Limit , will you use L'H rule ?

Calculus Level 5

f ( x ) = cos 1 ( cos x ) g ( x ) = { x } { x } 2 \displaystyle{f\left( x \right) =\cos ^{ -1 }{ \left( \cos { x } \right) } \\ g\left( x \right)=\sqrt { \left\{ x \right\} -{ \left\{ x \right\} }^{ 2 } } \\ }

For f : R R f: \mathbb{R}\rightarrow \mathbb{R} and g : R R g:\mathbb{R}\rightarrow \mathbb{R}

Find lim x 0 x f ( t ) d t 0 x g ( t ) d t \displaystyle{\lim _{ x\rightarrow \infty }{ \frac { \displaystyle \int _{ 0 }^{ x }{ f\left( t \right) \ \mathrm dt } }{ \displaystyle \int _{ 0 }^{ x }{ g\left( t \right) \ \mathrm dt } } } }

3 2 7 5 6 1 4 8

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

2 solutions

Wow! beautiful problem! We will use finding the area method to do this problem.

Observe that both the numerator and denominator tend to \infty . Also, f ( x ) f(x) is periodic with period 2 π 2\pi and g ( x ) g(x) is periodic with period 1 1 .

This means that we can partition the integrals into many parts of known area and add them up.

Let x = 2 m π x=2m\pi where m m is a positive integer.

0 x f ( t ) d t = m 0 2 π f ( t ) d t = m π 2 \Rightarrow \int _{ 0 }^{ x }{ f\left( t \right) dt } =m\int _{ 0 }^{ 2\pi }{ f\left( t \right) dt } =m{ \pi }^{ 2 } (easily obtained from graph)

Now,

0 x g ( t ) d t = x 0 1 g ( t ) d t = x 0 1 t 1 t d t = 2 x 0 1 / 2 t 1 t d t = x 0 1 / 2 1 4 t 2 d t = x π 8 \Rightarrow \int _{ 0 }^{ x }{ g\left( t \right) dt } =x\int _{ 0 }^{ 1 }{ g\left( t \right) dt } =x\int _{ 0 }^{ 1 }{ \sqrt { t } \sqrt { 1-t } dt } \\ =2x\int _{ 0 }^{ 1/2 }{ \sqrt { t } \sqrt { 1-t } dt } \\ =x\int _{ 0 }^{ 1/2 }{ \sqrt { 1-4{ t }^{ 2 } } dt } =x\frac { \pi }{ 8 }

Thus:

lim x 0 x f ( t ) d t 0 x g ( t ) d t = 8 m π 2 x π { \lim _{ x\rightarrow \infty }{ \frac { \int _{ 0 }^{ x }{ f\left( t \right) dt } }{ \int _{ 0 }^{ x }{ g\left( t \right) dt } } } }=\frac { 8m{ \pi }^{ 2 } }{ x\pi }

= 4 =\boxed{4} since x = 2 m π x=2m\pi

Note: π \pi is an irrational number hence, x = 2 m π x=2m\pi is technically incorrect, but it doesn't matter in this case as we are making a small approximation with respect to huge terms.

This is a really interesting question: Using the L'Hospital rule actually makes the limit harder to evaluate!

In fact, can someone post a solution that actually does use the L'Hospital rule, just for variety?

Shashwat Shukla - 6 years, 3 months ago

Log in to reply

I think using L-H rule will be scary ! isn't it ? :)

Deepanshu Gupta - 6 years, 3 months ago

Log in to reply

Definitely...But think about this: What if the question was to find the limit you get a f t e r after differentiating using L'Hospital?

Can we say that it's value is equal to the limit we found for this question? i.e: Is the converse of L'Hospital rule a valid line of approach? If so, under what conditions can it be applied?

@Deepanshu Gupta @Mvs Saketh @Raghav Vaidyanathan

Shashwat Shukla - 6 years, 3 months ago

Log in to reply

@Shashwat Shukla Yes I understand what are you saying , But truly speaking this question is given to me by my friend , and my first approach is L-H rule , and I try my best to calculate with L-H , But fails So Finally I also Use area method .

But If you still want , then we can discuss on this! Did you have any ideas ? @Shashwat Shukla @Mvs Saketh @Raghav Vaidyanathan

Deepanshu Gupta - 6 years, 2 months ago

Log in to reply

@Deepanshu Gupta Thank you for starting a discussion :)

Shashwat Shukla - 6 years, 2 months ago

@Deepanshu Gupta I do not think it is possible to do this via L'H. There are some rules for applying L'H, and as long as we don't know what they are, we cannot say anything.

Raghav Vaidyanathan - 6 years, 2 months ago

Log in to reply

@Raghav Vaidyanathan The rules being? Is it not just differentiability of the numerator and denominator?

Shashwat Shukla - 6 years, 2 months ago

Log in to reply

@Shashwat Shukla Wherever we can taylor expand, we can use L hospital, and here, in general it is not a continuous or differentiable at all points, (example, arccos(cos(x)) is not differentiable at the peak points , eg x=pi) and such peak points repeat themselves indefinitely, we cant even differentiate it at the lower points like x=2pi etc, and since in our approximations we assumed complete number of cycles (as raghav took x=2pim , and i neglected fractional part) , so i think we cannot use LH here as the conventional value of x we took was the point where function is not differentiable

Mvs Saketh - 6 years, 2 months ago

Log in to reply

@Mvs Saketh That makes sense. Thanks :)

Shashwat Shukla - 6 years, 2 months ago

@Mvs Saketh Thanks , for Explaining !

Deepanshu Gupta - 6 years, 2 months ago

@Mvs Saketh I agree that L'H fails in this case, but there's more to it than f not being smooth enough: we could alter the problem, making f (and g) as smooth as you like, and still have LH fail.

The deeper reason is that one of the requirements of LH (at least the form I have in mind) is precisely that the RHS limit must exist. In other words, L'H admits of cases where the left-hand side limit exists (e.g. our answer of "4"), but the RHS limits does not exist (e.g. undampened oscillation), which is just what we have here.

Peter Byers - 6 years, 2 months ago

Yes That is the intended way! Good +1 :)
I want to ask one thing have you guys not given INPHO and INCHO ? @Raghav Vaidyanathan @Shashwat Shukla

Deepanshu Gupta - 6 years, 3 months ago

Log in to reply

I did, and I qualified NSEP(not on merit but proportional representation clause). The exam was in Bangalore and I felt that I would not get it anyway. So I did not write INPhO.

Raghav Vaidyanathan - 6 years, 3 months ago

Log in to reply

Ohh ! I know it is very difficult to comprise and mannage time for olyimpiads and JEE , But still Congratulation for NSEP , May be Something better is waiting for you :) ! Actually I'am not eligible for olyimpiads this year , Since I passed 12th last year , So I feel bad after thinking that I can't give them this year ! But I feel good after thinking that may be I got Some other opportunities to Learn Higher studies , very soon....! And always hoping for Best :)

Deepanshu Gupta - 6 years, 2 months ago

No, I didn't give them...Mainly because our coaching teachers won't allow me to go for the camp and my parents are also not very supportive of other endeavors right before JEE.

Rather sad, that :/

Shashwat Shukla - 6 years, 3 months ago

Log in to reply

Ohh , bad luck , ! But don't worry brother you will surly get such lot more opportunities in your upcoming future ! You will then surly rock ! Good Luck Bro !

Deepanshu Gupta - 6 years, 2 months ago

Oh, bro but camp is after JEE advanced and no one can stop you then :)

Mvs Saketh - 6 years, 3 months ago

Log in to reply

@Mvs Saketh I didn't even know that while I could reply...Only came to know after I read up online about the schedule. regret

But anyways, you can rock the show for everyone who didn't make it (whatever be the reasons). Hope you have a blast! :D

Shashwat Shukla - 6 years, 3 months ago

Log in to reply

@Shashwat Shukla Ha ha, yes, i just wanted to get to the camp because i have heard legends that they teach really awesome stuff there :) also i heard that one is admitted (if he/she wants) into the department of atomic energy and CPI (chennai physics institute) directly if he qualifies for camp, is that true? any idea

Mvs Saketh - 6 years, 3 months ago

Log in to reply

@Mvs Saketh Not sure if you get a certificate guaranteeing admissions but a senior of mine did get a placement at the Homi Bhaba institute for research based s o l e l y solely (according to him)on this criteria.

At worst, it'll be like what KVPY is to IISc. Speaking of which, have you made up your mind in deciding between the IITs and IISc?

Shashwat Shukla - 6 years, 3 months ago

Log in to reply

@Shashwat Shukla Definitely IISC if i get into it, while i expect myself to be able to get in, it need not necessarily match reality and hence i have also opened other areas of admission like HBCSE, DAE and NISER,

But yours is almost guranteed as you cleared KVPY, much better for you :)

Mvs Saketh - 6 years, 3 months ago

Log in to reply

@Mvs Saketh The thing is, everywhere I turn, the one thing I get to hear about research is:

a)It takes way too long and things can go downhill anytime b)There aren't enough jobs c)The jobs don't pay well

While c) is of no concern to me as long as I have a roof over my head and a 'few' comforts, it's all good.

But b) is really worrying. Do you have a different take on it?

Shashwat Shukla - 6 years, 3 months ago

Log in to reply

@Shashwat Shukla Yes, i have similar takes on (C), as for (B), i dont think so bro, that applies to other BSc colleges perhaps, (where people go when they get nothing else), but not for these Reputed ones, like i think we will have a decent life provided we become a teacher at these institutes, Yes, we will have to work harder than most engineers, but then, No one said it was easy, but it is worth it and i am sure i will have no trouble finding a job atleast :)

Mvs Saketh - 6 years, 3 months ago

Log in to reply

@Mvs Saketh Wow...that is precisely what I also had in mind regarding the reputed institutes. Good to know you feel that way too :) Hope I do get to meet you in person. It'd be cool if we go to the same college :D

Shashwat Shukla - 6 years, 3 months ago

Log in to reply

@Shashwat Shukla Definitely, it will be :)

Mvs Saketh - 6 years, 3 months ago

@Raghav Vaidyanathan can you guide me how did you clear NSEP. I just passed 11th and going to 12th. How much time do I need to devote without hindering my JEE preparation?

Aditya Kumar - 6 years, 3 months ago

Log in to reply

Ask @Mvs Saketh . I didnt prepare at all for NSEP

Raghav Vaidyanathan - 6 years, 2 months ago

You have a typo (x instead of t in second integration.)

Ayush Verma - 5 years, 11 months ago

Log in to reply

corrected!

Raghav Vaidyanathan - 5 years, 11 months ago
Mvs Saketh
Mar 15, 2015

Nice problem bro,

Well, for large, 'x' , we can neglect the fractional difference in area between the actual area and the area of the nearest complete loop

(example- if x = 3 π 2 \frac {3 \pi }{2} , then area is π 2 + π 2 2 {\pi }^{2} + \frac {{\pi }^{2}}{2}

)

in general it is

[ x 2 π ] π 2 + ( s o m e f r a c t i o n o f π 2 ) [ x 2 π ] π 2 x 2 π π 2 = x 2 π \left[ \frac { x }{ 2\pi } \right] { \pi }^{ 2 }+(some\quad fraction\quad of\quad { \pi }^{ 2 })\simeq \left[ \frac { x }{ 2\pi } \right] { \pi }^{ 2 }\simeq \frac { x }{ 2\pi } { \pi }^{ 2 }=\frac { x }{ 2 } { \pi } for large 'x'

similarly the other function is

x π 8 + ( s o m e f r a c t i o n o f π 8 ) x 8 π \left\lfloor x \right\rfloor \frac { \pi }{ 8 } +(some\quad fraction\quad of\quad \frac { \pi }{ 8 } )\simeq \frac { x }{ 8 } { \pi } for large 'x'

dividing both we get 4

Nice approximations ! And bro Heartiest congo to you for INPHO , you have thunder at INPHO , and reallyYou Deserve that! :)

Deepanshu Gupta - 6 years, 3 months ago

Log in to reply

Thanks bro, thankyou very much :)

Mvs Saketh - 6 years, 3 months ago

Log in to reply

Just got to know that you qualified INPHO! Congrats man! You be destined for it!

Raghav Vaidyanathan - 6 years, 3 months ago

Congratulations!!

Shashwat Shukla - 6 years, 3 months ago

Congrats!!

Aditya Kumar - 6 years, 3 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...