Is this linear independence proof valid?

59 Views Asked by At

Let's say we have a 4x4 nilpotent matrix $V$ with an index of 4 and a vector $\vec{x}$ which is an element of $\mathcal{R}^4$, matrix $V$ and $\vec{x}$ are non-zero. How would we show that $V\vec{x}$ and $\vec{x}$ are linearly independent? The following is what I did but I personally think that my answer is somewhat iffy.

$$ \lambda_1\vec{x} + \lambda_2V\vec{x} = \vec{0} \\ \lambda_1(\vec{x_1}+\vec{x_2}+\vec{x_3}+\vec{x_4)} + \lambda_2V(\vec{x_1}+\vec{x_2}+\vec{x_3}+\vec{x_4}) = \vec{0} $$ Factorizing it according to the vectors $\vec{x_1}$ to $\vec{x_4}$: $$ (\lambda_1 + \lambda_2V)\vec{x_1}+(\lambda_1 + \lambda_2V)\vec{x_2}+(\lambda_1 + \lambda_2V)\vec{x_3}+(\lambda_1 + \lambda_2V)\vec{x_4} = \vec{0} $$ Solving this would give, all coefficients as 0 since they're all the same. I think I messed up somewhere in my understanding of linear independence or something but please help and explain it to me. Thank you in advance!

1

There are 1 best solutions below

0
On BEST ANSWER

As given, the statement you're asking to be proved is not necessarily true, because $\ V\vec{x}\ $ could be zero, even if both $\ V\ $ and $\ \vec{x}\ $ are non-zero, and if $\ V\vec{x}=\vec{0}\ $, then $\ V\vec{x}\ $ and $\ \vec{x}\ $ are not linearly independent.

Since $\ V\ $ is nilpotent, then there must be a smallest power, $\ V^n\ $, of $\ V\ $ that annihilates $\ \vec{x}\ $ (that is, $\ V^n\vec{x}=\vec{0}\ $ and $\ V^{n-1}\vec{x}\ne\vec{0}\ $), and if $\ V\vec{x}\ne\vec{0}\ $, then $\ n\ge2\ $. Suppose, in this case, that $$ \lambda_1 \vec{x}+ \lambda_2 V\vec{x}= \vec{0}\ . $$ Multiplying this equation by $\ V^{n-1}\ $ gives $$ \vec{0}=\lambda_1 V^{n-1} \vec{x}+ \lambda_2 V^n\vec{x}= \lambda_1 V^{n-1} \vec{x}\ . $$ Since $\ V^{n-1}\vec{x}\ne\vec{0}\ $, it follow that $\ \lambda_1=0\ $ and then that $\ \lambda_2 V\vec{x}= \vec{0}\ $. Since $\ V\vec{x}\ne\vec{0}\ $ it also follows that $\ \lambda_2=0\ $. Thus if $\ V\vec{x} \ne\vec{0}\ $ then $\ \vec{x}\ $ and $\ V\vec{x}\ $ are linearly independent.