Hi,
I need to compare non-linear fittings of 2 different experimental
distributions. I use nls() to the fit the 2 distribution and it works
pretty well.
The statistical comparison of 2 non linear fits is well described in
the book "Fitting Models to Biological Data using Linear and Nonlinear
Regression" from Harvey Motulsky and Arthur Christopoulos.
This test is based on the calculation of an F-value that compares the
residuals of
1°) the global fit of both data sets (with one value for each
parameter) on one hand,
2°) with the 2 fits of each separate data sets (with thus 2 values for
each parameter) on the other hand.
This methods allows to decide wether the 2 data sets belong to the
same curve. No problem this far.
But the next step is about comparing each parameter to decide wether
or not they are different between each data set. The comparison still
lies on the calculation of an F-value, but here I compare
1°) the residuals of the 2 fits of each separate data sets,
2°) with the residuals the 2 fits of each data sets which share a same
value of the parameter that is compared (i.e. the 2 curve fits find
the same value for the shared parameter).
This point is where I need some help : is there any function to make
non-linear fitting of (at least) 2 data sets with shared parameters ?
Hope someone will have some good news for me
Etienne Toffin
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
