| Author |
 Topic  |
|
|
Rasika Eranda
Sri Lanka
3 Posts |
 Posted - 11/23/2014 : 12:13:15 PM
|
Origin Ver. and Service Release (Select Help-->About Origin):
Operating System:
Dear Friends,
I need to use nonlinear fit for Thomas, Yoon-Nelson and Adams-Bohart adsorption models. I tried to build these equations using fitting function builder, but I fail to do so.
Equations are
Thomas model = y=1/(1+exp((K*q*M)/F-(K*C*x)))
Yoon-Nelson model= Y=exp(K*A-K*x)
Adams-Bohart model= Y= exp(K*C*x-K*N*F)
K is the respective model constant and it varies with different flow rates and bed height,
C is the initial concentration it is 50 mg L-1
x (independent variable) which is time it varies from 0 to 750 min
I tried to edit some default functions and use but it didn't work. I'm getting an error massage saying that, Fitting failed on all datasets"
Please help me on this
Rasika D |
|
|
Echo_Chu
China
Posts |
Posted - 11/24/2014 : 01:02:24 AM
|
Hi, Rasika
Could you please send the following files so that we can figure out what went wrong?
1) You functions definition file (*.FDF) at your User Files Folder.
2) Your Origin data to be fit, and which show the error message(fitting failed on all dataset).
3) Exact step-by-step procedure to reproduce your problem, including the initial parameter values you entered before starting fitting.
You can send the files to tech@originlab.com and quote the link of this post
Echo
Originlab Technical Service |
 |
|
|
Rasika Eranda
Sri Lanka
3 Posts |
Posted - 11/24/2014 : 10:10:03 AM
|
quote:
Originally posted by Echo_Chu
Hi, Rasika
Could you please send the following files so that we can figure out what went wrong?
1) You functions definition file (*.FDF) at your User Files Folder.
2) Your Origin data to be fit, and which show the error message(fitting failed on all dataset).
3) Exact step-by-step procedure to reproduce your problem, including the initial parameter values you entered before starting fitting.
You can send the files to tech@originlab.com and quote the link of this post
Echo
Originlab Technical Service
I sent you the email Please be kind enough to consider it
Rasika D |
|
|
|
Echo_Chu
China
Posts |
Posted - 11/25/2014 : 02:56:08 AM
|
Hi, Rasika
Thanks for sending us the files. I can see your fitting functions are properly built but the fitting keeps not converging
Please look at the page below for all possible reasons and solution that your fitting does not converge.
http://www.originlab.com/doc/Origin-Help/Fit-ConvergTolerance-Dependence#Over-parameterized_functions
In regard to your problem, the three functions are over-parameterized. I suggest you to rewrite your function as the link suggests.
Then the Thomas model can be rewritten and to fit as Slogistic3. (You can fix a = 1 during fitting)
http://www.originlab.com/doc/Origin-Help/SLogistic3-FitFunc
The Yoon-Nelson and Adams-Bohart and be rewritten and fit as Exp2PMod1
http://www.originlab.com/doc/Origin-Help/Exp2PMod1-FitFunc
I will send you the OPJ that how I fit your data with these two build-in function.
If you do want to fit with your function, you can set them proper initial values and fix some constant parameters and try again. (I am not sure what are the proper value of your constant in model so I did not try this way in the OPJ)
Echo |
|
|
|
Rasika Eranda
Sri Lanka
3 Posts |
Posted - 11/25/2014 : 03:51:25 AM
|
Thank you very much for your effort
Rasika D |
|
|
| |
Topic |
|
Non linear curve fit