remove outliers or not just outliers but wrong data

[DFGray]

Don,

I have coded this in LabVIEW a couple of times, but not while working here at NI, so I don't have the code any more.  Sorry LabVIEW does include a downhill simplex (if you have the right version), but it is limited to text functions and this application is a data fitting issue.  The easiest way to approach this is to make copies of all the downhill simplex VIs from vi.lib and then replace all the function evaluations with your error evaluation (for lorentzian fit, log of the absolute value of the difference between the data and the model).  Optionally, you can write the algorithm from scratch using the Numerical Recipes code.

I checked the links again and the PDFs do open, albeit slowly on IE7.  You can try right-click, save, and open manually.

-Damien

[mikeporter]

For a lot of good information on data smoothing - and a bunch of other good stuff - check out Tukey's book Exploratory Data Analysis (a.k.a "The Red Book").

Mike...