14.12.2012 Views

Complex Analysis - Maths KU

Complex Analysis - Maths KU

Complex Analysis - Maths KU

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

400 Chapter 7 Conformal Mappings<br />

Definition 7.2 Linear Fractional Transformation<br />

If a, b, c, and d are complex constants with ad − bc �= 0, then the complex<br />

function defined by:<br />

T (z) =<br />

az + b<br />

cz + d<br />

is called a linear fractional transformation.<br />

Linear fractional transformations are also called Möbius transformations<br />

or bilinear transformations. If c = 0, then the transformation T<br />

given by(1) is a linear mapping, and so a linear mapping is a special case of<br />

a linear fractional transformation. If c �= 0, then we can write<br />

az + b bc − ad 1 a<br />

T (z) = = + . (2)<br />

cz + d c cz + d c<br />

bc − ad<br />

Setting A = and B =<br />

c<br />

a<br />

, we see that the linear transformation T in<br />

c<br />

(2) can be written as the composition T (z) =f ◦g◦h(z), where f(z) =Az +B<br />

and h(z) =cz+d are linear functions and g(z) =1/z is the reciprocal function.<br />

The domain of a linear fractional transformation T given by(1) is the set<br />

of all complex z such that z �= −d/c. Furthermore, since<br />

T ′ ad − bc<br />

(z) =<br />

(cz + d) 2<br />

it follows from Theorem 7.1 and the requirement that ad − bc �= 0 that linear<br />

fractional transformations are conformal on their domains. The requirement<br />

that ad−bc �= 0 also ensures that the T is a one-to-one function on its domain.<br />

See Problem 27 in Exercises 7.2.<br />

Observe that if c �= 0, then (1) can be written as<br />

T (z) =<br />

az + b (a/c)(z + b/a) φ(z)<br />

= =<br />

cz + d z + d/c z − (−d/c) ,<br />

where φ(z) =(a/c)(z + b/a). Because ad−bc �= 0, we have that φ (−d/c) �= 0,<br />

and so from Theorem 6.12 of Section 6.4 it follows that the point z = −d/c is<br />

a simple pole of T .<br />

When c �= 0, that is, when T is not a linear function, it is often helpful to<br />

view T as a mapping of the extended complex plane. Since T is defined for<br />

all points in the extended plane except the pole z = −d/c and the ideal point<br />

∞, we need onlyextend the definition of T to include these points. We make<br />

this definition byconsidering the limit of T as z tends to the pole and as z<br />

tends to the ideal point. Because<br />

lim<br />

z→−d/c<br />

cz + d<br />

az + b =<br />

it follows from (25) of Section 2.6 that<br />

0<br />

a (−d/c)+b =<br />

lim<br />

z→−d/c<br />

az + b<br />

= ∞.<br />

cz + d<br />

0<br />

−ad + bc =0,<br />

(1)

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!