2
$\begingroup$

Suppose that $\{v_1,v_2,v_3\}$ is a linearly independent subset of a vector space $V$, with $\dim(V) = 4$, and that $v_4 \not\in \text{span}\{v_1,v_2,v_3\}$. Prove that $\{v_1,v_2,v_3,v_4\}$ is a basis for $V$.

My current answer would be:

Since $\dim(V) = 4$, it requires another vector (which is not linearly dependent) to form a basis.

Hence $v_4$ is not equal to $k_1 v_1+k_2 v_2+k_3 v_3$ (except if $k1=k2=k3 = 0$)

Since $v_4$ is not in $\text{span}\{v_1,v_2,v_3\}$, it is not linearly dependent and $\{v_1,v_2,v_3,v_4\}$ is a basis for $V$ (since vectors as a basis should be linearly independent).

Just wondering if you can help me whether my answer respond to the question well. Any help will be appreciated.

Thank you very much

$\endgroup$
  • $\begingroup$ Correct, absolutely fine. $\endgroup$ – астон вілла олоф мэллбэрг Apr 14 '16 at 1:30
  • $\begingroup$ You mean to write "and $v_4$ does not lie in the span of $v_1,v_2,v_3$" Of course a single vector cannot span a three dimensional space. $\endgroup$ – JMoravitz Apr 14 '16 at 1:36
  • 1
    $\begingroup$ I think you meant "is not in span..." instead of "does not span..." so I changed the wording when I edited to add typesetting. Please let me know if that is incorrect, or you can edit your post accordingly. $\endgroup$ – Bungo Apr 14 '16 at 1:36
  • $\begingroup$ Here is a Mathjax tutorial for future reference. It's easy to learn and allows you to typeset your posts properly. $\endgroup$ – Bungo Apr 14 '16 at 1:37
  • $\begingroup$ No need for a downvote. Here is an OP who does not blindly ask for an answer but who thinks, shares and wants verification...Hence my upvote $\endgroup$ – imranfat Apr 14 '16 at 1:39
2
$\begingroup$

The idea of your proof is correct, but it should be stated more clearly.

First, since $v_4 \not\in \text{span}\{v_1,v_2,v_3\}$, this means that there are no scalars $k_1,k_2,k_3$ for which $v_4 = k_1 v_1 + k_2 v_2 + k_3 v_3$, and this is remains true if $k_1 = k_2 = k_3 = 0$. You don't need to treat that as a special case. It is only when discussing linear independence that we need to exclude that case.

Then, to argue that $\{v_1,v_2,v_3,v_4\}$ is linearly independent, suppose for a contradiction that this is false. This means that there are scalars $k_1,k_2,k_3,k_4$, not all zero, such that $$k_1 v_1 + k_2 v_2 + k_3 v_3 + k_4 v_4 = 0$$ Now $k_4$ can't be zero, since otherwise this equation becomes $$k_1 v_1 + k_2 v_2 + k_3 v_3 = 0$$ which is only true if $k_1 = k_2 = k_3 = 0$ since $\{v_1,v_2,v_3\}$ is a linearly independent set. Therefore, since $k_4 \neq 0$, we can divide by $k_4$ and rearrange to obtain $$v_4 = -\frac{k_1}{k_4} v_1 - \frac{k_2}{k_4}v_2 - \frac{k_3}{k_4}v_3$$ But this expresses $v_4$ as a linear combination of $v_1,v_2,v_3$, so $v_4 \in \text{span}\{v_1,v_2,v_3\}$, which is a contradiction.

We conclude that $\{v_1,v_2,v_3,v_4\}$ must be linearly independent. Moreover, it is a linearly independent set of $4$ vectors, in a vector space of dimension $4$, hence it is a basis for $V$.

$\endgroup$
  • $\begingroup$ Thank you very much or your detailed answer. I'd like to have a question, please. if k4 = nonzero, can we still say that the set (v1,v2,v3,v4) is linearly independent? Since for linear independence should there only be the trivial solution for case AX=0? $\endgroup$ – Lucky Boy Apr 14 '16 at 2:06
  • 1
    $\begingroup$ @LuckyBoy If $k_1 v_1 + k_2 v_2 + k_3 v_3 + k_4 v_4$ where any of the $k_n$'s (including $k_4$) is nonzero, then this mean that $\{v_1,v_2,v_3,v_4\}$ is a linearly dependent set. $\endgroup$ – Bungo Apr 14 '16 at 2:14
  • $\begingroup$ I see, thanks a lot for your explanation $\endgroup$ – Lucky Boy Apr 14 '16 at 2:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.