Data Methodology
EatYay is built on a foundation of verified, objective nutrition facts. This document explains our data sources, mathematical scaling, Daily Value benchmarks, and reporting principles.
1. Data Provenance & Source Authority
Our foundational nutrition dataset is sourced directly from the U.S. Department of Agriculture (USDA) FoodData Central foundation and SR Legacy databases. Every whole food on EatYay retains its canonical USDA record identification (FDC ID), raw laboratory nutrient values, and official food description.
Restaurant nutrition items derive from verified brand disclosure documents, corporate nutritional guides, and published ingredient reports, cited on each individual menu item page.
2. Standardized Mass Basis (Per 100 Grams)
To enable scientific, un-biased side-by-side comparison across disparate foods, all underlying nutrient records are stored and calculated on a standardized 100-gram mass basis. Household portion sizes (e.g., 1 cup, 1 medium fruit, 1 oz, 1 tablespoon) scale dynamically from explicit USDA gram weights.
This standardized basis prevents distorted comparisons caused by arbitrary commercial serving sizes, allowing users to evaluate foods on a true nutrient-density footing.
3. Daily Value (%DV) Calculations
Percent Daily Value (%DV) figures indicate how much a nutrient in a serving of food contributes to a daily diet, based on the U.S. Food and Drug Administration (FDA) reference daily intakes for a 2,000-calorie daily diet:
4. Strict Null Semantics (Unknown vs. Zero)
EatYay enforces strict null semantics. If a nutrient value was not reported or tested by the USDA for a specific food item, it is represented as null (displayed as Not available (—)). It is never assumed to be zero.
In multi-ingredient recipe aggregation or side-by-side food comparisons, missing values propagate diagnostic status (Partial / Incomplete) rather than returning misleading zero sums. This prevents dangerous assumptions regarding nutrient intake.
5. Rounding Policy
Calculations maintain unrounded IEEE floating-point precision throughout internal processing. Display values round to one decimal place (0.1) for macro and micro values, and whole integers for calories and Daily Value percentages, ensuring consistent mathematical fidelity across the platform.