multivariate all-at-once fits

I read that a new feature for Igor Pro 6 is that "all-at-once fit functions can now implement multivariate fit functions." I understand how this is to be done with 1 dimensional data, but have been able to find instructions on how this should be implemented for multiple dimensions. Any help would be appreciated!
[quote=Rischaard]I read that a new feature for Igor Pro 6 is that "all-at-once fit functions can now implement multivariate fit functions." I understand how this is to be done with 1 dimensional data, but have been able to find instructions on how this should be implemented for multiple dimensions. Any help would be appreciated![/quote]

Hm.... it looks like the description is missing from the documentation!

The format is analogous to a regular multivariate fit function. With a univariate fit function you write a function like this:

Function myfunc(w, x) : FitFunc
  Wave w
  Variable x
 
  ...
end


To make it multivariate, you just add more independent variable inputs:

Function myfunc(w, x1, x2) : FitFunc
  Wave w
  Variable x1, x2
 
  ...
end


Similarly, the extension for all-at-once fitting functions is to simply add more independent variable inputs:

Function myAllAtOnceFunc(pw, xw1, xw2) : FitFunc
  Wave pw
  wave xw1, xw2
 
  ...
end


John Weeks
WaveMetrics, Inc.
support@wavemetrics.com