[R] disagreement in loglikelihood and deviace in GLM with weights leads to different models selected using step()

Marian K Talbert mtalbert at usgs.gov
Thu May 10 17:19:50 CEST 2012


An embedded and charset-unspecified text was scrubbed...
Name: not available
URL: <https://stat.ethz.ch/pipermail/r-help/attachments/20120510/7f584450/attachment.pl>


More information about the R-help mailing list