Showing that set of all vectors orthogonal to two vectors is a subspace.

1.1k Views Asked by At

Let $A_1$, $A_2$ be vectors in $R^{n}$. Show that the set of all vectors $B$ in $R^{n}$ such that $B$ is perpendicular both $A^{1}$ and $A^{2}$ is a subspace.

This was one of the tasks in S.L's linear algebra text book. I've constructed such proof:

Let $W=B$.

Notice that $O^TA_1=0=O^TA_2$ (where $O$ is a null vector), thus $O$ is in W.

Suppose $B$, $C$ are orthogonal to $A_1$ and $A_2$, then:

$$A_1^{T}(B+C)=A_1^{T}B+A_1^{T}C=0=A_2^{T}(B+C)=A_2^{T}B+A_2^{T}C$$

thus it's additive linear combination which is again in $W$.

Finally, suppose there is scalar c, then $(cB) \cdot A=c(A\cdot B)=0$.


Hence, If I'm correct, $B$ satisfies all three axioms to be a subspace.

Does this proof have any flaws? Also isn't $B$ also a kernel (null space) of $A_{1} \cup A_2$?

1

There are 1 best solutions below

3
On BEST ANSWER

The main confusion to me seems to be what is $B$? is it a vector or a subspace?

Perhaps define $$W=\{ b \in \mathbb{R}^n: A_1^Tb=A_2^Tb=0 \}$$

Also, for the closure under the addition, I would write

let $b, c \in W$, then for $i\in \{1,2\}$,

$$A_i^T(b+c)=A_i^Tb+A_i^Tc=0+0=0.$$ just to highlight that $A_i^Tb=0$ and $A_i^Tc=0$.

$W$ is the nullspace of $$\begin{bmatrix} A_1^T \\ A_2^T \end{bmatrix}$$

I do not comprehend the meaning of the union of two vectors, $A_1 \cup A_2$.