deprecates load_data_frame

docs/task_on_kart.rst

@@ -157,7 +157,7 @@ As an alternative, the `load` method loads individual task input by passing an i
 
 
 We can also omit the :func:`~gokart.task.TaskOnKart.requires` and write the task used by :func:`~gokart.parameter.TaskInstanceParameter`.
-Extensions include :func:`~gokart.task.TaskOnKart.load_data_frame` and :func:`~gokart.task.TaskOnKart.load_generator`. Please refer to :func:`~gokart.task.TaskOnKart.load`, :doc:`task_parameters`, and described later Advanced Features section.
+Also please refer to :func:`~gokart.task.TaskOnKart.load`, :doc:`task_parameters`, and described later Advanced Features section.
 
 
 TaskOnKart.dump
@@ -242,15 +242,6 @@ It is dumped and zipped with ``gensim.model.Word2Vec.save``.
 Please refer to :func:`~gokart.task.TaskOnKart.make_model_target`.
 
 
-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
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
@@ -308,4 +299,4 @@ Please note that :class:`~InMemoryTarget` is an experimental feature.
             redis_timeout=self.redis_timeout,
             raise_task_lock_exception_on_collision=False,
         )
-        return make_in_memory_target('dummy_path', task_lock_params, unique_id)
+        return make_in_memory_target('dummy_path', task_lock_params, unique_id)

gokart/task.py

@@ -5,17 +5,11 @@
 import inspect
 import os
 import random
-import sys
 import types
 from importlib import import_module
 from logging import getLogger
 from typing import Any, Callable, Dict, Generator, Generic, Iterable, List, Optional, Set, TypeVar, Union, overload
 
-if sys.version_info < (3, 13):
-    from typing_extensions import deprecated
-else:
-    from warnings import deprecated
-
 import luigi
 import pandas as pd
 from luigi.parameter import ParameterVisibility
@@ -330,32 +324,6 @@ 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:
-        def _flatten_recursively(dfs):
-            if isinstance(dfs, list):
-                return pd.concat([_flatten_recursively(df) for df in dfs])
-            else:
-                return dfs
-
-        dfs = self.load(target=target)
-        if isinstance(dfs, dict) and len(dfs) == 1:
-            dfs = list(dfs.values())[0]
-
-        data = _flatten_recursively(dfs)
-
-        required_columns = required_columns or set()
-        if data.empty and len(data.index) == 0 and len(required_columns - set(data.columns)) > 0:
-            return pd.DataFrame(columns=list(required_columns))
-        assert required_columns.issubset(set(data.columns)), f'data must have columns {required_columns}, but actually have only {data.columns}.'
-        if drop_columns:
-            data = data[list(required_columns)]
-        return data
-
     @overload
     def dump(self, obj: T, target: None = None, custom_labels: dict[Any, Any] | None = None) -> None: ...
 

gokart/testing/check_if_run_with_empty_data_frame.py

@@ -3,7 +3,6 @@
 from typing import List, Optional
 
 import luigi
-import pandas as pd
 from luigi.cmdline_parser import CmdlineParser
 
 import gokart
@@ -70,9 +69,8 @@ def _test_run_with_empty_data_frame(cmdline_args: List[str], test_run_params: te
     if test_run_params.namespace is not None:
         all_tasks = [t for t in all_tasks if t.task_namespace == test_run_params.namespace]
 
-    with patch('gokart.TaskOnKart.load_data_frame', new=lambda *args, required_columns=None, **kwargs: pd.DataFrame(columns=list(required_columns))):
-        with patch('gokart.TaskOnKart.dump', new=lambda *args, **kwargs: None):
-            test_status_list = [_run_with_test_status(t) for t in all_tasks]
+    with patch('gokart.TaskOnKart.dump', new=lambda *args, **kwargs: None):
+        test_status_list = [_run_with_test_status(t) for t in all_tasks]
 
     test_logger.info('gokart test results:\n' + '\n'.join(s.format() for s in test_status_list))
     if any(s.fail() for s in test_status_list):

test/test_task_on_kart.py

@@ -443,52 +443,6 @@ class DummyTaskAddConfiguration(gokart.TaskOnKart):
         mock_cmdline.return_value = luigi.cmdline_parser.CmdlineParser(['DummyTaskAddConfiguration', '--DummyTaskAddConfiguration-aa', '2'])
         self.assertEqual(DummyTaskAddConfiguration().aa, 2)
 
-    def test_load_list_of_list_pandas(self):
-        task = _DummyTask()
-        task.load = Mock(return_value=[pd.DataFrame(dict(a=[1])), [pd.DataFrame(dict(a=[2])), pd.DataFrame(dict(a=[3]))]])  # type: ignore
-
-        df = task.load_data_frame()
-        self.assertIsInstance(df, pd.DataFrame)
-        self.assertEqual(3, df.shape[0])
-
-    def test_load_single_value_dict_of_dataframe(self):
-        task = _DummyTask()
-        task.load = Mock(return_value={'a': pd.DataFrame(dict(a=[1]))})  # type: ignore
-
-        df = task.load_data_frame()
-        self.assertIsInstance(df, pd.DataFrame)
-        self.assertEqual(1, df.shape[0])
-
-    def test_load_data_frame_drop_columns(self):
-        task = _DummyTask()
-        task.load = Mock(return_value=pd.DataFrame(dict(a=[1], b=[2], c=[3])))  # type: ignore
-
-        df = task.load_data_frame(required_columns={'a', 'c'}, drop_columns=True)
-        self.assertIsInstance(df, pd.DataFrame)
-        self.assertEqual(1, df.shape[0])
-        self.assertSetEqual({'a', 'c'}, set(df.columns))
-
-    def test_load_data_frame_empty_input(self):
-        task = _DummyTask()
-        task.load = Mock(return_value=pd.DataFrame(dict(a=[], b=[], c=[])))  # type: ignore
-
-        df = task.load_data_frame(required_columns={'a', 'c'})
-        self.assertIsInstance(df, pd.DataFrame)
-        self.assertEqual(0, df.shape[0])
-        self.assertSetEqual({'a', 'b', 'c'}, set(df.columns))
-
-    def test_load_index_only_dataframe(self):
-        task = _DummyTask()
-        task.load = Mock(return_value=pd.DataFrame(index=range(3)))  # type: ignore
-
-        # connnot load index only frame with required_columns
-        self.assertRaises(AssertionError, lambda: task.load_data_frame(required_columns={'a', 'c'}))
-
-        df: pd.DataFrame = task.load_data_frame()
-        self.assertIsInstance(df, pd.DataFrame)
-        self.assertTrue(df.empty)
-        self.assertListEqual(list(range(3)), list(df.index))
-
     def test_use_rerun_with_inherits(self):
         # All tasks are completed.
         task_c = _DummyTaskC()