creating-analysis-projects

Structure a reproducible R project by separating read-only inputs in a `read/` directory, tracked pipeline code in `scripts/`, flat checkpoint intermediates in `checkpoints/`, and final outputs in `write/`. This strict separation prevents data corruption and ensures analysis reproducibility.

Organize Quarto QMD notebooks in a bioinformatics pipeline by using numbered file names within the `scripts/` directory. This enforces a sequential execution order and pairs with flat checkpoint handoffs to maintain clear data provenance across pipeline steps.

Yes, you can refactor existing R scripts by migrating them into numbered QMD notebooks within the `scripts/` directory. The scaffolding process audits and reorganizes code to ensure it follows strict directory conventions and checkpoint handoff mechanics.

Checkpoint intermediates in reproducible research pipelines work by saving flat data outputs in a dedicated `checkpoints/` directory after each numbered script step. This allows downstream QMD notebooks to load immutable intermediate states without re-running prior analysis stages.

You need integration with dependent skills for R code generation and Quarto document structure rather than specific R packages. The scaffolding enforces naming conventions and directory layout while relying on these external tools to generate the actual pipeline code.

Separate generated outputs into `write/figures/` and `write/tables/` directories in R projects to maintain per-notebook output provenance. This strict directory architecture ensures that final results are cleanly isolated from immutable inputs and intermediate checkpoint data.