Displaying 1 result from an estimated 1 matches for "ssexponen".
2008 Aug 29
1
nls() fails on a simple exponential fit, when lm() gets it right?
...LHS,data) {
xy=sortedXyData(mCall[["x"]],LHS,data)
r=0.01 #Just guess r=0.01 to start the fit
A=min(xy$y) #Use the minimum y value as an initial guess for A
value=c(r,A)
names(value)=mCall[c("r","A")]
return(value)
}
fmla=as.formula("~(A*exp(r*x))")
SSexponen=selfStart(fmla,initial=exponenInit,parameters=c("r","A"))
cat("Logarithmic fit (here, the self-start and the 'simpler' way match):\n")
tmp=getInitial(height~SSlogarithmic(dbh,aaa,bbb),data=data.frame(dbh,height))
cat("(Starting values for the logarithmic...