Wellness Predict

ML-based wellness prediction from raw space weather features.

Accepts raw space weather features and returns ML predictions for both condition classification (5 classes) and health impact score (0-100). Requires trained models from geomagnetic_wellness_trainer.py.

JSON body must include space weather features: kp, sr_amplitude, solar_wind_speed_km_s, dst_nt, proton_flux_pfu, local_mag_var_nt.

Optional circadian/seasonal features: hour_of_day_sin, hour_of_day_cos, day_of_year_sin, day_of_year_cos.