Limit point

a is limit point of set A if:

∀δ>0,∃x∈A such that 0<|x−a|<δ.\forall \, \delta > 0, \, \exists \, x \in A \text{ such that } 0 < |x - a| < \delta.

i.e. there are some x∈A around the limit point a.

Limit of a function

limx→af(x)=Lif∀ε>0,∃δ>0 such that ∀x∈A,0<|x−a|<δ⟹|f(x)−L|<ε.\begin{aligned} \lim_{x \to a} f(x) = L \\ \quad \text{if} \quad \forall \, \varepsilon > 0, \, \exists \, \delta > 0 \text{ such that } \\ \forall x \in A, \; 0 < |x - a| < \delta \implies |f(x) - L| < \varepsilon. \end{aligned}

To prove a limit using epsilon-delta

Technique: Reverse engineer from |f(x) - L| < ϵ.

Limit at infinity

limx→∞f(x)=L if ∀ε>0,∃N>0 such that ∀x>N,|f(x)−L|<ε.\lim_{x \to \infty} f(x) = L \text{ if } \forall \, \varepsilon > 0, \, \exists \, N > 0 \text{ such that } \forall x > N, \; |f(x) - L| < \varepsilon.

limn→∞an=L if ∀ε>0,∃N>0 such that ∀n>N,|an−L|<ε.\lim_{n \to \infty} a_n = L \text{ if } \forall \, \varepsilon > 0, \, \exists \, N > 0 \text{ such that } \forall n > N, \; |a_n - L| < \varepsilon.

Showing limit d.n.e

Limit tends to infinity

limx→af(x)=∞ if ∀M>0,∃δ>0 such that ∀x∈A,0<|x−a|<δ⟹f(x)>M.\lim_{x \to a} f(x) = \infty \text{ if } \forall \, M > 0, \, \exists \, \delta > 0 \text{ such that }\forall x \in A, \; 0 < |x - a| < \delta \implies f(x) > M.

limx→∞f(x)=∞ if ∀M>0,∃N>0 such that x>N⟹f(x)>M.\lim_{x \to \infty} f(x) = \infty \text{ if } \forall \, M > 0, \, \exists \, N > 0 \text{ such that } x > N \implies f(x) > M.

Construction of two sequences

If there are two sequences {an} and {bn} such that \text{If there are two sequences } \{a_n\} \text{ and } \{b_n\} \text{ such that }

an≠a,bn≠a, and both an→a as n→∞, but limn→∞f(an)≠limn→∞f(bn),a_n \ne a, \; b_n \ne a, \text{ and both } a_n \to a \text{ as } n \to \infty, \text{ but }\lim_{n \to \infty} f(a_n) \ne \lim_{n \to \infty} f(b_n),

then limx→af(x) does not exist.\text{then } \lim_{x \to a} f(x) \text{ does not exist.}

Take two sequences: an=1πn,bn=1π2+2πn.\text{Take two sequences: } a_n = \frac{1}{\pi n}, b_n = \frac{1}{\frac{\pi}{2} + 2\pi n}.

limn→∞f(an)=0,limn→∞f(bn)=1.\lim_{n \to \infty} f(a_n) = 0, \quad \lim_{n \to \infty} f(b_n) = 1.

Left-hand vs right-hand limit

limx→af(x)=L⇔limx→a−f(x)=limx→a+f(x)=L.\lim_{x \to a} f(x) = L \;\Longleftrightarrow\;\lim_{x \to a^-} f(x) = \lim_{x \to a^+} f(x) = L.

Asymptotes

x=a is a vertical asymptote if limx→a−f(x)=∞ or limx→a+f(x)=∞.x = a \text{ is a vertical asymptote if } \lim_{x \to a^-} f(x) = \infty \text{ or } \lim_{x \to a^+} f(x) = \infty.

y=L is a horizontal asymptote if limx→+∞f(x)=L or limx→−∞f(x)=L.y = L \text{ is a horizontal asymptote if } \lim_{x \to +\infty} f(x) = L \text{ or } \lim_{x \to -\infty} f(x) = L.

y=ax+b is an oblique asymptote if limx→∞(f(x)−(ax+b))=0.y = ax + b \text{ is an oblique asymptote if } \lim_{x \to \infty} \bigl(f(x) - (ax + b)\bigr) = 0.

Methods to find limits

Limit laws

For f:A1→ℝ,g:A2→ℝ, and a∈A1∩A2,if limx→af(x)=l and limx→ag(x)=m:\begin{aligned} \text{For } f: A_1 \to \mathbb{R}, \; g: A_2 \to \mathbb{R}, \text{ and } a \in A_1 \cap A_2,\\ \quad \text{if } \lim_{x \to a} f(x) = l \text{ and } \lim_{x \to a} g(x) = m: \end{aligned}

  1. lim⁡x→a(αf(x)+βg(x))=αl+βm.\lim_{x \to a} \bigl( \alpha f(x) + \beta g(x) \bigr) = \alpha l + \beta m.
  2. lim⁡x→a(f(x)g(x))=lm.\lim_{x \to a} \bigl( f(x) g(x) \bigr) = l m.
  3. lim⁡x→a(f(x)g(x))=lmif m≠0.\lim_{x \to a} \left( \frac{f(x)}{g(x)} \right) = \frac{l}{m}\quad \text{if } m \ne 0.
  4. lim⁡x→a(f(x))1/n=l1/n.\lim_{x \to a} \bigl( f(x) \bigr)^{1/n} = l^{1/n}.

Limits on RHS should exist for limit law to be applied.

**

If limx→af(x) does not exist and limx→ag(x)=m,\text{If } \lim_{x \to a} f(x) \text{ does not exist and } \lim_{x \to a} g(x) = m,

  1. lim⁡x→a(αf(x)+βg(x)) d.n.e.\lim_{x \to a} \bigl( \alpha f(x) + \beta g(x) \bigr) \text{ d.n.e.}
  2. If m≠0, then lim⁡x→a(f(x)g(x)) d.n.e.\text{If } m \ne 0, \text{ then } \lim_{x \to a} \bigl( f(x) g(x) \bigr) \text{ d.n.e.}

If m=0 or both limits don’t exist, we have no conclusion.

**

Continuous

Continuous f(x) at x=a: substitute x=a into equation.

Useful equations

  1. If lim⁡x→0f(x)=0, then lim⁡x→0(sin⁡(f(x))f(x))=1.\text{If } \lim_{x \to 0} f(x) = 0, \text{ then } \lim_{x \to 0} \left( \frac{\sin(f(x))}{f(x)} \right) = 1.

Derivation:

For the above unit circle:

Area of △ABE=12(BC)(AE)=12sin⁡α\text{Area of } \triangle ABE = \tfrac{1}{2} (BC)(AE) = \tfrac{1}{2} \sin \alpha

Area of sector ABE=12α\text{Area of sector } ABE = \tfrac{1}{2} \alpha

Area of △ADE=12(DE)(AE)=12tan⁡α\text{Area of } \triangle ADE = \tfrac{1}{2} (DE)(AE) = \tfrac{1}{2} \tan \alpha

Clearly, 12sin⁡α<12α<12tan⁡α\text{Clearly, } \tfrac{1}{2}\sin \alpha < \tfrac{1}{2}\alpha < \tfrac{1}{2}\tan \alpha

Divide by 12sin⁡α and take reciprocals:1>sin⁡αα>cos⁡α\text{Divide by } \tfrac{1}{2}\sin \alpha \text{ and take reciprocals:} \quad 1 > \frac{\sin \alpha}{\alpha} > \cos \alpha

limα→01=limα→0cos⁡α=1\lim_{\alpha \to 0} 1 = \lim_{\alpha \to 0} \cos \alpha = 1

So limα→0sin⁡αα=1\text{So } \lim_{\alpha \to 0} \frac{\sin \alpha}{\alpha} = 1

  1. Triangle inequality

|x+y|≤|x|+|y|.|x + y| \le |x| + |y|.

Reverse triangle inequality.

||x|−|y||≤|x−y|\Big| |x| - |y| \Big| \leq |x - y|

  1. For large x, factorial > exponential > polynomial > ln/lg.

For any p>0,ε>0,limx→+∞x!xx=0,limx→+∞eεxx!=0,limx→+∞xpeεx=0,andlimx→+∞(ln⁡x)pxε=0.\begin{aligned} \text{For any } p > 0, \, \varepsilon > 0, \quad \\ \lim_{x \to +\infty} \frac{x!}{x^x} = 0, \quad \\ \lim_{x \to +\infty} \frac{e^{\varepsilon x}}{x!} = 0, \quad \\ \lim_{x \to +\infty} \frac{x^p}{e^{\varepsilon x}} = 0, \quad \\ \text{and} \quad \\ \lim_{x \to +\infty} \frac{(\ln x)^p}{x^{\varepsilon}} = 0. \end{aligned}

Squeeze theorem

If g(x)≤f(x)≤h(x) for x∈I\{a}, and limx→ag(x)=limx→ah(x)=L, then limx→af(x)=L.\text{If } g(x) \le f(x) \le h(x) \text{ for } x \in I \setminus \{a\},\text{ and } \lim_{x \to a} g(x) = \lim_{x \to a} h(x) = L,\text{ then } \lim_{x \to a} f(x) = L.

limx→af(x)=L⇔limx→a|f(x)−L|=0.\lim_{x \to a} f(x) = L \;\Longleftrightarrow\;\lim_{x \to a} \bigl| f(x) - L \bigr| = 0.

If L = 0, we can use squeeze theorem on |f(x)|, and we instantly get a lower bound of 0.

E.g. lim⁡h→0hsin⁡1h\lim_{h \to 0} h \sin \frac{1}{h}

We can guess that L = 0, since h→0 and although sin(1/h) oscillates near 0, it is still bounded.

Thus 0≤|hsin1h|=|h|⋅|sin1h|≤|h|→0,as h→0,0 \le \left| h \sin \frac{1}{h} \right| = |h| \cdot \left| \sin \frac{1}{h} \right| \le |h| \to 0, \quad \text{as } h \to 0,

Other tricks