Why is computational reproducibility important?
summary
Explains why computational reproducibility — rerunning the same code and data to obtain the same results — matters by citing an ecology survey showing ~73% of papers are nonreproducible, with common problems including non‑open software (26%), missing data (56%), and omitted software versions (46%). Recommends concrete practices: use relative directories, record and cache sessions, adopt version control, and publish code on GitHub or similar so teams can reuse scripts, reviewers can verify analyses, and errors are caught early.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:10What is computational reproducibility
- 5:04Common problems in papers
- 7:16Why it matters
- 9:42Selfish benefits
- 12:33Research team recommendations
- 13:17Advantages for reviewers
- 14:22Open code publishing
- 15:24Call for cultural change
related talks
21mA guide to modern reproducible data science with R
48mCommon Data Science Mistakes
41mAssessing the Quality of Scientific Research and Its Data
21mEmpowering Reproducible Finance through Tidy Finance with R and Python
46mElegant R: Tips from a Software Developer
48mHuman data interactions: experiments in how humans analyze data
35mTowards Interpretable Data Science
22m. Cultivating Your Own R Ecosystem as a Solo Contributor