Now the weibull model has an automatic choice of starting values, fixing a bug that prevented

fitting the weibull function in most cases in the last version. Also, the log EC50 is reported
as estimated with nlm from the fitted weibull model. Because of the important changes, mainly 
in naming the different models which leads to incompatibility with the old drfit syntax,
I am increasing the minor version number to 0.04 with this release.



git-svn-id: http://kriemhild.uft.uni-bremen.de/svn/drfit@30 d1b72e20-2ee0-0310-a1c4-ad5adbbefcdc

1 files changed, 5 insertions, 2 deletions

diff --git a/man/drfit.Rd b/man/drfit.Rd
index 0d9ee6f..35d6d94 100644
--- a/man/drfit.Rd
+++ b/man/drfit.Rd
@@ -8,7 +8,7 @@
 \usage{
   drfit(data, startlogEC50 = NA, chooseone = TRUE, probit = TRUE, logit = FALSE,
     weibull = FALSE, linlogit = FALSE, linlogitWrong = NA, allWrong = NA, 
-    b0 = 2, f0 = 0)
+    s0 = 0.5, b0 = 2, f0 = 0)
 }
 \arguments{
   \item{data}{
@@ -55,9 +55,12 @@
     logit, probit, and the first model that produces a valid fit is used.
     If FALSE, all models that are set to TRUE and that can be fitted will be
     reported.}
+  \item{s0}{
+    If the weibull model is fitted, s0 gives the possibility to adjust the
+    starting value for the shape parameter of \code{\link{pweibull}}.}
   \item{b0,f0}{
     If the linearlogistic model is fitted, b0 and f0 give the possibility to
-    adapt the starting values for the parameters b and f.}
+    adjust the starting values for the parameters b and f.}
 }
 \value{
   \item{results}{