cleanup

Signed-off-by: Anatoly Myachev <[email protected]>

modin/core/storage_formats/pandas/query_compiler.py

@@ -3088,7 +3088,6 @@ def mapper(df: pandas.DataFrame):
                     )
                     # we have to keep other columns so setting their mask
                     # values with `False`
-                    # TODO: pyarrow backend?
                     mask = pandas.Series(
                         np.zeros(df.shape[1], dtype=bool), index=df.columns
                     )

modin/distributed/dataframe/pandas/partitions.py

@@ -90,7 +90,7 @@ def unwrap_partitions(
             f"Only API Layer objects may be passed in here, got {type(api_layer_object)} instead."
         )
 
-    modin_frame = api_layer_object._query_compiler._modin_frame
+    modin_frame = api_layer_object._query_compiler._modin_frame  # type: ignore[attr-defined]
     modin_frame._propagate_index_objs(None)
     if axis is None:
 
@@ -122,7 +122,7 @@ def get_block(partition: PartitionUnionType) -> np.ndarray:
                 ]
 
         actual_engine = type(
-            api_layer_object._query_compiler._modin_frame._partitions[0][0]
+            api_layer_object._query_compiler._modin_frame._partitions[0][0]  # type: ignore[attr-defined]
         ).__name__
         if actual_engine in (
             "PandasOnRayDataframePartition",

modin/logging/logger_decorator.py

@@ -19,7 +19,7 @@
 
 from functools import wraps
 from types import FunctionType, MethodType
-from typing import Any, Callable, Dict, Optional, Tuple, Type, TypeVar, Union, overload
+from typing import Any, Callable, Dict, Optional, Tuple, Type, Union
 
 from modin.config import LogMode
 
@@ -28,9 +28,6 @@
 _MODIN_LOGGER_NOWRAP = "__modin_logging_nowrap__"
 
 
-Fn = TypeVar("Fn", bound=Callable)
-
-
 def disable_logging(func: Callable) -> Callable:
     """
     Disable logging of one particular function. Useful for decorated classes.
@@ -49,17 +46,11 @@ def disable_logging(func: Callable) -> Callable:
     return func
 
 
-@overload
-def enable_logging(modin_layer: Fn) -> Fn:
-    # This helps preserve typings when the decorator is used without parentheses
-    ...
-
-
 def enable_logging(
-    modin_layer: Union[str, Fn, Type] = "PANDAS-API",
+    modin_layer: Union[str, Callable, Type] = "PANDAS-API",
     name: Optional[str] = None,
     log_level: LogLevel = LogLevel.INFO,
-) -> Callable[[Fn], Fn]:
+) -> Callable:
     """
     Log Decorator used on specific Modin functions or classes.
 
@@ -85,7 +76,7 @@ def enable_logging(
         # def func()
         return enable_logging()(modin_layer)
 
-    def decorator(obj: Fn) -> Fn:
+    def decorator(obj: Any) -> Any:
         """Decorate function or class to add logs to Modin API function(s)."""
         if isinstance(obj, type):
             seen: Dict[Any, Any] = {}

modin/numpy/indexing.py

@@ -214,7 +214,7 @@ def boolean_mask_to_numeric(indexer):
             # `itertools.compress` masks `data` with the `selectors` mask,
             # works about ~10% faster than a pure list comprehension
             itertools.compress(data=range(len(indexer)), selectors=indexer),
-            dtype="int64",
+            dtype=np.int64,
         )
 
 
@@ -585,7 +585,7 @@ def _compute_lookup(self, row_loc, col_loc):
                     # `Index.__getitem__` works much faster with numpy arrays than with python lists,
                     # so although we lose some time here on converting to numpy, `Index.__getitem__`
                     # speedup covers the loss that we gain here.
-                    axis_loc = np.array(axis_loc, dtype="int64")
+                    axis_loc = np.array(axis_loc, dtype=np.int64)
                 # Relatively fast check allows us to not trigger `self.qc.get_axis()` computation
                 # if there're no negative indices and so they don't not depend on the axis length.
                 if isinstance(axis_loc, np.ndarray) and not (axis_loc < 0).any():

modin/pandas/dataframe.py

@@ -3067,7 +3067,6 @@ def _validate_dtypes_min_max(self, axis, numeric_only) -> DataFrame:
         ):
             # check if there are columns with dtypes datetime or timedelta
             if all(
-                # TODO: pyarrow backend?
                 dtype != pandas.api.types.pandas_dtype("datetime64[ns]")
                 and dtype != pandas.api.types.pandas_dtype("timedelta64[ns]")
                 for dtype in self.dtypes
@@ -3104,7 +3103,6 @@ def _validate_dtypes_sum_prod_mean(
             not axis
             and numeric_only is False
             and any(
-                # TODO: pyarrow backend?
                 dtype == pandas.api.types.pandas_dtype("datetime64[ns]")
                 for dtype in self.dtypes
             )
@@ -3125,7 +3123,6 @@ def _validate_dtypes_sum_prod_mean(
         ):
             # check if there are columns with dtypes datetime or timedelta
             if all(
-                # TODO: pyarrow backend?
                 dtype != pandas.api.types.pandas_dtype("datetime64[ns]")
                 and dtype != pandas.api.types.pandas_dtype("timedelta64[ns]")
                 for dtype in self.dtypes