$A,B$ Matrices $5{\times}5$, $p(B)<p(A)$, dim solution space $Bx=0$ is $ 2$, prove: $BA \neq 0$

45 Views Asked by At

$A,B$ Matrices $5{\times}5$, $p(B)<p(A)$

Solution space of $Bx=0$: $Sp\{(1,-2,5,0,1),(4,0,-3,4,1)\}$

Prove: $BA \neq 0$.

*$p(A) = \operatorname{rank}(A)$


My attempt: (stuck in the middle)

We know that:

$\#n-p(B)=\dim P$

Where:

#n - num of varaibles

$p(B) - \operatorname{rank} B$

$\dim P$ - the dimension of the solution set for $Bx = 0$

Therefore, we conclude: $p(B) = 3$

$\Rightarrow 3<p(A) \leq 5 = 4 \ or \ 5$

I though maybe i could conclude somehow that $A$ is invertible ,therefore $BA \neq 0$, but could'nt make it. (The rank(A) may be 4).

hint?

*Edit:

As suggested by the answer i will write the way i understood it for people that will look at this question in the future.

If $p(A) = 5$, therefore A invertible, therefore for every B square matrix $5x5$: $BA \neq 0$ (invertible matrix proteprties).

If $p(A) = 4$, We will assume $BA = 0$ and get to a contradiction.

If $p(A) = 4$, there are 4 independent colums in A.

Choose 4 different vectors$x_1,x_2,x_3,x_4$, such that $Ax_1$ represent the first independent column, $Ax_2$ the second independent and so on.

We assume $BA = 0$, therefore, multiply from the right by $x_1$ get: $BAx_1 = 0$ its a unique, independent solution for $Bx = 0$.

Again, take $BA = 0$, multiply from the right by $x_2$ we get: $BAx_2 = 0$, its another independent solution for $B_x = 0$

And so on.

We get therefore 4 independent solutions for $Bx = 0$.

But we know that the dimension of the solution set of $Bx = 0$ is 2.

Contradiction.

$BA \neq 0$.

Thanks for the help.

1

There are 1 best solutions below

13
On BEST ANSWER

If rank of $A$ is $5$ and $BA=0$, it implies that $A$ is invertible and hence $B=0A^{-1}=0$, which is a contradiction, because nullity of $B$ is not $5$.

If rank of $A$ is $4$, then

$$B(Ax)=0\forall x\in\Bbb R^5$$ Hence, nullity of $B$ is atleast the rank of $A$, which is a contradiction (because kernel of $B$ contains atleast as many linearly independent vectors, as in the column space of $A$).