Eigenvalues & Eigenvectors: 2x2 Systems — Eigenvalues and Eigenvectors for 2x2 Systems (Intro) - Differential Equations

SECTION 1

Historical Context & Motivation

Imagine you need to predict how two quantities — say the populations of rabbits and foxes — change over time, and each quantity affects the other. Problems like these naturally lead to systems of differential equations, where multiple variables are intertwined. To untangle the system, mathematicians developed the concepts of eigenvalues and eigenvectors. The word "eigen" comes from German, meaning "own" or "characteristic." These ideas reveal the hidden, natural directions along which a system behaves most simply.

1743 • Euler & Rigid Bodies

Leonhard Euler studied rotating rigid bodies and encountered eigenvalue-like equations while analyzing principal axes of rotation — directions around which spinning is simplest.

1826 • Cauchy & Quadratic Forms

Augustin-Louis Cauchy formalized the idea of characteristic roots (eigenvalues) for symmetric matrices while studying geometry and the classification of conic sections.

1904 • Hilbert & Integral Equations

David Hilbert introduced the term "Eigenwert" (eigenvalue) in his work on integral equations, giving the concept its modern name and extending its use into infinite-dimensional spaces.

1960s–present • Computers & Modern Applications

With digital computing, eigenvalue algorithms became essential tools in engineering, physics, data science, and even internet search algorithms like Google's PageRank.

The central question that eigenvalues and eigenvectors answer is: when a matrix transforms a vector, are there special directions that stay the same — only getting stretched or flipped? Finding those directions and stretch factors is the key that unlocks the solution to a system of differential equations.

SECTION 2

Core Principles & Definitions

Before diving into calculations, let's build a solid understanding of what eigenvalues and eigenvectors actually mean. At the heart of these ideas is a simple relationship between a matrix and a vector. When you multiply a matrix A by a vector v, the result is usually a completely new vector pointing in a different direction. But for certain special vectors, the result points in the exact same direction — it's just scaled (stretched, shrunk, or flipped). Those are the eigenvectors, and the scale factor is the eigenvalue.

Eigenvalue (λ)

A scalar λ (lambda) such that Av = λv. It tells you how much the eigenvector is stretched or compressed. If λ is negative, the direction flips.

Eigenvector (v)

A nonzero vector v that satisfies Av = λv. It tells you which direction remains unchanged (except for scaling) when the matrix acts on it.

Characteristic Equation

The equation det(A − λI) = 0. For a 2×2 matrix, this produces a quadratic equation whose solutions are the eigenvalues.

Identity Matrix (I)

The 2×2 matrix with 1s on the diagonal and 0s elsewhere. It acts like the number 1 for matrices — multiplying any matrix or vector by I leaves it unchanged.

✦

KEY TAKEAWAY

Think of a matrix as a machine that pushes and rotates arrows (vectors). Most arrows come out pointing in a new direction. But an eigenvector is like a special arrow that passes through the machine and comes out pointing the same way — just longer, shorter, or flipped. The eigenvalue is the factor by which the machine stretches that arrow. It's like finding the one hallway in a funhouse where the mirrors only make you taller or shorter, but don't distort your shape.

SECTION 3

Visual Explanation

How a 2×2 Matrix Transforms Vectors

The diagram below shows what happens when the matrix A = [[2, 1], [1, 2]] acts on several vectors. Most vectors change direction, but two special vectors — the eigenvectors — only get scaled. The cyan arrow shows an eigenvector with eigenvalue λ = 3 (stretched to 3× its length), and the pink arrow shows an eigenvector with eigenvalue λ = 1 (unchanged in length).

The cyan arrows show the eigenvector [1, 1] — when A acts on it, the result is 3 × [1, 1], pointing the same direction but stretched by a factor of 3. The pink arrow shows eigenvector [1, −1], which stays the same since λ = 1. The violet dashed arrows show generic vectors that change direction after transformation — they are not eigenvectors.

Notice the critical difference. Generic vectors (shown in gray) get both stretched and rotated by the matrix, so their transformed versions (violet dashed) point in entirely new directions. The eigenvectors, however, stay on their original lines. This property is what makes them so powerful for solving differential equations: along eigenvector directions, complicated coupled systems reduce to simple, independent exponential growth or decay.

SECTION 4

Mathematical Framework

Now let's set up the algebra. Given a 2×2 matrix A, we want to find all scalars λ and nonzero vectors v such that Av = λv. This equation can be rearranged into a form that lets us use the determinant to solve for λ.

EIGENVALUE EQUATION

Av = λv

A = the 2×2 matrix, v = the eigenvector (nonzero), λ (lambda) = the eigenvalue (a scalar)

To find λ, rewrite Av = λv as Av − λv = 0, which factors as (A − λI)v = 0. Here, I is the 2×2 identity matrix. For a nonzero solution v to exist, the matrix (A − λI) must be singular — meaning its determinant equals zero.

CHARACTERISTIC EQUATION

det(A − λI) = 0

This equation in λ is called the characteristic equation. For a 2×2 matrix, it is always a quadratic equation in λ.

Let the general 2×2 matrix be A = [[a, b], [c, d]]. Then A − λI = [[a − λ, b], [c, d − λ]], and its determinant is:

DETERMINANT EXPANDED

(a − λ)(d − λ) − bc = 0

Expanding: λ² − (a + d)λ + (ad − bc) = 0. The quantity (a + d) is the trace of A, and (ad − bc) is the determinant of A.

COMPACT FORM

λ² − (tr A)λ + det A = 0

Use the quadratic formula: λ = [(tr A) ± √((tr A)² − 4·det A)] / 2. This gives two eigenvalues (which may be equal, real, or complex).

💡

Finding the Eigenvectors

Once you have each eigenvalue λ, plug it back into the equation (A − λI)v = 0 and solve for v. This typically involves setting up a simple system of two linear equations with two unknowns. Because the determinant is zero, the equations are dependent, giving you a family of solutions described by one free parameter — this is your eigenvector direction.

SECTION 5

Types of Eigenvalues for 2×2 Systems

The discriminant of the characteristic equation — that expression under the square root sign — determines what kind of eigenvalues you get. Each case leads to qualitatively different behavior in the corresponding system of differential equations. The diagram below summarizes the three main cases.

Three cases for eigenvalues of a 2×2 matrix, based on the discriminant Δ. Case 1 (Δ > 0) gives two distinct real eigenvalues and straight-line trajectories. Case 2 (Δ = 0) gives a repeated eigenvalue. Case 3 (Δ < 0) gives complex eigenvalues and oscillatory (spiraling) behavior.

In this introductory lesson we'll focus mainly on Case 1 — two distinct real eigenvalues — because it is the most common starting point and the easiest to visualize. Each eigenvalue pairs with its own eigenvector, giving you two independent directions that completely describe the system's behavior.

Discriminant	Eigenvalues	Behavior
Δ > 0	Two distinct real values λ₁ ≠ λ₂	Exponential growth/decay along two directions (node or saddle)
Δ = 0	One repeated real value λ₁ = λ₂	Degenerate node; may need generalized eigenvectors
Δ < 0	Complex pair α ± βi	Oscillation — spirals or circles in the phase plane

Summary of eigenvalue cases for 2×2 matrices

SECTION 6

Worked Example

Let's walk through a complete example. We'll find the eigenvalues and eigenvectors of the matrix A = [[3, 1], [0, 2]].

Find the Eigenvalues and Eigenvectors of A = [[3, 1], [0, 2]]

1
Step 1 — Write the Characteristic Equation

Compute A − λI = [[3 − λ, 1], [0, 2 − λ]]. Now set the determinant equal to zero: det(A − λI) = (3 − λ)(2 − λ) − (1)(0) = (3 − λ)(2 − λ) = 0.

(3 − λ)(2 − λ) = 0
2
Step 2 — Solve for the Eigenvalues

The quadratic factors neatly. Setting each factor to zero: 3 − λ = 0 gives λ₁ = 3, and 2 − λ = 0 gives λ₂ = 2. We have two distinct real eigenvalues.

λ₁ = 3, λ₂ = 2
3
Step 3 — Find the Eigenvector for λ₁ = 3

Substitute λ = 3 into (A − λI)v = 0. This gives [[0, 1], [0, −1]] · [v₁, v₂]ᵀ = [0, 0]ᵀ. Both rows say the same thing: v₂ = 0. Let v₁ = 1 (any nonzero value works). So the eigenvector is v₁ = [1, 0]ᵀ.

v₁ = [1, 0]ᵀ
4
Step 4 — Find the Eigenvector for λ₂ = 2

Substitute λ = 2 into (A − λI)v = 0. This gives [[1, 1], [0, 0]] · [v₁, v₂]ᵀ = [0, 0]ᵀ. The first row says v₁ + v₂ = 0, so v₁ = −v₂. Let v₂ = 1, then v₁ = −1. So the eigenvector is v₂ = [−1, 1]ᵀ.

v₂ = [−1, 1]ᵀ
5
Step 5 — Verify Your Answers

Check: A·v₁ = [[3,1],[0,2]]·[1,0]ᵀ = [3, 0]ᵀ = 3·[1, 0]ᵀ ✓. Check: A·v₂ = [[3,1],[0,2]]·[−1,1]ᵀ = [−3+1, 0+2]ᵀ = [−2, 2]ᵀ = 2·[−1, 1]ᵀ ✓. Both eigenvectors satisfy Av = λv.

Both verified: Av = λv ✓

💡

Pro Tip

Always verify your eigenvectors! Multiply A times your eigenvector and check that the result equals the eigenvalue times the eigenvector. This quick multiplication catches sign errors and algebraic mistakes.

SECTION 7

Strengths, Limitations & Common Pitfalls

The eigenvalue method is elegant and powerful, but like any tool, it has its sweet spots and its limitations. Understanding both will help you apply the technique correctly and know when to expect complications.

Strengths	Limitations / Pitfalls
Reduces a coupled system to independent equations — each eigenvector direction decouples.	Only works directly when the matrix has enough eigenvectors; repeated eigenvalues may require extra techniques (generalized eigenvectors).
Eigenvalues immediately reveal qualitative behavior: growth (λ > 0), decay (λ < 0), or oscillation (complex λ).	Complex eigenvalues require knowledge of complex numbers and Euler's formula, which may not yet be familiar.
For 2×2 systems, the characteristic equation is just a quadratic — you can always solve it by hand.	Arithmetic mistakes are common, especially sign errors when computing (A − λI) or the determinant.
Verification is easy: just check Av = λv.	Eigenvectors are not unique — any scalar multiple is also an eigenvector. Don't worry if your vector differs from the answer key by a constant factor.

Strengths and limitations of the eigenvalue/eigenvector method for 2×2 systems

✦

KEY TAKEAWAY

Eigenvectors are like the natural "compass directions" of a system. Just as north-south and east-west simplify navigation on a map, eigenvector directions simplify the motion of a system. Along each eigenvector direction, the behavior is just simple exponential growth or decay — no tangling with the other variable. The eigenvalue method works beautifully when you get two distinct real eigenvalues. Repeated or complex eigenvalues require additional tools, which you'll learn in upcoming lessons.

SECTION 8

Connection to Advanced Theory

The eigenvalue and eigenvector concepts you've learned for 2×2 systems are actually the foundation for much bigger ideas. The same approach scales up to 3×3, 4×4, or even n×n systems, and appears in nearly every branch of applied mathematics, physics, and engineering.

This Lesson (2×2 Intro)	Advanced Extensions
Characteristic equation is quadratic	For n×n matrices, it's a degree-n polynomial — may need numerical methods
Two eigenvalues, two eigenvectors	n eigenvalues and up to n eigenvectors; eigenspaces can be multi-dimensional
Constant coefficient systems: x' = Ax	Non-constant coefficient, nonlinear, and partial differential equations use local linearization around equilibria
Phase portraits: nodes, saddles	Stability analysis of equilibria in ecology, epidemiology, circuit design, and control systems

How 2×2 eigenvalue analysis connects to more advanced topics

In your next lessons, you'll see how eigenvalues and eigenvectors directly produce the general solution to a system of differential equations x' = Ax. The solution takes the form x(t) = c₁ $e^{λ₁t}$ v₁ + c₂ $e^{λ₂t}$ v₂, where each term represents motion along one eigenvector direction at a rate determined by its eigenvalue. Understanding this formula is the payoff for mastering the concepts introduced today.

SECTION 9

Practice Problems

Question 1: CONCEPTUAL

In your own words, explain what it means for a vector v to be an eigenvector of a matrix A. Why must v be nonzero?

Show answer

An eigenvector v of matrix A is a nonzero vector such that Av = λv for some scalar λ. This means that when A acts on v, the result is a vector pointing in the same (or exactly opposite) direction as v, just scaled by the factor λ. The vector v must be nonzero because the equation A·0 = λ·0 is always true for any λ, which provides no useful information. We need v ≠ 0 to identify the special directions where the matrix's action simplifies to pure scaling.

Question 2: BASIC CALCULATION

Find the eigenvalues of the matrix A = [[4, 0], [0, −1]]. (Hint: this is a diagonal matrix.)

Show answer

The characteristic equation is det(A − λI) = (4 − λ)(−1 − λ) − 0 = 0. Setting each factor to zero gives λ₁ = 4 and λ₂ = −1. For diagonal matrices, the eigenvalues are simply the diagonal entries. The eigenvector for λ₁ = 4 is [1, 0]ᵀ (along the x-axis), and the eigenvector for λ₂ = −1 is [0, 1]ᵀ (along the y-axis), since each axis direction is scaled independently.

Question 3: INTERMEDIATE

Find the eigenvalues and eigenvectors of A = [[1, 4], [2, 3]]. Show all steps.

Show answer

Step 1: tr(A) = 1 + 3 = 4, det(A) = (1)(3) − (4)(2) = 3 − 8 = −5. The characteristic equation is λ² − 4λ + (−5) = 0, or λ² − 4λ − 5 = 0. Step 2: Factor as (λ − 5)(λ + 1) = 0, giving λ₁ = 5 and λ₂ = −1. Step 3: For λ₁ = 5, solve (A − 5I)v = 0: [[−4, 4], [2, −2]] gives −4v₁ + 4v₂ = 0, so v₁ = v₂. Eigenvector: v₁ = [1, 1]ᵀ. For λ₂ = −1, solve (A + I)v = 0: [[2, 4], [2, 4]] gives 2v₁ + 4v₂ = 0, so v₁ = −2v₂. Eigenvector: v₂ = [−2, 1]ᵀ. Verify: A[1,1]ᵀ = [5, 5]ᵀ = 5[1,1]ᵀ ✓ and A[−2,1]ᵀ = [2, −1]ᵀ = −1[−2, 1]ᵀ ✓.

Question 4: APPLIED

Two competing bacterial colonies have populations x(t) and y(t) governed by the system x' = 3x + y, y' = x + 3y. Write the coefficient matrix A. Find its eigenvalues and eigenvectors, and describe what each eigenvalue tells you about the population dynamics.

Show answer

The coefficient matrix is A = [[3, 1], [1, 3]]. The characteristic equation is (3 − λ)² − 1 = 0, which expands to λ² − 6λ + 8 = 0. Factoring gives (λ − 4)(λ − 2) = 0, so λ₁ = 4 and λ₂ = 2. For λ₁ = 4: (A − 4I)v = 0 gives [[-1, 1],[1, -1]]v = 0, so v₁ = v₂; eigenvector v₁ = [1, 1]ᵀ. For λ₂ = 2: (A − 2I)v = 0 gives [[1, 1],[1, 1]]v = 0, so v₁ = −v₂; eigenvector v₂ = [1, −1]ᵀ. Both eigenvalues are positive, so both modes grow. Along [1,1] the colonies grow together (cooperating) at rate

e^{4t}

, and along [1,−1] they diverge from each other at rate

e^{2t}

. The faster-growing mode [1,1] dominates over time.

Question 5: CRITICAL THINKING

Consider a general 2×2 matrix A = [[a, b], [c, d]]. Prove that the sum of the eigenvalues equals the trace (a + d) and the product of the eigenvalues equals the determinant (ad − bc). What does this tell you if you already know the trace is 0?

Show answer

The characteristic equation is λ² − (a + d)λ + (ad − bc) = 0. By Vieta's formulas for a quadratic ax² + bx + c = 0, the sum of roots equals −b/a and the product equals c/a. Here the leading coefficient is 1, so λ₁ + λ₂ = (a + d) = tr(A) and λ₁ · λ₂ = (ad − bc) = det(A). If the trace is 0, then λ₁ + λ₂ = 0, which means λ₂ = −λ₁. The eigenvalues are equal in magnitude but opposite in sign. If both are real, this means one direction grows while the other decays at the same rate — this produces a saddle point. If det(A) > 0 and tr(A) = 0, then λ₁ · λ₂ > 0 combined with λ₁ = −λ₂ forces complex eigenvalues: λ = ±βi, giving pure oscillation (a center).

SUMMARY

Lesson Summary

Summary

An eigenvalue λ and eigenvector v of a 2×2 matrix A satisfy the equation Av = λv. To find the eigenvalues, set up the characteristic equation det(A − λI) = 0, which yields a quadratic in λ. Solve this quadratic (by factoring or the quadratic formula) to get the eigenvalues. Then substitute each eigenvalue back into (A − λI)v = 0 and solve the resulting system to find the corresponding eigenvector direction.

For a 2×2 matrix, the discriminant Δ = (tr A)² − 4·det A determines whether eigenvalues are distinct real (Δ > 0), repeated (Δ = 0), or complex conjugates (Δ < 0). Eigenvalues reveal the system's long-term behavior: positive eigenvalues mean growth, negative eigenvalues mean decay, and complex eigenvalues mean oscillation. Always verify by checking Av = λv!