Scientific Data Provenance in R: RDataTracker and DDG Explorer

Scientific data provenance is the information required to document the history of an item of data, including how it was created and how it was transformed. Data provenance has great potential to improve the transparency, reliability, and reproducibility of scientific results. However it has been little used to date by domain scientists because most systems that collect provenance require scientists to learn specialized software tools and jargon. This project is developing tools that allow scientists to collect, visualize, and query provenance directly from the R statistical language. The first tool (RDataTracker) is a library of R functions that can be downloaded and installed as an R package. RDataTracker allows the scientist to collect data provenance during an R console session or while executing an R script. The resulting provenance is stored on the scientist's computer as a DDG (data derivation graph) file. The second tool (DDG Explorer) is a stand-alone Java program that can be downloaded and run to visualize, store, and query DDGs. The third tool is an R script (DDGCheckpoint.R) may be used with RDataTRacker to create and restore checkpoints that store the R environment and user files. Documentation for all tools is included with the RDataTracker package or may be downloaded separately.