The Origin Forum
File Exchange
Try Origin for Free
The Origin Forum
Home | Profile | Register | Active Topics | Members | Search | FAQ | Send File to Tech support
Username:
 Password:
Save Password
Forgot your Password? | Admin Options

 All Forums
 Origin Forum
 Origin Forum
 Fitting with a distribution of parameters		  New Topic  Reply to Topic
 Printer Friendly
Author Previous Topic Topic Next Topic Lock Topic Edit Topic Delete Topic New Topic Reply to Topic

CStWue

5 Posts
Posted - 10/23/2019 :  09:04:24 AM  Show Profile  Edit Topic  Reply with Quote  View user's IP address  Delete Topic
Origin Ver. and Service Release (Select Help-->About Origin): OriginPro 2019b (64-bit)
Operating System: Windows 10

Hello,

I'm analyzing datasets of fluorescence lifetime experiments that contain discrete time channels in the X column and the signal and instrument response in two separate Y columns.

Currently, I'm fitting the data with a convolution of a single-exponential decay curve and the instrument response as shown in this tutorial for Origin C: https://www.originlab.com/doc/Tutorials/Fitting-Convolution

Now instead of the simple decay law f(x) = y0 + a * exp(-t*x) with individual parameters "a" and "t" I want to fit with a distribution of parameters such that my fitting function is:

f(x) = y0 + integral(0->tmax) P(t) * exp(-t*x) dt with P(t,A,tc,w) = A/(w*sqrt(pi/(4*ln(2)))) * exp(-4*ln(2)*(t-tc)^2/w^2)

"tmax", the upper limit of integration, will be a fixed value. The mean lifetime "tc", full width at half maximum "w" and area "A" of the distribution should be fitting parameters.

This function f(x) will then be convoluted with the instrument response as described in the link above.

How can I implement this idea with Origin?

Thank you!

YimingChen

1691 Posts
Posted - 10/23/2019 :  12:12:31 PM  Show Profile  Edit Reply  Reply with Quote  View user's IP address  Delete Reply
Hi,

To include integral into your fitting function, please refer to this page:
https://www.originlab.com/doc/Tutorials/Fitting-Integral-NAG

See if you can replace 
vSignal = A * exp( -t*vSample );
with an integral function.

James
Go to Top of Page

CStWue

5 Posts
Posted - 10/28/2019 :  08:51:26 AM  Show Profile  Edit Reply  Reply with Quote  View user's IP address  Delete Reply
Hello James,

thank you for your reply. I have looked at the tutorial you linked and this seems to be a good approach.

However, because the Origin C documentation is overall very limited, it is not clear to me how to combine the NAG integral function with the convolution that is linked in my initial post.

The NAG integral seems to take the x and values simply from the columns specified in the NLFit dialog. In the tutorial for the convolution fit these columns are accessed as datasets which are then passed to vectors, presumably for the fft convolution algorithm. These approaches seem fundamentally different two me, how could they be combined?
Go to Top of Page

YimingChen

1691 Posts
Posted - 10/28/2019 :  11:30:42 AM  Show Profile  Edit Reply  Reply with Quote  View user's IP address  Delete Reply
Hi,

I thought you wanted to fit a convolution function same as that in the tutorial except replacing the Exponential Decay function with your defined f(x). Since you mentioned:
"This function f(x) will then be convoluted with the instrument response as described in the link above."

Please send email to <tech@originlab.com> and provide more details on the approach you were using.

Thank you

James
Edited by - YimingChen on 10/28/2019 11:30:58 AM
Go to Top of Page
  Previous Topic Topic Next Topic Lock Topic Edit Topic Delete Topic New Topic Reply to Topic
 New Topic  Reply to Topic
 Printer Friendly
Jump To:
The Origin Forum © 2020 Originlab Corporation Go To Top Of Page
Snitz Forums 2000