I Approximating discrete sum by integral

Kashmir
Messages
466
Reaction score
74
I can't understand how this approximation works ##\sum_{k=0}^m\left(\frac{k}{m}\right)^n\approx\int_0^m\left(\frac{x}{m}\right)^ndx\tag{1}##Can you please help me
 
Physics news on Phys.org
Choose a moderate value of m, not too small, say m=20.
Then set n=1, and draw the graph of ##f(x) = (\frac xm)^n## between 0 and ##m##. On the same set of axes draw the series of m rectangles such that the k-th rectangle (k = 0, ..., m-1) is the set of points (x,y) with ##\frac km\le x\le \frac{k+1}m## and ##(\frac xm)^n\le y\le (\frac{x+1}m)^n##.

Then do the same thing on a new set of axes, for ##n=2##.

That should give you a strong sense for why the approximation works.
 
  • Like
Likes WWGD and Kashmir
andrewkirk said:
Choose a moderate value of m, not too small, say m=20.
Then set n=1, and draw the graph of ##f(x) = (\frac xm)^n## between 0 and ##m##. On the same set of axes draw the series of m rectangles such that the k-th rectangle (k = 0, ..., m-1) is the set of points (x,y) with ##\frac km\le x\le \frac{k+1}m## and ##(\frac xm)^n\le y\le (\frac{x+1}m)^n##.

Then do the same thing on a new set of axes, for ##n=2##.

That should give you a strong sense for why the approximation works.
We first think about the sum as the area under the rectangles having unit width.
This area is roughly the integral of the corresponding continuous function.

Am i right?
 
Kashmir said:
We first think about the sum as the area under the rectangles having unit width.
This area is roughly the integral of the corresponding continuous function.

Am i right?
Yes, that's correct. It is called a Riemann Sum. You can read more about them here, including some nice pictures.
 
  • Informative
  • Like
Likes Kashmir and WWGD
andrewkirk said:
Yes, that's correct. It is called a Riemann Sum. You can read more about them here, including some nice pictures.
Thank you :)
 

Similar threads

Back
Top