Thread Subject:

Romberg and Simpson combination

I am trying to help a friend out with a problem, but I can't seem to write the code. Can someone help me with this? That will be greatly appreciated!
The problem is as follows:

Develop an adaptive quadrature by combining (composite) Simpson's rule and Romberg integration as follows:
 On each interval [a; b], let h = (b- a)/4 and get the values of given function at 5 point
xj = a + j * h, with j = 0; 1; 2; 3; 4.
 Compute Simpson's rule to get Q1, and composite Simpson's rule to get Q2.
 Work out the correct Romberg integration rule for Simpson's rule; apply it Q1 and Q2 to
get a new estimate Q.
 When abs(Q - Q2) is less than given tolerance, accept Q as true integral value; otherwise
subdivide [a; b] into two intervals of equall length and recursively apply the same adaptive
quadrature to them.
 Make sure you do not ever calculate the given function more than once at any given point.
Your program should take tolerance as an input argument, and return the integral and number
of function evaluations. Report the results obtained from running your code on the function
f(x) = 10/x^2 *sin(10/x) on the interval [1; 2] with tolerance 10^-8.

"ze lu" <demijet01@yahoo.com> wrote in message <icajtc$hue$1@fred.mathworks.com>...
> I am trying to help a friend out with a problem, but I can't seem to write the code. Can someone help me with this? That will be greatly appreciated!
> The problem is as follows:
>
> Develop an adaptive quadrature by combining (composite) Simpson's rule and Romberg integration as follows:
>  On each interval [a; b], let h = (b- a)/4 and get the values of given function at 5 point
> xj = a + j * h, with j = 0; 1; 2; 3; 4.
>  Compute Simpson's rule to get Q1, and composite Simpson's rule to get Q2.
>  Work out the correct Romberg integration rule for Simpson's rule; apply it Q1 and Q2 to
> get a new estimate Q.
>  When abs(Q - Q2) is less than given tolerance, accept Q as true integral value; otherwise
> subdivide [a; b] into two intervals of equall length and recursively apply the same adaptive
> quadrature to them.
> Make sure you do not ever calculate the given function more than once at any given point.
> Your program should take tolerance as an input argument, and return the integral and number
> of function evaluations. Report the results obtained from running your code on the function
> f(x) = 10/x^2 *sin(10/x) on the interval [1; 2] with tolerance 10^-8.

Sorry. We tend not to do your homework for you.

Even if this is not your work (still undetermined)
why should we help you to help someone else
cheat on their homework?

If you have a specific question about something that
failed, then ask. But don't ask for someone to write
code for a homework problem.

john

hilarious. ming gu's student.

K <user@compgroups.net/> wrote in message <laudneU5pL6lAXTRnZ2dnUVZ_rOdnZ2d@giganews.com>...
> hilarious. ming gu's student.
>
I just finished the code. I was already in progress of coding when I posted the message. I didn't want to post my code here because I was afraid that other people would use it as their own... I was hoping someone would respond to help, and I would then privately message my code to them.

But what ever.

"david smith" wrote in message <icf6en$3dt$1@fred.mathworks.com>...
> K <user@compgroups.net/> wrote in message <laudneU5pL6lAXTRnZ2dnUVZ_rOdnZ2d@giganews.com>...
> > hilarious. ming gu's student.
> >
> I just finished the code. I was already in progress of coding when I posted the message. I didn't want to post my code here because I was afraid that other people would use it as their own... I was hoping someone would respond to help, and I would then privately message my code to them.
>
> But what ever.

Hey David, do u mind to email me this code if u still have it? thanks

Hey;
       Could you help me with this program, I will be really appreciate, my email address is ufozb919@hotmail.com. Thanks!

"jordan " <luyijiajordan0713@hotmail.com> wrote in message <jmd7p4$n55$1@newscl01ah.mathworks.com>...
> "david smith" wrote in message <icf6en$3dt$1@fred.mathworks.com>...
> > K <user@compgroups.net/> wrote in message <laudneU5pL6lAXTRnZ2dnUVZ_rOdnZ2d@giganews.com>...
> > > hilarious. ming gu's student.
> > >
> > I just finished the code. I was already in progress of coding when I posted the message. I didn't want to post my code here because I was afraid that other people would use it as their own... I was hoping someone would respond to help, and I would then privately message my code to them.
> >
> > But what ever.
>
> Hey David, do u mind to email me this code if u still have it? thanks