Rank and Nullity of a Matrix, Nullity of Transpose

Linear Algebra Problems and Solutions

Problem 140

Let $A$ be an $m\times n$ matrix. The nullspace of $A$ is denoted by $\calN(A)$.
The dimension of the nullspace of $A$ is called the nullity of $A$.
Prove the followings.

(a) $\calN(A)=\calN(A^{\trans}A)$.

(b) $\rk(A)=\rk(A^{\trans}A)$.

 
LoadingAdd to solve later

Sponsored Links


Hint.

For part (b), use the rank-nullity theorem and the result from (a).
The rank-nullity theorem says that for a $m\times n$ matrix,
\[\text{rank of }A+\text{ nullity of }A=n.\]

Proof.

(a)$\calN(A)=\calN(A^{\trans}A)$.

Show $\calN(A) \subset \calN(A^{\trans}A)$

Consider any $\mathbf{x} \in \calN(A)$. Then we have $A\mathbf{x}=\mathbf{0}$. Multiplying it by $A^{\trans}$ from the left, we obtain
\[A^{\trans}A\mathbf{x}=A^{\trans}\mathbf{0}=\mathbf{0}.\] Thus $\mathbf{x} \in \calN(A^{\trans}A)$, and hence $\calN(A) \subset \calN(A^{\trans}A)$.

Show $\calN(A) \supset \calN(A^{\trans}A)$

On the other hand, let $\mathbf{x} \in \calN(A^{\trans}A)$. Thus we have
\[A^{\trans}A\mathbf{x}=\mathbf{0}.\] Multiplying it by $\mathbf{x}^{\trans}$ from the left, we obtain
\[\mathbf{x}^{\trans}A^{\trans}A\mathbf{x}=\mathbf{x}^{\trans}\mathbf{0}=\mathbf{0}.\] This implies that we have
\[\mathbf{0}=(A\mathbf{x})^{\trans}(A\mathbf{x})=||A\mathbf{x}||^2\] and the length of the vector $A\mathbf{x}$ is zero, thus the vector $A\mathbf{x}=\mathbf{0}$. Hence $\mathbf{x} \in \calN(A)$, and we obtain $\calN(A) \supset \calN(A^{\trans}A)$.

(b) $\rk(A)=\rk(A^{\trans}A)$

We use the rank-nullity theorem and obtain
\begin{align*}
\rk(A)=n-\dim(\calN(A))=n-\dim(\calN(A^{\trans}A))=\rk(A^{\trans}A).
\end{align*}
(Note that the size of the matrix $A^{\trans}A$ is $n\times n$.)


LoadingAdd to solve later

Sponsored Links

More from my site

You may also like...

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

More in Linear Algebra
Problems and solutions in Linear Algebra
Sum of Squares of Hermitian Matrices is Zero, then Hermitian Matrices Are All Zero

Let $A_1, A_2, \dots, A_m$ be $n\times n$ Hermitian matrices. Show that if \[A_1^2+A_2^2+\cdots+A_m^2=\calO,\] where $\calO$ is the $n \times...

Close