Convergence and Divergence of Series: Calculus 2 Tutorial

Alternating series ∑(-1)^n a_n converges if a_n is decreasing and lim(n→∞) a_n = 0 (Alternating Series Test). Absolute convergence: ∑|a_n| converges. Conditional convergence: series converges but ∑|a_n| diverges. Absolute convergence implies conditional convergence.

Alternating Series Test: ∑(-1)^n b_n converges if b_n is non-increasing and lim b_n = 0. Absolute convergence: ∑|a_n| converges. Conditional convergence: series converges but ∑|a_n| diverges. Absolute convergence implies convergence, but conditional convergence does not.

A series is absolutely convergent if it converges and its series of absolute values also converges. A series is conditionally convergent if it converges but its series of absolute values diverges. For alternating series, if the absolute value series converges, the original series is absolutely convergent; otherwise, it is conditionally convergent. This distinction is crucial for exam questions asking whether a series is absolutely or conditionally convergent.

This section covers alternating series and absolute convergence: (1) Alternating series test (Leibniz test) - for ∑(-1)ⁿaₙ with aₙ > 0, if (aₙ) is monotonically decreasing and lim(n→∞) aₙ = 0, then the series converges; (2) Absolute convergence - a series ∑aₙ is absolutely convergent if ∑|aₙ| converges; (3) Conditional convergence - a series converges but not absolutely; (4) Absolute convergence implies convergence, but conditional convergence does not imply absolute convergence; (5) Examples demonstrating the distinction between absolute and conditional convergence; (6) Techniques for evaluating series with alternating signs. The section emphasizes that understanding the distinction between absolute and conditional convergence is crucial for analyzing series behavior.

This section covers absolute and conditional convergence. A series is absolutely convergent if ∑|a_n| converges. Absolute convergence implies regular convergence, but not vice versa. A series is conditionally convergent if it converges but is not absolutely convergent. The Alternating Series Test states that ∑(-1)^(n+1)b_n converges if b_n > 0, decreases monotonically, and approaches zero. The nth Term Test for Divergence states that if lim(a_n) ≠ 0, the series diverges.

To find the radius and interval of convergence for a power series, use the ratio test: if the limit equals zero, the series converges for all x (R=∞, interval=(-∞, ∞)); if the limit equals infinity, the series only converges at the center c (R=0, interval={c}); if the limit yields an expression like |x-c|/r < 1, solve for x to find the interval and then check the endpoints separately using the original series.

A power series is Σ (from n=0 to ∞) Cₙ(X - a)ⁿ, where X is the variable, Cₙ are coefficients, and 'a' is the center. The center is found by setting the expression inside parentheses to zero. Power series may converge for some X values and diverge for others. To find the interval of convergence, apply the ratio test: compute limₙ→∞ |aₙ₊₁/aₙ|. For Σ (X - 1)ⁿ/3ⁿ, this yields |X - 1|/3. Setting this < 1 gives radius R = 3 and interval |X - 1| < 3 → -2 < X < 4. Always test endpoints separately using the divergence test.

To find the radius and interval of convergence of a power series: (1) Use the formula R = lim(n→∞) |c_n/c_{n+1}| to find the radius; (2) Identify the center 'a' from (x - a); (3) Determine the interval as (a-R, a+R); (4) Test both endpoints separately using convergence tests like the Test for Divergence; (5) Include endpoints only if they converge. In this example, R = 2, center a = 3, interval (1, 5), with both endpoints diverging.

To find the interval of convergence: (1) Use Ratio Test to find radius R, (2) Solve |x-a| < R for the open interval, (3) Test endpoints separately. For ∑(-1)^n 4^n x^n/√n, Ratio Test gives R=1/4. Test x=1/4: ∑1/√n diverges (p=1/2<1). Test x=-1/4: ∑(-1)^n/√n converges conditionally. When lim(n→∞)|a_{n+1}/a_n| = 0 < 1 for all x, radius is infinite and interval is (-∞, ∞).

A power series is an infinite series of the form sum from n=0 to infinity of a_n(x - c)^n, where a_n are coefficients and c is the center. The simplest power series is the geometric series sum from n=0 to infinity of x^n = 1 + x + x^2 + ..., which equals 1/(1 - x) when |x| < 1. The radius of convergence R determines where the series converges and can be found using the ratio test formula R = lim(n→∞) |a_n/a_{n+1}| or the root test formula R = lim(n→∞) |a_n|^(1/n). The interval of convergence is found from |x - c| < R, giving x in (c - R, c + R). At the endpoints, the ratio and root tests fail, so convergence must be tested separately using other methods like the alternating series test or p-series test. Multiple examples demonstrate these concepts: sum from n=1 to infinity of (-1)^(n+1)x^n/n has R = 1, interval [-1, 1); sum from n=0 to infinity of x^n/n! has R = ∞, converging for all real x; sum from n=0 to infinity of (-1/2)^n(x - 2)^n has R = 1/2, interval (3/2, 5/2].

Taylor and Maclaurin polynomials are methods for approximating arbitrary functions using polynomials by matching the function's value and its derivatives at a specific point (typically 0 for Maclaurin series). The general form is p(x) = f(0) + f'(0)x + (f''(0)/2!)x² + (f'''(0)/3!)x³ + ... + (f^(n)(0)/n!)x^n, where each term includes the n-th derivative of the function evaluated at 0 divided by n factorial, multiplied by x^n. By adding more terms, the polynomial approximation becomes increasingly accurate near the center point, with the Maclaurin series being a special case of the Taylor series centered at 0.

Taylor and Maclaurin series convert complex functions into infinite polynomial series for simplification. Taylor series approximate functions around a center point x₀, with accuracy depending on proximity to x₀. The interval of convergence is typically (x₀ - 1, x₀ + 1). Maclaurin series is the special case where x₀ = 0. Standard series for basic functions include: e^x = 1 + x + x²/2! + ..., sin(x) = x - x³/3! + ..., cos(x) = 1 - x²/2! + ... For complex functions like √x or √(x-1), we calculate derivatives at the center point and substitute into the Taylor series formula. For implicit functions like y = x/y, we use implicit differentiation to find derivatives before applying the series.

Taylor and Maclaurin series are polynomials that can approximate other functions. The more terms included in the polynomial, the better the approximation becomes. For example, with e^x, as you include more terms in the series, the approximation gets closer to e^x. When extended to infinitely many terms, it provides a perfect approximation where plugging in any real value of x into both e^x and the infinite polynomial yields the same result.

When evaluating limits as x approaches 0, transcendental functions like sin(x), cos(x), e^x, and ln(1+x) can be approximated by their lowest-degree polynomial terms from Taylor series, where sin(x) ≈ x, cos(x) ≈ 1 - x²/2, e^x ≈ 1 + x, and ln(1+x) ≈ x; this approximation is valid because higher-order terms become negligible near zero, allowing complex limit problems to be simplified by replacing transcendental functions with their polynomial equivalents.

Taylor and Maclaurin series are infinite series that represent functions as sums of terms involving derivatives evaluated at a specific point, allowing complex functions to be approximated by polynomials; Maclaurin series are a special case of Taylor series where the expansion point is zero.

This concluding section develops the theoretical framework governing Fourier series behavior and demonstrates practical applications. The convergence theorem states that Fourier series converge to the function at continuity points and to the average of left/right limits at discontinuities. The video applies this to evaluate the Gregory series Σ [(-1)^(k+1)]/(2k-1) by substituting x = 1 into the Fourier series, yielding π/4. This demonstrates how Fourier series provide exact values for infinite series—a rare capability in calculus. Alternative solution methods leverage Fourier series linearity: decomposing f(x) = 3x + 3 into the Fourier series of 3x plus the constant function 3. Since the Fourier series of a constant equals the constant itself, this approach simplifies calculations. The section synthesizes Fourier series as powerful tools for both function approximation and exact infinite series evaluation, contrasting their interval-wide applicability against point-based methods like Taylor series.

This extensive section demonstrates how Fourier series serve as powerful analytical tools for computing infinite series sums. Three detailed examples are presented: (1) For f(x) = x² on [-π, π], the Fourier series yields Σₙ=1^∞ 1/n² = π²/6 through evaluation at x = π; (2) For the square wave function f(x) = ±1, the Fourier series leads to Σₖ=0^∞ (-1)^k/(2k+1) = π/4, providing an analytical proof of the Leibniz formula; (3) For f(x) = x on (-π, π), the Fourier series equals x for all x in (-π, π). Each example illustrates how verifying Dirichlet's conditions enables direct computation of series values, transforming difficult infinite sums into tractable Fourier series evaluations. This demonstrates the practical utility of Fourier analysis in mathematical computation.

Complex numbers simplify solving linear differential equations with constant coefficients. Assuming solutions of the form e^(λx) reduces equations to characteristic polynomial root-finding. For example, y'' + 2y' + 2y = 0 yields λ = -1 ± i, giving solutions e^(-x)(cos(x) ± i sin(x)), from which real solutions emerge. Fourier series benefit similarly: periodic functions f(x) can be expressed as ∑cₙe^(inx) over all integers n, replacing separate sine and cosine series. This complex exponential form simplifies calculations dramatically, as orthogonality conditions reduce to simple delta functions, making integration and coefficient computation much more straightforward.

Fourier series can be used to solve differential equations when the forcing function f(x) is periodic. The solution involves three steps: (1) finding the Fourier series of f(x), (2) assuming the solution y(x) has the same period and expressing it as a Fourier series, and (3) substituting the series and its derivatives into the differential equation to solve for the coefficients. This method works for differential equations of any order, with the key requirement being that f(x) must be periodic.

Fourier series represent periodic functions as infinite sums of trigonometric terms (sine and cosine), with coefficients calculated through integration; even functions yield only cosine terms while odd functions yield only sine terms, and these series can be used to evaluate infinite series by substituting specific values.