MCPcopy
hub / github.com/microsoft/qlib / _dataset

Method _dataset

qlib/data/cache.py:1084–1114  ·  view source on GitHub ↗
(
        self, instruments, fields, start_time=None, end_time=None, freq="day", disk_cache=1, inst_processors=[]
    )

Source from the content-addressed store, hash-verified

1082 )
1083
1084 def _dataset(
1085 self, instruments, fields, start_time=None, end_time=None, freq="day", disk_cache=1, inst_processors=[]
1086 ):
1087 if disk_cache == 0:
1088 # In this case, data_set cache is configured but will not be used.
1089 return self.provider.dataset(instruments, fields, start_time, end_time, freq)
1090 self.local_cache_path.mkdir(exist_ok=True, parents=True)
1091 cache_file = self.local_cache_path.joinpath(
1092 self._uri(
1093 instruments, fields, start_time, end_time, freq, disk_cache=disk_cache, inst_processors=inst_processors
1094 )
1095 )
1096 gen_flag = False
1097
1098 if cache_file.exists():
1099 if disk_cache == 1:
1100 # use cache
1101 df = pd.read_pickle(cache_file)
1102 return self.cache_to_origin_data(df, fields)
1103 elif disk_cache == 2:
1104 # replace cache
1105 gen_flag = True
1106 else:
1107 gen_flag = True
1108
1109 if gen_flag:
1110 data = self.provider.dataset(
1111 instruments, normalize_cache_fields(fields), start_time, end_time, freq, inst_processors=inst_processors
1112 )
1113 data.to_pickle(cache_file)
1114 return self.cache_to_origin_data(data, fields)
1115
1116
1117class DatasetURICache(DatasetCache):

Callers

nothing calls this directly

Calls 6

_uriMethod · 0.95
normalize_cache_fieldsFunction · 0.85
cache_to_origin_dataMethod · 0.80
datasetMethod · 0.45
existsMethod · 0.45
to_pickleMethod · 0.45

Tested by

no test coverage detected