|
|
Hi,
I am writing a little plugin to plot and correlate parameters/columns in the
Results table.
I have noticed that the CurveFitter returns similar polynomial coefficients to
those calculated with Excel and SPSS do, but only up to the 3rd degree
fitting. The 4th degree fitting differs quite a bit from Excel and SPSS (but
these 2 remain similar up to the 5th significant figure or so).
I have tried increasing the iterations of the CurveFitter and the number of
starts, but the differences remain.
I noted that there is a further parameter "error tolerance" but there is no
method to set this (for instance from another plugin).
Would reducing this value improve the fitting?
Many thanks,
Gabriel
|
Locked
