R help - Jul 2005 - how to solve the step halving factor problems in gnls and nls

Hi R-users,

Could you give me some advice in 
solving the problem of such error message from gnls and nls?
## begin error message

"Problem in gnls(y1 ~ glogit4(b, c, m, t, x), data.frame(x..: Step halving 
factor reduced below minimum in NLS step "

##and

"Problem in nls(y ~ 1/(1 + exp((xmid - x)/scal)), data = x..: step factor 
reduced below minimum "?

Thank you in advance.

Hanna
	[[alternative HTML version deleted]]

Yimeng Lu <yl2058 <at> columbia.edu> writes:
> Could you give me some advice in 
> solving the problem of such error message from gnls and nls?
> ## begin error message
> 
> "Problem in gnls(y1 ~ glogit4(b, c, m, t, x), data.frame(x..: Step
halving
> factor reduced below minimum in NLS step "
> 
Try to set nlsTol in the optional control parameter (gnlsControl) to a larger 
value, e.g. 0.1 instead of default 0.001, but be sure to check your results 
make sense. If this should help, you can try intermediate values.

Dieter Menne