The Student Room Group

Another math problem. integration

The question goes like this: y = f(x) is a decreasing function in the first quadrant, that passes through (0, 1).For any point P on f(x), the distance between P and the x-intercept of the tangent at P, is 1. Evaluate integration of f(x)dx from 0 to infinity.

I have worked out the relationship between y and x by integration according to the info that the length is 1. But it looks rather complicated (i do not know how to type it out but it is ln(a function of y)+square root of another function of y equals to x) which I have no idea how to further integrate it from 0 to infinity.

Reply 1

Original post
by u4ria
The question goes like this: y = f(x) is a decreasing function in the first quadrant, that passes through (0, 1).For any point P on f(x), the distance between P and the x-intercept of the tangent at P, is 1. Evaluate integration of f(x)dx from 0 to infinity.
I have worked out the relationship between y and x by integration according to the info that the length is 1. But it looks rather complicated (i do not know how to type it out but it is ln(a function of y)+square root of another function of y equals to x) which I have no idea how to further integrate it from 0 to infinity.
IMG_4345.jpeg
Here is a picture of what i have calculated and my problem is how to integrate it further.

Reply 2

Original post
by u4ria
IMG_4345.jpeg
Here is a picture of what i have calculated and my problem is how to integrate it further.

Where is the question from (and the other one)?

Reply 3

Have you used integration by substitution?

Reply 4

Original post
by mqb2766
Where is the question from (and the other one)?

Sorry for replying late. I do not know. I have after school lessons on math and that teacher gave it to me.

Reply 5

You mean which step? Because i have used one integration by substitution (using trigonometry) and got the equation in the pic. But now i have to integrate the y from 0 to infinity. Is that i apply integration by substitution in this step? How that could be?

Reply 6

Original post
by u4ria
Sorry for replying late. I do not know. I have after school lessons on math and that teacher gave it to me.

Wbf, it almost sounds like some sort of competition type question.

But you could flip the integration problem / variables here so integrate wrt y and its not too bad so a part circle and some arctrig-type terms.

Reply 7

Original post
by mqb2766
Wbf, it almost sounds like some sort of competition type question.
But you could flip the integration problem / variables here so integrate wrt y and its not too bad so a part circle and some arctrig-type terms.

Thanks. I think i worked in a complicated way before😓.

Reply 8

Original post
by mqb2766
Wbf, it almost sounds like some sort of competition type question.
But you could flip the integration problem / variables here so integrate wrt y and its not too bad so a part circle and some arctrig-type terms.

Ohno! I met some problems. I am not sure about the upper limit. The lower limit is given in the question that when x=0, y=1. But for the upper limit, i think that the derivative of y wrt x is a function of y, which is also a function of x. Since there is a negative sign in the derivative, as x increases, y will decreases. So as x goes to infinity, y goes to 0. Is that right? And in the question it mentions that in the first quadrant the function is decreasing. So here when i integrate, do i still consider the function is in the first quadrant?
Here is a picture of what i have worked out now.
IMG_4354.jpeg
(edited 11 months ago)

Reply 9

Original post
by u4ria
Ohno! I met some problems. I am not sure about the upper limit. The lower limit is given in the question that when x=0, y=1. But for the upper limit, i think that the derivative of y wrt x is a function of y, which is also a function of x. Since there is a negative sign in the derivative, as x increases, y will decreases. So as x goes to infinity, y goes to 0. Is that right? And in the question it mentions that in the first quadrant the function is decreasing. So here when i integrate, do i still consider the function is in the first quadrant?
Here os a picture of what i have worked out now.IMG_4354.jpeg

Its kind of hard to just see some of your working, but the suggestion about flipping the integration variables was based on

assume that x = ln((1+sqrt(1-y^2))/y) - sqrt(1-y^2) is correctf

noting that (0,1) is a point on the curve and were in quadrant 1. Also y->0 as x->inf. So
https://www.desmos.com/calculator/j56uofeoqq


So it doesnt matter whenther you consider the usual vertical riemann strips and integrate y(x) wrt x from x = 0 to inf, or consider horizontal reimann strips and integrate x(y) wrt y from y = 0 to 1. The integral isnt that bad and the ans is fairly simple.

But all this really is beyond a level and youre only posting snippets and saying its given by a teacher outside school sounds strange.

Reply 10

Original post
by mqb2766
Its kind of hard to just see some of your working, but the suggestion about flipping the integration variables was based on

assume that x = ln((1+sqrt(1-y^2))/y) - sqrt(1-y^2) is correctf

noting that (0,1) is a point on the curve and were in quadrant 1. Also y->0 as x->inf. So
https://www.desmos.com/calculator/j56uofeoqq


So it doesnt matter whenther you consider the usual vertical riemann strips and integrate y(x) wrt x from x = 0 to inf, or consider horizontal reimann strips and integrate x(y) wrt y from y = 0 to 1. The integral isnt that bad and the ans is fairly simple.
But all this really is beyond a level and youre only posting snippets and saying its given by a teacher outside school sounds strange.

Thanks. This question is really hard for me. My teacher may not expect me to work out the right answer. He just wants to give me some challenges and see what my idea is. And let me discuss it with other students. Actually there are a few weird questions he gave to us and i’m not sure.. just to let us think deeper i guess.

Reply 11

Original post
by u4ria
Thanks. This question is really hard for me. My teacher may not expect me to work out the right answer. He just wants to give me some challenges and see what my idea is. And let me discuss it with other students. Actually there are a few weird questions he gave to us and i’m not sure.. just to let us think deeper i guess.

Id check with your teacher about asking for hints from others.

Reply 12

Original post
by mqb2766
Id check with your teacher about asking for hints from others.

of course my teacher said so. He encourages us to discuss with others.

Reply 13

Guys. I listened to amazing solutions of this question from my classmate and my teacher.
The student thought it as an integration similar as what we do for y=f(theta) in polar coordinates . As we can observe that the y-intercept is 1 and the P could be any points on the curve, also (0,1). And at (0,1) the length happened to be one. So we can think of how the curve is drawn, which is like the length 1 is keep tipping and also as a tangent (the left graph shows this idea). Eventually it becomes horizontal. And this is same as the integration of area in polar coordinates. We integrate r^2/2 (formula for an arc) from 0 to pi/2 since it is in the first quadrant.
What my teacher said is quite similar. He considered the graph consists of many sectors from a quarter of a unit circle (shown in the right graphs). Although there are some margins, as we cut the sectors into infinity, these margins can be made up.image.jpg

Quick Reply