jtleek / swfdr Goto Github PK

Copyright (C) 2011 Jeffrey T. Leek (http://www.biostat.jhsph.edu/~jleek/contact.html) and Leah R. Jager ([email protected])

Note: These functions were written on a Mac and may have difficulties when read on Windows machines.

This file contains the code to scrape the P-values from pubmed (either run it first, or use the already calculated pvalueData.rda)

This file contains the function to estimate the science-wise false discovery rate

This file contains the code to reproduce the quantities and figures in the Jager/Leek paper

This file contains a helper function necessary for journalAnalysis.R

The pre-computed p-value data used for the Jager/Leek paper in .rda format.

A simulation study comparing our estimates to the truth when the assumptions hold and when they are badly violated.

Either: (1) Run getPvalues.R to obtain the p-value data, (2) Run journalAnalysis.R Or: (2) Run journalAnalysis.R using the pre-computed pvalueData.rda

To get the simulated results you should run sensitivity.R (supplementary sensitivity analysis) and simulation.R (main text sensivity analysis)

Note, because of the bootstrapping calculations, these functions may take a while (think order hours) to run.