| T O P I C R E V I E W |
| GeoGuy |
Posted - 02/26/2013 : 7:22:02 PM
Hey,
I'm having some trouble with a simple fit in Origin. I want to fit data from a thermal expansion experiment.
It's a cubic polynomial y=A+B*x+C*x^2+D*x^3 with an additional constraint, namely x=x-293. In the end, I want to get the calculated and fitted value of "B" from this equation.
I didn't see no problem first but it just won't work.
I appreciated any helpful suggestions.
Got access to OriginPro8G@Windows7.
Thank you.
A world build on bees. |
| 7 L A T E S T R E P L I E S (Newest First) |
| GeoGuy |
Posted - 03/06/2013 : 04:31:29 AM
I will try this out as soon as possible.
Thanks a lot!
A world build on bees. |
| Hideo Fujii |
Posted - 03/01/2013 : 10:27:05 AM
Hi GeoGuy,
I have defined a new function of your exact form:
y = A + B*(x-293.15) + C*(x-293.15)^2 + D*(x-293.15)^3
as Function Type="LabTalk Script" using Fitting Function Builder in my Origin9, and worked fine.
Unfortunately your version 8.0 doesn't have this intuitive tool to define a user-defined fit function,
but you should be able to do so by using Fitting Function Organizer as both tools can produce the .FDF
file at the end.
--Hideo Fujii
OriginLab |
| GeoGuy |
Posted - 03/01/2013 : 07:15:42 AM
Thank you, I guess the set column values tool was the right thing, results seem fine. I still don't understand why
y = A + B*(x-293.15) + C*(x-293.15)^2 + D*(x-293.15)^3
in the builder tool didn't work out.
Problem is solved anyways.
Thanks a lot!
A world build on bees. |
| Hideo Fujii |
Posted - 02/28/2013 : 4:08:44 PM
Hi GeoGuy,
For your application, maybe the easiest approaches are EITHER (though analytically you can get the new set of K-based parameters from the current C-based parameters):
1) Before fitting, convert your X values from Celsius data to Kelvin using Set Column Values tool. Then, fit it with the built-in Cubic function.
or
2) Define a completely new user-defined function:
y = A + B*(x-293.15) + C*(x-293.15)^2 + D*(x-293.15)^3
by Fitting Function Builder tool; then you can fit your data using this.
I hope I didn't miss your points.
--Hideo Fujii
OriginLab |
| GeoGuy |
Posted - 02/28/2013 : 2:38:11 PM
Well, that didn't work. Maybe I did not illustrate my problem. When it comes to functions, and math in general, it's over (:
I found the cubic polynomial, which I need. I simply need to modify the equation like this:
y = A + B*(x-293.15) + C*(x-293.15)^2 + D*(x-293.15)^3
I got Temperature data (°C) that I need to bring to °K somehow (god knows why I can't calculate °K values BEFORE i fit the data, I'm a Geology student and suddenly I find myself doing thermal expansion experiments with some crystollography guys).
If I enter the equation like that, something goes terribly wrong.
A world build on bees. |
| Hideo Fujii |
Posted - 02/27/2013 : 4:35:32 PM
Hi,
Is it actually like "A<=-293" ?
You can use the "Cubic" function in Polynomial category in the Non-Linear Curve fitter.
In it, you can set the boundary of the parameters.
Hope this is a good shot.
--Hideo Fujii
OriginLab |
| mikefeehts110 |
Posted - 02/26/2013 : 10:40:46 PM
Have you mis-typed the constraint? Wouldn't x=x-293 only have a solution if 0=-293? |
