Thanks a lot sir, I was doing integrals that included integration by parts and I don't have the idea to solve but your video and examples made it clear and consice!!!!!!!
Thank you so much for uploading this video! We're doing some slight review in Calculus 2, and this really helped me - loved your explanation / walkthrough of the problem! Thank you so much for making quality content.
Little tip for ibp. Whrn you make your "list" what is u and what is dv don't write the dv next to the u. I do it like this i note my u differentiate it and write its derivative below it. My du comes right to that. Than i integrate dv to get v and note that above it just right to my u. That way it is a clean table with all values/functions you need and if you have the formula your brain can take a lunchbreak from here
not gonna lie, kinda confused about deriving the formula. the notation confuses me because even though u is in the integral, it looks like we're integrating with respect to a completely different variable, as noted by the differential dv. and in reality u and v are u(x) and v(x), so i'm confused as to what variable we are integrating with respect to when i see dv. my pea brain finds the traditional leibniz notation easier to understand
Integration by parts makes so much sense, but all the weird notation confuses me. Why dV instead of dV/dx. Isn't it ad hoc to cancel the dx from the integral and derivative? I would want to avoid ad hoc methods as much as possible personally
At 8:56 he mentions something that would be related to your question. You would do Integration by Parts within Integration by Parts. Your "u" or your "v" would actually be a function that is a product of functions. If you are lucky, then you can split it and find something easy to work. If you are not lucky, then you do Integration by Parts multiple time. Sometimes, with Trigonometric functions, you still do the process multiple times and then either cancel or substitute a previous iteration to simplify.
I love you sir! is it possible if you can do some videos on IMO or maybe AMC, just an enquiry. A pretty sure a numbe rof people will be greatful me inclusive. You sir are a genius!!!!
choosing 'u' function by 'ILATE' rule I - inverse functions L - lograthmic functions A- algebraic functions T - trigonometric functions E - exponential functions than apply DI method. plz suggest me any integral calculus book
Use integration by parts. It's a little tricky, because you have to consider ln(x) as 1*ln(x), so that you choose ln(x) to derive and 1 to integrate. It blew my mind when first saw this, because in the end you end up just multiplying x (the integral of 1) by 1/x (the derivative of ln(x)) and everything cancels out beautifully
If you want to use the DI IBP method, that’s fine, the order for choosing what function to differentiate and integrate is by using the LIATE method, which stands for logs, inverse trig, algebraic, trig, and exponential. That is the word “detail” spelled backwards without the D. Here, ln(x) is the part where you differentiate it. The 1 is algebraic so that’s where you integrate it. Notice that you draw diagonally from ln(x) to x to multiply it, which is xln(x). Then, stop here at that point and draw horizontally to get that integral of a product, which is 1/x*x, or 1. The integral of 1 is x. Combining all of these together gives you xln(x)-x+C.
How about clearing up the apparent ambiguity that arises using this method when introducing the bounds of integration,i. e, the definite integral. Now that would be helpful. No offence, love the videos.
This is a topic where you get to the non-elementary function here, which is called error function denoted as erf(x). You won’t find this on your graphing calculator but can be typed it in as fnInt(function,x,lower limit of integration,upper limit of integration).
could you please solve this integration problem i had on an exam : ( e^x - 3)^2 --> it's exponential of x minus 3 in parathenis and squarred on the outside. I got stuck bc of e^x^2 and couldn't do it! could you try?
It's easy: (e^x-3)^2=( (e^x)^2=e^(2x) )-6*e^x+9 which we then integrate term by term; i.e. for the first term we substitute u=2x => du=2dx => dx=0.5du => e^(2x)dx=0.5(e^u)du=d(0.5*e^u + C1)=d(0.5*e^(2x)+C1) and the other two I won't need to explain, I think.
Isn't ((e^x)-3)^2 just ((e^x)-3) *((e^x)-3)? That is, int{e^(2x) +-6(e^x)+9}dx, which is (1/2)e^(2x) +6e^x+9x+C. I have a bit of a trick for you with the e^ax family of functions, actually: try differentiating e^ax and you get lim h->0 (e^(ax+ah) - e^(ax))/h =>1/h( e^(ax) *(e^(ah)-1), so multiply that e^(ax) out of the limit and you have e^ax * lim h->0 (e^(ah)-1)/h. If you recall from the derivative of a^x = a^x *ln(a), the limit as h->0 for (h-1)/h is ln(h), so you have e^(ax) * e^ln(a), which is just e^(ax) * a. Thus, for all functions in the family e^ax, their derivative is ae^(ax). So, if you want to undo that by integrating it, the integral for all e^(ax) must be (1/a)e^(ax)+C.
@A Prkr - I think you made some typo's: several times you wrote e^x when e^2x or e^ax was meant. (At least in "1/2e^x +6e^x+9x+C" and in "for all functions in the family e^ax, their derivative is ae^x". Could you please correct? @Divine Ligs: e^x squared is indeed e^(x*2), or e^(2*x) because multiplication is mutative, but that is *not* equal to e^x*x. The latter can be interpretated as x*e^x (
Given: integral dx/[x^2 * sqrt(4 + x^2)] Factor out a 4, from the square root function: sqrt(4 + x^2) = sqrt(4*(1 + (x/2)^2) = 2*sqrt(1 + (x/2)^2) let x/2 = tan(u) thus dx = 2*sec(u)^2 du Reconstruct in u-world: integral 2*sec(u)^2 / [2*(2*tan(u))^2 * sqrt(1 + tan(u)^2)] du = integral sec(u)^2 / [4*tan(u)^2 * sqrt(1 + tan(u)^2)] du Use trig identity: 1 + tan(u)^2 = sec(u)^2 Thus: integral sec(u)^2 / [4*tan(u)^2 * sqrt(sec(u)^2)] du integral sec(u)^2 / [4*tan(u)^2 * sec(u)] du integral sec(u) / [4*tan(u)^2 ] du Recall sec(u) = 1/cos(u), and tan(u) = sin(u)/cos(u). This means: sec(u)/tan(u)^2 = 1/cos(u) / [sin(u)/cos(u)]^2 = cos(u)^2/[sin(u)^2 * cos(u)] = cos(u)/sin(u)^2 The integral now becomes: 1/4 * integral cos(u)/sin(u)^2 du Let w = sin(u) Thus dw = cos(u) du Rewrite in w-world: 1/4*integral dw/w^2 = -1/4*1/w Translate back to u-world, then x-world: -1/(4*w) = -1/(4*sin(u)) = -1/(4*sin(arctan(x/2)) + C sin(arctan(x/2)) = x/sqrt(x^2 + 4) Solution: -sqrt(x^2 + 4)/(4*x) + C
Because the different families of functions, have different ways they relate to each other, and some are easier to integrate than others. Generally, logs and inverses are the hardest to integrate, since even integrating the primitive relies on integration by parts. While exponentials and the simplest of trig functions, are the easiest to integrate, since they remain at the same complexity. Algebraic functions could go either way, depending on what else is in the integrand. Exponentials and simple trig, have the property where they eventually loop as constant multiples of themselves, when integrated repeatedly. They don't get any more or less complicated, as the process continues. When integrated with a polynomial function (which is algebraic) out in front, the algebraic function eventually differentiates to zero, while the exponential or trig continues looping. This is the basis for the enders, where you differentiate the D-column to zero. When exponentials and trig are together, they both loop, and eventually produce a constant multiple of the original integral. After getting to the row where that happens, you can assign the letter I to be the original integral, and construct an IBP result that you can algebraically solve for I. This works as long as you don't produce a coefficient of +1 on the original integral, because then you'll end up dividing by zero, when trying to solve for I. Normally, you produce a negative coefficient when this happens. It turns out, that exponentials and trig are ultimately the same family of functions, as complex numbers reveal to us, which is why they both have this property. When inverse trig and logs are differentiated, they become algebraic. These end up being functions that you can regroup with an algebraic function after integration in the I-column, and find another method to integrate. So you get regrouper stops for IBP, with either a log or inverse trig, being integrated with an algebraic function in the original integrand.
When you integrate, you're multiplying everything by dx. Suppose the answer we're looking for, will ultimately end up equaling f(x), which can be differentiated by the product rule, to become a product of two functions, u & v. Take the derivative of u*v, with the product rule: d/dx u*v = u*dv/dx + v*du/dx Integrate every term by dx: integral d/dx (u*v) dx = integral u*dv + integral v*du Thus: u*v = integral u dv + integral v du Solve for integral u dv: integral u dv = u*v - integral v du Exactly what we were setting out to show.
If u and v are functions of x then (uv)dx is pretty much the same thing in different notation. If u or v isn't a function of x, then that's even better... Just factor the entire thing out, since it doesn't even depend on x
From what I know, this is one of the definition of the number e, and when the time came to evaluate this limit, the mathematician do did it, which I don't remember who was, made a numerical approximation. Basically, there's no way to prove why the limit is e, because it's a definition. I could be wrong though
So, what you can do here is to break this up into two integration by parts problems and then come back to solve them. First, treat int{((x^x)(x^(2x+1))}dx as its own f(x) where you have f1(x) , f1'(x), f2(x), and f2'(x) (you choose f1(x) and f2'(x) as specified in the video above) and solve for that integral. Then treat (ln(x+1)/((x^(4x+1)) the same way as g1(x) , g1'(x), g2(x), and g2'(x) (just treat the division sign as a multiplication of (1/(x^(4x+1)), and integrate that with u-substition with u= (x^(4x+1)). Then, having found F(x) = int{f1'(x)f2(x)}dx = (f1(x)f2(x) - int{f1(x)f2'(x)}dx, and G(x) = int{g1'(x)g2(x)}dx = (g1(x)g2(x) - int{g1(x)g2'(x)}dx, you integrate by parts again with F(x) and G(x) to complete the process. Also, please close your exponents next time, as e^(x+1) != e^(x) +1 and e^x+1 is ambiguous.
Bprp looking suave whilst teaching calculus
Showed this to my cat
My cat got an A+
Thank you sir
So it was your cat in the meme?
Finally you show us where the DI method comes from, I don't think you've ever shown us before
Thanks a lot sir, I was doing integrals that included integration by parts and I don't have the idea to solve but your video and examples made it clear and consice!!!!!!!
The process you are using is takes a lot of unnecessary complications away from solving these problems. It is very well organized; great job.
SEAN
Thank you so much for uploading this video! We're doing some slight review in Calculus 2, and this really helped me - loved your explanation / walkthrough of the problem! Thank you so much for making quality content.
Im learning integrals in my free time and these videos help a lot. Great work
অসাধারণ, এমনি শিক্ষক চাই। খুব ভালো।
शिक्षक हो तो इसतरा साबास। शुभकामनाएं।
Excellent, like your teaching . Good luck.
I actually know these. I just like watching this guy. He's fun.
The explanation as undoing the power rule is brilliant. It's like solving my biggest mystery in my like with 2 sentence. LOLOL
Great video. You are really good in teaching.
Awesome video! I could see you really wanted to do the DI Method.. Haha!
Just starting Calc 2 this week. Thanks!
#YAY I am starting to learn Calculus next week but thanks to your videos I am so far ahead that I know all of the content for the first year of school
Estos vídeos llegaron a mí como un milagro, ya que este semestre llevo ecuaciones diferenciales y necesito repasar integrales 😂😂
This man saving my whole career
Always a good review. Will definitely need to remember this going into Math Physics
and it gets even easier with the table method what cant this man do
If u can make a dumbo like me understand something like this uk you’ve achieved sth great . THANK YOU!!
Thank you. Straight to the point and made it simpler
Sometimes clever choice of constant of integration while integrating dv will help to simplify the integrand vdu
it is a good video for explaining concepts sir, thank you . keep on moving with this type of videos
You are the best teacher dudes
Amazing calcul 💞👌.. Its so easy for every one
You’re literally the best, thank you
Little tip for ibp. Whrn you make your "list" what is u and what is dv don't write the dv next to the u. I do it like this i note my u differentiate it and write its derivative below it. My du comes right to that. Than i integrate dv to get v and note that above it just right to my u. That way it is a clean table with all values/functions you need and if you have the formula your brain can take a lunchbreak from here
Thank u sir...it's very clear now although I will have to practice alot
superb saying man,good do more vedios and get good fame.
I use LATE method to choose u and dv
L= logarithmic
A= algebraic
T= trigonometric
E= exponential
not gonna lie, kinda confused about deriving the formula. the notation confuses me because even though u is in the integral, it looks like we're integrating with respect to a completely different variable, as noted by the differential dv. and in reality u and v are u(x) and v(x), so i'm confused as to what variable we are integrating with respect to when i see dv. my pea brain finds the traditional leibniz notation easier to understand
A very clear explanation! Many thanks!
Integration by parts makes so much sense, but all the weird notation confuses me. Why dV instead of dV/dx. Isn't it ad hoc to cancel the dx from the integral and derivative? I would want to avoid ad hoc methods as much as possible personally
integral of lnxdx is xlnx -x + c
Wow... Good explanation!
syafa azra thanks
Great introduction!
you're my new god
You're amazing. Thanks steve!
Sir and what if three or more functions are multiplied then how to compute integrals?
At 8:56 he mentions something that would be related to your question. You would do Integration by Parts within Integration by Parts.
Your "u" or your "v" would actually be a function that is a product of functions. If you are lucky, then you can split it and find something easy to work. If you are not lucky, then you do Integration by Parts multiple time.
Sometimes, with Trigonometric functions, you still do the process multiple times and then either cancel or substitute a previous iteration to simplify.
So clear. So clear.
You are great man, thanks
Really well explained. Thank you
I love you sir!
is it possible if you can do some videos on IMO or maybe AMC, just an enquiry. A pretty sure a numbe rof people will be greatful me inclusive.
You sir are a genius!!!!
one day i saw a cow vestida de uniforme!!!!!
Awesome teacher
such a clear explanation :)
u r the best
choosing 'u' function by 'ILATE' rule
I - inverse functions
L - lograthmic functions
A- algebraic functions
T - trigonometric functions
E - exponential functions
than apply DI method.
plz suggest me any integral calculus book
Pugh
when you take the derivative of (u•v) why don’t you divide by dx ? isn’t the derivative of (u•v) = d(u•v)/dx ?
good one
Thank you very much!!!
Does it undo the Quo Chen Lu???
i feel like it would be a lot more intuitive if it were more commonly shown as it is at 2:00
Thank you BPRP
Thank you so much
Thanks you bro
Really thanks
the table method is so much easier to understand imo
Is it possible to Integrate ln(x) and still get an answer? I keep trying, but get stuck
Use integration by parts. It's a little tricky, because you have to consider ln(x) as 1*ln(x), so that you choose ln(x) to derive and 1 to integrate.
It blew my mind when first saw this, because in the end you end up just multiplying x (the integral of 1) by 1/x (the derivative of ln(x)) and everything cancels out beautifully
If you want to use the DI IBP method, that’s fine, the order for choosing what function to differentiate and integrate is by using the LIATE method, which stands for logs, inverse trig, algebraic, trig, and exponential. That is the word “detail” spelled backwards without the D. Here, ln(x) is the part where you differentiate it. The 1 is algebraic so that’s where you integrate it. Notice that you draw diagonally from ln(x) to x to multiply it, which is xln(x). Then, stop here at that point and draw horizontally to get that integral of a product, which is 1/x*x, or 1. The integral of 1 is x. Combining all of these together gives you xln(x)-x+C.
Thank you!!!
me : oh no, the udv method, i’m scared 😱
Great thanks
Thanks 🙏 I👍
well I'm not very familliar whith calculus but I want to know why at firste when you differanted (u.v) you didn't multiply u'v+v'u by dx
Excellent I like you from Turkey how can I contact you
How about clearing up the apparent ambiguity that arises using this method when introducing the bounds of integration,i. e, the definite integral. Now that would be helpful. No offence, love the videos.
hey!!! How can I solve this integral: from 0 to 2 by (a-x)*e^(-(1/2)*(x-a)^2) dx ; with "a" real number, help me please(erf function????)
This is a topic where you get to the non-elementary function here, which is called error function denoted as erf(x). You won’t find this on your graphing calculator but can be typed it in as fnInt(function,x,lower limit of integration,upper limit of integration).
You're amazing ! #YAY
I shared to my chicken lol
could you please solve this integration problem i had on an exam : ( e^x - 3)^2 --> it's exponential of x minus 3 in parathenis and squarred on the outside. I got stuck bc of e^x^2 and couldn't do it! could you try?
It's easy: (e^x-3)^2=( (e^x)^2=e^(2x) )-6*e^x+9 which we then integrate term by term; i.e. for the first term we substitute u=2x => du=2dx => dx=0.5du => e^(2x)dx=0.5(e^u)du=d(0.5*e^u + C1)=d(0.5*e^(2x)+C1) and the other two I won't need to explain, I think.
THE EXPOSENT OF ANOTHER EXPOSENT ARE MULTIPLIED. So e^x squarred is e^x2 which makes it hard to solve even when you let in the form e^x*x
Can you please write the whole question in mathematical form with bracket ...etc. may be i could help you.
Isn't ((e^x)-3)^2 just ((e^x)-3) *((e^x)-3)? That is, int{e^(2x) +-6(e^x)+9}dx, which is (1/2)e^(2x) +6e^x+9x+C. I have a bit of a trick for you with the e^ax family of functions, actually: try differentiating e^ax and you get lim h->0 (e^(ax+ah) - e^(ax))/h =>1/h( e^(ax) *(e^(ah)-1), so multiply that e^(ax) out of the limit and you have e^ax * lim h->0 (e^(ah)-1)/h. If you recall from the derivative of a^x = a^x *ln(a), the limit as h->0 for (h-1)/h is ln(h), so you have e^(ax) * e^ln(a), which is just e^(ax) * a. Thus, for all functions in the family e^ax, their derivative is ae^(ax). So, if you want to undo that by integrating it, the integral for all e^(ax) must be (1/a)e^(ax)+C.
@A Prkr - I think you made some typo's: several times you wrote e^x when e^2x or e^ax was meant. (At least in "1/2e^x +6e^x+9x+C" and in "for all functions in the family e^ax, their derivative is ae^x". Could you please correct?
@Divine Ligs: e^x squared is indeed e^(x*2), or e^(2*x) because multiplication is mutative, but that is *not* equal to e^x*x. The latter can be interpretated as x*e^x (
I Love U😂❤❤❤👌👌
Integral of Xe^x
The answer is xe^x-e^x+C.
nice from morocco
Please help solve the integral of
dx/x^2 sqrt (4+x^2)
Given:
integral dx/[x^2 * sqrt(4 + x^2)]
Factor out a 4, from the square root function:
sqrt(4 + x^2) = sqrt(4*(1 + (x/2)^2) = 2*sqrt(1 + (x/2)^2)
let x/2 = tan(u)
thus dx = 2*sec(u)^2 du
Reconstruct in u-world:
integral 2*sec(u)^2 / [2*(2*tan(u))^2 * sqrt(1 + tan(u)^2)] du =
integral sec(u)^2 / [4*tan(u)^2 * sqrt(1 + tan(u)^2)] du
Use trig identity:
1 + tan(u)^2 = sec(u)^2
Thus:
integral sec(u)^2 / [4*tan(u)^2 * sqrt(sec(u)^2)] du
integral sec(u)^2 / [4*tan(u)^2 * sec(u)] du
integral sec(u) / [4*tan(u)^2 ] du
Recall sec(u) = 1/cos(u), and tan(u) = sin(u)/cos(u). This means:
sec(u)/tan(u)^2 = 1/cos(u) / [sin(u)/cos(u)]^2 = cos(u)^2/[sin(u)^2 * cos(u)] = cos(u)/sin(u)^2
The integral now becomes:
1/4 * integral cos(u)/sin(u)^2 du
Let w = sin(u)
Thus dw = cos(u) du
Rewrite in w-world:
1/4*integral dw/w^2 = -1/4*1/w
Translate back to u-world, then x-world:
-1/(4*w) = -1/(4*sin(u)) =
-1/(4*sin(arctan(x/2)) + C
sin(arctan(x/2)) = x/sqrt(x^2 + 4)
Solution:
-sqrt(x^2 + 4)/(4*x) + C
Why do we use ILATE?
Because the different families of functions, have different ways they relate to each other, and some are easier to integrate than others.
Generally, logs and inverses are the hardest to integrate, since even integrating the primitive relies on integration by parts. While exponentials and the simplest of trig functions, are the easiest to integrate, since they remain at the same complexity. Algebraic functions could go either way, depending on what else is in the integrand.
Exponentials and simple trig, have the property where they eventually loop as constant multiples of themselves, when integrated repeatedly. They don't get any more or less complicated, as the process continues. When integrated with a polynomial function (which is algebraic) out in front, the algebraic function eventually differentiates to zero, while the exponential or trig continues looping. This is the basis for the enders, where you differentiate the D-column to zero.
When exponentials and trig are together, they both loop, and eventually produce a constant multiple of the original integral. After getting to the row where that happens, you can assign the letter I to be the original integral, and construct an IBP result that you can algebraically solve for I. This works as long as you don't produce a coefficient of +1 on the original integral, because then you'll end up dividing by zero, when trying to solve for I. Normally, you produce a negative coefficient when this happens. It turns out, that exponentials and trig are ultimately the same family of functions, as complex numbers reveal to us, which is why they both have this property.
When inverse trig and logs are differentiated, they become algebraic. These end up being functions that you can regroup with an algebraic function after integration in the I-column, and find another method to integrate. So you get regrouper stops for IBP, with either a log or inverse trig, being integrated with an algebraic function in the original integrand.
Look at you in a suit and all fancy!
Explain why you can just ignore all the "/dx"es?
When you integrate, you're multiplying everything by dx.
Suppose the answer we're looking for, will ultimately end up equaling f(x), which can be differentiated by the product rule, to become a product of two functions, u & v.
Take the derivative of u*v, with the product rule:
d/dx u*v = u*dv/dx + v*du/dx
Integrate every term by dx:
integral d/dx (u*v) dx = integral u*dv + integral v*du
Thus:
u*v = integral u dv + integral v du
Solve for integral u dv:
integral u dv = u*v - integral v du
Exactly what we were setting out to show.
#Yay! :D
What you told is "integration (u)dv "
But how to integrate this
"integration (u.v)dx"
Did you watch the rest of the video? You get to choose u and dv, then you need to compute du and v.
If u and v are functions of x then (uv)dx is pretty much the same thing in different notation.
If u or v isn't a function of x, then that's even better... Just factor the entire thing out, since it doesn't even depend on x
@@RitobanRoyChowdhury
Thank u for your reply ...
But still I couldn't get it
why no DI method 😞
It works, but he is just showing you how to use the up IBP method.
#YAY
DI method.
question:
how to prove the limit:
lim((1+(1/n))^n=e
when n->infinity
From what I know, this is one of the definition of the number e, and when the time came to evaluate this limit, the mathematician do did it, which I don't remember who was, made a numerical approximation. Basically, there's no way to prove why the limit is e, because it's a definition. I could be wrong though
That is the def of e but if we take that as the def of e we should first prove it convergent first but i dont know how
DI y :)
Un día vi una vaca sin cola vestida de uniforme 🐌
There is other way for solution
Help!!!! How to integrate (x^x(x^2x + 1)(lnx + 1))/(x^4x + 1) dx...
#yay
Let u =x^x and use partial fraction
So, what you can do here is to break this up into two integration by parts problems and then come back to solve them. First, treat int{((x^x)(x^(2x+1))}dx as its own f(x) where you have f1(x) , f1'(x), f2(x), and f2'(x) (you choose f1(x) and f2'(x) as specified in the video above) and solve for that integral. Then treat (ln(x+1)/((x^(4x+1)) the same way as g1(x) , g1'(x), g2(x), and g2'(x) (just treat the division sign as a multiplication of (1/(x^(4x+1)), and integrate that with u-substition with u= (x^(4x+1)). Then, having found F(x) = int{f1'(x)f2(x)}dx = (f1(x)f2(x) - int{f1(x)f2'(x)}dx, and G(x) = int{g1'(x)g2(x)}dx = (g1(x)g2(x) - int{g1(x)g2'(x)}dx, you integrate by parts again with F(x) and G(x) to complete the process. Also, please close your exponents next time, as e^(x+1) != e^(x) +1 and e^x+1 is ambiguous.
you are very handsome
I shared this video with my dog. he threw up. probably not good for dogs … beware.
I'll keep it in mind
Thanks so much !!!
#YAY
#YAY
#YAY
#YAY
#YAY
#YAY