🔬 Scientific Research · Reproducibility
Make computational results reproducible
Environment pinning, deterministic pipelines, and one-command reruns from raw data.
intermediate~40 minResearchersPhD StudentsPostdocs
Steps
- 1Pin environments: lockfiles plus container image recorded per result
- 2Version control everything including parameter files and seeds
- 3Structure pipeline raw → processed → figures with explicit stages
- 4Never edit intermediate files manually; regenerate instead
- 5Test one-command rerun on clean machine before submission
- 6Publish code/data under DOIs via archival repositories
Common Pitfalls
- ▲'It works on my laptop' dependency webs
- ▲Figures hand-tweaked after generation
Commands
Install with skills CLI
$ npx skills add aniruddhaadak80/skills --skill reproducibility-computational-reproducibilityInstall globally
$ npx skills add aniruddhaadak80/skills --skill reproducibility-computational-reproducibility -gTags
#reproducibility#pipelines#research-code#scientific-research