Mobius composition and matrices
Composition of Mobius transformations has a particularly pleasant form, and connects naturally with the important topic of matrix multiplication from linear algebra.
In this step, you will
learn how to multiply matrices
see the connection between matrix multiplication and composition of Mobius functions.
Mobius transformations and matrices
We have seen that a Mobius transformation, or fractional linear transformation, is a function of the form
where the numbers , , and can be arbitrary (subject to the common sense requirement that the fraction does not end up with a zero in the denominator!) So the function can be encoded by the matrix
This kind of mathematical object is very important in linear algebra.
In this somewhat advanced step, you’ll see how the composition of fractional linear transformations connects directly with matrix multiplication.
Q1 (E): What is the corresponding matrix for the Mobius transformation
How to multiply matrices
First of all here is how you multiply a row vector by a column vector (the three line is used to signify a definition):
When we multiply matrices, that is four operations of this kind where we multiply the rows of the first matrix by the columns of the second matrix, and arrange the result in a matrix as follows:
So for example:
Q2 (M): For the matrices
find the products and Note that matrix multiplication is not in general commutative!
Composition of Mobius transformations
Suppose we have two Mobius transformations
Then the composition of these two transformations is
This is exactly the Mobius transformation corresponding the matrix product
We have demonstrated something interesting here. Composition of Mobius transformations corresponds exactly to multiplication of matrices!
Q3(M): Find the composition of the Mobius transformations
A1. The matrix for the Mobius transformation
A2. One product is
The other product is
A3. You could find this directly, but we’ll use matrix multiplication to solve this question. The matrix for the Mobius transformation
Similarly the matrix for the Mobius transformation
So, the matrix corresponding to the Mobius transformation
Therefore, the composition is
© UNSW Australia 2015