fix data provider init bug

qlib/__init__.py

@@ -35,6 +35,7 @@ def init(default_conf="client", **kwargs):
     if _logging_config:
         set_log_with_config(_logging_config)
 
+    # FIXME: this logger ignored the level in config
     LOG = get_module_logger("Initialization", level=logging.INFO)
     LOG.info(f"default_conf: {default_conf}.")
 

qlib/config.py

@@ -13,7 +13,8 @@ Two modes are supported
 
 import copy
 from pathlib import Path
-import re, os
+import re
+import os
 
 
 class Config:

qlib/data/data.py

@@ -1035,10 +1035,7 @@ def register_all_wrappers():
 
     _calendar_provider = init_instance_by_config(C.calendar_provider, module)
     if getattr(C, "calendar_cache", None) is not None:
-        _calendar_cache_config = {}
+        _calendar_provider = init_instance_by_config(C.calendar_cache, module, provide=_calendar_provider)
-        _calendar_cache_config.update(C.calendar_cache)
-        _calendar_cache_config["kwargs"].update(provider=_calendar_provider)
-        _calendar_provider = init_instance_by_config(_calendar_cache_config, module)
     register_wrapper(Cal, _calendar_provider, "qlib.data")
     logger.debug(f"registering Cal {C.calendar_provider}-{C.calenar_cache}")
 
@@ -1054,19 +1051,13 @@ def register_all_wrappers():
         # This provider is unnecessary in client provider
         _eprovider = init_instance_by_config(C.expression_provider, module)
         if getattr(C, "expression_cache", None) is not None:
-            _expression_cache_config = {}
+            _eprovider = init_instance_by_config(C.expression_cache, module, provider=_eprovider)
-            _expression_cache_config.update(C.expression_cache)
-            _expression_cache_config["kwargs"].update(provider=_eprovider)
-            _eprovider = init_instance_by_config(C.expression_cache, module)
         register_wrapper(ExpressionD, _eprovider, "qlib.data")
         logger.debug(f"registering ExpressioneD {C.expression_provider}-{C.expression_cache}")
 
     _dprovider = init_instance_by_config(C.dataset_provider, module)
     if getattr(C, "dataset_cache", None) is not None:
-        _dataset_cache_config = {}
+        _dprovider = init_instance_by_config(C.dataset_cache, module, provider=_dprovider)
-        _dataset_cache_config.update(C.dataset_cache)
-        _dataset_cache_config["kwargs"].update(provider=_dprovider)
-        _dprovider = init_instance_by_config(_dataset_cache_config, module)
     register_wrapper(DatasetD, _dprovider, "qlib.data")
     logger.debug(f"registering DataseteD {C.dataset_provider}-{C.dataset_cache}")
 

qlib/model/base.py

@@ -14,7 +14,7 @@ class BaseModel(Serializable, metaclass=abc.ABCMeta):
         pass
 
     def __call__(self, *args, **kwargs) -> object:
-        """ levarge Python syntactic sugar to make the models' behaviors like functions """
+        """ leverage Python syntactic sugar to make the models' behaviors like functions """
         return self.predict(*args, **kwargs)
 
 

qlib/utils/__init__.py

@@ -195,7 +195,7 @@ def get_cls_kwargs(config: Union[dict, str], module) -> (type, dict):
 
 
 def init_instance_by_config(
-    config: Union[str, dict], module=None, accept_types: Union[type, Tuple[type]] = tuple([])
+        config: Union[str, dict], module=None, accept_types: Union[type, Tuple[type]] = tuple([]), **kwargs
 ) -> object:
     """
     get initialized instance with config
@@ -229,8 +229,8 @@ def init_instance_by_config(
     if module is None:
         module = get_module_by_module_path(config["module_path"])
 
-    klass, kwargs = get_cls_kwargs(config, module)
+    klass, cls_kwargs = get_cls_kwargs(config, module)
-    return klass(**kwargs)
+    return klass(**cls_kwargs, **kwargs)
 
 
 def compare_dict_value(src_data: dict, dst_data: dict):

qlib/workflow/expm.py

@@ -9,7 +9,7 @@ from .exp import MLflowExperiment
 from .recorder import MLflowRecorder
 from ..log import get_module_logger
 
-logger = get_module_logger("workflow", "Warning")
+logger = get_module_logger("workflow", "WARN")
 
 
 class ExpManager: