Ur, another integration-by-parts problem! I have g(a) = the integral from -a to a of e^((-x^2)/9) (x^4)dx. I know two things: I need to integrate twice, and the answer will model the erf(q) equation.
But what I don't get is how to split up the variables to do my intergration-by-parts. Is it u= x^4, dv= e^((-x^2)/2), or vice versa?
either way, I have been struggling with this and not getting a satisfactory answer
Definitely make u equal to the x^4 ! Use the acronym LIATE to figure out what should be u, so that way you don't always guess it out. L - logarithmic I - inverse trig A - algebraic T - trig E - exponential So since algebraic comes before exponential, use the x^4 as your u.
I think we can make use of the same trick we did last time since this function is even
yes we can, we just have to do it twice
Hi guys, just left for a quick sec
So, I was looking at that: can I split the x^4 up? I was trying, and making two intergrations at the moment
but I get\[\int\limits_{-a}^{a}w^2 *e^-w/9\]?
I'm guessing I can do u = w^2, du = 2w; dv = e^(-w/9), v = e^(-w/9) at this point?
that's what I'm trying right now...
\[\Large\int_{-a}^ax^4e^{-\frac{x^2}9}dx=\int_0^ax^3(2xe^{-\frac{x^2}9})dx\]\[u=x^3\implies du=3x^2dx\]\[dv=2xe^{-\frac{x^2}9}dx\implies v=-9e^{-\frac{x^2}9}\]\[\int_0^ax^3(2xe^{-\frac{x^2}9})dx=-9x^3e^{-\frac{x^2}9}+27\int_0^ax^2e^{-\frac{x^2}9}dx\]\[=-9x^3e^{-\frac{x^2}9}+27\int_0^ax(xe^{-\frac{x^2}9})dx\]\[u=x\implies du=dx\]\[dv=xe^{-\frac{x^2}9}dx\implies v=-\frac92e^{-\frac{x^2}9}\]so the integral is now\[-9x^3e^{-\frac{x^2}9}+27(-\frac92xe^{-\frac{x^2}9}+\frac92\int e^{-\frac{x^2}9}dx)\]crap I got stuck...
wait, that's fine I think: I need to integrate twice, and then write it in terms of the erf(q) function I mentioned above. I have similar work like yours, up to a point...
oh you can use the error function ?! I missed that the last integral is the error function
so then my answer is right :)
Wait, so this is how I reasoned the first substution: did you do something like w = x^2, dw = 2xdx? and then so it would be the integeral 2xwe^-wdx before you did intergration by parts?
no I didn't think like that I used the even function thing to get it into the form I have on the first line then did u=x^3, dv=2xe^(-x^2/9)
hmm..
I guess I'm having trouble intergrating e^((-x^2)/9) once I try to do the first intergration by parts
\[\Large\int_{-a}^ax^4e^{-\frac{x^2}9}dx=\int_0^ax^3(2xe^{-\frac{x^2}9})dx\]\[u=x^3\implies du=3x^2dx\]\[dv=2xe^{-\frac{x^2}9}dx\implies v=-9e^{-\frac{x^2}9}\]\[\int_0^ax^3(2xe^{-\frac{x^2}9})dx=-9a^3e^{-\frac{a^2}9}+27\int_0^ax^2e^{-\frac{x^2}9}dx\]\[=-9a^3e^{-\frac{a^2}9}+27\int_0^ax(xe^{-\frac{x^2}9})dx\]\[u=x\implies du=dx\]\[dv=xe^{-\frac{x^2}9}dx\implies v=-\frac92e^{-\frac{x^2}9}\]so the integral is now\[-9a^3e^{-\frac{a^2}9}+27(-\frac92ae^{-\frac{a^2}9}+\frac92\int_0^a e^{-\frac{x^2}9}dx)\]\[-9a^3e^{-\frac{a^2}9}+27\left(-\frac92ae^{-\frac{a^2}9}+\frac{27}4\sqrt\pi\text{erf}(\frac a3)\right)\]I'll let you simplify that ;)
okay your problem is that you \(cannot\) integrate\[\int e^{-\frac{x^2}9}dx\]in terms of traditional functions, the answer to the integral is\[\int e^{-\frac{x^2}9}dx=\frac32\sqrt\pi\text{erf}(\frac x3)\]so that is where you are supposed to be stuck :P
right; so I am having issues with doing intergration by parts while being stuck there.
It seems that you don't try to intergrate it at all, you just kind of go with it
you can prove that integral is what it is if you know a bit about the gamma function and laplace transforms
But you still take the intergal of dv...how did you do the integral of 2xe^((-x^2)/9)?
let \[u=e^{-\frac{x^2}9}\implies du=-\frac29xe^{-\frac{x^2}9}dx\]integral becomes\[-9\int du\]
gotta do those simple u-subs in your head as you move through calculus
yea, I get that so far
hold on
\[\int\limits_{}^{} e^(x^2/9) dx\] = ? (2x/9)*e^((-x^2)/9)?
I don't think it's as easy as that---or I'm doing something wrong
wait, you don't intergrate/derive the x^2 anymore, only the 1/9 in your work
I am confused what you mean ...btw to keep the exponents where they are supposed to be enclose them in {} not in ()
as I said before you cannot integrate e^(-x^2/9)dx, but you can integrate 2xe^(-x^2/9)dx the 2x outside allows us to do a simple substitution
oh!
the trick is realizing how similar 2xe^(-x^2/9) is to the derivative of e^(-x^2/9)
I think it clicked, yes?
I'll brb...
I got it! Thanks so much! It finally clicked!
I'm glad, I think I will celebrate this rather difficult integration by parts problem with a beer see ya!
see yah!
The main trick is to know when to separate the powers of x to fit your purposes, I guess. This will probably appear on an exam.
Join our real-time social learning platform and learn together with your friends!