← ⛰️ Maxima, Minima & Taylor Series
Subtopic 6 of 6

3.6Functions of two variables: partial derivatives and saddle points

The idea

With two variables, a partial derivative changes one variable while freezing the other, and the Hessian test classifies stationary points.

Think of it like this: Standing on a hill and asking the slope east, then the slope north, separately.

Why it exists

It extends max/min ideas to surfaces: bowls, hills and saddles (a mountain pass).

How it works

  1. Compute fx and fy and solve fx = fy = 0.
  2. At each solution compute D = fxxfyy − fxy².
  3. D > 0 with fxx > 0: minimum; D > 0 with fxx < 0: maximum.
  4. D < 0: saddle. D = 0: no decision.

Worked example: f = x³ − 3xy + y³

fx = 3x² − 3y and fy = 3y² − 3x vanish at (0, 0) and (1, 1). At (1, 1): fxx = 6, fyy = 6, fxy = −3, D = 36 − 9 = 27 > 0, so a local minimum. At (0, 0): D = −9, a saddle.

Now play with it

  • →Move the point between (0, 0) and (1, 1) and watch the verdict change.
  • →Try x² − y² to see a saddle in the heat map.
Heat map of f on [−3, 3]². Blue = low, red = high. White dot = (x₀, y₀).
Gradient (fx, fy)(0, 0)
fxx, fyy, fxy6, 6, -3
D = fxx·fyy − fxy²27
✓ Stationary point: local minimum

Stationary at (0, 0) (saddle) and at (1, 1) (local minimum: fxx = 6, D = 36 − 9 = 27 > 0). Move the point to see both.

Check yourself

For f = x² − y², the origin is

Remember: D = fxx·fyy − fxy². D < 0 means saddle.
Exam notes (the dense version)

Partial derivatives fx, fy differentiate with respect to one variable, treating the other as a constant. A stationary point has fx = fy = 0. With D = fxxfyy − (fxy)²: D > 0 and fxx > 0 ⇒ local min, D > 0 and fxx < 0 ⇒ local max, D < 0 ⇒ saddle, D = 0 ⇒ test fails. Mixed partials are equal when they are continuous: fxy = fyx.

Previous year questions: Maxima, Minima & Taylor Series (3)

PYQs are grouped by topic, so this is the full set for Maxima, Minima & Taylor Series; try the ones that match this subtopic. Open all 3 on the topic page.