The Hmisc and Design libraries have been updated respectively to versions 1.4-2
and 1.1-1. New versions for Linux/Unix/Windows may be obtained from
http://hesweb1.med.virginia.edu/biostat/s/library/r . Web sites for the
libraries are http://hesweb1.med.virginia.edu/biostat/s/Hmisc.html and
http://hesweb1.med.virginia.edu/biostat/s/Design.html .
Thanks to Xiao Gang Fan for porting the libraries to Windows once again.
I still have not had time to fix details in the documentation that will allow R
CMD check to be passed, which would allow me to submit the libraries to CRAN.
Change logs may be obtained from
http://hesweb1.med.virginia.edu/biostat/s/library/common
Several bugs have been fixed, and new features added. In Hmisc, better use as
been made of "units" attributes of variables, and units are
incorporated into tables and plots produced by summary.formula. For plots,
plotmath is used, especially for annotating plots with test statistics produced
by summary(..., test=TRUE). A "matrix tapply" function (mApply) has
been added. A new LaTeX table formatting option, ctable, has been added to
latex.default. A new function getHdata makes it easy to download and import
datasets and their descriptions from our web site. latex functions were changed
to allow file="" to be specified, for inline inclusion of LaTeX code
in Sweave. Problems with sas.get and [.factor have been fixed. fit.mult.impute
was enhanced to compute Rubin's degrees of freedom for t-statistics for
scalar parameter tests, as well as to compute a missing information index.
In Design, glmD works much more reliably. plotmath is now used by plot.Design.
Bill Pikounis' <v_bill_pikounis at merck.com> improvements have been
incorporated in bootcov to allow both cluster and group to be specified.
offsets now work in lrm when NAs are present, and predict.Design now works with
offsets. bootcov now works with glmD, and summary and print work for glmD
(thanks: Fredrik Lundgren <fredrik.lundgren at
norrkoping.mail.telia.com>). validate.tree had a bug fix.
Extended documentation for the libraries, and an introduction to the S language
have been updated also (http://hesweb1.med.virginia.edu/biostat/s/doc/splus.pdf)
and now include more R-specific information.
Thanks to those who have reported bugs and fixes, and Happy New Year to all.
----
Frank E Harrell Jr Prof. of Biostatistics & Statistics
Div. of Biostatistics & Epidem. Dept. of Health Evaluation Sciences
U. Virginia School of Medicine http://hesweb1.med.virginia.edu/biostat
