Max and Min Values- Partial Derivatives Guide

Max and Min Values: Partial Derivatives Made Simple

You're probably here because partial derivatives are kicking your butt. Or you need to actually understand them for an exam. Either way, this guide cuts the nonsense and gets straight to what works.

Partial derivatives let you find how a function changes when you nudge just one variable—while keeping the others locked in place. Sounds simple. The math can still be brutal if you don't have a system.

Here's that system.

What Are Partial Derivatives, Exactly?

Regular derivatives work with single-variable functions. You have f(x), you take dx, done. Partial derivatives handle multivariable functions like f(x, y). You pick one variable and differentiate while treating every other variable as a constant.

Notation you'll see:

• ∂f/∂x — partial derivative with respect to x
• ∂f/∂y — partial derivative with respect to y
• fₓ — same thing, shorthand
• fᵧ — same thing, shorthand

That's it. The symbols look scary but the operation is straightforward: apply normal differentiation rules, freeze everything else.

Why Finding Max and Min Values Matters

Optimization problems. That's why. You need to find the highest or lowest point of a surface. A business wants maximum profit. An engineer wants minimum stress on a material. A model wants the best fit parameters.

Partial derivatives are your tool for finding these points on multivariable surfaces. Without them, you're guessing.

The Critical Points Method

Critical points are where a function's surface could have a peak, valley, or saddle. Finding them is your first step.

Step 1: Find the First Partial Derivatives

Take ∂f/∂x and ∂f/∂y. These tell you the slope in each direction.

Step 2: Set Both Equal to Zero

Solve the system:

• ∂f/∂x = 0
• ∂f/∂y = 0

The point(s) that satisfy both equations are your critical points. This is where the surface levels out horizontally—where a peak or valley could exist.

Step 3: Classify the Critical Point

Not every critical point is a max or min. Some are saddle points—slopes that cancel out but the surface keeps going up in one direction and down in another.

The Second Derivative Test for Two Variables

This is where most textbooks lose people. Here's the test that actually classifies critical points.

Calculate these values at your critical point (x₀, y₀):

• A = fₓₓ (second partial with respect to x)
• B = fᵧᵧ (second partial with respect to y)
• C = fₓᵧ (mixed partial, same result either order)

Then compute: D = AB - C²

What D Tells You

This table is your classification cheat sheet. Keep it handy.

Example: Working Through It

Let's classify the critical points of f(x, y) = x² + y² - 4x - 2y + 5

Step 1: First Partial Derivatives

fₓ = 2x - 4
fᵧ = 2y - 2

Step 2: Set Equal to Zero

2x - 4 = 0 → x = 2
2y - 2 = 0 → y = 1

Critical point at (2, 1)

Step 3: Second Partial Derivatives

fₓₓ = 2
fᵧᵧ = 2
fₓᵧ = 0

Step 4: Apply the Test

D = (2)(2) - (0)² = 4

D > 0 and A > 0, so (2, 1) is a local minimum

Plug it back in: f(2, 1) = 2² + 1² - 4(2) - 2(1) + 5 = 4 + 1 - 8 - 2 + 5 = 0

The minimum value is 0 at point (2, 1)

Getting Started: Your Checklist

When you face a new max/min problem, run through this:

1. Verify the function — make sure you have the right expression
2. Compute fₓ and fᵧ — first partials
3. Solve the system — find all critical points
4. Compute second partials — fₓₓ, fᵧᵧ, fₓᵧ
5. Calculate D — AB - C² at each critical point
6. Classify — use the table above
7. Evaluate the function — get actual max/min values when needed

That's the whole process. Practice it three times and it'll stick.

Common Mistakes That Cost Points

• Forgetting the product rule — when functions multiply, you can't just differentiate one factor
• Mixing up signs in D — D = AB - C², not AB + C²
• Assuming D > 0 means max — it only confirms an extremum exists; check A's sign
• Skipping boundary analysis — local max/mins aren't always global; check edges of your domain

Functions With Three Variables

The method extends directly. For f(x, y, z), you take ∂f/∂x, ∂f/∂y, and ∂f/∂z. Set all three to zero. You get a system of three equations. Solve for x, y, z.

The second derivative test gets more complicated with three variables—you'll need a matrix of second partials called the Hessian. For most undergraduate problems, two variables covers what you need.

When the Test Fails (D = 0)

Some functions have critical points that can't be classified with the second derivative test. The classic example is f(x, y) = x⁴ + y⁴ at (0, 0). The test gives D = 0, but (0, 0) is clearly a minimum.

In these cases, you examine the function behavior directly. Check values along different paths approaching the point. Or look at the function's form—if it's a sum of even powers, you're probably looking at a minimum.

What You're Actually Solving For

Partial derivatives aren't just textbook math. They show up everywhere:

• Machine learning — gradient descent uses partial derivatives to find optimal parameters
• Physics — finding equilibrium points in systems with multiple forces
• Economics — optimizing production with multiple input variables
• Engineering — stress analysis across multiple dimensions

The mechanics stay the same. You identify where the function flattens out, then check whether that flat spot is a peak, valley, or neither.

The Bottom Line

Finding max and min values with partial derivatives comes down to three moves: take first partials, set them to zero, then apply the second derivative test with D = AB - C².

Everything else is just practice. Work through problems until the process is automatic. The notation fades once you see enough examples. The logic clicks faster than you expect.