NASA RUL project - windowing the data

data_windowing.py

WINDOW_SIZE = 20

def get_windowed_dataframes(df):
  df_groups = df.sort_values(['unit_number', 'time']).groupby('unit_number')
  all_rollings = []
  for _, group_df in df_groups:
    group_df_rolling = group_df.rolling(window=WINDOW_SIZE)  
    all_rollings.extend([wnd for wnd in group_df_rolling if len(wnd) == WINDOW_SIZE])
  return all_rollings


def get_windowed_xy(all_rollings):
  all_rollings_X = np.array([wnd[[x for x in wnd.columns if 'sensor_' in x]].values for wnd in all_rollings])
  all_rollings_y = np.array([wnd['RUL'].iloc[-1] for wnd in all_rollings]).clip(max=125)
  return all_rollings_X, all_rollings_y

all_rollings = get_windowed_dataframes(train_df)
X_train_rolling, y_train_rolling = get_windowed_xy(all_rollings)
X_test_rolling = np.array(test_df.sort_values(['unit_number', 'time']).groupby('unit_number').\
                          apply(lambda group_df: group_df[[x for x in test_df.columns if 'sensor_' in x]].\
                                iloc[-WINDOW_SIZE:].values).tolist())

for i, rolling_df in enumerate(all_rollings):
  rolling_df['instance_id'] = i
  rolling_df.drop(columns=['unit_number'], inplace=True)
all_rollings_df = pd.concat(all_rollings)