| peak finding with Guassfit [message #86891] |
Tue, 10 December 2013 09:57  |
Adele
Messages: 4
Registered: September 2013
|
Junior Member |
|
|
I am working with 1D data, need to find two peak positions of each data, these two peaks close to each other.
Currently, I've been using a gaussian fit with 6 parameters, and using
the center of the Gaussian as the peak location. For each data set,the proximate peak location and fit width were given, gaussfit was used in the range of (peak-width, peak+width). For some data set the results are good, but for other data set the center of gaussian was shifted even out of the given range. These data sets do not have significant difference.
Does anyone have any idea about this situation? Thank you very much!
|
|
|
|
|
|
| Re: peak finding with Guassfit [message #86908 is a reply to message #86899] |
Wed, 11 December 2013 06:36  |
Adele
Messages: 4
Registered: September 2013
|
Junior Member |
|
|
On Tuesday, December 10, 2013 6:23:24 PM UTC-5, Phillip Bitzer wrote:
> On Tuesday, December 10, 2013 11:57:49 AM UTC-6, Adele wrote:
>
>> I am working with 1D data, need to find two peak positions of each data, these two peaks close to each other.
>
>>
>
>> the center of the Gaussian as the peak location.
>
>>
>
>
>
> I'm confused here - does the data have two Gaussians or one?
>
> the data has two Gaussians which close to each other.
>
> If one, do you need all six terms? In other words, do you have any a priori information on the background?
I extended the fit width then Gaussian fit works well for both data. Thanks.
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
