cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
Showing results for 
Search instead for 
Did you mean: 

Community Tip - When posting, your subject should be specific and summarize your question. Here are some additional tips on asking a great question. X

Variational method

aayyappan
1-Newbie

Variational method

Hello to all,

I hope I can get some help here. I created a worksheet in mathcad for variational method.

Actually I am know only the basic about mathcad. In my worksheet , the minimization couldnot be able to do,

it take too much of time to do the minimization. Finally I am not able to get the answer.

So please see my attachement and help me to get the answer.

Thank you


8 REPLIES 8

Have you ever tried to evaluate your function E(a) for any single position of a?

Yes, I have tried that. But the forth term of E, which shows some error.

so coudnot get the answer. is that a problem for minimization?. If it means What shall I do? see the attachment i have tried for a=2.

I've simplified your integrals as much as possible; replaced them by discrete summations (with a really coarse resolution!) ; and they still take a while to evaluate! See attached. Incidentally the kernels of your inner integrals in E are independent of the variable over which you are integrating, so these inner integrals can just be replaced by 2pi.

See attached.

Alan

Alan Thank you for time consideration to solve my problem. You did a great work for me.

you replaced -Zinf by 2 and Pinf by 10. But in my problem it should be infinity only.

Anitha Ayyappan wrote:

Alan Thank you for time consideration to solve my problem. You did a great work for me.

you replaced -Zinf by 2 and Pinf by 10. But in my problem it should be infinity only.

Yes, as it stands you will probably have to wait for an infinite time for the program to work if you use infinite (or very large) limits. The important question is: do the results change as these limits get bigger? Once the results stop changing (if they do) then you can safely use finite limits.

I did a little more work and simplified a few more things in your worksheet - see attached. Now you can use slightly larger limits and increase the resolution of the discrete sum representation of E (again, you need to keep increasing the resolution until the results don't change).

Calculation speed has increased, but is still fairly low!

Alan

Helo alan,

Thank you for continous rply for my questions. I went through the work sheet , which you made it for me.

I have some questions on this worksheet. please clarify that for me.

1.you have plot some graph between z and k. and k and p(row). wht it is necessary?

2. integration was changed into summazion. i cannot understand that clearly. you have any notes regarding that modification. because i am so confused with that np and nz.

please see the attachment.

Anitha Ayyappan wrote:

1.you have plot some graph between z and k. and k and p(row). wht it is necessary?

It isn't necessary, but it shows the motivation for limiting zinf and rhoinf - see attached.

2. integration was changed into summazion. i cannot understand that clearly.

Integration is just summation with the intervals taken to the limit of zero width. All I've done is to say, this makes the calculation take too much time, so I'll approximate the integral by using finite-width intervals. It's a coarse approximation, and you would need to play around with the resolution (i.e. size of width). It is quite possible that my approximations are too coarse and of no practical use. However, I don't think you'll get a result using your original, exact formulation (not in Mathcad anyway)!

Alan

-MFra-
21-Topaz II
(To:aayyappan)

Hi Anitha Ayyappan,

perhaps modifying your trial wave function (without discontinuities) you can obtain some results as in the following immage:

aayyappan.jpg

PS: you have to modify the normalization integral.

Bye

FM

Top Tags