UniMaths
Free chapterTry everything here: cards, proofs and practice. Sign up free to save your progress and unlock all 7 chapters.Sign up free
CALC1 · 1

Functions and limits

12 concepts · 49 questions

Standard functions (exponentials, logs, inverse trig and hyperbolic functions), limits and how to compute them, standard limits, continuity and asymptotes.

Overview 1The plan of this section

Calculus is about how functions change, so we first need a reliable stock of functions and a working notion of limit.

  1. The standard functions: exponentials and logarithms (Def. 1.2), inverse trig (Def. 1.3) and hyperbolic functions (Def. 1.4, Prop. 1.5).
  2. Limits (Def. 1.6): what f(x)f(x) approaches, computed with the limit laws (Technique 1.7) and a few standard limits (Prop. 1.8).
  3. Continuity (Def. 1.9) and the behaviour at infinity (Def. 1.10). The rigorous ε\varepsilon–δ\delta theory is in Analysis I; here the aim is fluent computation.

Toolkit first: exp/log, inverse trig, hyperbolic functions.

Then limits: laws, standard limits, continuity, asymptotes.

Everything later (derivatives, integrals, series) is built from limits.

Definition 1.1Functions, domains, composition and inverses

A real function ff assigns to each xx in its domain D⊆RD\subseteq\mathbb{R} one value f(x)f(x); its range is {f(x):x∈D}\{f(x): x\in D\}. The natural domain of a formula is the largest set where it makes sense (no division by 00, square roots of negatives, logs of non-positives). The composite (f∘g)(x)=f(g(x))(f\circ g)(x) = f(g(x)) is defined where g(x)g(x) lies in the domain of ff. If ff is one-to-one (strictly increasing or decreasing on its domain), it has an inverse f−1f^{-1} with f−1(f(x))=xf^{-1}(f(x)) = x; its graph is the reflection of ff's in y=xy = x. ff is even if f(−x)=f(x)f(-x) = f(x) and odd if f(−x)=−f(x)f(-x) = -f(x).

Natural domain: avoid ÷0\div0, negative\sqrt{\text{negative}}, ln⁡(non-positive)\ln(\text{non-positive}).

Inverse: swap xx and yy and solve; graph reflected in y=xy = x. Needs ff one-to-one.

Even (symmetric in the yy-axis): x2,cos⁡xx^2, \cos x. Odd (rotational symmetry): x3,sin⁡xx^3, \sin x.

Definition 1.2Exponentials and logarithms

exe^x is the exponential function (e≈2.71828e\approx2.71828); it is positive, strictly increasing, and ea+b=eaebe^{a+b} = e^ae^b, (ea)b=eab(e^a)^b = e^{ab}. Its inverse is the natural logarithm ln⁡:(0,∞)→R\ln: (0, \infty)\to\mathbb{R}, so eln⁡x=xe^{\ln x} = x and ln⁡ex=x\ln e^x = x. Laws: ln⁡(ab)=ln⁡a+ln⁡b\ln(ab) = \ln a + \ln b, ln⁡ab=ln⁡a−ln⁡b\ln\frac ab = \ln a - \ln b, ln⁡ar=rln⁡a\ln a^r = r\ln a. General powers: ax=exln⁡aa^x = e^{x\ln a} for a>0a > 0, and log⁡ax=ln⁡xln⁡a\log_ax = \frac{\ln x}{\ln a}.

ln⁡\ln undoes exe^x. ln⁡\ln turns products into sums and powers into multiples.

ax=exln⁡aa^x = e^{x\ln a}: every exponential is ee to something.

Change of base: log⁡ax=ln⁡x/ln⁡a\log_ax = \ln x/\ln a.

Definition 1.3Inverse trigonometric functions

Restrict the trig functions to intervals where they are one-to-one, and invert:

  • arcsin⁡:[−1,1]→[−π2,π2]\arcsin: [-1, 1]\to[-\frac\pi2, \frac\pi2], the inverse of sin⁡\sin on [−π2,π2][-\frac\pi2, \frac\pi2];
  • arccos⁡:[−1,1]→[0,π]\arccos: [-1, 1]\to[0, \pi], the inverse of cos⁡\cos on [0,π][0, \pi];
  • arctan⁡:R→(−π2,π2)\arctan: \mathbb{R}\to(-\frac\pi2, \frac\pi2), the inverse of tan⁡\tan on (−π2,π2)(-\frac\pi2, \frac\pi2). So sin⁡(arcsin⁡x)=x\sin(\arcsin x) = x for x∈[−1,1]x\in[-1, 1], but arcsin⁡(sin⁡x)=x\arcsin(\sin x) = x only for x∈[−π2,π2]x\in[-\frac\pi2, \frac\pi2].

arcsin⁡,arctan⁡\arcsin, \arctan give angles in [−π2,π2][-\frac\pi2, \frac\pi2]; arccos⁡\arccos in [0,π][0, \pi].

Trap: arcsin⁡(sin⁡3π4)=π4\arcsin(\sin\frac{3\pi}4) = \frac\pi4, not 3π4\frac{3\pi}4.

Draw a triangle to simplify things like cos⁡(arcsin⁡x)=1−x2\cos(\arcsin x) = \sqrt{1 - x^2}.

Definition 1.4Hyperbolic functions

cosh⁡x=ex+e−x2,sinh⁡x=ex−e−x2,tanh⁡x=sinh⁡xcosh⁡x.\cosh x = \frac{e^x + e^{-x}}2,\qquad\sinh x = \frac{e^x - e^{-x}}2,\qquad\tanh x = \frac{\sinh x}{\cosh x}.

cosh⁡\cosh is even with minimum cosh⁡0=1\cosh0 = 1; sinh⁡\sinh is odd and strictly increasing; tanh⁡\tanh is odd with range (−1,1)(-1, 1). (A hanging chain has the shape of a cosh⁡\cosh curve, the catenary.)

cosh⁡=\cosh = even part of exe^x, sinh⁡=\sinh = odd part: ex=cosh⁡x+sinh⁡xe^x = \cosh x + \sinh x.

Shapes: cosh⁡\cosh is a U (min 11), sinh⁡\sinh an S through 00, tanh⁡\tanh squashed between ±1\pm1.

Proposition 1.5Hyperbolic identities and inverse hyperbolic functions

(i) cosh⁡2x−sinh⁡2x=1\cosh^2x - \sinh^2x = 1; sinh⁡2x=2sinh⁡xcosh⁡x\sinh2x = 2\sinh x\cosh x; cosh⁡2x=cosh⁡2x+sinh⁡2x\cosh2x = \cosh^2x + \sinh^2x. (ii) Osborn's rule: a trig identity becomes a hyperbolic one by replacing cos⁡→cosh⁡\cos\to\cosh, sin⁡→sinh⁡\sin\to\sinh and changing the sign of each product of two sines. (iii) arsinh⁡x=ln⁡(x+x2+1)\operatorname{arsinh}x = \ln\big(x + \sqrt{x^2 + 1}\big) for all xx; arcosh⁡x=ln⁡(x+x2−1)\operatorname{arcosh}x = \ln\big(x + \sqrt{x^2 - 1}\big) for x≥1x\ge1; artanh⁡x=12ln⁡1+x1−x\operatorname{artanh}x = \frac12\ln\frac{1 + x}{1 - x} for ∣x∣<1|x| < 1.

cosh⁡2−sinh⁡2=1\cosh^2 - \sinh^2 = 1 (compare cos⁡2+sin⁡2=1\cos^2 + \sin^2 = 1).

Inverses are logs: arsinh⁡x=ln⁡(x+x2+1)\operatorname{arsinh}x = \ln(x + \sqrt{x^2 + 1}).

Method: set y=sinh⁡xy = \sinh x, get a quadratic in exe^x, keep the positive root.

Proof

Definition 1.6Limits

lim⁡x→af(x)=L\lim_{x\to a}f(x) = L means f(x)f(x) can be made as close to LL as we like by taking xx close enough to aa (but x≠ax\ne a). Precisely: for every ε>0\varepsilon > 0 there is δ>0\delta > 0 with ∣f(x)−L∣<ε|f(x) - L| < \varepsilon whenever 0<∣x−a∣<δ0 < |x - a| < \delta. One-sided limits x→a−x\to a^-, x→a+x\to a^+ approach from one side; the limit exists iff both exist and agree. The value f(a)f(a) itself plays no role.

Limit = where f(x)f(x) is heading as x→ax\to a, regardless of f(a)f(a).

Exists iff left and right limits agree.

Example: x2−1x−1→2\frac{x^2 - 1}{x - 1}\to2 as x→1x\to1, though it is undefined at 11.

Technique 1.7Computing limits

Limit laws: limits of sums, products, quotients (non-zero denominator limit) and compositions with continuous functions are what you expect. If direct substitution gives a number, that's the limit. For 00\frac00 forms: (1) factorise and cancel; (2) rationalise (multiply by a conjugate) for square roots; (3) use a standard limit (see Prop. 1.8, next); (4) squeeze: if g≤f≤hg\le f\le h near aa and g,h→Lg, h\to L then f→Lf\to L; (5) later, L'Hôpital or Taylor series.

Try substituting. If 00\frac00: factorise, rationalise, use standard limits, or squeeze.

Squeeze example: ∣xsin⁡1x∣≤∣x∣→0|x\sin\frac1x|\le|x|\to0. Never "cancel" a zero factor without checking it's 00\frac00.

Proposition 1.8Standard limits

lim⁡x→0sin⁡xx=1,lim⁡x→01−cos⁡xx2=12,lim⁡x→0ex−1x=1,lim⁡x→0ln⁡(1+x)x=1,lim⁡n→∞(1+an)n=ea.\lim_{x\to0}\frac{\sin x}x = 1,\qquad\lim_{x\to0}\frac{1 - \cos x}{x^2} = \frac12,\qquad\lim_{x\to0}\frac{e^x - 1}x = 1,\qquad\lim_{x\to0}\frac{\ln(1 + x)}x = 1,\qquad\lim_{n\to\infty}\Big(1 + \frac an\Big)^n = e^a.

Also, as x→∞x\to\infty: xkex→0\frac{x^k}{e^x}\to0 and ln⁡xxk→0\frac{\ln x}{x^k}\to0 for every k>0k > 0 (exponentials beat powers, powers beat logs).

sin⁡xx→1\frac{\sin x}x\to1, ex−1x→1\frac{e^x - 1}x\to1, ln⁡(1+x)x→1\frac{\ln(1 + x)}x\to1, 1−cos⁡xx2→12\frac{1 - \cos x}{x^2}\to\frac12.

(1+an)n→ea(1 + \frac an)^n\to e^a.

Growth: ln⁡x≪xk≪ex\ln x\ll x^k\ll e^x.

Proof

Definition 1.9Continuity and the intermediate value theorem

ff is continuous at aa if lim⁡x→af(x)=f(a)\lim_{x\to a}f(x) = f(a), and continuous on an interval if continuous at every point. Polynomials, rational functions (where defined), exe^x, ln⁡\ln, trig and their combinations are continuous. Intermediate value theorem (proved in Analysis I): if ff is continuous on [a,b][a, b] and cc is between f(a)f(a) and f(b)f(b), then f(x)=cf(x) = c for some x∈[a,b]x\in[a, b]. In particular a sign change means a root.

Continuous = no jumps, holes or blow-ups: lim⁡f(x)=f(a)\lim f(x) = f(a).

IVT: continuous + sign change ⇒ a root in between.

Piecewise functions: check the glue points.

Definition 1.10Limits at infinity and asymptotes

lim⁡x→∞f(x)=L\lim_{x\to\infty}f(x) = L means f(x)f(x) gets close to LL for large xx; then y=Ly = L is a horizontal asymptote. If f(x)→±∞f(x)\to\pm\infty as x→ax\to a, x=ax = a is a vertical asymptote. If f(x)−(mx+c)→0f(x) - (mx + c)\to0, y=mx+cy = mx + c is an oblique asymptote. For rational functions, divide by the highest power of xx in the denominator; polynomial division gives oblique asymptotes.

Rational functions at ∞\infty: compare top and bottom degrees.

Vertical asymptotes: zeros of the denominator that don't cancel.

Oblique: polynomial division, e.g. x2+1x=x+1x\frac{x^2 + 1}x = x + \frac1x, asymptote y=xy = x.

Check 1.11Check: evaluating a limit

  1. Substitute. A number? Done. Non-zero over zero? Infinite limit (check signs on each side).
  2. 00\frac00: factorise/cancel, rationalise, or reduce to a standard limit.
  3. x→∞x\to\infty: divide by the dominant term; remember ln⁡x≪xk≪ex\ln x\ll x^k\ll e^x.
  4. Oscillating factors (sin⁡1x\sin\frac1x): squeeze.
  5. Check numerically with a value close to aa if unsure.

Substitute → 00\frac00 tricks → dominant terms → squeeze.

Sanity check: plug in x=a+0.001x = a + 0.001.