2 but a one-dimensional eigenspace.
Check: J(1,0)ᵀ=2(1,0)ᵀ, but J(0,1)ᵀ=(1,2)ᵀ is not a multiple of (0,1)ᵀ.
Recall the method later with the formula’s flashcards.
10 flashcards · English materials
Already have an account? Sign in for flashcardsAvailable from Essential. A ready-made formula set, no AI credits required. Article exercises are not all included.
Real eigenvalues of a 2×2 matrix
The specialized calculator returns the larger or smaller real eigenvalue of an entered 2×2 matrix. It checks the discriminant sign using canonical decimal values and explicitly rejects complex cases. The article obtains eigenvectors by solving a nullspace; the scalar calculator does not pretend to return an entire vector. Repeated values must not be interpreted as proof of two independent eigenvectors.[1]
Numerically, subtracting nearly equal terms in the quadratic formula can cause cancellation. A stable evaluation first computes the larger-magnitude root, then obtains the other through their product ad−bc when the first is nonzero. This changes neither eigenvalue and is not a general eigensolver for arbitrary matrix sizes.[1]
Apply the definition and handle special cases
Calculate · Foundation
A=diag(3,−2). Find eigenvalues and basis eigenvectors.
Show hint
Multiply the two standard basis vectors.
Learning goal: Interpret negative eigenvalues geometrically.
Show solution and calculation
Ae1=3e1 and Ae2=−2e2.
Thus λ1=3 with e1=(1,0), λ2=−2 with e2=(0,1).
The second arrow reverses.
Error analysis · Foundation
Is zero an eigenvector because A0=λ0?
Show hint
The definition contains an extra condition.
Learning goal: Distinguish zero vector from zero eigenvalue.
Show solution and calculation
No.
Eigenvectors must be nonzero; otherwise zero would satisfy the equation for every λ without providing information.
Eigenvalue λ=0 is allowed when a nonzero vector lies in the kernel.
Interpret · Application
A=2I also has repeated eigenvalue 2. How does its eigenspace differ from the second example?
Show hint
Here A−2I is the zero matrix.
Learning goal: Separate multiplicity from eigenspace dimension.
Show solution and calculation
Every nonzero vector is an eigenvector; the eigenspace is two-dimensional.
The same repeated eigenvalue can therefore have different eigenspace dimensions.
Transfer · Application
The matrix with rows (0,−1),(1,0) rotates by 90°. Does it have real eigenvectors?
Show hint
Form det(A−λI).
Learning goal: Distinguish real and complex scope.
Show solution and calculation
λ2+1=0 has no real root.
There are no nonzero real eigenvectors; over the complex numbers the eigenvalues are ±i.
The real-only calculator rejects the case.
Calculate · Application
For A with rows (4,2),(0,1), check v=(2,−3) for eigenvalue 1.
Show hint
Calculate both components of Av.
Learning goal: Verify a proposed eigenpair directly.
Show solution and calculation
Av=(8−6,−3)=(2,−3)=1·v.
The nonzero vector is indeed an eigenvector.
The triangular matrix’s other eigenvalue is 4.
Check your understanding with the formula’s reference test.
8 test questions · English materials
Already have an account? Sign in for the testAvailable from Essential. A ready-made formula set, no AI credits required. Article exercises are not all included.
A complete diagonalization calculation
An n×n matrix is diagonalizable over the chosen field when it has n independent eigenvectors. Place them as columns of P and put their matching eigenvalues in the same order in D. Then AP=PD and invertibility gives A=PDP−1. Repeated eigenvalues do not alone prevent diagonalization; eigenspace dimensions matter.
Columns (1,0) and (1,1) correspond to eigenvalues 2 and 3.
The check gives AP=[[2,3],[0,3]]=PD. With P−1=[[1,−1],[0,1]], PDP−1 recovers [[2,1],[0,3]]. Powers satisfy Aᵏ=PDᵏP−1; at k=2 this gives [[4,5],[0,9]], matching direct multiplication. A real symmetric matrix has an orthonormal eigenbasis, so P can be orthogonal and P−1=Pᵀ. For this nonsymmetric example P is not orthogonal.
Counterexample: [[1,1],[0,1]] has only eigenvalue 1 and a one-dimensional eigenspace. No invertible matrix of two independent eigenvectors exists. A formal D=I therefore cannot reproduce the matrix.
Sources for this explanation
Quanta · Updated November 6, 2026
Editorial responsibility: Quanta Study editorial team
Review scope
This explanation, 2 worked examples and 5 exercises are available in German and English. Structure, source references and mathematical notation undergo automated checks. Calculations, model assumptions and both language versions receive a separate AI review.
An independent, named human subject review has not yet been documented. Our exercise numbers are worked examples, rather than measurements published by the cited sources.
Editorial responsibility, sources and methodsChange history
- Authored a distinct DE/EN article with model limits, two worked examples and five varied exercises.
- A complete diagonalization calculation
Cite this reference page
Quanta. Eigenvalues and Eigenvectors. Updated November 6, 2026. https://quanta-study.de/en/formula-sheet/eigenwerte-vektoren
When citing a specific scientific statement, also cite the specialist source linked to that statement.