Linear Algebra in DEs — Linear Algebra in Differential Equations (Eigenvalues Connection—Intro) - Linear Algebra

SECTION 1

Historical Context & Motivation

Imagine you want to predict how a population of rabbits grows, how a hot cup of cocoa cools down, or how a bridge vibrates in the wind. All of these real-world situations involve differential equations — equations that describe how things change. For centuries, mathematicians tried to solve these equations one at a time, each requiring its own clever trick.

Then came a breakthrough: what if we could use the tools of linear algebra — matrices, vectors, and especially eigenvalues — to solve many differential equations at once using a single, organized method? This idea transformed science and engineering forever.

1690s • Newton & Leibniz Create Calculus

Isaac Newton and Gottfried Leibniz independently invent calculus, giving us the language of rates of change and the first differential equations.

1750s • Euler Solves Differential Equations Systematically

Leonhard Euler develops methods for solving differential equations, noticing patterns that hint at deeper structure beneath the solutions.

1826 • Cauchy Formalizes Eigenvalues

Augustin-Louis Cauchy introduces the idea of characteristic roots of a matrix — what we now call eigenvalues — laying the groundwork for connecting linear algebra to differential equations.

1900s • Matrices Meet Differential Equations

Mathematicians realize that systems of differential equations can be written as matrix equations, and eigenvalues provide the key to unlocking their solutions.

Today • Eigenvalue Methods Power Modern Technology

From climate modeling to video game physics to medical imaging, eigenvalue-based methods for solving differential equations are everywhere in modern technology.

The central question this lesson addresses is: How do eigenvalues help us solve differential equations? We will start from the basics and build up to this powerful connection step by step.

SECTION 2

Core Principles & Definitions

Before we connect eigenvalues to differential equations, let's make sure we understand the key building blocks. Each idea below is simple on its own. The magic happens when we combine them.

Differential Equation (DE)

An equation that involves a function and its derivative (rate of change). For example, dy/dt = 3y says "the rate of change of y is 3 times y itself."

Matrix

A rectangular grid of numbers arranged in rows and columns. Matrices let us organize and solve systems (groups) of equations all at once.

Eigenvalue (λ)

A special number associated with a matrix. When a matrix multiplies a certain vector, the result is just that vector scaled (stretched or shrunk) by the eigenvalue λ (lambda).

Eigenvector

The special vector that only gets scaled (not rotated) when multiplied by a matrix. Every eigenvalue has at least one eigenvector partner.

Exponential Function eλt

The function e^λt is the natural solution to the simplest differential equation. It describes growth (when λ > 0) or decay (when λ < 0).

✦

KEY TAKEAWAY

Think of eigenvalues like a cheat code for solving differential equations. Normally, solving a system of DEs is like trying to untangle a ball of yarn — everything is connected. Finding eigenvalues is like finding the individual strands: once you separate them, each strand (equation) becomes simple to handle on its own. The eigenvalue tells you how fast each strand grows or shrinks.

SECTION 3

Visual Explanation — The Big Picture

The diagram below shows how the key ideas connect. On the left, we start with a system of differential equations. In the center, we extract a matrix and find its eigenvalues. On the right, those eigenvalues give us the solution directly.

The three-step pipeline: a system of DEs becomes a matrix, whose eigenvalues plug directly into the solution. The bottom row shows why: eigenvectors separate entangled directions, eigenvalues give growth rates, and exponential functions tie it all together.

Notice the key insight in the diagram: the eigenvalues λ₁ = 3 and λ₂ = 1 appear directly as the exponents in the solution. This is not a coincidence — it is the entire point! The eigenvalues tell us how fast each part of the solution grows or decays. A positive eigenvalue means growth (getting bigger over time), and a negative eigenvalue means decay (shrinking toward zero).

SECTION 4

Mathematical Framework

Let's build up the math piece by piece. Don't worry — each step is simple. The power comes from putting them together.

The Simplest Differential Equation

SIMPLEST DE

dy/dt = λy

This says: "the rate of change of y is proportional to y itself." The constant λ (lambda) controls how fast y changes. The solution is the exponential function below.

SOLUTION TO SIMPLEST DE

y(t) = Ce^(λt)

Here C is a constant determined by the starting value of y, e ≈ 2.718 is Euler's number, and λ appears in the exponent. If λ > 0, the function grows; if λ < 0, it decays.

Writing a System as a Matrix Equation

When we have two or more linked differential equations, we can write them in a compact form using a matrix. For a system like dx/dt = 2x + y and dy/dt = x + 2y, we write this as a single matrix equation.

MATRIX FORM OF A SYSTEM

d𝐱/dt = A𝐱, where A = [2 1; 1 2] and 𝐱 = [x; y]

The vector 𝐱 bundles our unknowns (x and y) together, and the matrix A captures how x and y influence each other's rates of change.

The Eigenvalue Equation

EIGENVALUE EQUATION

A𝐯 = λ𝐯

This says: when matrix A multiplies the special vector 𝐯 (the eigenvector), the result is just λ (the eigenvalue) times that same vector. The direction doesn't change — it only stretches or shrinks.

💡

The Big Connection

For the system d𝐱/dt = A𝐱, if λ is an eigenvalue and 𝐯 is its eigenvector, then 𝐱(t) = $e^{λt}$ 𝐯 is a solution. This is the golden link: eigenvalues become the exponents in the solution, and eigenvectors give the direction of that solution.

SECTION 5

Eigenvalue Types and Solution Behavior

Different eigenvalues produce different behaviors in the solution. Understanding this is crucial because it lets you predict what a system will do without even solving the equation completely. Below is a visual guide to the three main cases.

Three eigenvalue cases: positive eigenvalues (λ > 0) produce exponential growth, negative eigenvalues (λ < 0) produce exponential decay, and zero eigenvalues (λ = 0) produce a constant (steady-state) solution.

Eigenvalue Sign	Solution Behavior	Real-World Example
λ > 0 (positive)	Exponential growth — solution gets bigger and bigger	Unchecked population growth, viral spread
λ < 0 (negative)	Exponential decay — solution shrinks toward zero	Radioactive decay, cooling objects
λ = 0 (zero)	Constant / steady state — nothing changes	Balanced chemical reaction, equilibrium
λ is complex (a + bi)	Oscillation — solution waves up and down (advanced topic)	Spring vibrations, electrical circuits

Eigenvalue types and their corresponding solution behaviors

SECTION 6

Worked Example — Solving a System of DEs

Let's walk through a complete example. We'll solve the system dx/dt = 3x + 0y and dy/dt = 0x + (−2)y. This system describes two quantities x and y that change over time independently.

Solving a 2×2 System Using Eigenvalues

1
Step 1 — Write the System as a Matrix Equation

We write the system d𝐱/dt = A𝐱 where A = [3, 0; 0, −2]. The matrix A is a diagonal matrix (numbers only on the diagonal), which makes this example a great starting point.

A = [3, 0; 0, −2]
2
Step 2 — Find the Eigenvalues

For a diagonal matrix, the eigenvalues are simply the numbers on the diagonal. So λ₁ = 3 and λ₂ = −2. In general, you would solve det(A − λI) = 0, but diagonal matrices give us the answer immediately.

λ₁ = 3, λ₂ = −2
3
Step 3 — Find the Eigenvectors

For λ₁ = 3: we solve (A − 3I)𝐯 = 0. This gives [0, 0; 0, −5]𝐯 = 0, so 𝐯₁ = [1; 0]. For λ₂ = −2: we solve (A + 2I)𝐯 = 0. This gives [5, 0; 0, 0]𝐯 = 0, so 𝐯₂ = [0; 1]. The eigenvectors point along the x-axis and y-axis.

𝐯₁ = [1; 0], 𝐯₂ = [0; 1]
4
Step 4 — Write the General Solution

Each eigenvalue-eigenvector pair gives one piece of the solution: $e^{λt}$ × 𝐯. The general solution combines both pieces with arbitrary constants c₁ and c₂.

𝐱(t) = c₁ $e^{3t}$ [1; 0] + c₂ $e^{−2t}$ [0; 1]
5
Step 5 — Interpret the Answer

This means x(t) = c₁ $e^{3t}$ and y(t) = c₂ $e^{−2t}$ . The x-component grows exponentially (because λ₁ = 3 is positive), while the y-component decays toward zero (because λ₂ = −2 is negative). The eigenvalues predicted exactly this behavior!

x(t) = c₁ $e^{3t}$ , y(t) = c₂ $e^{−2t}$

SECTION 7

Strengths and Limitations

The eigenvalue method is powerful, but like any tool, it has situations where it shines and situations where it needs extra care. Here's a balanced comparison.

Strengths	Limitations
Provides a systematic, step-by-step method — no guessing needed	Only works directly for linear (straight-line-relationship) systems
Eigenvalues instantly reveal growth, decay, or oscillation	Repeated eigenvalues require extra techniques (generalized eigenvectors)
Works for systems with many equations (not just two)	Finding eigenvalues of large matrices can be computationally expensive
Connects beautifully to many areas of science and engineering	Complex eigenvalues require knowledge of trigonometry (sin, cos)

Comparing the strengths and limitations of the eigenvalue method

✦

KEY TAKEAWAY

The eigenvalue method is like a universal remote control for systems of differential equations. It works on many different "channels" (types of linear systems) with the same sequence of buttons (find matrix, find eigenvalues, build solution). But just like a universal remote, it doesn't control everything — non-linear systems need different tools.

SECTION 8

Connection to Advanced Theory

What you've learned in this lesson is just the beginning. As you move into more advanced math and science courses, the eigenvalue-differential equation connection becomes even richer. Here is a preview of where these ideas lead.

What You Learned Here	Where It Leads Next
2×2 systems with real eigenvalues	n×n systems with complex eigenvalues and oscillating solutions
Constant coefficient matrices	Matrix exponentials ( $e^{At}$ ) for time-varying coefficients
Eigenvalues predict growth or decay	Stability analysis — predicting whether bridges collapse or circuits overload
Linear systems only	Linearization — using eigenvalues to approximate non-linear systems near equilibrium

From introductory to advanced applications

In college-level courses like ordinary differential equations (ODEs) and dynamical systems, eigenvalues become your primary tool for understanding how systems evolve over time. The foundation you build here — understanding that eigenvalues are the exponents in the solution — will carry you through all of these advanced topics.

SECTION 9

Practice Problems

Question 1: CONCEPTUAL

A system of differential equations has eigenvalues λ₁ = 5 and λ₂ = −3. Without solving the system, describe what the solution will look like over time. Will it grow, decay, or do both?

Show answer

The solution will exhibit both growth and decay. The component associated with λ₁ = 5 (positive) will grow exponentially as

e^{5t}

, meaning it gets bigger and bigger over time. Meanwhile, the component associated with λ₂ = −3 (negative) will decay as

e^{−3t}

, shrinking toward zero. Over long periods of time, the growing part will dominate, and the solution will appear to grow overall. The eigenvalues tell us the growth and decay rates without needing the full solution.

Question 2: BASIC CALCULATION

Given the diagonal matrix A = [4, 0; 0, −1], write down the eigenvalues and the general solution to the system d𝐱/dt = A𝐱.

Show answer

For a diagonal matrix, the eigenvalues are the diagonal entries: λ₁ = 4 and λ₂ = −1. The corresponding eigenvectors are 𝐯₁ = [1; 0] and 𝐯₂ = [0; 1] (the standard basis vectors). Using the formula 𝐱(t) = c₁

e^{λ₁t}

𝐯₁ + c₂

e^{λ₂t}

𝐯₂, the general solution is x(t) = c₁

e^{4t}

and y(t) = c₂

e^{−t}

. The x-component grows while the y-component decays.

Question 3: INTERMEDIATE

For the matrix A = [2, 1; 0, 3], find the eigenvalues by solving det(A − λI) = 0. Then write the general form of the solution (you don't need to find the eigenvectors).

Show answer

We compute A − λI = [2−λ, 1; 0, 3−λ]. The determinant is (2−λ)(3−λ) − (1)(0) = (2−λ)(3−λ). Setting this equal to zero: (2−λ)(3−λ) = 0, which gives λ₁ = 2 and λ₂ = 3. Since this is an upper triangular matrix, the eigenvalues are just the diagonal entries — a useful shortcut! The general solution will have the form 𝐱(t) = c₁

e^{2t}

𝐯₁ + c₂

e^{3t}

𝐯₂, where 𝐯₁ and 𝐯₂ are the eigenvectors. Both eigenvalues are positive, so both parts of the solution grow over time.

Question 4: APPLIED

A scientist models two interacting chemical concentrations with the system dx/dt = −x and dy/dt = −4y. Write this as a matrix equation, find the eigenvalues, and predict: which chemical disappears faster?

Show answer

The matrix form is d𝐱/dt = A𝐱 where A = [−1, 0; 0, −4]. This is a diagonal matrix, so the eigenvalues are λ₁ = −1 and λ₂ = −4. Both are negative, confirming that both chemicals decay over time. The solutions are x(t) = c₁

e^{−t}

and y(t) = c₂

e^{−4t}

. Since |λ₂| = 4 is larger than |λ₁| = 1, the second chemical (y) decays four times faster than the first. After just 1 second, y has dropped to

e^{−4}

≈ 1.8% of its original value, while x is still at

e^{−1}

≈ 36.8%. The magnitude of the eigenvalue tells us the speed of decay.

Question 5: CRITICAL THINKING

Suppose a 3×3 matrix A has eigenvalues λ₁ = 2, λ₂ = 0, and λ₃ = −5. Describe the long-term behavior of the system d𝐱/dt = A𝐱. Which eigenvalue dominates the solution as t → ∞? What happens to the component with λ₂ = 0?

Show answer

The general solution is 𝐱(t) = c₁

e^{2t}

𝐯₁ + c₂

e^{0·t}

𝐯₂ + c₃

e^{−5t}

𝐯₃ = c₁

e^{2t}

𝐯₁ + c₂𝐯₂ + c₃

e^{−5t}

𝐯₃. As t → ∞, the term with λ₁ = 2 dominates because

e^{2t}

grows without bound. The term with λ₃ = −5 dies away very quickly (

e^{−5t}

→ 0). The most interesting part is the λ₂ = 0 term: since

e^{0·t}

= e⁰ = 1, this component stays constant forever. It represents a steady-state contribution — a part of the solution that neither grows nor decays. In the long run, the system is dominated by the largest positive eigenvalue, but the zero eigenvalue leaves a permanent 'fingerprint' in the solution.

SUMMARY

Lesson Summary

Summary

In this lesson, you learned that differential equations describe how quantities change over time, and that systems of these equations can be written compactly using matrices. The golden connection is that eigenvalues of the coefficient matrix become the exponents in the exponential solution $e^{λt}$ , while eigenvectors determine the direction of each solution component.

The sign of each eigenvalue tells you the behavior: positive eigenvalues produce growth, negative eigenvalues produce decay, and zero eigenvalues produce a constant. The general solution is built by combining all eigenvalue-eigenvector pairs: 𝐱(t) = c₁ $e^{λ₁t}$ 𝐯₁ + c₂ $e^{λ₂t}$ 𝐯₂ + … This systematic method, connecting linear algebra to differential equations, is one of the most powerful tools in all of applied mathematics.