Dear R-listers, I would like to know if there is a way to programmatically generate parameter start values for the model y~(a*exp(b*x)+c*exp(d*x)) in R. I've scoured the help files and archives for nls() and similar searches, and have read Fox 2002 - the best advice has been to make estimates from a priori knowledge of the data. However, in the Matlab CurveFit tool, reliable start values are somehow estimated 'heuristically', as is claimed in their help files. Is there a similar R routine than may produce reasonable start values? Many thanks, Paul Brewin Paul E Brewin (PhD) Center for Research in Biological Systems University of California San Diego 9500 Gilman Drive MC 0505 La Jolla CA, 92093-0505 USA Ph: 858-822-0871 Fax: 858-822-3610
This has been discussed a number of times in the past, e.g. see this thread: http://tolstoy.newcastle.edu.au/R/e2/help/06/09/0980.html On 10/20/06, Paul Brewin <pebrewin at sdsc.edu> wrote:> Dear R-listers, > > I would like to know if there is a way to programmatically generate > parameter start values for the model y~(a*exp(b*x)+c*exp(d*x)) in R. > I've scoured the help files and archives for nls() and similar searches, > and have read Fox 2002 - the best advice has been to make estimates from > a priori knowledge of the data. However, in the Matlab CurveFit tool, > reliable start values are somehow estimated 'heuristically', as is > claimed in their help files. Is there a similar R routine than may > produce reasonable start values? > > Many thanks, > Paul Brewin > > Paul E Brewin (PhD) > Center for Research in Biological Systems > University of California San Diego > 9500 Gilman Drive MC 0505 > La Jolla CA, 92093-0505 > USA > > Ph: 858-822-0871 > Fax: 858-822-3610 > > ______________________________________________ > R-help at stat.math.ethz.ch 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. >
On 10/19/06, Paul Brewin <pebrewin at sdsc.edu> wrote:> Dear R-listers, > > I would like to know if there is a way to programmatically generate > parameter start values for the model y~(a*exp(b*x)+c*exp(d*x)) in R. > I've scoured the help files and archives for nls() and similar searches, > and have read Fox 2002 - the best advice has been to make estimates from > a priori knowledge of the data. However, in the Matlab CurveFit tool, > reliable start values are somehow estimated 'heuristically', as is > claimed in their help files. Is there a similar R routine than may > produce reasonable start values??SSbiexp or example(SSbiexp)