| T O P I C R E V I E W |
| humeyra |
Posted - 10/06/2004 : 05:30:18 AM
I have a problem on using deconvolution function of origin 6.0. I have been unable to deconvolute the gel permeation chromatograms.
i have only the data of time and response. When i write this data; data 1b must have an odd number of points error occurred and following the response data set Data1-B must be less than half the size of Data1-A.
I really don't know how can i solve this problem. It would be excellent if someone helps me;
humeyramert@yahoo.com
Origin Version (Select Help-->About Origin):
Operating System:
Edited by - humeyra on 10/06/2004 05:42:55 AM |
| 6 L A T E S T R E P L I E S (Newest First) |
| nintwjf |
Posted - 09/01/2005 : 01:08:37 AM
To"humeyra":
Now I met the same problem as you met ago,I want to ask you for the answer to it.Thanks a lot.
Thank for your reading!
nintwjf
|
| humeyra |
Posted - 10/11/2004 : 02:50:42 AM
Hi;
Good news! I succeded to deconvolute my peak on weekend!:)
I tried many operations and now i'm trying to find out how did i do:)Thanks for your attention!
Humeyra
|
| easwar |
Posted - 10/07/2004 : 1:51:03 PM
Hi,
If what you mean by deconvolution is to fit multiple peaks to your data and thus determine what contribution to the data came from each peak, then fitting with multiple gaussians (or with any other peak function) is the right thing to do.
If on the other hand you are taking about your signal shape, and you want to take out the effect of say your instrument response from your measured data, then that is the kind of deconvolution I was refering to. (For example, the true signal may be a narrow gaussian and your instrument adds a spread to it due to the broad response curve of your instrument, and you want to "remove" the instrument effect from the measurement).
Looks like what you are looking for is the first one, in which case multiple peak fit is the right thing to do.
If you need further help, please contact tech support.
Easwar
OriginLab
|
| humeyra |
Posted - 10/07/2004 : 04:49:26 AM
Hi again:)
Also i would like to learn the difference between deconvolution and fit multi peaks (gaussian) functions.
I performed the second one and the programme gave me the area of the overlapped two peaks seperately. May i trust this result? Or do i have to try to find out how to use deconvolution function?
Thanks for reading!
|
| humeyra |
Posted - 10/07/2004 : 04:13:05 AM
I really thank u for your attention!
I set my signal data (belong to my polymer) in B coloumn and response data (belong to referance polymer) in C coloumn. When i choose B and C coloumn for deconvolution i have taken the same alert ' data c should be smaller than half....' then i have shorten my C coloumn by erasing some of the data. But there is no difference i took the same alert.
What is my fault???
|
| easwar |
Posted - 10/06/2004 : 10:34:26 AM
Hi,
In order to perform deconvolution, you need to have the signal data and the response data. In addition you may have a time axis (X) data. So you need at least two Y data sets.
Then, when the operation is performed from the GUI, there are the limitations that the response has to be odd number of points, should be smaller than half the size of the signal, and should be symmetric. This limitation will be resolved in a future version.
In versions 7.0/7.5, one can instead use the NAG library to perform the deconvolution and code for doing such is posted in this thread:
http://www.originlab.com/forum/topic.asp?TOPIC_ID=3151
In versions 6.1 and earlier, there is currently no work around to the limitations.
Easwar
OriginLab
|
