Lecture from: 29.05.2024 | Video: Video ETHZ
Clicker Question: A Famous Improper Integral
What is the value of ? Answer:
Reasoning
Strategy: Split, Evaluate, Combine! To tackle an integral over the entire real line , we first need to check if it “behaves” at both ends. The standard way is to split it into two pieces at some convenient point (like ), and then evaluate each piece as an improper integral. If both pieces converge to finite values, then the original integral converges to their sum.
An antiderivative (“Stammfunktion”) of is .
Let’s evaluate the two halves:
-
Integral from to : We know . As , (the arctangent function approaches its horizontal asymptote). So, . This part converges.
-
Integral from to : As , . So, . This part also converges.
Combining the pieces: According to our definition for improper integrals that are open/unbounded on both sides, we pick an arbitrary point (let’s use here, as we effectively did above) and sum the two improper integrals: Using : Since both parts converge, the whole integral converges to . This is a well-known and beautiful result!
The Integral Test for Series Convergence
Connecting Infinite Sums and Infinite Areas This theorem, often called the Integral Test, provides a powerful bridge between the convergence of an infinite series and the convergence of an improper integral . It applies to positive, decreasing functions. The intuition is that the sum can be seen as an approximation of the area under the curve (using rectangles of width 1), and vice-versa.
Theorem: The Integral Test
Let be a monotonically decreasing function (so and if ).
Then the infinite series converges if and only if the improper integral converges.
Proof Idea (Using “Staircase” Functions)
The core idea is to compare the series with the integral by “sandwiching” the function between two simpler, “step-like” functions built from the series terms.
Let be the floor function (the greatest integer less than or equal to ). Define two auxiliary functions for :
- (This function is constant on intervals , taking the value ).
- (This function is constant on intervals , taking the value ).
In the image:
- On the interval , . This forms an “upper” rectangle.
- On the interval , . This forms a “lower” rectangle.
Since is monotonically decreasing: For any , we have . Because is decreasing, . This translates to: for all .
Now, let’s integrate these functions from to some large integer : .
Let’s evaluate the integrals of and :
- . (The integral of from to is the sum of the areas of the “upper” rectangles, which corresponds to the sum of the series terms ).
- . (The integral of from to is the sum of the areas of the “lower” rectangles, corresponding to ).
So, we have the crucial chain of inequalities:
Now, we consider the two directions of the “if and only if”
-
If converges: This means the sequence of partial sums is bounded above. From the right inequality, (which is a finite number). The function is monotonically increasing (since ). Since is monotonically increasing and bounded above, must converge.
-
If converges: This means is a finite number. From the left inequality, (which is finite). So, the partial sums are bounded above. Since , the series is a series of non-negative terms with bounded partial sums, so it must converge. If converges, then also converges.
This completes the argument.
Example: The p-Series
When does the series converge (for )? The function is positive and monotonically decreasing for if . So we can apply the Integral Test. We need to check the convergence of the corresponding improper integral .
We’ve already answered this question! (In Example 5.8.2 (2) from the previous lecture notes). The integral converges if and only if .
Therefore, by the Integral Test, the series converges if and only if . This is a very important result for series convergence!
Example: Series involving
When does the series converge? (We start at because ). The function is positive and decreasing for (if , or if and is large enough). We would compare this to the integral . Let , so . The integral becomes . From our previous p-series integral result (replacing with and with ), this integral converges if and only if . So, the series converges if and only if . (See script for more details on checking the conditions for ).
Conclusion: Stirling’s Formula - Approximating Factorials
A Remarkable Approximation for Factorials grow incredibly fast. Stirling’s formula gives an amazing approximation for when is large, connecting it to powers of , , and . It often arises from approximating with .
Definition: Asymptotic Equivalence
Two sequences of positive real numbers, and , are called asymptotically equivalent if Notation: . This means that for large , is very close to in a relative sense.
Theorem: Stirling’s Formula (Asymptotic Version)
This means for large .
Proof Idea (Sketch using Integral Approximation of )
The proof is quite involved and often uses techniques like approximating sums by integrals, or more advanced methods. A common starting point is to consider : .
This sum can be approximated by the integral . We know . So, .
The sum can be more precisely related to using the trapezoidal rule for approximating integrals, or by carefully comparing the sum to areas of rectangles above and below the curve.
The area of the trapezoid under the graph of from to is .
Summing these trapezoids from to : (since ).
Let . . It can be shown that this sequence (which represents the accumulated difference between the integral and the trapezoidal sum) converges to a constant as .
The proof involves showing is non-negative and monotonically increasing (or decreasing, depending on the precise setup) and bounded. This means .
Exponentiating this gives . The constant turns out to be . Determining this exact constant usually requires a more sophisticated argument, often involving Wallis’s product for : By substituting the asymptotic form into Wallis’s product, one can solve for and find .
This leads to the full Stirling’s formula: .
Summary of the Course (Chapters 1-5)
Let’s take a moment to look back at the major themes and results from the first five chapters. This helps consolidate what we’ve learned and see how the ideas connect.
Chapter 1: Real Numbers, Euclidean Spaces, Complex Numbers
- Real Numbers (): Field axioms (how addition/multiplication work), order axioms (how comparison works).
- Key Difference between and : is order complete (or has the least upper bound property). This is fundamental!
- Consequences of Completeness: Existence of square roots (e.g., is real but not rational), supremum and infimum for bounded sets.
- as a Real Vector Space: Vectors, addition, scalar multiplication.
- Scalar (Dot) Product in , Cross Product in .
- Complex Numbers (): Definition , operations, polar form .
- Roots of Unity (Kreisteilungsgleichung), Fundamental Theorem of Algebra (“All polynomials over can be fully factored into linear terms”).
Chapter 2: Sequences and Series
- Sequences: Definition, convergence, limits.
- Limit Laws for sequences (sum, product, quotient).
- Weierstrass Theorem for Monotonic Sequences: “Monotonic bounded sequences converge.”
- Limes Superior, Limes Inferior: Greatest/smallest accumulation points of a sequence.
- Cauchy Sequences: Sequences whose terms eventually get arbitrarily close to each other. (In and , Cauchy convergent).
- Nested Interval Theorem.
- Bolzano-Weierstrass Theorem: “Bounded sequences (in ) have convergent subsequences.”
- Sequences in and .
- Series:
- Convergence of series (as limits of partial sums).
- Geometric series.
- Comparison tests.
- Absolute Convergence ( converges) convergence of .
- Alternating series, Leibniz criterion.
- Rearrangements of series (absolutely convergent series can be rearranged freely).
- Ratio and Root tests for convergence.
- Power Series and their radius of convergence (absolute convergence inside, divergence outside).
- Double series, Cauchy product of series.
- Exponential Series and its addition theorem .
Chapter 3: Continuous Functions
- Real-valued functions: Domain, codomain, boundedness, monotonicity.
- Continuity:
- definition.
- Sequential continuity: ”.”
- Intermediate Value Theorem: “Continuous functions on intervals take on every value between any two function values.”
- Min-Max Theorem (Extreme Value Theorem): “Continuous functions on compact (closed and bounded) intervals attain their maximum and minimum values.”
- Continuity of Inverse Functions (especially for monotonic functions).
- Real Exponential Function () and its Properties.
- Natural Logarithm () as its inverse.
- Convergence of Function Sequences:
- Pointwise vs. Uniform convergence.
- Continuity of the limit function under uniform convergence.
- Continuity of Power Series (within their radius of convergence).
- Trigonometric Functions (), the number .
- Limits of Functions:
- Definition via sequence convergence.
- Limit laws, examples.
- One-sided limits ().
Chapter 4: Differentiable Functions
- Definition via limit of difference quotients: .
- Tangent lines to functions.
- Differentiation Rules: Linearity, Product Rule, Quotient Rule, Chain Rule.
- Derivative of Inverse Functions: .
- Necessary Condition for Local Extrema: (for interior points).
- Rolle’s Theorem.
- Mean Value Theorem (MVT): ” such that .”
- Connection between Monotonicity and the First Derivative.
- Arcus Functions (inverse trig) and their derivatives.
- L’Hôpital’s Rule for limits of "" or "" forms (“differentiate numerator and denominator”).
- Convex Functions.
- Higher Order Derivatives and smooth functions.
- Differentiation of Power Series (term by term).
- Taylor Approximation / Taylor Polynomials: .
- Taylor’s Theorem with Remainder: .
- Sufficient Criterion for Local Extrema (Higher Derivative Test: if and , then odd no extremum; even and local min; even and local max).
Chapter 5: The Riemann Integral
- Definition via partitions, lower/upper Darboux sums.
- Mesh size of partitions.
- Riemann Sums: , with .
- Integrable Functions:
- Fundamental operations preserve integrability ( etc.).
- Linearity of the integral.
- Continuous functions are integrable.
- Monotonic functions are integrable.
- Monotonicity and Triangle Inequality for Integrals.
- Cauchy-Schwarz Inequality for Integrals: .
- Mean Value Theorem for Integrals.
- Fundamental Theorem of Calculus (Hauptsatz):
- Existence of antiderivatives for continuous functions: .
- Calculation of definite integrals using antiderivatives: .
- Integration by Parts.
- Substitution Rule.
- Area of Circles, Volume of Spheres.
- Indefinite Integrals, Techniques of Integration.
- Partial Fraction Decomposition (for rational functions: case “all real roots in denominator,” case “also complex roots in denominator”).
- Function Sequences and Integration (uniform convergence allows swapping and ).
- Integration of Power Series (term by term).
- Improper Integrals (unbounded intervals or functions).
- Comparison Test between Series and Improper Integrals (Integral Test).
- Stirling’s Formula: .