0
$\begingroup$

I was reading a book which says that in case we are dealing with homogeneous linear equations,if A matrix has size of $m\times n$ and if $m<n$, then it will have infinite solutions other than trivial solution (where $x_1=x_2=x_3=\dots=0$), where $x_1,x_2,x_3$ are the variables in equation.

But i think the trivial solution will always possible even in case my matrix rank $<n$ ?Regardless of my rank trivial solution is always possible and in case of infinite solution one of the solution is trivial solution only? Am i correct?

$\endgroup$
  • 1
    $\begingroup$ Can you please adjust your question? It seems like something is missing. $\endgroup$ – Alberto Andrenucci Jun 22 '17 at 18:28
  • $\begingroup$ Yes.Doing it.It is showing half the question in preview itself. $\endgroup$ – rahul sharma Jun 22 '17 at 18:29
1
$\begingroup$

Trivial solution is always possible. Imagine that you take a standard matrix and multiply it by the $0$ vector. It will always give you the $0$ vector.

In fact, in a linear application, you have that $f(0)=0$.

What the book is trying to say is: if you have $n$ incognitas but just $m$ equations, with $m<n$, you will have $n-m$ degrees of freedom.

In fact, if you take $m$ equations, they give you conditions only on $m$ components of your vector. The other ones, $n-m$, are at your choice.

Let's do an example: you have just one equation, saying you that $x_1=4$. Considering that you are in $\mathbb{R}^3$ you have a lot of vectors satisfying this condition. In fact, you just have to put the first coordinate equal to $4$, for the other two you have free choice. In fact, having $2$ degrees of freedom, the $\text{Span}$ generated by vectors satisfying this condition has dimension $2$:

$$\text{Span}\left(\begin{pmatrix}4\\0\\1 \end{pmatrix}, \begin{pmatrix}4\\1\\0 \end{pmatrix} \right)$$

$\endgroup$
  • $\begingroup$ Got it. I took the wrong interpretation from the book. Thanks a lot. $\endgroup$ – rahul sharma Jun 22 '17 at 20:33
  • $\begingroup$ No problem, glad to help! $\endgroup$ – Alberto Andrenucci Jun 22 '17 at 20:33

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.