similar to: nls(): Levenberg-Marquardt, Gauss-Newton, plinear - PI curve fitting

Hi, It seems to me that the most suitable method in R for curve-fitting is the use of nls, which uses a Gauss-Newton (GN) algorithm, while the use of the Levenberg-Marquardt (LM) algorithm does not seem to be very stressed in R. According to this [1] by Ripley, 'Levenberg-Marquardt is hardly competitive these days' which could imply the low emphasize on LM in R. The position of LM is, to

Hi, I have a problem fitting the following Weibull Model to a set of data. The model is this one: a-b*exp(-c*x^d) If I fitted the model with CurveExpert I can find a very nice set of coefficients which create a curve very close to my data, but when I use the nls.lm function in R I can't obtain the same result. My data are these: X Y 15 13 50 13 75 9 90 4 With the commercial

Hi All, Is the Levenberg-Marquardt algorithm available in R. This method combines the steepest descent algorithm and Newton's method. Thanks in Advance, Dermot MacSweeney. ************************************************************** Dermot MacSweeney NMRC, Email: dsweeney at nmrc.ucc.ie Lee Maltings, Tel: +353 21 904178 Prospect Row, Fax: +353 21 270271 Cork, WWW:

Hello, I want to fit some non linear models with the Levenberg-Marquardt algorithm. It doesn''t seem to exist any function to do this in R ( well, maybe one does, but I''m a new user, and the only documentation I have is "An introduction to R"). I''d like to know if this function exists, maybe throught an additionnal package. I''d also like to know if if

Dear all, I would like to use the Levenberg-Marquardt algorithm for non-linear least-squares regression using function nls.lm. Can anybody help me to find a a way to compute confidence intervals on the fitted parameters as it is possible for nls (using confint.nls, which does not work for nls.lm)? Thank you for your help Michael

Hello, how can I use the function "cor()" with x and y in function "aggregate()" or "by()"? The data are like this: x y group 1 4 B 2 4 B 3 5 C I would like obtain the correlation between x and y for each subset. I don't want to use the workaround with the function subset(), because I have many groups. Thanks in advance. Christfried Kunath

Hi All. I've run into a problem with the plinear algorithm in nls that is confusing me. Assume the following reaction time data over 15 trials for a single unit. Trials are coded from 0-14 so that the intercept represents reaction time in the first trial. trl RT 0 1132.0 1 630.5 2 1371.5 3 704.0 4 488.5 5 575.5 6 613.0 7 824.5 8 509.0 9

Dear R gurus, As part of finding initial values for a much more complicated fit I want to fit a function of the form y ~ a + bx + cx^d to fairly "noisy" data and have hit some problems. To demonstrate the specific R-related problem, here is an idealised data set, smaller and better fitting than reality: # idealised data set aDF <- data.frame( x= c(1.80, 9.27, 6.48, 2.61, 9.86,

hello all i havnt had a chance to read through the references provided for the "nls" function (since the libraries are closed now). can anyone shed some light on how the "plinear" algorithm works? also, how are the fitted values obtained? also, WHAT DOES THE ".lin" below REPRESENT? thanking you in advance ###################################### i have a quick

I'm wondering what packages exist to implement nonlinear least squares regression in R other than 'nls'. Are there packages which implement methods to estimate the optimum values of the parameters which do not use the Gauss-Newton algorithm e.g. use Nelder Mead. In particular, I'd be interested where this is done where the methods of the plinear algorithm are also used (the initial

Dear sirs It seems like there is a bug in `profile.nls` with `algorithm = "plinear"` when a matrix is supplied on the right hand side. Here is the bug and a potential fix ##### # example where profile.nls does not work with `plinear` but does with # `default` require(graphics) set.seed(1) DNase1 <- subset(DNase, Run == 1) x <- rnorm(nrow(DNase1)) f1 <- nls(density ~ b1/(1 +

I was wondering if there is way to place constraints upon the "plinear" algorithm of nls, or rather is there a manner in which this can be achieved because nls does not allow this to be done. I only want to place constraints on one of the nonlinear parameters, a, such that it is between 0 and 1. I have attempted to use a=pnorm(a*) , but then the fitting procedure becomes

I want to fit a nonlinear model of the form: Y=A+B*X1+C*X2+log(X3/(X3+D)) I think that the best way is to use the plinear algorithm, but I don't know how to specify the formula in the nls function. I've tried: Y~cbind(rep(1,times=length(Y)),X1,X2,log(X3/(X3+D))) But this fits the model: Y=A+B*X1+C*X2+D1*log(X3/(X3+D)) How can I specify the formula correctly? Thanks, M

Dear r-Help readers, 1) Is there support for NSUR in some R package yet? 2) Is there a general function of applying the Gauss-Newton or Marquard method, in which the function of calculating the partial derivatives can be specified by the user? Contemporaneously correlated non-linear models (NSUR) is a method to fit a system of non-linear equations. I want to use to fit several non-linear

Hello, following functions doesnt work correct with my data: median(), geo.mean(). My datafiles contain more than 10.000 lines and six columns from a flow-cytometer-measurment. I need the arithmetic and geometric mean and median. For the calculation of the geometric mean i wrote following function: fix(geo.mean) function(x) { n<-length(x)

I have found some "issues" (bugs?) with nls confidence intervals ... some with the relatively new "port" algorithm, others more general (but possibly in the "well, don't do that" category). I have corresponded some with Prof. Ripley about them, but I thought I would just report how far I've gotten in case anyone else has thoughts. (I'm finding the code

Hi! I'm translating some code from Matlab to R and I found a problem. I need to translate Matlab's function 'lsqnonlin' (http://www-ccs.ucsd.edu/matlab/toolbox/optim/lsqnonlin.html) into R, and at the beginning I thought it would be the same as R's 'optim'. But then I looked at the definition of 'lsqnonlin' and I don't quite see how to make

In the course of revising a paper I have had occasion to attempt to maximize a rather complicated log likelihood using the function nlm(). This is at the demand of a referee who claims that this will work better than my proposed use of a home- grown implementation of the Levenberg-Marquardt algorithm. I have run into serious hiccups in attempting to apply nlm(). If I provide gradient and

Hi everyone, I'm trying to perform a bi exponential Fit with the package NLS. the plinear algorithm seems to be a good choice see: p<-3000 q<-1000 a<--0.03 b<--0.02 t<-seq(0:144);t y<-p*exp(a*t) + q*exp(b*t)+rnorm(t,sd=0.3*(p* exp(a*t) + q*exp(b*t))) fittA <- nls(y~cbind(exp(a*t), exp(b*t)), algorithm="plinear",start=list(a=-.1, b=-0.2), data=list(y=y, t=t),

Dear Mr Graves, Thank you very much for your response. Nobody else from this mailing list ventured to reply to me for the two weeks since I posted my question. "nlminb" and "optim" are just optimization procedures. What I need is not just optimization, but a nonlinear CURVE FITTING procedure. If there is some way to perform nonlinear curve fitting with the