Marko Djordjevic 40d6d1739e fix(ml): add windowed feature flattening for inference parity ... The model was trained on 94-candle sliding windows flattened to 2820 features (94 candles x 30 features). Inference was sending raw per-candle features (27 columns). Changes: - Rewrite preprocessing to return (X, window_times) tuple - Add sliding window creation with correct feature ordering - Fill missing columns (average, barCount) with 0 for feature parity - Fill NaN from indicator warmup with 0 instead of dropping rows - Always compute all indicators (including MFI) for feature parity - Update predict and batch predict endpoints for new signature		2026-02-15 22:07:06 +01:00
..
ml	fix(ml): add windowed feature flattening for inference parity	2026-02-15 22:07:06 +01:00