Chapter 5 — Sequences and Series of Functions
Functions built as limits of other functions — power series, Fourier series, iterates. When does the limit inherit continuity, integrals and derivatives? The answer is one idea: uniform convergence.
In 1821 Cauchy's influential course of analysis stated, with a proof, that a convergent series of continuous functions always has a continuous sum. Five years later Abel pointed out, in a footnote, that the series $\sin x-\tfrac12\sin2x+\tfrac13\sin3x-\cdots$ seems to contradict it: every term is continuous, the series converges at every real $x$, and yet its sum equals $\tfrac x2$ for $-\pi\lt x\lt\pi$ and drops to $0$ at $x=\pm\pi$ — a jump. Neither the theorem nor the example was wrong in an obvious way; what was missing was a distinction. A sequence of functions can converge at every point while the rate of convergence gets worse and worse from point to point. In the 1840s Seidel and Stokes isolated the difference, and Weierstrass made uniform convergence — convergence at one common rate for all points — a central notion of analysis.
This chapter is about that distinction and its rewards. Uniform convergence is exactly what is needed to pass continuity, integrals and (with care) derivatives to the limit. It makes power series behave like polynomials inside their interval of convergence, it explains Abel's theorem at the edge of that interval, and it frames one of the most beautiful facts of analysis: Weierstrass's theorem that every continuous function on a closed bounded interval is a uniform limit of polynomials. The same ideas reappear throughout the course — in the sup norm on $C[a,b]$ (Chapter 7), in the convergence theorems of Lebesgue integration (§4.5), and in the analytic functions of complex analysis (Unit 2).
- find the pointwise limit of a sequence of functions and decide, by computing $\sup_D|f_n-f|$ or by exhibiting bad points, whether the convergence is uniform on a given set;
- prove uniform convergence of series of functions with the Cauchy criterion, the Weierstrass M-test and the Dirichlet and Abel tests — and recognise when the M-test is silent;
- apply Dini's theorem and name the counterexample that shows each of its hypotheses is needed;
- state with full hypotheses, prove and apply the theorems on continuity, integration and differentiation of limits, and use the standard counterexamples to spot invalid interchanges;
- compute radii and intervals of convergence (Cauchy–Hadamard, ratio formula), differentiate and integrate power series term by term, and sum series with them and with Abel's theorem;
- prove the Weierstrass approximation theorem with Bernstein polynomials and use it, for example to show that a continuous function with vanishing moments is zero.
How this is examined
This chapter is one of the richest sources of CSIR NET questions. Part B typically asks for the supremum of $|f_n(x)-f(x)|$ over a set, whether a named sequence (the sliding hump $nxe^{-nx}$, the family $x^n$, a rational expression in $n$ and $x$) converges uniformly on a given interval, the radius of convergence of a power series with factorial or $n$-th power coefficients, or the value of $\lim_n\int f_n$ when the naive answer is wrong. Part C, where several options may be correct, prefers statements: "the limit function is continuous", "the series may be differentiated term by term", "the convergence is uniform on $[0,\infty)$ but not on $(0,\infty)$", "every continuous function on $(0,1)$ is a uniform limit of polynomials", "if all moments vanish then $f=0$". Each false option is one of the counterexamples in the ⚠️ boxes of the two sections, and each true option is a theorem whose hypotheses you must check. GATE MA asks the same material, often as numerical-answer questions on radii and limits of integrals.
Sections in this chapter
Pointwise and Uniform Convergence
Pointwise versus uniform limits, the sup-norm criterion, the standard examples, the Cauchy criterion, the Weierstrass M-test, Dirichlet and Abel tests for series of functions, and Dini's theorem.
Uniform Convergence, Power Series and Approximation
Continuity, integration and differentiation of limits; radius of convergence and term-by-term calculus of power series; Abel's theorem; Bernstein polynomials and the Weierstrass approximation theorem.
You will use convergence of real sequences and the completeness of $\mathbb{R}$ (§2.1), $\limsup$ and the comparison of ratios and roots (§2.2), the convergence tests for series — especially the Cauchy criterion, the root test, Leibniz, Dirichlet and Abel, and summation by parts (§2.3), compactness and the Heine–Borel theorem (§3.1), uniform continuity (§3.2), the mean value theorem (§3.3) and Darboux sums with Riemann's criterion (§4.1). Above all, be ready to read quantifiers carefully: the whole chapter turns on the difference between "for every $x$ there is $N$" and "there is $N$ for every $x$".