def test_find_processed_hourly_files_non_madis(self):
"""Ensures correct output from find_processed_hourly_files.
In this case, primary data source is non-MADIS.
"""
these_file_names, _ = raw_wind_io.find_processed_hourly_files(
start_time_unix_sec=PERIOD_START_TIME_UNIX_SEC,
end_time_unix_sec=PERIOD_END_TIME_UNIX_SEC,
primary_source=NON_MADIS_PRIMARY_SOURCE,
top_directory_name=TOP_DIRECTORY_NAME, raise_error_if_missing=False)
self.assertTrue(
these_file_names == PROCESSED_HOURLY_FILE_NAMES_NON_MADIS)
def _read_wind_observations(storm_object_table,
max_time_before_storm_start_sec,
max_time_after_storm_end_sec, top_directory_name):
"""Reads wind observations from one or more files.
:param storm_object_table: pandas DataFrame created by _read_storm_tracks.
:param max_time_before_storm_start_sec: Max wind time before beginning of
storm cell. If wind observation W occurs >
max_time_before_storm_start_sec before the first time in storm cell S, W
cannot be linked to S.
:param max_time_after_storm_end_sec: Max wind time after end of storm cell.
If wind observation W occurs > max_time_after_storm_end_sec after the
last time in storm cell S, W cannot be linked to S.
:param top_directory_name: Name of top-level directory with wind
observations (files created by `raw_wind_io.write_processed_file`).
:return: wind_table: pandas DataFrame with columns documented in
`raw_wind_io.write_processed_file`.
"""
min_wind_time_unix_sec = numpy.min(storm_object_table[
tracking_io.TIME_COLUMN].values) - max_time_before_storm_start_sec
max_wind_time_unix_sec = numpy.max(storm_object_table[
tracking_io.TIME_COLUMN].values) + max_time_after_storm_end_sec
wind_file_names, _ = raw_wind_io.find_processed_hourly_files(
start_time_unix_sec=min_wind_time_unix_sec,
end_time_unix_sec=max_wind_time_unix_sec,
primary_source=WIND_DATA_SOURCE,
top_directory_name=top_directory_name,
raise_error_if_missing=True)
num_files = len(wind_file_names)
list_of_wind_tables = [None] * num_files
for i in range(num_files):
print('Reading wind observations from file {0:d}/{1:d}:'
' "{2:s}"...').format(i + 1, num_files, wind_file_names[i])
list_of_wind_tables[i] = raw_wind_io.read_processed_file(
wind_file_names[i])[REQUIRED_WIND_COLUMNS]
if i == 0:
continue
list_of_wind_tables[i], _ = list_of_wind_tables[i].align(
list_of_wind_tables[0], axis=1)
return pandas.concat(list_of_wind_tables, axis=0, ignore_index=True)