T O P I C R E V I E W |
Ibex |
Posted - 06/09/2009 : 12:09:43 PM Origin Ver. and Service Release (Select Help-->About Origin): Pro 8 SR5 Operating System:Win XP
I think there is a bug in the multiple curve fitting function. I fitted 4 different curves with: - different X ranges and stepsizes - instrumental weighting - Boltzmann function
When I do a "concatenate fit" everything's fine. But when I switch to "global fit" and share all 4 Variables it results in the fitting curve for curve 1 and not in a combined fit. When I try to bring the parameters to the right values and fix them, chi-sqr is rising.
Besides that problem Origin crashes from time to time when including 4 ranges in the X-Function dialog.
Regards Ibex
|
7 L A T E S T R E P L I E S (Newest First) |
goodwill221 |
Posted - 07/09/2009 : 10:53:54 AM quote: Originally posted by Echo_Chu
Hi, Ibex,
Thanks for reporting this problem! This is a bug in SR5 and we will get it fix in our SR6.
Echo OriginLab
Well I am having the same problem and what you advise me? Is it possible to remove this bug..
|
Ibex |
Posted - 07/02/2009 : 04:47:24 AM Hi Larry, I'm glad, that the bug is found, but when approximately will the SR6 be released? Is it a question of days, weeks, or months? Thanks Ibex |
larry_lan |
Posted - 07/02/2009 : 03:49:40 AM Hi Ibex:
We finally found that this is a bug intruduced in SR5, which related to the weighting method. We have fixed it, and it will be OK in the next patch. Sorry for the inconvenience
Thanks Larry OriginLab Technical Services |
Echo_Chu |
Posted - 07/02/2009 : 03:49:09 AM Hi, Ibex,
Thanks for reporting this problem! This is a bug in SR5 and we will get it fix in our SR6.
Echo OriginLab
|
Ibex |
Posted - 06/29/2009 : 09:56:31 AM Hi, are there any news about this topic?
Maybe I have an additional hint to the origin of that problem. After fitting several other Curves with some shared variables I found in the report sheet under statistics that the value for "Residual Sum of Squares" is 132.5 for the first curve and approx. 0.5 for all other curves although all curves are very similar. I have the impression that there is something wrong with the calculation of chi-square when doing weighted fits.
Ibex |
Ibex |
Posted - 06/10/2009 : 05:47:48 AM Thanks Echo for your fast response. I Uploaded the data to the FTP Server as MCF_bug_ibex.opj.
Ibex |
Echo_Chu |
Posted - 06/10/2009 : 05:04:49 AM Hi, Ibex
It looks we can not reproduce your problem with our sample data. Would you mind to provide us your data so that we can do further testing?
You can send your data/opj from this page: http://www.originlab.com/index.aspx?s=1&lm=123&pid=752
Echo OriginLab |