How benilton/wcs-rio works
Project Type and Purpose
• It provides helper scripts (in R and Python/Jupyter), input data files, and Quarto notebooks that render an HTML website (in the docs/ folder).
Key Architectural Components
a. Data Sources
• GSM4339769_C141_filtered_feature_bc_matrix.h5 (primary HDF5 data)
• degs_c2xc0.csv.gz (precomputed CSV data)
• data_source.txt (data provenance notes)
b. Analysis Code
• R scripts: helpers_wcs.R, setup_wcs.R
• Quarto R Markdown (*.qmd): de_annot.qmd, norm_clust.qmd
• Python/Jupyter notebook: de_annot.ipynb
c. Build & Rendering Engine
• Quarto (configured via _quarto.yml)
• Site assets (CSS, reveal.js plugins in docs/site_libs/)
d. Output Website
• Static HTML files (docs/*.html)
• Static assets (docs/style.css, JS, fonts)
e. Version Control & Deployment
• Git repo (managed by .gitignore, .nojekyll)
• Deployed via GitHub Pages or similar (serving docs/)
Component Interactions & Data Flow
• Raw Data → Analysis Code
→ Intermediate Results (CSV, R objects, etc.)
→ Quarto Rendering
→ Static HTML + Assets in docs/
→ Served as a website (e.g., GitHub Pages)
Architectural Patterns & Principles
• Workflow Pipeline: data ingestion → processing → documentation
• Separation of concerns:
• Data layer (files)
• Compute layer (scripts & notebooks)
• Presentation layer (Quarto → HTML)
• Reproducibility: All steps and dependencies declared in scripts and _quarto.yml.
• Clear labels on all components (Data, Scripts, Renderer, Output)
• Callouts for key config files (_quarto.yml, .gitignore) and deployment target (GitHub Pages)