From Friday, April 19th (11:00 PM CDT) through Saturday, April 20th (2:00 PM CDT), 2024, ni.com will undergo system upgrades that may result in temporary service interruption.
We appreciate your patience as we improve our online experience.
From Friday, April 19th (11:00 PM CDT) through Saturday, April 20th (2:00 PM CDT), 2024, ni.com will undergo system upgrades that may result in temporary service interruption.
We appreciate your patience as we improve our online experience.
01-23-2015 12:09 PM
Hi
I am running an experiement where I calculate the power spectrum of a signal and then fit a Gaussian to the Power Spectrum in order to measure a peak.
When I use labview to fit to my data, I get a peak with the correct centre, but an overly large standard deviation.
If I analyse the same data in Origin Lab, I am able to fit a very good Gaussian with low R^2 value.
Attached is an exmple of my data.
I would appreciate any suggestions as to how to get labview to perform a better fit.
01-23-2015 12:14 PM
We cannot help if you do not provide the Labview code you used to make the fit. The data would be helpful as well.
01-23-2015 02:07 PM - edited 01-23-2015 02:13 PM
Seems you are using Gaussian Peak Fit. If you don't wire the parameter bounds input, the offset is held fixed at zero by default (upper and lower bound=0, see the help). Seems to be happening in your case.
Fixes:
Wire the parameter bounds input (and at least make the offset to also be -inf..+inf).
01-24-2015 05:29 PM
Hi,
Sorry, I thought I had added the code as well as my sample graph.
Here is the code attached.
01-24-2015 08:04 PM
AlexMalm wrote:Here is the code attached.
We are missing a subVI.
Anyway, my guess was right. Why in the world would you use an inner subVI of the plain Gaussian Fit tool from the palette?? As you can easily tell, the one you are using cannot fit for an offset, but you data clearly shows a baseline that is not zero!
Use the real Gaussian Fit from the fitting palette and follow my recommendations in my earlier post. Simple as that! 😄
01-26-2015 02:29 AM