Skip to Content
Math416Complex Variables (Lecture 2)

Math416 Lecture 2

Review?

z1=r1(cosθ1+isinθ1)=r1cis(θ1)z_1=r_1(\cos\theta_1+i\sin\theta_1)=r_1\text{cis}(\theta_1) z2=r2(cosθ2+isinθ2)=r2cis(θ2)z_2=r_2(\cos\theta_2+i\sin\theta_2)=r_2\text{cis}(\theta_2) z1z2=r1r2cis(θ1+θ2)z_1z_2=r_1r_2\text{cis}(\theta_1+\theta_2) nZ,zn=rncis(nθ)\forall n\in \mathbb{Z}, z^n=r^n\text{cis}(n\theta)

De Moivre’s Formula

nZ,zn=rncis(nθ)\forall n\in \mathbb{Z}, z^n=r^n\text{cis}(n\theta)

New Fancy stuff

Claim:

nZ,z1n=rncis(1nθ)\forall n\in \mathbb{Z}, z^{\frac{1}{n}}=\sqrt[n]{r}\text{cis}\left(\frac{1}{n}\theta\right)

Proof:

Take an nnth power, De Moivre’s formula holds \forall rational kQk\in \mathbb{Q}.

Example:

we calculate 1131^{\frac{1}{3}}

1=cis(2kπ)1=\text{cis}\left(2k\pi\right) 113=cis(2kπ3)1^{\frac{1}{3}}=\text{cis}\left(\frac{2k\pi}{3}\right)

When k=0k=0, we get 11

When k=1k=1, we get cis(2π3)=12+i32\text{cis}\left(\frac{2\pi}{3}\right)=-\frac{1}{2}+i\frac{\sqrt{3}}{2}

When k=2k=2, we get cis(4π3)=12i32\text{cis}\left(\frac{4\pi}{3}\right)=-\frac{1}{2}-i\frac{\sqrt{3}}{2}

Strange example

Let p(x)=a3x3+a2x2+a1x+a0p(x)=a_3x^3+a_2x^2+a_1x+a_0 be a polynomial with real coefficients.

Without loss of generality, Let a3=1a_3=1, x=yβx=y-\beta

We claim β=a23\beta=\frac{a_2}{3}

p(x)=(yβ)3+a2(yβ)2+a1(yβ)+a0=y3+(a23β)y2+(a13β22a2β)y+(a03β33a1βa2β2)\begin{aligned} p(x)&=(y-\beta)^3+a_2(y-\beta)^2+a_1(y-\beta)+a_0\\ &=y^3+\left(a_2-3\beta\right)y^2+\left(a_1-3\beta^2-2a_2\beta\right)y+\left(a_0-3\beta^3-3a_1\beta-a_2\beta^2\right)\\ \end{aligned}

It’s sufficient to know how to solve real cubic equations.

q(x)=x3+ax+bq(x)=x^3+ax+b

Let x=w+cwx=w+\frac{c}{w}

Solve

(w+cw)3+a(w+cw)+b=0w3+3wcw+3c2w2+aw+acw+b=0\begin{aligned} (w+\frac{c}{w})^3+a(w+\frac{c}{w})+b=0\\ w^3+3w\frac{c}{w}+3\frac{c^2}{w^2}+aw+\frac{ac}{w}+b=0\\ \end{aligned}

We choose cc such that 3c+a=03c+a=0, c=a3c=-\frac{a}{3}

w3+3c2w+b=0w6+bw3+c2=0\begin{aligned} w^3+3\frac{c^2}{w}+b=0\\ w^6+bw^3+c^2=0\\ \end{aligned}

Notice that w6+bw3+c2=0w^6+bw^3+c^2=0 is a quadratic equation in w3w^3.

w3=b±b24c32w^3=\frac{-b\pm\sqrt{b^2-4c^3}}{2}

So ww is a cube root of b±b24c32\frac{-b\pm\sqrt{b^2-4c^3}}{2}

x=w+cw=wa3wx=w+\frac{c}{w}=w-\frac{a}{3w}

Example:

p(x)=x33x+1=0p(x)=x^3-3x+1=0

a=3a=-3, b=1b=1, c=a3=33=1c=-\frac{a}{3}=-\frac{-3}{3}=1

w3=b±b24c32=1±142=1±3i2\begin{aligned} w^3&=\frac{-b\pm\sqrt{b^2-4c^3}}{2}\\ &=\frac{-1\pm\sqrt{1-4}}{2}\\ &=\frac{-1\pm\sqrt{3}i}{2}\\ \end{aligned}

To take cube root of ww,

w3=cis(2π3+2kπ)w^3=\text{cis}\left(\frac{2\pi}{3}+2k\pi\right)

So

Case 1:

w=cis(2π9+2kπ3)w=\text{cis}\left(\frac{2\pi}{9}+\frac{2k\pi}{3}\right)

It is sufficient to check k=0,1,2k=0,1,2 by nth root of unity.

When k=0k=0, w=cis(2π9)w=\text{cis}\left(\frac{2\pi}{9}\right)

When k=1k=1, w=cis(8π9)w=\text{cis}\left(\frac{8\pi}{9}\right)

When k=2k=2, w=cis(14π9)w=\text{cis}\left(\frac{14\pi}{9}\right)

Case 2:

w=cis(4π9+2kπ3)w=\text{cis}\left(\frac{4\pi}{9}+\frac{2k\pi}{3}\right)

When k=0k=0, w=cis(4π9)w=\text{cis}\left(\frac{4\pi}{9}\right)

When k=1k=1, w=cis(10π9)w=\text{cis}\left(\frac{10\pi}{9}\right)

When k=2k=2, w=cis(16π9)w=\text{cis}\left(\frac{16\pi}{9}\right)

So the final roots are:

w+cw=w+1ww+\frac{c}{w}=w+\frac{1}{w} cis(θ)+1cis(θ)=cis(θ)+cis(θ)=2cos(θ)\text{cis}(\theta)+\frac{1}{\text{cis}(\theta)}=\text{cis}(\theta)+\text{cis}(-\theta)=2\cos(\theta)

So the final roots are:

2cos(2π9),2cos(8π9),2cos(14π9),2cos(4π9),2cos(10π9),2cos(16π9)2\cos\left(\frac{2\pi}{9}\right), 2\cos\left(\frac{8\pi}{9}\right), 2\cos\left(\frac{14\pi}{9}\right), 2\cos\left(\frac{4\pi}{9}\right), 2\cos\left(\frac{10\pi}{9}\right), 2\cos\left(\frac{16\pi}{9}\right)

Remember cos(2πθ)=cos(θ)\cos(2\pi-\theta)=\cos(\theta)

So the final roots are:

2cos(2π9),2cos(8π9),2cos(14π9)2\cos\left(\frac{2\pi}{9}\right), 2\cos\left(\frac{8\pi}{9}\right), 2\cos\left(\frac{14\pi}{9}\right)

Compact

A set KRnK\in \mathbb{R}^n is compact if and only if it is closed and bounded. Compact Theorem in Math 4111 

If {xn}K\{x_n\}\in K, then there must be some point ww such that every disk D(w,ϵ)D(w,\epsilon) contains infinitely many points of KK. Infinite Point Theorem about Compact Set in Math 4111 

Unfortunately, C\mathbb{C} is not compact.

Riemann Sphere and Complex Projective Space

Let CR2R3\mathbb{C}\sim \mathbb{R}^2\subset \mathbb{R}^3

We put a unit sphere on the origin, and project the point on sphere to R2\mathbb{R}^2 by drawing a line through the north pole and the point on the sphere.

So all the point on the north pole is mapped to outside of the unit circle in R2\mathbb{R}^2.

all the point on the south pole is mapped to inside of the unit circle in R2\mathbb{R}^2.

The line through (0,0,1)(0,0,1) and (ξ,η,z)(\xi,\eta,z) intersects the unit sphere at (x,y,0)(x,y,0)

Line (tx,ty,1t)(tx,ty,1-t) intersects z2z^2 at t2x2+t2y2+(1t)2=1t^2x^2+t^2y^2+(1-t)^2=1

So t=21+x2+y2t=\frac{2}{1+x^2+y^2}

z=x+iy11+z2(2Re(z),2Im(z),z21)z=x+iy\mapsto \frac{1}{1+|z|^2}(2Re(z),2Im(z),|z|^2-1) (ξ,η,z)ξ+iη1z(\xi,\eta,z)\mapsto \frac{\xi+i\eta}{1-z}

This is a homeomorphism. C{}S2\mathbb{C}\setminus\{\infty\}\simeq S^2

Derivative of a function

Suppose Ω\Omega is an open subset of C\mathbb{C}.

A function f:ΩCf:\Omega\to \mathbb{C}‘s derivative is defined as

f(z0)=limzz0f(z)f(z0)zz0f'(z_0)=\lim_{z\to z_0}\frac{f(z)-f(z_0)}{z-z_0}

f=u+ivf=u+iv, u,v:ΩRu,v:\Omega\to \mathbb{R}

How are ff' and derivatives of uu and vv related?

  1. Differentiation and complex linearity applies to ff

Chain rule applies

ddz(f(g(z)))=f(g(z))g(z)\frac{d}{dz}(f(g(z)))=f'(g(z))g'(z)

Polynomials

ddzzn=nzn1\frac{d}{dz}z^n=nz^{n-1}
Last updated on