Proof of Identity for Transpose of a Product of Matrices

Theorem
If  
\[A\]
  and  
\[B\]
  are matrices that can be multiplied, then  
\[(AB)^{T} = B^T A^T\]

Proof
Let  
\[A=(a_{ij})\]
  be an  
\[n \times m\]
  matrix and  
\[B=(b_{ij})\]
  be an  
\[n \times r\]
  matrix.
Then  
\[C=(c_{ij})\]
  where
\[c_{ij} = a_{i1}b_{1j}+ a_{i2}b_{2j}+ ...+ a_{in}b_{nj} =\sum_{k=1}^{k=m} a_{ik}b_{kj} \]

\[c_{ij}\]
  is the element in the  
\[i\]
^{th} row and  
\[j\]
^{th} column of  
\[C\]
  and the element in the  
\[j\]
^{th} row and  
\[i\]
^{th} column of  
\[C^T =(AB)^T\]
.
The elements  
\[b_{1j}, \: b_{2j} ,..., \: b_{mj}\]
  form the  
\[j\]
^{th} column of  
\[B\]
  and the  
\[j\]
^{th} row of  
\[B^T\]
.
The elements  
\[a_{i1}, \: a_{i2} ,..., \: a_{in}\]
  form the  
\[i\]
^{th} row of  
\[A\]
  and the  
\[i\]
^{th} column of  
\[A^T\]
.
Thus the element in the  
\[j\]
^{th} row and  
\[i\]
^{th} column of  
\[B^T A^T\]
  is
\[b_{1j} a_{i1} +b_{2j} a_{i2}+...+ b_{mj} a_{m1}= \sum_{k=1}^{k=m} a_{ik}b_{kj}\]
.
Hence  
\[(AB)^{T} = B^T A^T\]

Add comment

Security code
Refresh