data_fetcher.py 文件源码

def output(self):
        ''' 
        Generate data wrapper for Mahali temperatures

        @return Mahali temperature data wrapper
        '''

        # Function to extract date from filename (only month/day/year, no hours/minutes/seconds)
        def toDateTime(in_filename):
            return pd.to_datetime(pd.to_datetime(in_filename[7:25]).strftime('%Y-%m-%d'))


        # Read in file list:
        mahali_temperature_info = resource_filename('skdaccess', os.path.join('support','mahali_temperature_info.txt'))
        filenames = pd.read_csv(mahali_temperature_info,header=None,
                                names=('station','filename'),
                                skipinitialspace=True)


        # Create a columns of dates
        filenames['date'] = filenames['filename'].apply(toDateTime)

        # Need to grab day before as data can spill over
        adjusted_start_date = self.start_date - pd.to_timedelta('1d')
        adjusted_end_date = self.end_date + pd.to_timedelta('1d')


        station_list = self.ap_paramList[0]()
        # Get data for each selected station one day before until one day afte requested date
        index_to_retrieve = np.logical_and.reduce([filenames.loc[:, 'station'].apply(lambda x: x in station_list),
                                                   filenames.loc[:, 'date'] >= adjusted_start_date,
                                                   filenames.loc[:, 'date'] <= self.end_date])

        all_temperature_data = self.retrieveOnlineData(filenames[index_to_retrieve])

        # Due to data spillover, cut each data frame in dictionary
        for station in all_temperature_data.keys():
            all_temperature_data[station] = all_temperature_data[station].loc[adjusted_start_date:adjusted_end_date]

        # Return table wrapper of data
        return TableWrapper(all_temperature_data, default_columns = ['Temperature'])