[R] Regression and fitting

From: Gunther Höning <gunther.hoening_at_ukmainz.de>
Date: Tue, 01 Jul 2008 14:40:44 +0200

Dear list,

I have some values like
Time 2 4 8 24 48 72

```UTR	82543	169105	207615	96633		31988	7005
UTRs	82687	172934	205541	101842	31898	6950
```
of a twice repeated meassurement.

I know that the underlying function is of :

f(x) = SO/(sqrt(2*pi)*sigma *x) * exp(-(S0*ln x - mu)^2/(2*sigma^2)).

How can I determine the value of S0, sigma and mu.

I tried nls:

lognormal <- function(x,S0,sigma,mu) {S0/(sqrt(2*pi)*sigma *x) * exp(-(S0*log (x) - mu)^2/(2*sigma^2))}
x <- rep(c(2,4,6,24,48,72),2)
y <-
c(82543,169105,207615,96633,31988,7005,82687,172934,205541,101842,31898,6950 )
dat <- data.frame(x,y)
regfit <- nls(y ~ lognormal(x,S0,sigma,mu) ,data =dat)

Output:

Error in nlsModel(formula, mf, start, wts) :

singular gradient matrix at initial parameter estimates Warning message:
No starting values specified for some parameters. Intializing 'S0', 'sigma', 'mu' to '1.'. Consider specifying 'start' or using a selfStart model in: nls(y ~ lognormal(x, S0, sigma, mu), data = dat)

Can anybody help on this topic please ?

Gunther

R-help_at_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. Received on Tue 01 Jul 2008 - 12:54:20 GMT

Archive maintained by Robert King, hosted by the discipline of statistics at the University of Newcastle, Australia.
Archive generated by hypermail 2.2.0, at Tue 01 Jul 2008 - 14:31:42 GMT.

Mailing list information is available at https://stat.ethz.ch/mailman/listinfo/r-help. Please read the posting guide before posting to the list.