MPFX version of ExpModGauss
josh.liptonduffin
I'm creating a custom fitting routine using MultiPeakFit2. My fitting model uses exponentially modified Gauss line shapes to fit some spectroscopy data. The fitting is working fine, but I am running into problems getting the peaks out for display back to the user when I call MPFXEMGPeak.
If I execute the following code
make/o testcoefs={300,1,1,.5} //some typical seed coefficients make/n=128 ywave //create a y wave to have the shape setscale /I x, 280,315,"units",ywave duplicate/o ywave, xwave //make an explicit x wave to match the dataset xwave=x display ywave vs xwave //plot the data MPFXEMGPeak(testcoefs,ywave,xwave) //attempt to make an EMG out of the data
the last line appears to do nothing at all. No error message, no updating of the wave... just nothing.
If I instead execute
EMGPeak(testcoefs,ywave,xwave) //use the old-style EMG peak function
it appears to work.
Obviously MPFXEMGPeak must be working in some capacity, since it works in the multipeak fit package as distributed; why can't I make it work on its own?? I'm trying to make this routine as generic as possible so I don't want to hard code the old version because I'm looking up MPF results using peakfuncinfo calls.
Try
If that doesn't make a visible difference (like, it remains a flat line) then we have a problem.
John Weeks
WaveMetrics, Inc.
support@wavemetrics.com
March 3, 2016 at 10:24 am - Permalink
yields no peak. If I follow those commands with
I get a peak as expected.
Some details: I am running v6.37 on MacOS 10.10.5 on a mid-2014 macbook pro with 16Gb RAM and an i7 processor at 3GHz.
I also have Igor 7.00B5 installed, but I find the display to be not super snappy on this machine so I'm not concentrating on adapting to it for now.
Is it possible that my v6.37 is somehow looking at a different MPFXOP because the beta is installed? Is it a Mac vs Windows compile issue?
EDIT: it works in beta 7! I guess I can use that for now.
March 3, 2016 at 03:00 pm - Permalink