import pandas as pd
import numpy as np

# File paths
input_file = r''
output_file = r''

# Load the data
data = pd.read_csv(input_file)

# Define Hamming window parameters
window_duration = 10  # in seconds
sampling_rate = 100  # Assuming 100 Hz sampling rate; adjust if different
window_size = window_duration * sampling_rate

# Apply Hamming window
hamming_window = np.hamming(window_size)


def apply_window_to_segment(segment):
    return segment * hamming_window[:len(segment)]


# Process each sensor column
sensor_columns = [
    col for col in data.columns if col not in ['TimeStamps', 'Subject', 'Activity', 'Trial', 'Tag']
]


def apply_hamming_window(df):
    windowed_data = []

    for i in range(0, len(df), window_size):
        segment = df.iloc[i:i + window_size]
        if len(segment) == window_size:
            for col in sensor_columns:
                segment.loc[:, col] = apply_window_to_segment(segment[col].values)
            windowed_data.append(segment)

    return pd.concat(windowed_data, ignore_index=True)


windowed_data = apply_hamming_window(data)

# Save the output
windowed_data.to_csv(output_file, index=False)