GT-AWT: Game-Theoretic Adaptive Wavelet Thresholding for H-Infinity Robust Filtering Under Non-Stationary Unbounded Disturbances — Code and Data

Code and data accompanying a manuscript on Game-Theoretic Adaptive Wavelet Thresholding (GT-AWT), a method that reformulates wavelet thresholding as a zero-sum minimax game coupled to an H-infinity robust filter for signal estimation under non-stationary, unbounded disturbances. Includes: (1) real financial market data , daily prices and log-returns for 64 JSE-listed equities, 2015–2026; (2) real South African Eskom load-shedding stage records, 2022–2025; (3) the full Python implementation of the proposed method and nine baseline denoising methods it is compared against; (4) the experimental and simulation scripts used to produce all reported results. See the included README for full dataset provenance and reproducibility notes.