Integrating for approximation of a sum

In summary: Thank you for your help!In summary, the problem requires finding a value for N such that the infinite series ##∑^{\infty}_{n=1}\frac{log(n)}{n^2}## lies between ##∑^{N}_{n=1}\frac{log (n)}{n^2}## and ##∑^{N}_{n=1}\frac{log(n)}{n^2}+0.005.## Using definite integration and solving numerically, N is found to be 1687.
  • #1
goraemon
67
4

Homework Statement



Find an N so that ##∑^{\infty}_{n=1}\frac{log(n)}{n^2}## is between ##∑^{N}_{n=1}\frac{log (n)}{n^2}## and ##∑^{N}_{n=1}\frac{log(n)}{n^2}+0.005.##

Homework Equations


Definite integration


The Attempt at a Solution


I began by taking a definite integral: ##\int^{\infty}_{N}\frac{log(n)}{n^2}dn## and, using integration by parts, arrived at the following answer: ##\frac{log(N)+1}{N}##. (Is this right? If not, I could post the steps I used to try to see where I made an error)

Next we need ##\frac{log(N)+1}{N}## to be within 0.005 as given by the problem, so:

##\frac{log(N)+1}{N}=0.005=\frac{1}{200}##

But I'm having trouble how to solve for N algebraically. Would appreciate any help.
 
Physics news on Phys.org
  • #2
I actually got -1 times what you have. remember: ##\int \frac{1}{n^2} = -\frac{1}{n}##
 
  • #3
BiGyElLoWhAt said:
I actually got -1 times what you have. remember: ##\int \frac{1}{n^2} = -\frac{1}{n}##

Right but since we're taking a definite integral on the interval from N to ∞, and since ##\frac{-log(N)-1}{N}## as N approaches infinity equals zero, shouldn't the definite integral work out to:
##0-\frac{-log(N)-1}{N}=\frac{log(N)+1}{N}##?
 
  • #4
goraemon said:

Homework Statement



Find an N so that ##∑^{\infty}_{n=1}\frac{log(n)}{n^2}## is between ##∑^{N}_{n=1}\frac{log (n)}{n^2}## and ##∑^{N}_{n=1}\frac{log(n)}{n^2}+0.005.##

Homework Equations


Definite integration


The Attempt at a Solution


I began by taking a definite integral: ##\int^{\infty}_{N}\frac{log(n)}{n^2}dn## and, using integration by parts, arrived at the following answer: ##\frac{log(N)+1}{N}##. (Is this right? If not, I could post the steps I used to try to see where I made an error)

Your integral is correct.

Next we need ##\frac{log(N)+1}{N}## to be within 0.005 as given by the problem, so:

##\frac{log(N)+1}{N}=0.005=\frac{1}{200}##

The graph of [itex]x^{-2}\log(x)[/itex] has a maximum in [itex][1,2][/itex], so it's difficult to say whether the integral is an over-estimate or under-estimate of the sum. To be safe I'd require that [tex]\frac{\log(N) + 1}{N} < \frac{1}{400}[/tex].

But I'm having trouble how to solve for N algebraically. Would appreciate any help.

Equations of that type have to be solved numerically. You're looking for zeroes in [itex]x > 0[/itex] of [tex]f(x) = \frac{x}{200} - 1 - \log x[/tex]. The derivative of this function is [tex]
f'(x) = \frac{1}{200} - \frac{1}{x}[/tex] so the second derivative ([itex]x^{-2}[/itex]) is everywhere positive. There is a minimum at [itex]x = 200[/itex] where [itex]f(200) < 0[/itex]. Since [itex]f(x) \to +\infty[/itex] as [itex]x \to 0[/itex] or [itex]x \to \infty[/itex] there are exactly two solutions. Logic dictates you want the larger, since if [itex]N[/itex] works then any larger [itex]M[/itex] should also work.
 
  • Like
Likes 1 person
  • #5
Thanks, I tried solving it numerically and came up with N = 1687, which the textbook confirms is correct.
 

FAQ: Integrating for approximation of a sum

What is integrating for approximation of a sum?

Integrating for approximation of a sum is a mathematical method used to estimate the value of a sum by using integrals. It involves dividing the sum into smaller parts and approximating the area under the curve of each part using integration.

Why is integrating for approximation of a sum useful?

Integrating for approximation of a sum allows us to estimate the value of a sum when it is not possible to find its exact value. It also helps in solving complex problems involving sums and integrals.

What are the steps involved in integrating for approximation of a sum?

The first step is to divide the sum into smaller parts. Then, find the function that represents the sum and integrate it. Next, plug in the limits of integration and add up the individual approximations to get the overall approximation of the sum.

What are the limitations of integrating for approximation of a sum?

Integrating for approximation of a sum may not always give an accurate result. It is an estimation method and the accuracy depends on the number of parts the sum is divided into and the complexity of the function used for integration.

Can integrating for approximation of a sum be used for all types of sums?

No, integrating for approximation of a sum is most effective for infinite sums or sums with a large number of terms. It may not be useful for finite sums with a small number of terms as it may not provide a significant improvement in accuracy compared to other methods.

Similar threads

Replies
3
Views
955
Replies
8
Views
1K
Replies
6
Views
815
Replies
1
Views
919
Replies
6
Views
839
Replies
16
Views
1K
Back
Top