Wishes for the CurveFitter

[jumpfunky]

Dear list,

the CurveFitter is really cool but I'm missing the possiblity to give
different data points a different weight. Any chance that it could be
extended by such a weighting?

Some question for a work-around:

Given 4 Points P1, P2, P3, P4 and I would like to fit a regression line
 (or some other line) and I have the unnormalized weights (4,3,2,1),
would it be the same to an weighted fit to make an unweighted fit
though the points:

P1,P1,P1,P1,P2,P2,P2,P3,P3,P4?


Cheers,
Thorsten

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Bill Christens-Barry-2]

Thorsten,

If you can live with weights that are restricted to positive integer values, you can increase the weight of a point by making multiple copies of that point in the list. If a point appears n times, it will have a weight of n. Otherwise, I don't know of a general solution.

Bill Christens-Barry

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html