test-intelligence-algorithms

Validating sports science algorithms involves testing VDOT, TSS, TRIMP, and VO2max calculations against research-based reference values using deterministic tests with synthetic athlete data to ensure mathematical correctness and physiological plausibility.

Testing VDOT and FTP calculations after updates requires running intelligence test fixtures via the Rust toolchain, using synthetic athlete data with known outcomes to verify that race predictions remain within expected physiological ranges.

Yes, a Rust toolchain is required to test sports science recovery and nutrition calculations, as the validation process relies on running deterministic tests through cargo test commands with intelligence test fixtures.

Running deterministic tests for VO2max and TRIMP calculations is done by executing `cargo test intelligence -- --nocapture`, which validates algorithm outputs against synthetic athlete data and research-based reference values.

Validating TSS and nutrition algorithms includes testing edge cases and error conditions against known benchmarks, verifying mathematical correctness for extreme inputs and ensuring physiological plausibility across diverse synthetic athlete profiles.

Yes, verifying VDOT race prediction accuracy for intermediate runners is achieved by running intelligence tests that compare 5K race prediction outputs against expected physiological ranges using deterministic synthetic athlete data fixtures.