NASA RUL project - automatic feature generation

auto_feature_generation.py

all_rollings_df_X = all_rollings_df.drop(columns=['RUL'])

extracted_features = extract_features(all_rollings_df_X,
                                      column_id="instance_id", column_sort="time",
                                      n_jobs=4, default_fc_parameters=tsfresh.feature_extraction.settings.MinimalFCParameters())
impute(extracted_features)
features_filtered = select_features(extracted_features, y_train_rolling, n_jobs=4)

def get_last_window_from_unit(group_df):
  res_df = group_df[[x for x in test_df.columns if 'sensor_' in x]].iloc[-10:]
  res_df['time'] = np.arange(1, len(res_df) + 1)
  return res_df

test_df_X = test_df.sort_values(['unit_number', 'time']).groupby('unit_number').\
                          apply(get_last_window_from_unit).reset_index().drop(columns=['level_1'])
extracted_features_test = extract_features(test_df_X, column_id="unit_number", column_sort="time",
                                           n_jobs=4, default_fc_parameters=tsfresh.feature_extraction.settings.MinimalFCParameters())
extracted_features_test_min = extracted_features_test[features_filtered.columns]

xgbr_features_automatic = XGBRegressor()
xgbr_features_automatic.fit(features_filtered.values, y_train_rolling)

print_train_test_results(features_filtered.values,
                         extracted_features_test_min.values,
                         y_train_rolling, y_test, xgbr_features_automatic)