Ask your own question, for FREE!
Mathematics 9 Online
OpenStudy (kainui):

Is there a simple way to express this? \[\Large (A+B)^k\]

OpenStudy (kainui):

\[\Large (A+B)^k\] A and B are matrices and k is positive integer. We can't assume that A and B commute.

OpenStudy (dan815):

how can A+B not commute

OpenStudy (kainui):

They're matrices...

OpenStudy (perl):

matrices commute under addition

OpenStudy (kainui):

\[\Large (A+B)^2 \ne A^2+2AB+B^2\]

OpenStudy (dan815):

Oh i thought u meant it cant be written (B+A)^k

OpenStudy (dan815):

how about diagonlizing both

OpenStudy (kainui):

Hmm well that's a good idea but technically I'm not really interested in matrices I'm interested in noncommuting operators and thought this question would be a simple analogy to see if we could figure out a binomial theorem to use instead. So can we do it without diagonalizing them, but still have some kind of binomial theorem form? haha

OpenStudy (perl):

what happens when you expand ( A + B ) * ( A + B ) using matrix multiplication rules distributivity

OpenStudy (perl):

is that equal to (A+B) * A + (A+B)*B A*A + B*A + A*B + B*B ?

OpenStudy (anonymous):

Ok so we can conclude this Can ::) assune they are diagonalizable?

OpenStudy (dan815):

u can do that perl

OpenStudy (kainui):

My real question is finding the general form of this derivative, I expressed it as this \[\Large \dfrac{d^k}{dx^k}\left( e^{f(x)} \right)=e^{f(x)}(f'(x)\cdot+\frac{d}{dx}\cdot)^k\] So when k=1 we have: \[\large e^{f(x)} (f'(x)+\frac{d}{dx})1=e^{f(x)} (f'(x)*1+\frac{d}{dx}1)=e^{f(x)}f'(x)\]

OpenStudy (dan815):

how to write a general form for (A+B)^k is what we gotta find

OpenStudy (kainui):

Yeah you are right @perl on distributing it out, it is A^2+AB+BA+B^2

OpenStudy (perl):

ok i see , so we can't assume A, and B commute under multiplication, so it doesn't simplify to A^2 + 2AB + B^2

OpenStudy (kainui):

For comparison, \[e^{f(x)} (f'(x)+\frac{d}{dx})(f'(x)+\frac{d}{dx})1 = e^{f(x)}(f'(x)^2+f''(x)+f'(x)\frac{d}{dx}+\frac{d^2}{dx^2}) \\ e^{f(x)}(f'(x)^2+f''(x))=\frac{d^2}{dx^2}(e^{f(x)})\]

OpenStudy (anonymous):

\(e^A=\phi(A)\)

OpenStudy (kainui):

So yeah, the A and B thing is really a lot simpler than all this nonsense, unless you can figure out how to diagonalize either of these, multiplication by f'(x) or the derivative operator.

OpenStudy (anonymous):

such that \(\phi(A)\) is a fundamental solution matrices

OpenStudy (kainui):

Hmm well here f(x) is just a regular function of x, not a matrix at all.

OpenStudy (anonymous):

hmm :\

OpenStudy (kainui):

Really, the general form of this derivative is what I'm after where f(x) is a function. \[\Large \frac{d^k}{dx^k}(e^{f(x)})\] So if you can help me solve this or the matrix problem would be helping me the same haha.

OpenStudy (thomas5267):

I think repeated logarithmic differentiation would work pretty well here.

OpenStudy (kainui):

Oh that sounds promising, can you explain what you mean by that?

OpenStudy (thomas5267):

Do you know what logarithmic differentiation is?

OpenStudy (kainui):

Yes, but I'm not sure how I would apply it to this situation.

OpenStudy (kainui):

Oh another fact, we can assume I already know all the derivatives of f(x), meaning f'(x), f''(x), f'''(x), etc...

OpenStudy (thomas5267):

Nothing special. Just brute force some logarithmic derivative for a few times and try to find a closed form solution lol.

OpenStudy (kainui):

Yeah brute force how, I'm not really seeing that lol.

OpenStudy (anonymous):

the only way i see is expanding :|

OpenStudy (kainui):

Oh like a power series? I'll try that

OpenStudy (anonymous):

and we can state special cases:- (it simplify it very good) A,B diagonalizable A,B upper triangle (or the other type) A,B diagonal

OpenStudy (anonymous):

i'll take snap for my work

OpenStudy (kainui):

Hmm well I don't really have a clue how we could say multiplication by f'(x) is a diagonal matrix or that the derivative of a function is a upper triangular matrix or something like that... Unless maybe we somehow use a polynomial with an infinite matrix...

OpenStudy (anonymous):

hmm no it work with n*n hmm but i see f(x) and f'(x) as vectors not scalar

OpenStudy (kainui):

By the way, the function f(x) has its derivatives defined as this: \[\Large f^{(k)}(x) = \frac{a}{(x+b)^k}\] for k>0

OpenStudy (anonymous):

too much work -.-

OpenStudy (kainui):

Wait show me what you have @Marki it might be helpful since I figured out a way to make an upper triangular matrix out of the derivative operator

OpenStudy (anonymous):

i just dumbly expanded it

OpenStudy (thomas5267):

For reference: \[ \begin{align*} \frac{d}{dx}e^{g(x)}=&\,\,e^{g(x)} g'(x)\\ \frac{d^2}{dx^2}e^{g(x)}=&\,\,e^{g(x)} g''(x)+e^{g(x)} g'(x)^2\\ \frac{d^3}{dx^3}e^{g(x)}=&\,\,e^{g(x)} g^{(3)}(x)+e^{g(x)} g'(x)^3+3 e^{g(x)} g'(x) g''(x)\\ \frac{d^4}{dx^4}e^{g(x)}=&\,\,e^{g(x)} g^{(4)}(x)+3 e^{g(x)} g''(x)^2+e^{g(x)} g'(x)^4+4 e^{g(x)} g^{(3)}(x) g'(x)\\&+6 e^{g(x)} g'(x)^2 g''(x)\\ \frac{d^5}{dx^5}e^{g(x)}=&\,\,e^{g(x)} g^{(5)}(x)+e^{g(x)} g'(x)^5+5 e^{g(x)} g^{(4)}(x) g'(x)\\ &+10 e^{g(x)} g^{(3)}(x) g''(x)+10 e^{g(x)} g^{(3)}(x) g'(x)^2+10 e^{g(x)} g'(x)^3 g''(x)\\ &+15 e^{g(x)} g'(x) g''(x)^2\\ \frac{d^6}{dx^6}e^{g(x)}=&\,\,e^{g(x)} g^{(6)}(x)+10 e^{g(x)} g^{(3)}(x)^2+15 e^{g(x)} g''(x)^3+e^{g(x)} g'(x)^6\\ &+6 e^{g(x)} g^{(5)}(x) g'(x)+15 e^{g(x)} g^{(4)}(x) g''(x)+15 e^{g(x)} g^{(4)}(x) g'(x)^2\\ &+20 e^{g(x)} g^{(3)}(x) g'(x)^3+15 e^{g(x)} g'(x)^4 g''(x)+45 e^{g(x)} g'(x)^2 g''(x)^2\\ &+60 e^{g(x)} g^{(3)}(x) g'(x) g''(x) \end{align*} \]

OpenStudy (kainui):

\[\Large \left[\begin{matrix}0 & 1 &0 & 0 \\ 0 & 0 &2 & 0 \\ 0 & 0 &0 & 3 \\ 0 & 0 &0 & 0 \end{matrix}\right] \left[\begin{matrix}1 \\x \\ x^2 \\ x^3\end{matrix}\right]=\left[\begin{matrix}1 \\x \\ x^2 \\ x^3\end{matrix}\right]\] Ok I don't really know how to make the notation correct to saywhat I mean but I have seen this before but never really used it. I think I need to switch these out with the basis and put the coefficients in here, but that's kind of difficult

OpenStudy (usukidoll):

omg the latex is high

OpenStudy (anonymous):

for triangle (A+B)^k |dw:1422785712510:dw|

OpenStudy (anonymous):

typo |dw:1422785945094:dw|

Can't find your answer? Make a FREE account and ask your own questions, OR help others and earn volunteer hours!

Join our real-time social learning platform and learn together with your friends!
Can't find your answer? Make a FREE account and ask your own questions, OR help others and earn volunteer hours!

Join our real-time social learning platform and learn together with your friends!