How do you change a vector in a non-canonical base to another non-canonical base?

269 Views Asked by At

I'm a little confused by this. I've been watching some videos on change of bases, but they tend to talk about the change of basis matrix.

Let's say we have coefficients of a vector V in base $B = (e1,e2,e3)$. To get the coefficients of $V$ in the canonical basis, we do the following

$[e1 e2] * [V]_B = [V]_C$ (where $C$ is the canonical basis). Is this right so far?

Ok, so this is the online videos I've found have explained how to change a basis. But how do I do it if I want to convert my vector so that it's in another canonical base? Would I just have to turn it into the standard base, and then back into a different base, or is there an easier method? Thanks.

1

There are 1 best solutions below

0
On

You express the coordinates of the target basis into the original basis and then it's the same method that you use for the canonical basis.

For instance if you have a basis $\mathcal{B_1} = (e_1,e_2,e_3)$ and a basis $\mathcal{B_2} = (f_1,f_2,f_3)$ and let's say $V_1$ is the column representation of a vector in $\mathcal{B_1}$ and $V_2$ the same vector in $\mathcal{B_2}$.

You have a certain relation between the basis. Namely, there exists 9 scalars such that: $$\begin{cases} f_1 = \alpha_1e_1 + \alpha_2e_2 + \alpha_3e_3 \\ f_2 = \beta_1e_1 + \beta_2e_2 + \beta_3e_3 \\ f_3 = \gamma_1e_1 + \gamma_2e_2 + \gamma_3e_3 \end{cases}$$

Then the change of basis matrix is: $$ P = \begin{pmatrix} \alpha_1 & \beta_1 & \gamma_1 \\ \alpha_2 & \beta_2 & \gamma_2 \\ \alpha_3 & \beta_3 & \gamma_3 \end{pmatrix}$$

And then $V_2 = P^{-1}V_1$