Taylor Series For Two Variables Examples

The Taylor series for two variables is a powerful tool for approximating functions of two independent variables using polynomials. This series extends the concept of the single-variable Taylor series to higher dimensions, allowing us to analyze and approximate complex functions around a specific point. It's widely used in various fields, including physics, engineering, and computer graphics, for tasks like optimization, sensitivity analysis, and surface modeling.

Understanding the Taylor Series for Two Variables

The Taylor series expansion for a function f(x, y) around a point (a, b) provides a polynomial approximation of the function in the neighborhood of that point. The series is constructed using the partial derivatives of f evaluated at (a, b).

The general form of the Taylor series for two variables is:

f(x, y) = f(a, b) + f_x(a, b)(x - a) + f_y(a, b)(y - b)
         + 1/2! [f_xx(a, b)(x - a)^2 + 2f_xy(a, b)(x - a)(y - b) + f_yy(a, b)(y - b)^2]
         + 1/3! [f_xxx(a, b)(x - a)^3 + 3f_xxy(a, b)(x - a)^2(y - b) + 3f_xyy(a, b)(x - a)(y - b)^2 + f_yyy(a, b)(y - b)^3]
         + ...

Where:

• f(x, y) is the function you want to approximate.
• (a, b) is the point around which you are expanding the series.
• f_x(a, b), f_y(a, b), f_xx(a, b), f_xy(a, b), f_yy(a, b), and so on are the partial derivatives of f with respect to x and y, evaluated at the point (a, b).
• n! represents the factorial of n.

The subscripts denote the partial derivatives. For example:

• f_x represents the first partial derivative of f with respect to x.
• f_y represents the first partial derivative of f with respect to y.
• f_xx represents the second partial derivative of f with respect to x twice.
• f_xy represents the second partial derivative of f with respect to x and then y.
• f_yy represents the second partial derivative of f with respect to y twice.

Simplified Notation and Compact Form

The above form can be cumbersome to write out for higher-order terms. A more compact notation uses multi-indices. However, for clarity in the examples, we will primarily use the expanded form.

Key Concepts

• Partial Derivatives: The Taylor series relies heavily on partial derivatives. Understanding how to calculate these is crucial. Remember that when taking a partial derivative with respect to one variable, you treat the other variable as a constant.
• Point of Expansion (a, b): The accuracy of the Taylor series approximation is generally best in the immediate vicinity of the point (a, b). As you move further away, the approximation may become less accurate.
• Order of the Approximation: The order of the Taylor series refers to the highest-order derivatives included in the expansion. A first-order Taylor series includes only first derivatives, a second-order includes first and second derivatives, and so on. Higher-order approximations generally provide better accuracy but are also more computationally expensive.

Examples of Taylor Series Expansion for Two Variables

Let's walk through several examples to illustrate how to apply the Taylor series formula.

Example 1: Expanding f(x, y) = x² + xy + y² around (1, 1)

1. Define the function and the point:

• f(x, y) = x² + xy + y²
• (a, b) = (1, 1)

2. Calculate the partial derivatives:

• f_x = 2x + y
• f_y = x + 2y
• f_xx = 2
• f_xy = 1
• f_yy = 2
• All higher-order partial derivatives are zero.

3. Evaluate the function and its derivatives at (1, 1):

• f(1, 1) = 1² + (1)(1) + 1² = 3
• f_x(1, 1) = 2(1) + 1 = 3
• f_y(1, 1) = 1 + 2(1) = 3
• f_xx(1, 1) = 2
• f_xy(1, 1) = 1
• f_yy(1, 1) = 2

4. Construct the Taylor series:

Since the higher-order derivatives are zero, the Taylor series is a polynomial of degree 2:

f(x, y) ≈ f(1, 1) + f_x(1, 1)(x - 1) + f_y(1, 1)(y - 1)
         + 1/2! [f_xx(1, 1)(x - 1)^2 + 2f_xy(1, 1)(x - 1)(y - 1) + f_yy(1, 1)(y - 1)^2]

Substitute the calculated values:

f(x, y) ≈ 3 + 3(x - 1) + 3(y - 1) + 1/2 [2(x - 1)^2 + 2(1)(x - 1)(y - 1) + 2(y - 1)^2]

Simplify:

f(x, y) ≈ 3 + 3x - 3 + 3y - 3 + (x - 1)^2 + (x - 1)(y - 1) + (y - 1)^2
f(x, y) ≈ 3x + 3y - 3 + x^2 - 2x + 1 + xy - x - y + 1 + y^2 - 2y + 1
f(x, y) ≈ x^2 + xy + y^2

In this particular case, the Taylor series expansion exactly reconstructs the original function. This is because the original function was a polynomial of degree 2, and the Taylor series included all derivatives up to that degree.

Example 2: Expanding f(x, y) = sin(x)cos(y) around (0, 0) – Up to Second Order

1. Define the function and the point:

• f(x, y) = sin(x)cos(y)
• (a, b) = (0, 0)

2. Calculate the partial derivatives:

• f_x = cos(x)cos(y)
• f_y = -sin(x)sin(y)
• f_xx = -sin(x)cos(y)
• f_xy = -cos(x)sin(y)
• f_yy = -sin(x)cos(y)

3. Evaluate the function and its derivatives at (0, 0):

• f(0, 0) = sin(0)cos(0) = 0
• f_x(0, 0) = cos(0)cos(0) = 1
• f_y(0, 0) = -sin(0)sin(0) = 0
• f_xx(0, 0) = -sin(0)cos(0) = 0
• f_xy(0, 0) = -cos(0)sin(0) = 0
• f_yy(0, 0) = -sin(0)cos(0) = 0

4. Construct the Taylor series (up to second order):

f(x, y) ≈ f(0, 0) + f_x(0, 0)(x - 0) + f_y(0, 0)(y - 0)
         + 1/2! [f_xx(0, 0)(x - 0)^2 + 2f_xy(0, 0)(x - 0)(y - 0) + f_yy(0, 0)(y - 0)^2]

Substitute the calculated values:

f(x, y) ≈ 0 + 1(x) + 0(y) + 1/2 [0(x)^2 + 2(0)(x)(y) + 0(y)^2]

Simplify:

f(x, y) ≈ x

Therefore, the second-order Taylor series approximation of sin(x)cos(y) around (0, 0) is simply x. This is a reasonable approximation when x and y are close to zero. The cos(y) term is close to 1, and sin(x) is approximately x for small x.

Example 3: Expanding f(x, y) = e^(x+y) around (0, 0) – Up to Second Order

1. Define the function and the point:

• f(x, y) = e^(x+y)
• (a, b) = (0, 0)

2. Calculate the partial derivatives:

• f_x = e^(x+y)
• f_y = e^(x+y)
• f_xx = e^(x+y)
• f_xy = e^(x+y)
• f_yy = e^(x+y)

3. Evaluate the function and its derivatives at (0, 0):

• f(0, 0) = e^(0+0) = 1
• f_x(0, 0) = e^(0+0) = 1
• f_y(0, 0) = e^(0+0) = 1
• f_xx(0, 0) = e^(0+0) = 1
• f_xy(0, 0) = e^(0+0) = 1
• f_yy(0, 0) = e^(0+0) = 1

4. Construct the Taylor series (up to second order):

f(x, y) ≈ f(0, 0) + f_x(0, 0)(x - 0) + f_y(0, 0)(y - 0)
         + 1/2! [f_xx(0, 0)(x - 0)^2 + 2f_xy(0, 0)(x - 0)(y - 0) + f_yy(0, 0)(y - 0)^2]

Substitute the calculated values:

f(x, y) ≈ 1 + 1(x) + 1(y) + 1/2 [1(x)^2 + 2(1)(x)(y) + 1(y)^2]

Simplify:

f(x, y) ≈ 1 + x + y + 1/2 (x^2 + 2xy + y^2)
f(x, y) ≈ 1 + x + y + 1/2 (x + y)^2

This is the second-order Taylor series approximation of e^(x+y) around (0, 0). It provides a good approximation when x and y are close to zero.

Example 4: Expanding f(x, y) = xy² around (2, 3)

1. Define the function and the point:

• f(x, y) = xy²
• (a, b) = (2, 3)

2. Calculate the partial derivatives:

• f_x = y²
• f_y = 2xy
• f_xx = 0
• f_xy = 2y
• f_yy = 2x
• All higher-order partial derivatives are zero.

3. Evaluate the function and its derivatives at (2, 3):

• f(2, 3) = (2)(3)² = 18
• f_x(2, 3) = 3² = 9
• f_y(2, 3) = 2(2)(3) = 12
• f_xx(2, 3) = 0
• f_xy(2, 3) = 2(3) = 6
• f_yy(2, 3) = 2(2) = 4

4. Construct the Taylor series:

Since the higher-order derivatives are zero, the Taylor series is a polynomial of degree 2:

f(x, y) ≈ f(2, 3) + f_x(2, 3)(x - 2) + f_y(2, 3)(y - 3)
         + 1/2! [f_xx(2, 3)(x - 2)^2 + 2f_xy(2, 3)(x - 2)(y - 3) + f_yy(2, 3)(y - 3)^2]

Substitute the calculated values:

f(x, y) ≈ 18 + 9(x - 2) + 12(y - 3) + 1/2 [0(x - 2)^2 + 2(6)(x - 2)(y - 3) + 4(y - 3)^2]

Simplify:

f(x, y) ≈ 18 + 9x - 18 + 12y - 36 + 6(x - 2)(y - 3) + 2(y - 3)^2
f(x, y) ≈ 9x + 12y - 36 + 6(xy - 3x - 2y + 6) + 2(y^2 - 6y + 9)
f(x, y) ≈ 9x + 12y - 36 + 6xy - 18x - 12y + 36 + 2y^2 - 12y + 18
f(x, y) ≈ -9x - 12y + 18 + 6xy + 2y^2

Note that this Taylor series approximation is not equal to the original function xy². This is because we stopped at the second-order terms. To get the exact function back, you would need to consider higher-order terms (which are all zero in this particular case, meaning our second-order approximation is actually the exact representation, although it may not look like it at first glance). Let's verify:

Original Function at (x,y): xy^2 Taylor Series Approximation: -9x - 12y + 18 + 6xy + 2y^2

Let x = 2 + dx and y = 3 + dy, where dx and dy are small deviations from the point (2,3). Then:

Original Function: (2+dx)(3+dy)^2 = (2+dx)(9 + 6dy + dy^2) = 18 + 12dy + 2dy^2 + 9dx + 6dxdy + dxdy^2

Taylor Series: -9(2+dx) - 12(3+dy) + 18 + 6(2+dx)(3+dy) + 2(3+dy)^2 = -18 - 9dx - 36 - 12dy + 18 + 6(6 + 2dy + 3dx + dxdy) + 2(9 + 6dy + dy^2) = -36 - 9dx - 12dy + 18 + 36 + 12dy + 18dx + 6dxdy + 18 + 12dy + 2dy^2 = 18 + 9dx + 12dy + 6dx*dy + 2dy^2

Notice that when we substitute x = 2+dx and y = 3+dy into both the original function and the Taylor Series, we obtain equivalent expressions (ignoring the dx*dy^2 term, which would come from a higher-order Taylor Series expansion). This confirms that our Taylor Series is indeed a valid approximation of the original function around the point (2,3).

Key Takeaways from the Examples

• Accuracy: The Taylor series provides a good approximation of the function near the point of expansion (a, b). The accuracy generally decreases as you move further away from (a, b).
• Order: Higher-order Taylor series generally provide better accuracy but require more computation.
• Polynomial Functions: If the original function is a polynomial, the Taylor series expansion will eventually exactly match the original function when you include all terms up to the degree of the polynomial.
• Non-Polynomial Functions: For non-polynomial functions (like trigonometric or exponential functions), the Taylor series provides an approximation. The more terms you include, the better the approximation will be within a certain radius of convergence.

Applications of Taylor Series for Two Variables

The Taylor series for two variables has numerous applications across various fields:

• Optimization: In optimization problems, especially those involving functions of two or more variables, the Taylor series can be used to approximate the function near a stationary point. This can help in finding local minima or maxima. The second-order Taylor expansion is often used to determine the nature of a critical point (whether it's a minimum, maximum, or saddle point) using the Hessian matrix.
• Sensitivity Analysis: The Taylor series can be used to analyze how the output of a function changes in response to small changes in the input variables. The first-order Taylor series provides a linear approximation of the function's sensitivity.
• Numerical Analysis: The Taylor series is fundamental to many numerical methods for solving differential equations and approximating integrals.
• Computer Graphics: Taylor series are used in computer graphics for surface modeling and approximation. They can be used to create smooth curves and surfaces by approximating them with polynomial patches.
• Physics and Engineering: In physics, Taylor series are used to approximate solutions to complex physical systems. For example, they can be used to linearize equations of motion for small oscillations. In engineering, they are used in control systems, signal processing, and many other areas.
• Machine Learning: While less direct, Taylor series concepts are related to gradient-based optimization methods used in machine learning. Understanding Taylor series can provide a deeper understanding of how these optimization algorithms work.

Limitations of Taylor Series

While powerful, the Taylor series has limitations:

• Convergence: The Taylor series does not always converge to the original function for all values of x and y. The series may only converge within a certain radius of convergence.
• Computational Cost: Calculating higher-order derivatives can be computationally expensive, especially for complex functions.
• Approximation Error: The Taylor series provides an approximation, not an exact representation of the function (except for polynomials). The approximation error can become significant as you move further away from the point of expansion.
• Requirement of Differentiability: The function must be sufficiently differentiable at the point of expansion. If the function has discontinuities in its derivatives, the Taylor series may not exist or may not converge properly.

FAQ

Q: What is the difference between a Taylor series and a Maclaurin series?

A: A Maclaurin series is simply a Taylor series expanded around the point (0, 0). It's a special case of the Taylor series.

Q: How do I determine the order of the Taylor series to use?

A: The order of the Taylor series depends on the desired accuracy and the complexity of the function. Higher-order series generally provide better accuracy but are more computationally expensive. You may need to experiment to find the right balance.

Q: How do I handle functions with singularities (points where they are not differentiable)?

A: The Taylor series cannot be used directly at singularities. You may need to use other techniques, such as Laurent series, which are designed for functions with singularities.

Q: Can I use Taylor series for functions of more than two variables?

A: Yes, the Taylor series can be generalized to functions of any number of variables. The formula becomes more complex as the number of variables increases, but the underlying principle remains the same.

Q: How can I use software to compute Taylor series expansions?

A: Many mathematical software packages (e.g., Mathematica, Maple, MATLAB, SymPy in Python) have built-in functions for computing Taylor series expansions. These tools can significantly simplify the process, especially for complex functions.

Conclusion

The Taylor series for two variables is a valuable tool for approximating functions and solving problems in various fields. By understanding the underlying principles and practicing with examples, you can effectively apply this technique to analyze and approximate complex functions. While it has limitations, the Taylor series remains a fundamental concept in calculus and its applications. Remember to consider the point of expansion, the order of the approximation, and the convergence properties of the series to obtain accurate and meaningful results. Experiment with different functions and points to gain a deeper understanding of how the Taylor series works and its potential applications in your own work.