Re: [R] Error 'singular gradient' in nonlinear model fitting

2008-02-16 Thread Douglas Bates
On Feb 15, 2008 11:50 AM, Gabor Grothendieck <[EMAIL PROTECTED]> wrote: > Try fitting regressing log(WEIGHT) against log(TOTAL) using lm > and then transform the results into starting values for nls (or > possibly that is sufficient and you don't need the nls results). Also, set the trace argument

Re: [R] Error 'singular gradient' in nonlinear model fitting

2008-02-15 Thread Gabor Grothendieck
Try fitting regressing log(WEIGHT) against log(TOTAL) using lm and then transform the results into starting values for nls (or possibly that is sufficient and you don't need the nls results). On Fri, Feb 15, 2008 at 12:41 PM, HongSheng Liao <[EMAIL PROTECTED]> wrote: > > w.age.female.2004 <- nls(W

Re: [R] Error 'singular gradient' in nonlinear model fitting

2008-02-15 Thread Peter Dalgaard
HongSheng Liao wrote: > w.age.female.2004 <- nls(WEIGHT ~ (alpha*TOTAL^beta)/454, > start=list(alpha=1, beta=3), > data=spottedseatrout2004.female.data) > > I am trying to fit above model to length-weight data of a fish species > (spotted seatrout) by

Re: [R] Error 'singular gradient' in nonlinear model fitting

2008-02-15 Thread Ravi Varadhan
ilto:[EMAIL PROTECTED] On Behalf Of HongSheng Liao Sent: Friday, February 15, 2008 12:42 PM To: r-help@r-project.org Subject: [R] Error 'singular gradient' in nonlinear model fitting w.age.female.2004 <- nls(WEIGHT ~ (alpha*TOTAL^beta)/454, start=lis

[R] Error 'singular gradient' in nonlinear model fitting

2008-02-15 Thread HongSheng Liao
w.age.female.2004 <- nls(WEIGHT ~ (alpha*TOTAL^beta)/454, start=list(alpha=1, beta=3), data=spottedseatrout2004.female.data) I am trying to fit above model to length-weight data of a fish species (spotted seatrout) by year (1999-2006). The converge