| Author |
 Topic  |
|
|
Kieran
Australia
2 Posts |
 Posted - 10/07/2020 : 04:02:37 AM
|
The standard "derived parameters" in the Logistic.fdf are
EC10=10^(log(x0) - log(9)/p)
EC20=10^(log(x0) + log(0.25)/p)
EC50=x0
EC80=10^(log(x0) + log(4)/p)
EC90=10^(log(x0) + log(9)/p)
These do not provide accurate estimates. I can see that these can be altered in the "Fitting Function Organiser" but have not been able to insert the correct equation (either syntax is wrong or equation is wrong)
The equation is "y=(A1-A2)/(1 + (x/x0)^p) + A2"
I am in need of help to rearrange the equation to solve for x
Any ideas would be greatly appreciated |
|
|
easwar
USA
1965 Posts |
Posted - 10/07/2020 : 09:44:13 AM
|
Hello,
Yes, these derived parameter expressions assume the data is normalized between 0 and 100 and further the fit is done with the asymptotes fixed to values of 0 and 100.
There is a Find X from Y feature in the fitter where you can ask for X values to be computed, after the fit, for your specific Y values. You will find the control under Settings Tab, named "Find X/Y".
Please try this and see if you get the values you expect.
This help page has details on the algorithm:
https://www.originlab.com/doc/Origin-Help/FitRef-FindVal-StandardCurve
Easwar
OriginLab |
 |
|
|
Kieran
Australia
2 Posts |
Posted - 10/07/2020 : 6:20:32 PM
|
quote:
Originally posted by easwar
Hello,
Yes, these derived parameter expressions assume the data is normalized between 0 and 100 and further the fit is done with the asymptotes fixed to values of 0 and 100.
There is a Find X from Y feature in the fitter where you can ask for X values to be computed, after the fit, for your specific Y values. You will find the control under Settings Tab, named "Find X/Y".
Please try this and see if you get the values you expect.
This help page has details on the algorithm:
https://www.originlab.com/doc/Origin-Help/FitRef-FindVal-StandardCurve
Easwar
OriginLab
Thanks for your response Easwar.
The find x from y function does compute the EC figures correctly [same as graphed] but I cannot find a way to use this as a source to populate the table when I include Derived parameters in printout.
Can you advise on where I can fix the asymptotes to 0 & 100? This should fix the problem, I hope! |
|
|
|
easwar
USA
1965 Posts |
Posted - 10/07/2020 : 9:15:14 PM
|
Hi Kieran,
You can fix the parameter values in the Parameters tab of the NLFit dialog:
Easwar |
|
|
| |
Topic |
|
|
|
Derived parameters in Logistic curve fit