Version 0.4-0 of the `multcomp' package has just been uploaded to CRAN.
Known bugs have been fixed and, inspired by some questions posted on
r-help, methods for `lm' and `glm' objects for the generics `simint' and
`simtest' have been implemented. Thanks to all bug-reporters!
Torsten
>From the CHANGES file:
0.4-0 (13.08.2003)
`simint' and `simtest' now have methods for `lm' and `glm'
the baseline level for Dunnett contrasts can be determined by the
`base' argument
`contr.Dunnett' computes the MP-inverse for use with `contrasts'
code cleanup in internal `parseformula'
`subset' and `na.action' work now (unused factor levels are removed)
we now check if the contrasts passed to `simint' and `simtest'
are estimable (and give a warning if this is not the case)
______________________________________________
[EMAIL PROTECTED] mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-help
