Why does Armstrong use the transpose in his linear transformation definition?

2 replies · 2K views · last reply
Math Amateur
Gold Member
MHB
Messages
3,920
Reaction score
48
At the start of Chapter 9, M. A. Armstrong in his book, "Groups and Symmetry" (see text below) writes the following:

" ... ... Each matrix $$A$$ in this group determines an invertible linear transformation $$f_A: \mathbb{R} \to \mathbb{R}$$ defined by $$f_A(x) = x A^t$$ ... ... "I know that one may define entities how one wishes ... but why does Armstrong define $$f$$ in terms of the transpose of $$A$$ rather than just simply $$A$$ ... there must be some reason or advantage to this ... but what is it? Can someone help to explain ...

I note in passing that Kristopher Tapp in his book, "Matrix Groups for Undergraduates" (Chapter 1, Section 5) ... see text below ... defines the action of a linear transformation ( multiplication by a matrix $$A$$) as $$R_A = X \cdot A$$ ... thus not using the transpose of $$A$$ ...Hope that someone can help ...

Peter=======================================================================================

The above post refers to the start of Ch. 9 of M. A. Armstrong's book, "Groups and Symmetry" ... so I am providing the relevant text ... as follows:View attachment 9568
The above post also refers to Chapter 1, Section 5 of Kristopher Tapp's book, "Matrix Groups for Undergraduates" ... so I am providing the relevant text ... as follows:View attachment 9569Note that Tapp uses $$\mathbb{K}$$ to refer to one of the real numbers, the complex numbers or the quaternions ...Hope that helps,

Peter
 
Attachments
  • Printed textbook page introducing the general linear group GL_n: the set of invertible n x n real matrices under multiplication, with the product entry formula, the map f_A(x) = xA^t, and the statement that matrix multiplication is non-commutative for n >= 2.
    Armstrong - Start of Chapter 9 ... .png
    39.3 KB · Views: 157
  • Printed textbook page, section 5 'Matrices as linear transformations': Definition 1.9 of K-linear maps, Definition 1.10 defining R_A(X) = X·A and L_A(X) = (A·X^T)^T, a 2x2 example with A = [[1,2],[3,4]] and its explicit formulas, then Proposition 1.11.
    Tapp - Chater 5, Section 1... .png
    35 KB · Views: 186
Peter said:
At the start of Chapter 9, M. A. Armstrong in his book, "Groups and Symmetry" (see text below) writes the following:

" ... ... Each matrix $$A$$ in this group determines an invertible linear transformation $$f_A: \mathbb{R} \to \mathbb{R}$$ defined by $$f_A(x) = x A^t$$ ... ... "I know that one may define entities how one wishes ... but why does Armstrong define $$f$$ in terms of the transpose of $$A$$ rather than just simply $$A$$ ... there must be some reason or advantage to this ... but what is it?
The answer is given in the text from Armstrong's book that you posted. If $f_A(x)$ is defined to be $xA^T$ then the map $A\mapsto f_A$ preserves multiplication: $f_{AB} = f_Af_B$. If the transpose does not occur in the definition then the map would reverse the order and you would get $f_{AB} = f_Bf_A$.
 
Opalg said:
The answer is given in the text from Armstrong's book that you posted. If $f_A(x)$ is defined to be $xA^T$ then the map $A\mapsto f_A$ preserves multiplication: $f_{AB} = f_Af_B$. If the transpose does not occur in the definition then the map would reverse the order and you would get $f_{AB} = f_Bf_A$.
Thanks for the help, Opalg ...

Peter