deprecate load_data_frame

m3dev · Dec 20, 2024 · d302976 · d302976
1 parent b1f0463
commit d302976
Show file tree

Hide file tree

Showing 3 changed files with 10 additions and 5 deletions.
diff --git a/docs/for_pandas.rst b/docs/for_pandas.rst
@@ -37,19 +37,17 @@ This is useful when dataframe has nullable columns because pandas auto-conversio
 Easy to Load DataFrame
 ----------------------
 
-The :func:`~gokart.task.TaskOnKart.load_data_frame` method is used to load input ``pandas.DataFrame``.
+The :func:`~gokart.task.TaskOnKart.load` method is used to load input ``pandas.DataFrame``.
 
 .. code:: python
 
     def requires(self):
         return MakeDataFrameTask()
 
     def run(self):
-        df = self.load_data_frame(required_columns={'colA', 'colB'}, drop_columns=True)
+        df = self.load()
 
-This allows us to omit ``reset_index`` and ``drop`` when loading. If there is a missing column in an example above, ``AssertionError`` will be raised. This feature is useful for pipelines based on pandas.
-
-Please refer to :func:`~gokart.task.TaskOnKart.load_data_frame`.
+Please refer to :func:`~gokart.task.TaskOnKart.load`.
 
 
 Fail on empty DataFrame

diff --git a/docs/task_on_kart.rst b/docs/task_on_kart.rst
@@ -247,6 +247,9 @@ TaskOnKart.load_data_frame
 
 Please refer to :doc:`for_pandas`.
 
+.. warning::
+    This function is deprecated. Please use :func:`~gokart.task.TaskOnKart.load` instead.
+
 
 TaskOnKart.fail_on_empty_dump
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

diff --git a/gokart/task.py b/gokart/task.py
@@ -6,6 +6,7 @@
 from importlib import import_module
 from logging import getLogger
 from typing import Any, Callable, Dict, Generator, Generic, Iterable, List, Optional, Set, TypeVar, Union, overload
+from typing_extensions import deprecated
 
 import luigi
 import pandas as pd
@@ -311,6 +312,9 @@ def _load(targets):
 
         return _load(self._get_input_targets(target))
 
+    @deprecated("""This function is deprecated. use `load` instead.
+If you want to specify `required_columns` and `drop_columns`, please extract the columns after loading. ex: `load()[['colA', 'colB']]`
+""")
     def load_data_frame(
         self, target: Union[None, str, TargetOnKart] = None, required_columns: Optional[Set[str]] = None, drop_columns: bool = False
     ) -> pd.DataFrame: