[R-sig-dyn-mod] Fitting parameters in a Verlhust population growth model
Thomas Petzoldt
Thomas.Petzoldt at tu-dresden.de
Fri Nov 26 20:23:18 CET 2010
Hello Jose,
this looks indeed like the typical order-of-magnitude-problem. Have you
already tried to rescale the parameters? Section 2.4.3 in the
"simecol-Howto" vignette explains how this can be done.
The vignette is included in the package, but can also be found online:
http://cran.r-project.org/web/packages/simecol/vignettes/b-simecol-howtos.pdf
The following seems to work:
resultado <- fitOdeModel(verlhust,
obstime=censo$ano,
yobs=subset(censo,select=poblacion),
fn=ssqOdeModel,
initialize=FALSE,
method="PORT", scale=c(1/0.01, 1/1e8),
lower=c(alfa=0,L=0))
Note also that setting solver="ode" speeds up by about 50%.
Thomas P.
More information about the R-sig-dynamic-models
mailing list