if someone is interested in (last night I dreamed):
...
SH <- -3.8
fn.2 <- function(p){
for (i1 in ilong){
ex[i1] <- (1E-19*p[1]*n*L*voigt(u,v,f[i1],foN,p[2],p[3])[[1]])
}
sum((log(tt)-ex)^2)
}
out.2 <-nlm(fn.2, p = c(SH,GG,GL), hessian = TRUE,
steptol = 1e-6, iterlim = 1000,print.level=2)
SN <- out.2$estimate[1]*(1E-19)
GGN <- out.2$estimate[2]
GLN <- out.2$estimate[3]
...
which works:
...iteration = 10 Parameter: [1] -3.800003499 0.005181922 0.006242639 Function Value [1] 1.068854 Gradient: [1] -0.01561930 0.02665086 -0.01232618
Successive iterates within tolerance. Current iterate is probably solution. ...
Another thing: Since I receive the R-help digest
I have this in my ~/.emacs:
(custom-set-faces
'(font-lock-string-face ((t (:foreground "green3"))))
'(font-lock-keyword-face ((t (:foreground "#f939ff"))))
)
(defun highlight-R-help ()
(interactive)
(highlight-regexp "Message:" 'font-lock-keyword-face)
(highlight-regexp "Date:" 'font-lock-string-face)
(highlight-regexp "Subject:" 'font-lock-string-face)
(highlight-regexp "From:" 'font-lock-string-face))
(defun unhighlight-R-help () (interactive) (unhighlight-regexp "Message:" ) (unhighlight-regexp "Date:" ) (unhighlight-regexp "Subject:") (unhighlight-regexp "From:")) ;;;
regards Thomas
______________________________________________ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
