Hreneee/west-africa-nutrition-ranking

# West Africa Nutrition Ranking Project This project targets a resource-constrained population health problem by ranking foods in West Africa. The goal is not to produce a generic nutrient-density score. The pipeline combines food composition data, PubMed-derived deficiency signals, and TOPSIS multi-criteria decision analysis so rankings reflect region-relevant nutrition priorities. The PubMed component is a structured literature scan, not a systematic review. The output should be interpreted as a transparent prioritization tool, not a definitive policy recommendation. ## Current Pipeline Status The current pipeline uses dashed, threshold-agnostic filenames. Latest verified default run: - Total WAFCT foods screened: `1028` - Excluded by eligibility filter: `26` - Pending manual review and excluded until resolved: `2` - Eligible before complete-case filtering: `1000` - Missingness threshold: `10%` - Foods removed for missing retained nutrient data: `166` - Foods ranked: `834` - Nutrients used in TOPSIS: `17` The latest run summary is saved at: ```text outputs/pipeline-run-summary.txt ``` ## Directory Layout ```text west-africa-nutrition-ranking/ data/ raw/ wafct_raw.xlsx processed/ dashed pipeline outputs outputs/ figures/ pipeline-run-summary.txt removed-columns-log.csv scripts/ config.py pipeline.py run-pipeline.py 01_clean_wafct.py ... 10_generate_figures.py ``` ## Requirements Use the project virtual environment if it already exists: ```bash .venv/bin/python scripts/run-pipeline.py ``` If you need to recreate the environment: ```bash python3 -m venv .venv source .venv/bin/activate pip install -r requirements.txt ``` The pipeline uses the cached PubMed file: ```text data/processed/pubmed-west-africa-deficiency-2015-present.csv ``` Do not refetch PubMed unless you intentionally want to update the literature record set. If PubMed must be refetched, set an NCBI contact email in the environment before running the fetch script: ```bash export NCBI_E …