def create_features_and_label()

in 11_realtime/flightstxf/flights_transforms.py [0:0]

• 36 lines of code
• 10 McCabe index (conditional complexity)

def create_features_and_label(event, for_training):
    try:
        model_input = {}

        if for_training:
            model_input.update({
                'ontime': 1.0 if float(event['ARR_DELAY'] or 0) < 15 else 0,
            })

        # features for both training and prediction
        model_input.update({
            # same as in ch9
            'dep_delay': event['DEP_DELAY'],
            'taxi_out': event['TAXI_OUT'],
            # distance is not in wheelsoff
            'distance': approx_miles_between(event['DEP_AIRPORT_LAT'], event['DEP_AIRPORT_LON'],
                                             event['ARR_AIRPORT_LAT'], event['ARR_AIRPORT_LON']),
            'origin': event['ORIGIN'],
            'dest': event['DEST'],
            'dep_hour': to_datetime(event['DEP_TIME']).hour,
            'is_weekday': 1.0 if to_datetime(event['DEP_TIME']).isoweekday() < 6 else 0.0,
            'carrier': event['UNIQUE_CARRIER'],
            'dep_airport_lat': event['DEP_AIRPORT_LAT'],
            'dep_airport_lon': event['DEP_AIRPORT_LON'],
            'arr_airport_lat': event['ARR_AIRPORT_LAT'],
            'arr_airport_lon': event['ARR_AIRPORT_LON'],
            # newly computed averages
            'avg_dep_delay': event['AVG_DEP_DELAY'],
            'avg_taxi_out': event['AVG_TAXI_OUT'],

        })

        if for_training:
            model_input.update({
                # training data split
                'data_split': get_data_split(event['FL_DATE'])
            })
        else:
            model_input.update({
                # prediction output should include timestamp
                'event_time': event['WHEELS_OFF']
            })

        yield model_input
    except Exception as e:
        # if any key is not present, don't use for training
        logging.warning('Ignoring {} because: {}'.format(event, e), exc_info=True)
        pass