Eigenvalues and Eigenvectors
Eigenvalues λ and eigenvectors v satisfy Av=λv — the matrix scales v without rotating it. For 2×2: λ²−tr(A)λ+det(A)=0. Sum of eigenvalues = trace; product = determinant. Essential for PageRank, PCA, vibration analysis.
Why This Mathematical Concept Matters
Why: Eigenvalues determine matrix behavior: stability of dynamical systems, principal directions in PCA, natural frequencies in vibration. Google PageRank uses the dominant eigenvector.
How: Solve det(A−λI)=0 for eigenvalues. For each λ, solve (A−λI)v=0 for eigenvectors. 2×2: λ=(tr±√(tr²−4·det))/2. Trace and determinant give quick checks.
- ●Eigenvectors define directions that the matrix only scales.
- ●Complex eigenvalues occur for rotation-like transformations.
- ●Diagonalizable matrices: A=PDP⁻¹ where D has eigenvalues.
📐 Real-World Examples — Click to Load
Matrix Entries
Eigenvalue Magnitudes
Trace Decomposition (Eigenvalue Contribution)
📐 Calculation Breakdown
Eigenvectors
⚠️For educational and informational purposes only. Verify with a qualified professional.
🧮 Fascinating Math Facts
PageRank: dominant eigenvector (λ=1) of link matrix ranks pages.
PCA: eigenvectors of covariance matrix are principal components.
— ML
📋 Key Takeaways
- • Eigenvalues λ satisfy Av = λv — the matrix scales the eigenvector without rotating it
- • For 2×2: λ² - (a+d)λ + (ad-bc) = 0 where trace = a+d, determinant = ad-bc
- • Sum of eigenvalues = trace; product = determinant
- • Complex eigenvalues occur when discriminant < 0 (e.g., rotation matrices)
💡 Did You Know?
📖 How It Works
The eigenvalue equation Av = λv means: when you multiply the matrix A by an eigenvector v, you get a scalar multiple of v. The scalar λ is the eigenvalue. Geometrically, A only stretches or shrinks v; it does not rotate it.
To find eigenvalues: solve det(A - λI) = 0. For 2×2, this gives the quadratic λ² - tr(A)λ + det(A) = 0. Use the quadratic formula. For each eigenvalue, solve (A - λI)v = 0 to get eigenvectors.
🎯 Expert Tips
Symmetric Matrices
Real symmetric matrices have real eigenvalues and orthogonal eigenvectors. Essential for PCA.
Rotation Matrices
90° rotation [[0,-1],[1,0]] has eigenvalues ±i — complex eigenvalues indicate rotation.
Stability
For dynamical systems: |λ| < 1 means stable; |λ| > 1 means growing/unstable.
Trace & Determinant
Quick check: λ₁+λ₂=tr(A), λ₁λ₂=det(A). Use to verify your eigenvalues.
📊 Quick Reference
| Property | Formula |
|---|---|
| Trace | tr(A) = a₁₁ + a₂₂ |
| Determinant (2×2) | det(A) = a₁₁a₂₂ - a₁₂a₂₁ |
| Characteristic (2×2) | λ² - tr·λ + det = 0 |
| Eigenvalue sum | λ₁ + λ₂ = tr(A) |
| Eigenvalue product | λ₁ · λ₂ = det(A) |
❓ FAQ
What does Av = λv mean?
The eigenvector v, when multiplied by A, is only scaled by λ. The direction of v is unchanged (or reversed if λ < 0).
When do complex eigenvalues occur?
When the discriminant tr² - 4·det < 0. Common for rotation matrices. They come in conjugate pairs a±bi.
How is this used in data science?
PCA: eigenvectors of the covariance matrix are principal components; eigenvalues give variance explained.
What is the geometric interpretation?
Eigenvectors are directions that the matrix only stretches. Eigenvalues are the stretch factors.
Why does Google use eigenvalues?
PageRank finds the eigenvector for λ=1 of the link matrix — the steady-state distribution of "importance."
Can a matrix have no eigenvalues?
Over the reals, 2×2 can have 0 or 2 (real) eigenvalues. Over complex numbers, every n×n matrix has n eigenvalues.
📚 Official Sources
⚠️ Disclaimer: This calculator supports 2×2 matrices (full) and 3×3 (characteristic polynomial, numerical eigenvalues). For research or production, verify with MATLAB, NumPy, or similar. Complex eigenvector display is simplified for 2×2.