Difficulty: Intermediate to Advanced | Prerequisites: Calculus I fundamentals (limits, derivatives, integrals), basic set theory, comfort with logical quantifiers (∀, ∃).
Real analysis is the rigorous backbone behind everything you learned in Calculus I. Where calculus gives you the tools (derivatives, integrals, limits), real analysis asks why those tools work and when they break. This material also introduces the logical and foundational frameworks that underpin modern mathematics, including measure theory, distributions, and formal proof. If you skipped the epsilon-delta proofs in earlier courses, now is the time to revisit them.
Real analysis formalises the intuitive ideas of calculus (limits, continuity, convergence) using rigorous definitions built on the real number system. Mathematical logic and proof provide the reasoning framework. Measure theory generalises length, area, and volume. Distributions extend the concept of differentiation to functions that are not classically differentiable.
Real analysis (the theory of functions of a real variable)
A branch of mathematical analysis dealing with the real numbers and real-valued functions. It covers the analytic properties of real functions and sequences, convergence and limits, continuity, smoothness, and the calculus of the real numbers.
In simple terms, this is where you prove all the things calculus told you to take on faith.
Mathematical proof
A deductive argument for a mathematical statement, built from previously established theorems and ultimately traceable to axioms (self-evident or assumed truths). A proof must show a statement is always true, not just true in many observed cases.
Think of it as the difference between "I checked a thousand examples" and "I showed it works for every possible case."
Measure theory (measure)
A framework invented by Lebesgue to define integrals for all but the most pathological functions. A measure on a set is a systematic way to assign a number to each suitable subset, interpreted as its "size." It generalises the concepts of length, area, and volume.
In simple terms, it answers the question: how do you measure the size of something when the usual ruler does not work?
Distributions (generalised functions)
Introduced by Schwartz. Objects that generalise the classical notion of functions in analysis. They make it possible to differentiate functions whose derivatives do not exist in the classical sense. Any locally integrable function has a distributional derivative. Widely used in the theory of partial differential equations, where distributional solutions may exist even when classical solutions do not.
Think of it as a way to take derivatives of things that, classically, have no derivative.
Non-standard analysis
A reformulation of calculus using a logically rigorous notion of infinitesimal numbers, originated in the 1960s by Robinson. It uses mathematical logic to augment the real number system with infinitesimal and infinite numbers (called hyperreal numbers), recovering the original Newton-Leibniz style of reasoning.
In simple terms, it makes "infinitely small" a mathematically legal concept, not just a hand-wavy one.
Mathematical logic
A subfield of mathematics exploring the applications of formal reasoning to mathematics. It bears close connections to metamathematics, foundations of mathematics, and theoretical computer science. Its unifying themes are the expressive power of formal systems and the deductive power of formal proof systems.
Think of it as the study of what you can and cannot prove, and how to prove it properly.
Infinity (infinite, symbol: ∞)
An abstract concept describing something without any bound. In mathematics, infinity is often treated as if it were a number ("an infinite number of terms"), but it is not the same sort of number as a natural or real number.
In simple terms, infinity is not a number you can reach by counting; it is a concept describing unboundedness.
The analytic properties of real functions and sequences
Convergence and limits of sequences of real numbers
The calculus of real numbers (differentiation and integration, done rigorously)
Continuity, smoothness, and related properties of real-valued functions
Lebesgue's framework for defining integrals on a broader class of functions than Riemann integration allows
A measure assigns a "size" (a non-negative number) to each suitable subset of a given set
This generalises length (in one dimension), area (in two), and volume (in three)
A set is "suitable" if it belongs to a sigma-algebra, the collection of subsets the measure is defined on
The Riemann integral struggles with functions that are highly discontinuous. The Lebesgue integral handles these by measuring the "size" of the set where the function takes each value, rather than slicing the domain into small intervals. This makes convergence theorems (like the Dominated Convergence Theorem) far cleaner.
A proof is a deductive argument, not an empirical one. Checking many examples is not a proof.
Proofs trace back to axioms: self-evident or assumed starting points.
Two main strategies:
Direct proof: assume the hypotheses, derive the conclusion step by step
Proof by contradiction: assume the conclusion is false, derive a contradiction
Exhaustive case analysis counts as proof only when every possible case is listed and verified.
Studies formal reasoning and the strict rules of valid inference applied to mathematics
Key connections: metamathematics (reasoning about mathematical systems), foundations of mathematics (what mathematics is built on), theoretical computer science (decidability, computability)
Two central themes:
The expressive power of formal systems: what can a given formal language say?
The deductive power of formal proof systems: what can a given set of rules prove?
Infinity (∞) is not a real number. It describes unboundedness.
In mathematics, infinity is treated as if it were a number in certain contexts ("an infinite number of terms"), but it does not obey the usual arithmetic rules.
Distinguishing countable infinity (the naturals, ℕ) from uncountable infinity (the reals, ℝ) is a foundational result of set theory (Cantor's diagonal argument).
Introduced by Laurent Schwartz to extend differentiation to functions that are not classically differentiable
A distribution is not a function in the traditional sense; it is a linear functional acting on test functions (smooth functions with compact support)
Any locally integrable function has a distributional derivative
In the theory of partial differential equations (PDEs), it is often easier to establish the existence of distributional (or "weak") solutions than classical solutions
The blanks in the source material: "any locally integrable function has a distributional (adj. form) derivative" and "establish the existence of distributional (adj. form) solutions"
Originated in the 1960s by Abraham Robinson
Makes the notion of infinitesimal numbers logically rigorous, rather than treating them as informal shorthand
Uses the transfer principle from mathematical logic to augment the real number system with infinitesimal and infinite numbers
The resulting extended number system is called the hyperreal numbers (denoted *ℝ)
Hyperreals allow a Leibniz-style development of calculus: instead of epsilon-delta limits, you work directly with infinitesimals, then take the "standard part" to recover a real-number answer
This approach does not replace standard analysis; it provides an alternative foundation that some find more intuitive for certain arguments
\lim_{n \to \infty} a_n = L \quad \Leftrightarrow \quad \forall \varepsilon > 0, \; \exists N \in \mathbb{N} : n > N \Rightarrow |a_n - L| < \varepsilonThe epsilon-delta definition of a limit of a sequence. This is the foundational definition you will use throughout real analysis.
\mu\left(\bigcup_{i=1}^{\infty} A_i\right) = \sum_{i=1}^{\infty} \mu(A_i) \quad \text{(for pairwise disjoint } A_i \text{)}Countable additivity of a measure. This property is what distinguishes a measure from a simpler "size" assignment.
Notation to know: ℝ (the reals), ℕ (the naturals), ∞ (infinity), *ℝ (the hyperreals in non-standard analysis), ∀ (for all), ∃ (there exists), σ-algebra (the collection of "measurable" sets).
Measure theory is the mathematical backbone of probability theory. Every probability space is built on a measure (the probability measure), which is why Lebesgue integration appears throughout statistics, machine learning, and quantitative finance.
Distributions are essential in physics and engineering. The Dirac delta "function" (which is a distribution, not a classical function) is used to model point charges, impulse forces, and signal processing filters.
Students often think "proof by example" is valid. It is not. Showing something works for ten, a hundred, or a million cases does not constitute a mathematical proof. You must show it holds for all cases.
Students often confuse infinity with a very large number. Infinity is not a number at all; it is a concept describing unboundedness. You cannot do ordinary arithmetic with it.
Students often assume all infinite sets are the same "size." They are not. The natural numbers (ℕ) are countably infinite; the real numbers (ℝ) are uncountably infinite, a strictly larger kind of infinity.
Students often think the Dirac delta is a function. It is not a function in the classical sense; it is a distribution. It assigns a value through integration against test functions, not through pointwise evaluation.
⚠️ Be able to state and apply the epsilon-delta definition of a limit. This is the single most tested foundational concept in real analysis.
⚠️ Know the difference between a proof and an empirical argument. Exam questions frequently ask you to prove a statement, not verify it with examples.
⚠️ Understand what a measure is and why countable additivity matters. Expect at least one question connecting measure theory to integration.
⚠️ Know what "distributional derivative" means and why it matters for PDEs. The fill-in-the-blank format in the source material suggests this is exam-relevant vocabulary.
⚠️ Be clear on what hyperreal numbers are and how non-standard analysis relates to the standard epsilon-delta approach.
True or False: Showing that a statement holds for 1,000 cases constitutes a mathematical proof. (False. A proof must cover all cases.)
Fill in the blank: A measure on a set is a systematic way to assign a ______ to each suitable subset of that set. (number, intuitively interpreted as its size)
True or False: Infinity is a real number. (False. It is a concept describing unboundedness.)
Fill in the blank: Any locally integrable function has a ______ derivative. (distributional)
True or False: Non-standard analysis was developed in the 1800s. (False. It was developed in the 1960s by Robinson.)
Q: What distinguishes a mathematical proof from an empirical argument?
A: A proof uses deductive reasoning from axioms and previously established theorems to show a statement is always true. An empirical argument only checks individual cases and cannot guarantee the statement holds universally.
Q: What problem did Lebesgue's measure theory solve?
A: It provided a rigorous framework for defining integrals of a much broader class of functions than the Riemann integral could handle, particularly highly discontinuous functions.
Q: What are distributions, and why were they introduced?
A: Distributions (generalised functions) were introduced by Schwartz to extend differentiation to functions whose classical derivatives do not exist. They are especially useful in the theory of PDEs, where distributional solutions may exist even when classical solutions do not.
Q: What are hyperreal numbers?
A: The number system produced by non-standard analysis, which augments the real numbers with infinitesimal and infinite quantities. They allow a rigorous version of the Leibniz-style calculus that works directly with infinitesimals.
Q: Is infinity a number? Explain.
A: No. Infinity is an abstract concept describing unboundedness. While it is sometimes treated as if it were a number in mathematical notation (e.g., limits "approaching infinity"), it does not obey the ordinary rules of arithmetic and is not an element of the real numbers.
This material connects directly to probability theory, because every probability space is built on a measure. If you go on to study stochastic processes or statistics at a graduate level, measure theory is the language you will work in.
Mathematical logic connects to theoretical computer science, particularly computability theory and the limits of what algorithms can decide. Godel's incompleteness theorems, a landmark of mathematical logic, show that any sufficiently powerful formal system contains true statements it cannot prove.
Distributions connect to Fourier analysis and signal processing. The Fourier transform of a distribution is itself a distribution, which is how engineers handle signals that are not square-integrable.
Real analysis, theory of functions of a real variable, mathematical proof, deductive reasoning, axioms, measure theory, Lebesgue integral, Lebesgue measure, sigma-algebra, distributions, generalised functions, Schwartz distributions, distributional derivative, weak solution, non-standard analysis, Robinson, hyperreal numbers, infinitesimals, transfer principle, mathematical logic, formal systems, formal proof, metamathematics, foundations of mathematics, infinity, countable, uncountable, Cantor, epsilon-delta, convergence, continuity, smoothness, MATH 104, calculus foundations, Penn State calculus