4.1 Metric Spaces in Complex

Assuming we are in R and C

Let S be R or C and d(x,y) be the usual distance in those spaces

{R:d(x,y)=|xy|C:d(x,y)=d(x,y)=((x)(y))2+((x)(y))2

Properties:

d(x,y)0d(x,y)=0x=yd(x,y)=d(y,x)d(x,z)d(x,y)+d(y,z)B(x,δ)={yS:d(x,y)<δ}

Neighbourhood

NS is a neighbourhood of xS if it contains a ball B(x,δ)

Open Set

A is open if zA, Bδ(z)A

Every ball Bδ(x) is open:

For x,yBδ(x), let δ=δd(x,y),

Take zBδ(y),

d(z,y)<δd(z,y)<δd(x,y)d(x,z)<δzBδ(x)

Intersection of Finite Open Sets

Is open;

Let O1,,On be open
Take zi=1nOi, i,zOi
There exists δi such that Bδi(z)Oi
just take min{δ1,,δn}

Union of Any Open Sets

Is open
Bδi(z)OiOi

Closed Set

The complement of an open set

Union of Finite Amount of Closed Sets

Is closed

Take x(nCi)c, therefore for any δ, Bδ(x)nCi

xCii    xi=1nCic using De Morgans rule.
Each Ci is closed Cicis open.

Thus, for each i there exists δi>0 such that

Bδi(x)Cic

Set δ=min{δ1,,δn}>0

Intersection of any Closed Set

Is closed
Take x(Ci)c, therefore for any δ,

Bδ(x)Ci

xnCic where Cic is open.

xCicBδ(x)Cic(Ci)c so its open

Terminology

Interior

The interior of X (int(X) or X) is the union of all open sets inside X

X is open iff X=X

Closure

The closure of X (cl(X) or X¯ ) is the intersection of all closed sets that contain X

X is closed iff X¯=X

Boundary

Is the closure of X minus the interior of X. A bound is a boundary (Bd(X) or X) iff all of its neighbourhoods intersect both X and Xc.

Exterior

The exterior of X is the interior of Xc=(Xc)

Isolated point

xX is an isolated point of X if x has a neighbourhood whose intersection with X is only x

Accumulation Point / limit point

xX is an accumulation point if any neighbourhood x contains infinitely many points from X.

Take A={accumulation points of X}

Take yAc, Bδ(y) where Bδ(y)X={a1,a2,,an} is finite

let δ=12min{d(a1,y),,d(an,y)}

Bδ(y)Ac

So its complement is open. Therefore A is closed.

Connectedness

A subset of K of C or (R) is called connected if:

KU1U2

where U1 and U2 are open and U1U2= and KU1 or KU2.
(K cannot be subset of two disjoint open sets)

Families of Connected Sets

If (Kα)αI is a family of connected sets αIKα then αIKα is connected.

Assume A and B, are open and partition αIKα
Take Ki assume KiA. Take Kj, ji.

Assume zKiKj, zA

KjA and KjB are disjoint but Kj is connected. Since zA KjA

Connected Set Decomposition

Any KC can be uniquely decomposed by

K=Uα

Where each Uα is connected.

Connected Intermediate Value Theorem

If K is connected and f is continuous on K then f(K) is connected

Proof:

Lemma:
If f:XY is cts, if AY is open, f1(A) is also open

f1(A)={xY|f(x)A}

Take bf1(A)f(b)A
ε>0 such that Bε(f(b))A since A is open

If f is continuous it means δ>0 s.t d(b,x)<δ for any d(f(b),f(x))<ε, for any ε. Therefore f(x)Bε(f(b)) and Bδ(b)f1(A)

IVT
Assume f(K)=AB which are disjoint non-empty open sets and K is connected

Kf1(A)f1(B)f1(A)f1(B)=

Which contradicts K being connected

Path-Connectedness

Arc/Path

A continuous γ:[a,b]C is called an arc or a path from γ(a) to γ(b)

Path-Connect

KC is path-connected if z1,z2K then there exists γ:[a,b]K so that γ(t) is in K for any t[a,b]

Theorems


B(z,δ) and B(z,δ) are path connected

Let UC be open, then U is connected iff it is path connected


A path-connected subset of C is always connected, but not all connected subsets of C are path connected


If UC is open then each of its connected components is open

Regions

A region is an open connected subset of C
Analytic functions on a region D is represented by A(D)

Constant Functions on Regions

If fA(D) and f(z)=0 for any zD then f is constant in D

Analytic Definition

f is analytic on KC if there exists an open set U, KU,

F:UC,FA(U), so that F|K=f

A complex-valued function f(z) defined on a region Ω is to be analytic in Ω if it has a derivative at each point of Ω


Compactness

If KαIUα, all Uα are open, then there is a finite number of Uα, such that KU1U2Un (finite open-cover)

All are equivalent:

  1. K is closed and bounded subset of C
  2. If {zn}K then there exists a subsequence znk st limkznk=zK
  3. If f:KR is continuous then f(K) is closed and bounded. The values maxf(z) and minf(z) are finite and they are achieved at some point on K
  4. If f:KC is continuous then f(K) is bounded and closed
  5. If f:KR(or C) is continuous, then f is uniformly continuous. ε>0,δ>0 such that if |z1z2|<δ|f(z1)f(z2)|<ε

If for every n, Kn is compact and KnKn+1 then 1Kn