linefit( line_shape )

where line_shape is Lorentz / Gauss used to determine the fitted
line-shape (Lorentzian or Gaussian).
Performs a line fitting of the current 1D spectrum (sometimes
called incorrectly deconvolution) based on the
current peak-table contents.

The parameter of the line will be modified
to the best fit of the current spectrum using the Levenberg-
Marquardt method. The fit is estimated with a chi squared value,
computing on the displayed region, using the value of the noise.
Optimum fit is btained for a chi squared equal to 1. This value is
return in the $CHI2 variable. The program will fit the currently
displayed region, using the values in the peak table as starting
values, and returning the fitted values in the same peak table. Uses
the values of TOLERANCE and ITER for the run. All the values in the
peak-table will be used, but only the line in the displayed region
will be adjusted. Unknown line shapes will be assumed to be
Lorentzian.

see also : noise linefitd fitexp get

NPK Documentation, version 0.995 - 2009-10-14