Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

There is an added condition of

Suppose $u_1,u_2,\ldots,u_k$ are linearly independent. Show that if A is invertible, then $Au_1,Au_2,\ldots,Au_k$ are linearly independent.

my solution

since $u_1,u_2,\ldots,u_k$ are linearly independent, then $c_1 u_1+c_2 u_2+\cdots+c_k u_k=0$ implies $c_1,\ldots$ are all zero.

by multiplying A on both sides, we have $c_1 A u_1+c_2 A u_2+\cdots =c_k A u_k=0$ since all those c can only be 0, therefore $Au_1,Au_2,\ldots,Au_k$ are linearly independent.

Notice how i didnt use the supposition if A is invertive and still get the answer. I am wondering if my answer is right or wrong...

share|cite|improve this question
You have to start with the second equation. Is there some way back to the first, so that you can apply lin.ind. of the $u_i$? – Lord_Farin Oct 18 '12 at 13:55
I get what you mean, but if i could get the answer without using the suposition, isn tthat better? That is, so i have shown that no matter if a is invertible or not, the answer will always be linearly independent – Yellow Skies Oct 18 '12 at 14:01
@SingaporeanDude.: But that result is clearly false (take $A=0$ for example), so your reasoning must be flawed. – Chris Eagle Oct 18 '12 at 14:02
Im still unsure how is it flawed :( – Yellow Skies Oct 18 '12 at 14:04
Just because the coefficients have to be $0$ for the $\mathbf{u_i}$ to sum to $\mathbf{0}$ doesn't mean the same holds for $A\mathbf{u_i}$. – EuYu Oct 18 '12 at 14:07
up vote 2 down vote accepted

Hint: Assume by negation they are linear dependent and write a non-trivial linear combination that sums to zero.

Now this is the step where if the $A$ weren't there then you had a non-trivial linear combination of the $u_{i}$'s that sums to zero, can you get rid of the $A$ ? remember that $A$ is invertible means there exist $A^{-1}$

ADDED: Your "proof" is incorrect. What you did is take something of the form $0+...+0=0$ and multiplied it by $A$ (each $c_i$ is zero so basicly you summed zeros).

You started out with all the coefficients as zero (before multiplying by $A$), this does not show how any linear combination is zero.

To see this more clearly say I give you all the coefficient as $1$, that is : $Au_{1}+...+Au_{k}$ and I claim that $Au_{1}+...+Au_{k}=0$ can you trace from your proof that this can not happen ?

Clearly it is wrong for take $A=0$ , then it is clear that $\{Au_{i}\}_{i=1}^{i=k}$ is linear dependent since all the vectors are $0$. The assumption about $A$ is important

share|cite|improve this answer
Thanks that makes things extremely clear!! – Yellow Skies Oct 18 '12 at 14:17
@SingaporeanDude. - I'm very glad it helps! :) – Belgi Oct 18 '12 at 14:21

The problem is that you haven’t paid enough attention to just what it is that you’re trying to prove. You want to show that the vectors $Au_1,\dots,Au_k$ are linearly independent. Write down exactly what this means: you want to show that if some linear combination $c_1Au_1+\ldots+c_kAu_k=0$, then $c_1=\ldots=c_k=0$. The most straightforward approach to proving an if-then statement is to assume the hypothesis and work your way to the conclusion. It isn’t always the best approach, or even a feasible approach, but it’s the most straightforward and therefore worth a look.

Assume, then, that $c_1Au_1+\ldots+c_kAu_k=0$. How can you reasonably hope to show that $c_1=\ldots=c_k=0$? One way would be somehow to find linearly independent vectors $v_1,\dots,v_k$ such that $c_1v_1+\ldots+c_kv_k=0$; then you could use the linear independence of the $v_i$ to conclude that $c_1=\ldots=c_k=0$. But where might these $v_i$ come from? Well, we know that $u_1,\dots,u_k$ are linearly independent, and they’re the only vectors about which we’ve been given any information, so perhaps we should try to show that $c_1u_1+\ldots+c_ku_k=0$. Is there any way to deduce that from the hypothesis that $c_1Au_1+\ldots+c_kAu_k=0$? Here is where you’ll use the information that $A$ is invertible.

share|cite|improve this answer
Thanks that is very helpful – Yellow Skies Oct 18 '12 at 14:19

Suppose $c_1Au_1+\ldots+c_kAu_k=0$. Then $A(c_1u_1+\ldots+c_ku_k)=0$. Since $A$ is invertible, we have $c_1u_1+\ldots+c_ku_k=0$. Since $u_1,\ldots,u_k$ are linearly independent, we have $c_1=\ldots=c_k=0$. Hence $Au_1,\ldots,Au_k$ are linearly independent.

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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