Need a little help setting up Batch Multipeak function

Share icon

Hi,

 

I have used the multipeak procedure on a test wave and am getting fits to my liking after using some manual starting points.  What is necessary to use the resulting fit as the starting point for the batch fitting?  The fit has three Gaussians and I can define that in the custom string.  What is necessary to use the fit coefficients as the starting points?  It is my understanding that I need to include some waves in the result_0 folder - which I guess I need to create before running the batch fit.  What waves should include in the folder?

Andy

I think I figured it out.

Copied and renamed "Peak 0 Coefs" as "peakcoef_0", ... for the three gaussian peaks and "Baseline Coefs" as "blcoef".

 

One possible improvement/suggestion is to allow an existing fit folder to be pointed to as starting point.

 

Andy

June 3, 2025 at 09:52 am - Permalink

Andy- The Multipeak Fit package in Igor 9 now has a UI for batch MPF. Does that help? If you are trying to use your own code to do batch MPF: DisplayHelpTopic "MPF2_AutoMPFit()"

June 3, 2025 at 11:39 am - Permalink

Andy, if I understand you correctly, you are trying to initialize MPF batch fitting with a Multipeak Fit set you have done manually. There are convenient tools for this as well. John already pointed you to the AutoMPFit help. Please have a look at sub-chapter titled "Converting AutoMPFit Results to Multipeak Fit Sets and Back". There you will find that you can load additional tools via:

#include <AutoMPFit to Fit Set Converter>

Then, in your case you want to set up an AutoMPFit folder from a Multipeak Fit set, so you have to use:

MPF2_CopyFromMPFSetFolder(resultDFBase, peakType, PeakCoefWaveFormat, BLType, BLCoefWaveName, yWaveList, xWaveList, InitialGuessOptions)

For example, after doing the first demo fit within the Multipeak Fit demo file, I can run this:

MPF2_CopyFromMPFSetFolder("result_", "Gauss", "PeakCoefs%d", "LogCubic", "blcoef", "root:'data set 1':'my y data'", "", 0)

This will give a folder 'result_0' which is set up to be used with AutoMPFit() in combination with initial guess options 0 or 1.

June 3, 2025 at 10:04 pm - Permalink