run_preprocessing.py

import os
import numpy as np
import config
from data_pipeline.preprocessing import HAPTPreprocessor, UMAFallPreprocessor, process_dataset

def main():
    os.makedirs(config.HAPT_OUTPUT_DIR, exist_ok=True)
    os.makedirs(config.UMAFALL_OUTPUT_DIR, exist_ok=True)

    print("Processing HAPT Dataset")
    hapt_preprocessor = HAPTPreprocessor(config.HAPT_RAW_DIR, config.HAPT_LABELS_FILE)
    hapt_df = hapt_preprocessor.load_data()
    
    if not hapt_df.empty:
        print(f"HAPT data loaded. Shape: {hapt_df.shape}")
        X_hapt, y_hapt = process_dataset(hapt_df, config.HAPT_WINDOW_SIZE, config.HAPT_STEP_SIZE)
        np.save(os.path.join(config.HAPT_OUTPUT_DIR, "X_hapt.npy"), X_hapt)
        np.save(os.path.join(config.HAPT_OUTPUT_DIR, "y_hapt.npy"), y_hapt)
        print(f"HAPT features saved. X shape: {X_hapt.shape}, y shape: {y_hapt.shape}")
    else:
        print("HAPT data loading failed or returned empty.")

    print("\nProcessing UMAFall Dataset")
    umafall_preprocessor = UMAFallPreprocessor(config.UMAFALL_RAW_DIR, config.UMAFALL_LABEL_DICT)
    umafall_df = umafall_preprocessor.load_data()

    if not umafall_df.empty:
        print(f"UMAFall data loaded. Shape: {umafall_df.shape}")
        X_uma, y_uma = process_dataset(umafall_df, config.UMAFALL_WINDOW_SIZE, config.UMAFALL_STEP_SIZE)
        np.save(os.path.join(config.UMAFALL_OUTPUT_DIR, "X_umafall.npy"), X_uma)
        np.save(os.path.join(config.UMAFALL_OUTPUT_DIR, "y_umafall.npy"), y_uma)
        print(f"UMAFall features saved. X shape: {X_uma.shape}, y shape: {y_uma.shape}")
    else:
        print("UMAFall data loading failed or returned empty.")

if __name__ == "__main__":
    main()