Basic utility functions to support reproducible research. Based on ideas published by Christopher Gandrud. This package relies on four stages of data analysis:
- Gather - collect data
- Cook - alter raw data so that it is fit for analysis
- Analyze - construct models or other analysis
- Present - synthesize the analysis so that it may be presented to an end-user or decision maker
A sample workflow
represtools::NewResearch("Baseball")
represtools::Gather("Hitters")
# write some code
represtools::Cook("Hitters")
# write some code
represtools::Analyze("Handedness")
# write some code
represtools::Present("Handedness", title = "On the quality of right-handed batters", output = "html")
# write some code
represtools::Make()
This workflow ensures that anything which changes at an earlier stage of the process must force reconstruction of later stages.
More information is available here: http://pirategrunt.com/represtools/.