RSS Matters

Hmisc: An Add-on Library for S-Plus

By Rich Herrington, Research and Statistical Support Services

This month we examine the freely available add-on library, Hmisc, for the S-Plus statistical environment. Libraries such as Hmisc greatly extend the functionality of S-Plus. These libraries are archived at the Carnegie Mellon University Website: http://lib.stat.cmu.edu/S/ . The Hmisc library can also found at it's official Website: http://hesweb1.med.virginia.edu/biostat/s/Hmisc.html

Description

The Hmisc library contains many functions useful for data analysis, high-level graphics, utility operations, functions for computing sample size and power, translating SAS datasets into S-PLUS, imputing missing values, advanced table making, variable clustering, character string manipulation, recoding variables, and bootstrap repeated measures analysis. Most of these functions were written by Frank Harrell, but a few were collected from statlib and from s-news. This collection of functions includes all of Harrell's submissions to statlib other than the functions in the Design and display libraries. A few of the functions do not have help documentation. The Hmisc library is extensively documented in a 286 page manual which can be obtained as a .pdf file from:  http://hesweb1.med.virginia.edu/biostat/s/doc/splus.pdf . The Win98/NT installation files can be found in zipped form at:  http://hesweb1.med.virginia.edu/biostat/s/win/ The Hmisc library and it's accompanying library, design, have been installed on the ACSLAB sever installation of S-Plus 2000 and is available for researchers on campus to access.

Using the Hmisc Library in S-Plus 2000

Once S-Plus has been started, the libraries can be accessed:

Invoking the Hmisc and design library will create pull down help menus on the main menu Help:

These helps include many examples which can be "cut" from the help windows back into the "command" window. This allows users to experiment with a function before using it with their own data. The following example is a simulation study using the repeated measures bootstrapping function rm.boot:

The following output is produced:

References

An Introduction to S-Plus and the Hmisc and Design Libraries.  Carlos Alzola and Frank Harrell. 
http://hesweb1.med.virginia.edu/biostat/s/doc/splus.pdf