Department of Biostatistics Seminar/Workshop Series
Some New Directions in Statistical Computing
Frank E. Harrell, Jr., PhD
Professor & Chair, Department of Biostatistics, Vanderbilt University School of Medicine
AND
Jeff Horner, B.S.
Computer Systems Analyst III, Department of Biostatistics, Vanderbilt University School of Medicine
Wednesday, December 5, 1:30-2:30pm, MRBIII Room 1220
In the first part of this seminar, Frank Harrell will briefly discuss the current state of the art in making reproducible statistical reports, using R, knitr, and LaTeX. He will show one way to extend this structure by using a macro preprocessor to allow one to easily create repetitive sections of a document when the analysis code and sentences change very little. This pertains, for example, to the analysis of a series of dependent variables, one per LaTeX \section or \chapter. Then he will discuss his ideas for new areas of emphasis for the Department of Biostatistics related to statistical computing and study design.
In the second half of the seminar, Jeffrey Horner will demonstrate the currently most promising model for the implementation and dissemination of web-based statistical computing tools: shiny (see
http://www.rstudio.com/shiny). The shiny platform, including the R shiny package, is ideal for implementing study design tools and power, sample size, and precision calculators. Jeffrey will demonstrate how to use shiny to implement an interactive survival analysis power module that allows for complex study design parameters such as delayed treatment effect, drop-in, and drop-out. But uses of shiny may go farther than that.