| T O P I C R E V I E W |
| federicolan |
Posted - 03/04/2012 : 12:25:11 PM
Origin Ver . and Service Release (Select Help-->About Origin): OriginPro 8 SR0 v8.0724 (B724)
Operating System: Windows Xp
Hi everybody, here is my problem:
I'm studying how different treaments act on the dimensions of the ice crystals of a frozen sample, either by retarding their growth or lowering their initial size.
I perform a frequency count on the basis of the area of the objects, and then I do a Gaussian Fit to get informations on the distribution of the population (which I assume follow a normal distribution).
Of course the quality of the fit will be determined by the size of the intervals that I've choosen for the frequency count.
Is it possible to let Origin choose the appropiate inteval size for the frequency count, on the basis of the quality of the Gaussin fit (in terms of the best R2 coefficent).
I really don't know how to start so I hope I did't posted my question in the wrong section.
Thanks in advance
Federico |
| 1 L A T E S T R E P L I E S (Newest First) |
| Sam Fang |
Posted - 03/05/2012 : 04:03:14 AM
Fitting may be not appropriate for the problem. You can plot histogram and choose distribute curve (Uncheck Scale to Maximum (%)) in the Plot Details' dialog to get information on the distribution of the population.
Origin8.6's distribution curve provides a normalization method to fit the histogram. It uses the scale factor N*dt for a probability density function, where N is the number of points, and dt is the bin size. If the histogram follows a statistics distribution, the distribution curve will agree with the histogram. And the distribution curve also updates as the histogram's bin size changes.
However Origin8.0 doesn't support this normalization method, and it only supports to scale to the maximum of the histogram. Maybe you can try this feature in Origin8.6.
Sam
OriginLab Technical Services |
|
|
