Higher-Dimensional Linear Approximation
What you'll learn: How to extend linear approximation to functions of three or more variables using the gradient vector.
The Core Idea
You've already learned how to approximate a function of two variables near a point using a tangent plane. The exact same principle works for functions of three, four, or any number of variables—you just can't visualize it as easily!
For a function of two variables, the linear approximation at point is:
For a function of three variables , the linear approximation at follows the same pattern:
Notice the pattern: you add one term for each variable, using the corresponding partial derivative times the displacement in that direction.
The Gradient Connection
This is really the gradient at work. The gradient collects all partial derivatives into a single vector. For three variables:
The linear approximation can be written compactly as:
where is the displacement vector from your base point.
This formula works in any dimension—just include as many partial derivatives as you have variables!
Analogy
Think of it like asking for directions. In 2D (a city), you need north-south and east-west guidance. In 3D (a skyscraper), you also need elevator instructions. Each dimension requires one additional piece of rate-of-change information.
Key Takeaway: Linear approximation in higher dimensions uses the gradient: add one term for each variable, multiplying its partial derivative by the displacement in that direction.