Skip to Content
Math416Complex Variables (Lecture 5)

Math416 Lecture 5

Review

Let ff be a complex function. that maps R2\mathbb{R}^2 to R2\mathbb{R}^2. f(x+iy)=u(x,y)+iv(x,y)f(x+iy)=u(x,y)+iv(x,y).

Df(x+iy)=(uxuyvxvy)=(αβσδ)Df(x+iy)=\begin{pmatrix} \frac{\partial u}{\partial x} & \frac{\partial u}{\partial y}\\ \frac{\partial v}{\partial x} & \frac{\partial v}{\partial y} \end{pmatrix}=\begin{pmatrix} \alpha & \beta\\ \sigma & \delta \end{pmatrix}

So,

fz=12(ux+vy)i12(vx+uy)=12(α+δ)i12(βσ)\begin{aligned} \frac{\partial f}{\partial z}&=\frac{1}{2}\left(u_x+v_y\right)-i\frac{1}{2}\left(v_x+u_y\right)\\ &=\frac{1}{2}\left(\alpha+\delta\right)-i\frac{1}{2}\left(\beta-\sigma\right)\\ \end{aligned} fz=12(ux+vy)+i12(vx+uy)=12(αδ)+i12(β+σ)\begin{aligned} \frac{\partial f}{\partial \overline{z}}&=\frac{1}{2}\left(u_x+v_y\right)+i\frac{1}{2}\left(v_x+u_y\right)\\ &=\frac{1}{2}\left(\alpha-\delta\right)+i\frac{1}{2}\left(\beta+\sigma\right)\\ \end{aligned}

When ff is conformal,

Df(x+iy)=(αββα)Df(x+iy)=\begin{pmatrix} \alpha & \beta\\ -\beta & \alpha \end{pmatrix}

So,

fz=12(α+α)+i12(β+β)=a\frac{\partial f}{\partial z}=\frac{1}{2}(\alpha+\alpha)+i\frac{1}{2}(\beta+\beta)=a fz=12(αα)+i12(ββ)=0\frac{\partial f}{\partial \overline{z}}=\frac{1}{2}(\alpha-\alpha)+i\frac{1}{2}(\beta-\beta)=0

Less pain to represent a complex function using four real numbers.

Chapter 3: Linear fractional Transformations

Let a,b,c,da,b,c,d be complex numbers. such that adbc0ad-bc\neq 0.

The linear fractional transformation is defined as

ϕ(z)=az+bcz+d\phi(z)=\frac{az+b}{cz+d}

If we let ψ(z)=ezfgz+h\psi(z)=\frac{ez-f}{-gz+h} also be a linear fractional transformation, then ϕψ\phi\circ\psi is also a linear fractional transformation.

New coefficients can be solved by

(abcd)(efgh)=(klmn)\begin{pmatrix} a & b\\ c & d \end{pmatrix} \begin{pmatrix} e & f\\ g & h \end{pmatrix} = \begin{pmatrix} k&l\\ m&n \end{pmatrix}

So ϕψ(z)=kz+lmz+n\phi\circ\psi(z)=\frac{kz+l}{mz+n}

Complex projective space

RP1\mathbb{R}P^1 is the set of lines through the origin in R2\mathbb{R}^2.

We defined (a,b)(c,d),(a,b),(c,d)R2{(0,0)}(a,b)\sim(c,d),(a,b),(c,d)\in\mathbb{R}^2\setminus\{(0,0)\} if t0,tR{0}\exists t\neq 0,t\in\mathbb{R}\setminus\{0\} such that (a,b)=t(c,d)(a,b)=t(c,d).

RP1=S1{±x}S1R\mathbb{P}^1=S^1\setminus\{\pm x\}\cong S^1

Equivalently,

CP1\mathbb{C}P^1 is the set of lines through the origin in C\mathbb{C}.

We defined (a,b)(c,d),(a,b),(c,d)C{(0,0)}(a,b)\sim(c,d),(a,b),(c,d)\in\mathbb{C}\setminus\{(0,0)\} if t0,tC{0}\exists t\neq 0,t\in\mathbb{C}\setminus\{0\} such that (a,b)=(tc,td)(a,b)=(tc,td).

So, zC{0}\forall z\in\mathbb{C}\setminus\{0\}:

If a0a\neq 0, then (a,b)(1,ba)(a,b)\sim(1,\frac{b}{a}).

If a=0a=0, then (0,b)(0,b)(0,b)\sim(0,-b).

So, CP1\mathbb{C}P^1 is the set of lines through the origin in C\mathbb{C}.

Linear fractional transformations

Let M=(abcd)M=\begin{pmatrix} a & b\\ c & d \end{pmatrix} be a 2×22\times 2 matrix with complex entries. That maps C2\mathbb{C}^2 to C2\mathbb{C}^2.

Suppose MM is non-singular. Then adbc0ad-bc\neq 0.

If M(z1z2)=(ω1ω2)M\begin{pmatrix} z_1\\ z_2 \end{pmatrix}=\begin{pmatrix} \omega_1\\ \omega_2 \end{pmatrix}, then M(tz1tz2)=(tω1tω2)M\begin{pmatrix} tz_1\\ tz_2 \end{pmatrix}=\begin{pmatrix} t\omega_1\\ t\omega_2 \end{pmatrix}.

So, MM induces a map ϕM:CP1CP1\phi_M:\mathbb{C}P^1\to\mathbb{C}P^1 defined by M(z1)=(az+bcz+d1)M\begin{pmatrix} z\\ 1 \end{pmatrix}=\begin{pmatrix} \frac{az+b}{cz+d}\\ 1 \end{pmatrix}.

ϕM(z)=az+bcz+d\phi_M(z)=\frac{az+b}{cz+d}.

If we let M2=(efgh)M_2=\begin{pmatrix} e &f\\ g &h \end{pmatrix}, where adbc0ad-bc\neq 0 and ehfg0eh-fg\neq 0, then ϕM2(z)=ez+fgz+h\phi_{M_2}(z)=\frac{ez+f}{gz+h}.

So, M2M1=(abcd)(efgh)=(z1)M_2M_1=\begin{pmatrix} a&b\\ c&d \end{pmatrix}\begin{pmatrix} e&f\\ g&h \end{pmatrix}=\begin{pmatrix} z\\ 1 \end{pmatrix}.

This also gives (kz+lmz+n)(kz+lmz+n1)\begin{pmatrix} kz+l\\ mz+n \end{pmatrix}\sim\begin{pmatrix} \frac{kz+l}{mz+n}\\ 1 \end{pmatrix}.

So, if abcd0ab-cd\neq 0, then M1\exists M^{-1} such that M2M1=IM_2M_1=I.

So non-constant linear fractional transformations form a group under composition.

When do two matrices gives the t0t_0 same linear fractional transformation?

M21M1=αIM_2^{-1}M_1=\alpha I

We defined GL(2,C)GL(2,\mathbb{C}) to be the group of general linear transformations of order 2 over C\mathbb{C}.

This is equivalent to the group of invertible 2×22\times 2 matrices over C\mathbb{C} under matrix multiplication.

Let FF be the function that maps MM to ϕM\phi_M.

F:GL(2,C)Homeo(CP1)F:GL(2,\mathbb{C})\to\text{Homeo}(\mathbb{C}P^1)

So the kernel of FF is the set of matrices that represent the identity transformation. kerF={αI},αC{0}\ker F=\left\{\alpha I\right\},\alpha\in\mathbb{C}\setminus\{0\}.

Corollary of conformality

If ϕ\phi is a non-constant linear fractional transformation, then ϕ\phi is conformal.

Proof

Know that ϕ0ϕ(z)=z\phi_0\circ\phi(z)=z,

Then ϕ(z)=ϕ01ϕϕ0(z)\phi(z)=\phi_0^{-1}\circ\phi\circ\phi_0(z).

So ϕ(z)=az+bcz+d\phi(z)=\frac{az+b}{cz+d}.

ϕ:C{}C{}\phi:\mathbb{C}\cup\{\infty\}\to\mathbb{C}\cup\{\infty\} which gives ϕ()=ac\phi(\infty)=\frac{a}{c} and ϕ(dc)=\phi(-\frac{d}{c})=\infty.

So, ϕ\phi is conformal.

Proposition 3.4 of Fixed points

Any non-constant linear fractional transformation except the identity transformation has 1 or 2 fixed points.

Proof

Let ϕ(z)=az+bcz+d\phi(z)=\frac{az+b}{cz+d}.

Case 1: c=0c=0

Then \infty is a fixed point.

Case 2: c0c\neq 0

Then ϕ(z)=az+bcz+d\phi(z)=\frac{az+b}{cz+d}.

The solution of ϕ(z)=z\phi(z)=z is cz2+(da)zb=0cz^2+(d-a)z-b=0.

Such solutions are z=(da)±(da)2+4bc2cz=\frac{-(d-a)\pm\sqrt{(d-a)^2+4bc}}{2c}.

So, ϕ\phi has 1 or 2 fixed points.

Proposition 3.5 of triple transitivity

If z1,z2,z3CP1z_1,z_2,z_3\in\mathbb{C}P^1 are distinct, then there exists a non-constant linear fractional transformation ϕ\phi such that ϕ(z1)=z2\phi(z_1)=z_2 and ϕ(z3)=\phi(z_3)=\infty.

Proof as homework.

Theorem 3.8 Preservation of clircles

We defined clircle to be a circle or a line.

If ϕ\phi is a non-constant linear fractional transformation, then ϕ\phi maps clircles to clircles.

Proof continue on next lecture.

Last updated on