Mastering Taylor Series in Multiple Variables

In your undergraduate maths journey, you have likely mastered the Taylor series for a single variable. However, as you progress into multivariable calculus, you will encounter functions where the output depends on several inputs, such as $f(x, y)$. Understanding how to approximate these functions near a specific point is essential for solving problems in optimisation, physics, and numerical analysis.

This article will guide you through the transition from single-variable Taylor series to the multivariable case. You will learn how to construct linear and quadratic approximations, which are vital for your exams and for understanding the local behaviour of surfaces in higher dimensions.

The Concept of Multivariable Taylor Expansion

Just as a single-variable Taylor series approximates a function $f(x)$ near $x=a$ using derivatives, a multivariable Taylor series approximates $f(x, y)$ near a point $(a, b)$. The core idea remains the same: we use the value of the function and its partial derivatives at the point $(a, b)$ to build a polynomial that mimics the function's behaviour locally.

For a function $f(x, y)$, the Taylor expansion about $(a, b)$ is given by:

$$f(x, y) \approx f(a, b) + f_x(a, b)(x-a) + f_y(a, b)(y-b) + \frac{1}{2!}[f_{xx}(a, b)(x-a)^2 + 2f_{xy}(a, b)(x-a)(y-b) + f_{yy}(a, b)(y-b)^2]$$

This expansion includes the constant term, the first-order (linear) terms, and the second-order (quadratic) terms.

First-Order (Linear) Approximation

The first-order Taylor polynomial is simply the equation of the tangent plane to the surface at $(a, b)$. It is defined as:

$$L(x, y) = f(a, b) + f_x(a, b)(x-a) + f_y(a, b)(y-b)$$

This is the best linear approximation of the function near the point. It is widely used in physics to linearise complex systems.

Worked Example 1: Linear Approximation

Find the linear approximation of $f(x, y) = x^2y + 3y^3$ at the point $(1, 2)$.

1. Calculate the function value: $f(1, 2) = (1)^2(2) + 3(2)^3 = 2 + 24 = 26$.
2. Find partial derivatives: $f_x = 2xy$ and $f_y = x^2 + 9y^2$.
3. Evaluate at $(1, 2)$: $f_x(1, 2) = 2(1)(2) = 4$ and $f_y(1, 2) = 1^2 + 9(2)^2 = 1 + 36 = 37$.
4. Assemble the linear approximation: $L(x, y) = 26 + 4(x-1) + 37(y-2)$.

Simplifying, we get $L(x, y) = 4x + 37y - 48$.

Second-Order (Quadratic) Approximation

To improve accuracy, we include second-order partial derivatives. This accounts for the curvature of the surface. The quadratic approximation is:

$$Q(x, y) = L(x, y) + \frac{1}{2}[f_{xx}(a, b)(x-a)^2 + 2f_{xy}(a, b)(x-a)(y-b) + f_{yy}(a, b)(y-b)^2]$$

Worked Example 2: Quadratic Approximation

Find the quadratic approximation of $f(x, y) = e^x \sin(y)$ at $(0, 0)$.

1. Function value: $f(0, 0) = e^0 \sin(0) = 0$.
2. First derivatives: $f_x = e^x \sin(y) \implies f_x(0, 0) = 0$; $f_y = e^x \cos(y) \implies f_y(0, 0) = 1$.
3. Second derivatives: $f_{xx} = e^x \sin(y) \implies f_{xx}(0, 0) = 0$; $f_{yy} = -e^x \sin(y) \implies f_{yy}(0, 0) = 0$; $f_{xy} = e^x \cos(y) \implies f_{xy}(0, 0) = 1$.
4. Assemble: $Q(x, y) = 0 + 0(x-0) + 1(y-0) + \frac{1}{2}[0(x-0)^2 + 2(1)(x-0)(y-0) + 0(y-0)^2]$.

Result: $Q(x, y) = y + xy$.

Common Mistakes

1. Forgetting the Factorial: Students often forget the $\frac{1}{2!}$ coefficient for the second-order terms, which leads to incorrect scaling of the curvature.
2. Mixed Partial Derivative Confusion: Remember that the cross-term is $2f_{xy}(x-a)(y-b)$. Forgetting the factor of 2 is a frequent error.
3. Incorrect Point Evaluation: Always ensure you evaluate the derivatives at the specific point $(a, b)$ before substituting them into the polynomial formula.
4. Sign Errors: Be careful with signs when differentiating trigonometric or exponential functions, especially when evaluating at zero.

Frequently Asked Questions

What is the difference between a Taylor series and a Taylor polynomial? A Taylor polynomial is a finite sum (an approximation), whereas a Taylor series is an infinite sum that, if convergent, represents the function exactly.

Why do we use Taylor series in multiple variables? They allow us to approximate complex, non-linear surfaces with simple polynomials, making it easier to perform calculations like finding local extrema or solving differential equations.

Does the order of differentiation matter? For smooth functions, Clairaut's Theorem states that $f_{xy} = f_{yx}$, so the order of partial differentiation does not change the result.

Conclusion

Mastering Taylor series in multiple variables is a cornerstone of advanced calculus. By breaking down complex surfaces into manageable linear and quadratic components, you gain a deeper insight into the geometry of functions. For more practice and to see these concepts come to life, visit MathInstructor AI to generate a free animated lesson on this topic.