Tutorial: Oil vs Chips Matrix Effects¶
- Data:
examples/data/oil_synthetic.csv,examples/data/chips_synthetic.csv - Notebook:
examples/notebooks/02_oil_vs_chips_matrix_effects.ipynb - Protocol:
examples/protocols/Chips_vs_Oil_MatrixEffects_v1.yml - Focus: Divergence markers, effect sizes, interpretation of ratios.
Tutorial – Oil vs chips matrix effects¶
What this tutorial covers¶
- Problem: Matrix effects—markers may behave differently in pure oils vs chips.
- Datasets:
examples/data/oils.csvandexamples/data/chips.csv(matching peak/ratio columns, metadata). - Protocol:
examples/protocols/oil_vs_chips.yaml(divergence analysis).
Why it matters (theory)¶
Matrix components (starch/protein) can alter mean ratios, CV, and heating trends. Identifying matrix-robust vs matrix-sensitive markers is critical for QA. See rq_engine_theory.md.
CLI workflow¶
foodspec-run-protocol \
--input examples/data/oils.csv \
--input examples/data/chips.csv \
--protocol examples/protocols/oil_vs_chips.yaml \
--output-dir runs/oil_vs_chips_demo
foodspec-publish runs/oil_vs_chips_demo/<timestamp> --fig-limit 6
figures/ for divergence plots and trend charts; tables/ for oil_vs_chips summaries.
Example figure (from run bundle)¶

How to read the results¶
- Divergence tables: look for significant differences (post-FDR) in mean, CV, or trends.
- Effect sizes (e.g., Cohen’s d, slope deltas) quantify practical impact.
- Interpretation tags (e.g., “stable in oil, unstable in chips”) highlight matrix sensitivity.
- Prefer matrix-robust markers for cross-matrix QA; avoid markers that flip behavior between oil and chips.
Cross-links¶
- Cookbook: cookbook_rq_questions.md and cookbook_preprocessing.md
- Theory: rq_engine_theory.md