MATH 409 Lecture 9

From Notes
Jump to navigation Jump to search

« previous | Tuesday, September 24, 2013 | next »

Lecture Slides

Review

Limit Supremum and Infimum

Let {xn} be a bounded sequence of real numbers. For any n∈ℕ, let En denote the set of numbers of the form xk, where k≥n. (En starts from the nth term in the sequence; note that En+1 is nested within En, and En has at most one more element than En+1; Another xn could occur later in the sequence).

The set En is bounded, hence sup⁡En and inf⁡En exist. Observe that the sequence {sup⁡En} is decreasing, the sequence {inf⁡En} is increasing (since E1, E2, … are nested sets), and both are bounded. Therefore both sequences are convergent.

The limit of {sup⁡En} is called the limit supremum of the sequence {xn} and is denoted lim supn→∞xn.

The limit of {inf⁡En} is called the limit infimum of the sequence {xn} and is denoted lim infn→∞xn.

If the sequence is not bounded above, then lim supn→∞=+∞. If the sequence is not bounded below, then lim infn→∞=−∞.

Properties

lim infn→∞xn≤lim supn→∞xn

Note we have inf⁡En≤sup⁡En for all subsequences En, therefore lim infn→∞≤lim supn→∞En if they exist.

lim infn→∞xn and lim supn→∞xn are limit points of the sequence {xn}.

All limit points of {xn} are contained in the interval [lim infn→∞xn,lim supn→∞xn]

The sequence {xn} converges to a limit a if and only if lim infn→∞xn=lim supn→∞xn=a.


Limits of Functions

Let I⊂ℝ be an open interval and a∈I. Suppose f:E→ℝ is a function defined on a set E⊃I∖{a}.

We say that the function f converges to a limit L∈ℝ at the point a if for every ϵ>0 there exists δ=δ(ϵ)>0 such that

0<|x−a|<δ⟹|f(x)−L|<ϵ

Notation: L=limx→af(x) or f(x)→L as x→a.

The set (a−δ,a)∪(a,a+δ) is called the punctured δ-neighborhood of a. Convergence to L means that, given ϵ>0, the image of this set under the map f is contained in the ε-neighborhood (L−ϵ,L+ϵ) of L provided that δ is small enough.


Vs. Limits of Sequences

Theorem. Let I be an open interval containing a point a∈ℝ and f be a function defined on I∖{a}. Then f(x)→L as x→a if and only if for any sequence {xn} of elements of I∖{a},

limn→∞xn=a⟹limn→∞f(xn)=L

Proof. Suppose that f(x)→L as x→a. Consider an arbitrary sequence {xn} of elements of the set I∖{a} converging to a. For any ϵ>0, there exists δ>0 such that 0<|x−a|<δ implies |f(x)−L|<ϵ for all x∈ℝ. Further, there exists N∈ℕ such that |xn−a|<δ for all n≥N. Then if |f(xn)−L|<ϵ for all n≥N [1]. Then |f(xn)−L|<ϵ for all n≥N. We conclude that f(xn)→L as n→∞.

Conversely, suppose that f(x)↛L as x→a. Then there exists ϵ>0 such that for any δ>0, the image of the punctured neighborhood (a−δ,a)∪(a,a+δ) of the point a under the map f is not contained in (L−ϵ,L+ϵ). In particular, for any n∈ℕ there exists a point xn∈(a−1n)∪(a,a+1n) such that xn∈I and |f(xn)−L|≥ϵ. We have that hte sequence {xn} converges to a and xn∈I∖{a}. However, f(xn)↛L as n→∞.

quod erat demonstrandum

Using this sequential characterization of limits, we can derive limit theorems for convergence of functions from analogous theorems dealing with convergence of sequences.


Limit Theorems

Squeeze Theorem

If limx→af(x)=limx→ag(x)=L and f(x)≤h(x)≤g(x) for all x in a punctured neighborhood of the point a, then limx→ah(x)=L.

Comparison Theorem

If limx→af(x)=L and limx→ag(x)=M

Arithmetic Theorems

If limx→af(x)=L and limx→ag(x)=M, then

limx→a(f+g)(x)=L+Mlimx→a(f−g)(x)=L−Mlimx→a(fg)(x)=LM

If, additionally, M≠0, then

limx→a(fg)(x)=LM


Divergence to Infinity

Let I⊂ℝ be an open interval and a∈I. Suppose f:E→ℝ is a function defined on the set E⊃I∖{a}.

We say that the function f diverges to +∞ at the point a if for every C∈ℝ there exists δ=δ(C)>0 such that

0<|x−a|<δ⟹f(x)>C

Notation: limx→af(x)=+∞ or f(x)→+∞ as x→a.

Similarly, divergence to −∞ at the point a


One-Sided Limits

Let f:E→ℝ be a function defined on a set E⊂ℝ.

We say that f converges to a right-hand limit L∈ℝ at a point a∈ℝ if the domain E contains an interval (a,b) and for every ϵ>0 there exists δ=δ(ϵ)>0 such that

a<x<a+δ⟹|f(x)−L|<ϵ

Notation: L=limx→a+f(x).

Similarly, we define the left-hand-limit limx→a−f(x).


Theorem. f(x)→L as x→a if and only if limx→a+f(x)=limx→a−f(x)=L.


Limits at Infinity

Let f:E→ℝ be a function defined on a set E⊂ℝ.

We say that f converges to a limit L∈ℝ as x→+∞ if the domain E contains an interval (a,+∞) and for every ϵ>0, there exists C=C(ϵ)∈ℝ such that

x>C⟹|f(x)−L|<ϵ

Notation: L=limx→+∞f(x) or f(x)→L as x→+∞

Similarly we define the limit limx→−∞f(x).


Examples

Constant function. f(x)=c for all x∈ℝ and some c∈ℝ.

limx→af(x)=c for all a∈ℝ. Also, limx→±∞f(x)=c.

Identity function. f(x)=x for x∈ℝ.

limx→af(x)=a for all a∈ℝ. Also limx→±∞f(x)=±∞.

Heaviside function. Prof. defines it as f(x)={1x>00x≤0

limx→0+=0, and limx→0−=1.

Harmonic function. f:ℝ∖{0}→ℝ, where f(x)=1x

  • limx→af(x)=1a for all a≠0.
  • limx→0+=+∞
  • limx→0−=−∞
  • limx→±∞f(x)=0

Sine. f:ℝ∖{0}→ℝ, where f(x)=sin⁡1x

limx→0+f(x) does not exist since f((0,δ))=[−1,1] for any δ>0.

f:ℝ∖{0}→ℝ, where f(x)=xsin⁡1x.

limx→0=0 by squeeze theorem between f(x)=x and f(x)=−x.

Dirichlet function. f(x)={1x∈ℚ0x∈ℝ∖ℚ

limx→af(x) does not exist since f((c,d))={0,1} for any interval (c,d). In other words, both rational and irrational points are dense in ℝ.

Riemann function. f(x)={1qx=pq a reduced fraction0x∈ℝ∖ℚ.

limx→af(x)=0 for all a∈ℝ indeed, for any n∈ℕ and a bounded interval (c,d), there are only finitely many points x∈(c,d) such that f(x)≥1n.

On the other hand, limx→+∞f(x) and limx→−∞f(x) do not exist.

Footnotes

  1. ↑ N depends on δ, which in turn depends on ϵ.