Take the 2-minute tour ×
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.

Let's say we have a basis $B$ of $\mathbb{R}^n$ consisting of vectors $\vec{v}_1$ through $\vec{v}_n$, and some other basis $C$ of $\mathbb{R}^n$. Then, would $[\vec{v}_1]_C$ through $[\vec{v}_n]_C$ be a basis for $\mathbb{R}^n$ as well? It seems obvious, but I am not sure how to go about this.

I am trying to prove this by creating a matrix $S$ that is a change of basis matrix, so it has $[\vec{v}_1]_C$ through $[\vec{v}_n]_C$ as its columns. So I get $C$$S$=$B$. I know that kernels of both $C$ and $B$ are equal to zero, since they are square matrices with independent columns as they are bases, but I am not sure what that tells me about $S$. Maybe there is a way to show that a matrix with independent columns ($C$) times a matrix with not-independent columns ($S$) can't (always/sometimes?) yield another matrix with independent columns ($B$).

share|improve this question

2 Answers 2

up vote 3 down vote accepted

Yes, the coordinate vectors of $\mathbf{v}_1,\ldots,\mathbf{v}_n$ with respect to the matrix $C$ will be a basis for $\mathbb{R}^n$.

Note that you are abusing notation somewhat: if $B$ and $C$ are bases, then they aren't matrices; what you really want is to have a basis $\beta$, and $B$ is the matrix whose columns are the vectors in $\beta$; and another basis $\gamma$, and $C$ is the matrix whose columns are the vectors in $\gamma$.

From what you have: since $CS = B$, and the kernel of $B$ is zero, then so is the kernel of $S$: if $\mathbf{x}$ lies in the kernel of $S$, then $$\mathbf{0} = C\mathbf{0} = C(S\mathbf{x}) = (CS)\mathbf{x} = B\mathbf{x}.$$ But since the kernel of $B$ is zero, that means that $\mathbf{x}=\mathbf{0}$. So the kernel of $S$ is zero, which means that its columns are linearly independent, hence a basis. But the columns of $S$ are the coordinate vectors of the elements of $\beta$ written in terms of $\gamma$, which gives your result.

For your final question, yes, you are correct. Here's a general statement you can prove (using the same technique as above):

If $A$ is $n\times p$, $B$ is $p\times m$, and $C$ is $n\times m$ with $AB=C$, then $\mathrm{ker}(B)\subseteq\mathrm{ker}(C)$. That is, if $\mathbf{x}\in\mathbb{R}^m$ is in the kernel of $B$, then it is also in the kernel of $C$.

Another way of attacking the problem is to prove that $[\mathbf{v}_1]_{\gamma},\ldots,[\mathbf{v}_n]_{\gamma}$ are linearly independent directly. But this is easy: suppose that $$\alpha_1[\mathbf{v}_1]_{\gamma} + \cdots + \alpha_n[\mathbf{v}_n]_{\gamma} = \mathbf{0}.$$ Then we have: $$\mathbf{0} = \alpha_1[\mathbf{v}_1]_{\gamma} + \cdots + \alpha_n[\mathbf{v}_n]_{\gamma} = [\alpha_1\mathbf{v}_1+\cdots+\alpha_n\mathbf{v}_n]_{\gamma}.$$ But this means that $\alpha_1\mathbf{v}_1+\cdots+\alpha_n\mathbf{v}_n=\mathbf{0}$, and since $\mathbf{v}_1,\ldots,\mathbf{v}_n$ are linearly independent, that means $\alpha_1=\cdots=\alpha_n=0$.

In essence, it's the same argument as you have above, but playing directly with the vectors instead of the associated change-of-basis matrices.

share|improve this answer

Yes. If $SB = C$ then $S = CB^{-1}$, which is also an invertible matrix since $B$ and $C$ are invertible.

If $B$ and $C$ are basis, there is a basis change matrix $S$ (or rather, a linear application $f$ associated to $S$) that changes $B$ into $C$. But $f$ is not only the application that turns one basis into another, it is an application that turns any basis into another basis. If $(e_1,\ldots,e_n)$ is the canonical basis of $\mathbb{R}^n$, $f$ turns this basis into a new basis, whose vectors are the column vectors of its matrix representation $S$.

Those applications $f$ who turn basis into basis are precisely the bijective linear applications from $\mathbb{R}^n$ to $\mathbb{R}^n$. It is easy to check that if $f$ is a change of basis, then it is bijective ; and if $f$ is bijective, it turns any basis into another basis.

share|improve this answer

Your Answer

 
discard

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.