Matrix Representation of Differential Operator w/ 3 Basis Vectors

jeebs
Messages
314
Reaction score
5
I have 3 basis vectors:
e_1 = sin^2(x), e_2 = cos^2(x), e_3 = sin(x)cos(x)

I am told that the combination rule is just normal addition, and that the differential operator is defined by Dp(x) = p'(x).

My task is to show that D = \left(\begin{array}{ccc}0&0&-1\\0&0&1\\2&-2&0\end{array}\right) in this basis. So, what I've done is calculated the derivative of each of the basis vectors:
\frac{d}{dx}e_1 = 2cos(x)sin(x) = 2e_3 ... \frac{d}{dx}e_2 = -2cos(x)sin(x) = -2e_3 ... \frac{d}{dx}e_3 = cos^2(x) - sin^2(x) = e_2 - e_1

Now I'm looking at this vector p(x) being transformed into its 1st derivative p'(x). If I'm not mistaken, we can write p(x) as some arbitrary linear combination of the basis vectors, like:

p(x) = p_1e_1 + p_2e_2 + p_3e_3 = \sum_k p_ke_k = \left(\begin{array}{c}p_1e_1 \\ p_2e_2 \\ p_3e_3 \end{array}\right)... p'(x) = p_1\frac{d}{dx}e_1 + p_2\frac{d}{dx}e_2 + p_3\frac{d}{dx}e_3 = \left(\begin{array}{c}p_1e_1' \\ p_2e_2' \\ p_3e_3' \end{array}\right)
and I need to somehow find the coefficients Dij for p'(x) = Dp(x) = \left(\begin{array}{ccc}D_{11}&D_{12}&D_{13}\\D_{21}&D_{22}&D_{23}\\D_{31}&D_{32}&D_{33} \end{array}\right) \left(\begin{array}{c}p_1e_1 \\ p_2e_2 \\ p_3e_3 \end{array}\right) = \left(\begin{array}{c}D_{11}p_1e_1 + D_{12}p_2e_2 + D_{13}p_3e_3\\ D_{21}p_1e_1 + D_{22}p_2e_2 + D_{23}p_3e_3 \\ D_{31}p_1e_1 + D_{32}p_2e_2 + D_{33}p_3e_3 \end{array}\right) = \left(\begin{array}{c}p_1e_1' \\ p_2e_2' \\ p_3e_3' \end{array}\right)

My problem is that I cannot quite figure out how I can get from what I currently know to having the coefficients of the D matrix. I know this is probably trivial but I've been sat staring at this for ages now...

For instance, take, the first component of p'(x). we have p_1e_1' = D_{11}p_1e_1 + D_{12}p_2e_2 + D_{13}p_3e_3 = 2p_1e_3. Since the coefficients pi are arbitrary, it means we can set D11 = D12 = 0, hence p1e1' = D13p3e3 therefore D13 = p1e1' / p3e3, and since the pi factors are arbitrary, they can be set = 1.

thus D13 = e1' / e3, but this does not give me the D13 = -1 that I require - it gives me D13 = 2e3 / e3 = 2.
What am I doing wrong here? I notice there is a 2 in the matrix, but it's in the wrong corner - D31 rather than D13. I can't see where my mistake is though, unless that switching of corners is just a fluke.
 
Last edited:
Physics news on Phys.org
hi jeebs! :smile:

i can't make out what you're doing :confused:

to find D, you know that …

D(1,0,0) has to be (0,0,2)

D(0,1,0) has to be (0,0,-2)

D(0,0,1) has to be (-1,1,0) …

doesn't that make it clear what D is? :wink:
 
Right I partially get why you said that, but don't fully understand. We have:
e1' = 2e3
e2' = 2e3
e3' = e2 - e1

So, for some reason (I'm not sure why this is allowed), we've decided that:
e_1 = \left(\begin{array}{c}1\\ 0 \\ 0\end{array}\right) ... e_2 = \left(\begin{array}{c}0\\ 1 \\ 0\end{array}\right) ... e_3 = \left(\begin{array}{c}0\\ 0 \\ 1\end{array}\right)
Why are we allowed to say this? Isn't this only true for the Cartesian unit vectors? *I get the feeling there's something big about this that I doo't understand yet *.
Anyway, from there we get:
e_1' = \left(\begin{array}{c}0\\ 0 \\ 2\end{array}\right) ... e_2' = \left(\begin{array}{c}0\\ 0 \\ -2\end{array}\right) ... e_3' = \left(\begin{array}{c}0\\ 1 \\ 0\end{array}\right) - \left(\begin{array}{c}1\\ 0 \\ 0\end{array}\right) = \left(\begin{array}{c}-1\\ 1 \\ 0\end{array}\right)

It's still not obvious to me where these matrix coefficients come from either. What I tried from this is to say that the vector p could be, say,
p = e1 + e2 + e3 and p' = e1' + + e2' + e3'. In other words, we get:

p' = \left(\begin{array}{ccc}D_{11}&D_{12}&D_{13}\\D_{21}&D_{22}&D_{23}\\D_{31}&D_{32}&D_{33} \end{array}\right) \left(\begin{array}{c}1 \\ 1 \\ 1 \end{array}\right) = \left(\begin{array}{c}-1 \\ 1 \\ 0 \end{array}\right)

(noting that e3' = e3' + e1' + e2') .So from this I would have 3 equations to find 9 unknowns, which aint happening. What's going on here?
 
Last edited:
(abc;def;ghi)(100) = (adg)
(abc;def;ghi)(010) = (beh)
(abc;def;ghi)(001) = (cfi)
jeebs said:
So, for some reason (I'm not sure why this is allowed), we've decided that:
e_1 = \left(\begin{array}{c}1\\ 0 \\ 0\end{array}\right) ... e_2 = \left(\begin{array}{c}0\\ 1 \\ 0\end{array}\right) ... e_3 = \left(\begin{array}{c}0\\ 0 \\ 1\end{array}\right)
Why are we allowed to say this?

but that's the definition of e1 e2 and e3
 
PS. I just added a bit on right at the bit that you just mentioned, just before you responded. Why is that the definition of all basis vectors? not just the Cartesian i,j,k?
 
jeebs said:
Why is that the definition of all basis vectors? not just the Cartesian i,j,k?

Because it is.

That's what a basis is (in a vector space).

What did you think it is? :confused:
 
I don't know, is it the law that all basis vectors have to have a length of 1?
also, doesn't writing them as (1,0,0), (0,1,0) and (0,0,1) imply they are orthogonal, which I did not think was necessary for a set of basis vectors?
 
Back
Top