def test_duplicate_init(self):
_id = "test_duplicate_init"
_ = DataReader(self.path1, _id=_id, c_conf="some_conf")
with self.assertRaises(ValueError):
_ = DataReader(self.path2, _id=_id)
# with self.assertWarns(DuplicateSettingWarning):
_ = DataReader(self.path1, self.path2, _id=_id)
with self.assertRaises(ValueError):
_ = DataReader(self.path1, _id=_id, c_conf="some_else")
with self.assertRaises(ValueError):
_ = DataReader(self.path1,
_id=_id,
c_conf="some_conf",
c_conf_2="some_2")
with self.assertLogs(level=logging.INFO):
_ = DataReader(self.path1, _id=_id)
with self.assertLogs(level=logging.INFO):
_ = DataReader(self.path1, _id=_id, c_conf="some_conf")
with self.assertLogs(level=logging.INFO):
_ = DataReader(_id=_id, c_conf="some_conf")
with self.assertRaises(ValueError):
_ = DataReader(read_func=pd.read_excel, _id=_id)
with self.assertLogs(level=logging.INFO):
_ = DataReader(read_func=pd.read_csv, _id=_id)
def test_delay_set_path_and_read(self):
_id = "test_delay_set_path_and_read"
reader = DataReader(_id=_id)
with self.assertRaises(AttributeError):
_ = reader.train()
reader.train_path = self.path1
_ = reader.train()
def test_duplicate_csv_reading(self):
_id = "test_duplicate_csv_reading"
reader1 = DataReader(self.path1, _id=_id)
reader2 = DataReader(self.path1, _id=_id)
reader3 = DataReader(_id=_id)
self.assertTrue(reader1.train() is not reader2.train())
self.assertTrue(reader2.train() is not reader3.train())
def test_singleton(self):
reader1 = DataReader()
reader2 = DataReader()
self.assertTrue(reader1 is reader2)
reader3 = DataReader(_id="test_singleton")
self.assertTrue(reader1 is not reader3)
reader4 = DataReader(_id="test_singleton")
self.assertTrue(reader3 is reader4)
def test_get_set_path_with_init(self):
train_path = self.path1
test_path = self.path2
val_path = self.path3
reader = DataReader(train_path,
test_path,
val_path,
_id="test_get_set_path_with_init")
self.assertEqual(reader.train_path, train_path)
with self.assertRaises(ValueError):
reader.train_path = test_path
with self.assertLogs(level=logging.INFO):
reader.train_path = train_path
self.assertEqual(reader.test_path, test_path)
with self.assertRaises(ValueError):
reader.test_path = train_path
with self.assertLogs(level=logging.INFO):
reader.test_path = test_path
self.assertEqual(reader.val_path, val_path)
with self.assertRaises(ValueError):
reader.val_path = train_path
with self.assertLogs(level=logging.INFO):
reader.val_path = val_path
def test_singleton_for_multi_file(self):
reader_1 = get_reader_1()
reader_2 = get_reader_2()
reader_3 = DataReader(_id="test_singleton_for_multi_file")
self.assertTrue(reader_1 is reader_2)
self.assertTrue(reader_2 is reader_3)
def test_initial_read_func_and_read_kwargs(self):
_id = "test_initial_read_func_and_read_kwargs"
reader = DataReader(self.path1, _id=_id + "1", conf_a="a", conf_b="B")
self.assertTrue(reader._DataReader__read_func is pd.read_csv)
self.assertDictEqual(reader._DataReader__read_kwargs, {
"conf_a": "a",
"conf_b": "B"
})
reader = DataReader(self.path1,
_id=_id + "2",
read_func=pd.read_clipboard)
self.assertTrue(reader._DataReader__read_func is pd.read_clipboard)
with self.assertRaises(AssertionError):
_ = DataReader(self.path1, _id=_id + "3", read_func="A string.")
def test_get_set_path_without_init(self):
train_path = self.path1
test_path = self.path2
val_path = self.path3
reader = DataReader(_id="test_get_set_path_without_init")
with self.assertRaises(AttributeError):
_ = reader.train_path
# with self.assertWarns(DuplicateSettingWarning):
reader.train_path = train_path
with self.assertRaises(ValueError):
reader.train_path = test_path
self.assertEqual(reader.train_path, train_path)
with self.assertRaises(AttributeError):
_ = reader.test_path
# with self.assertWarns(DuplicateSettingWarning):
reader.test_path = test_path
with self.assertRaises(ValueError):
reader.test_path = train_path
self.assertEqual(reader.test_path, test_path)
with self.assertRaises(AttributeError):
_ = reader.val_path
# with self.assertWarns(DuplicateSettingWarning):
reader.val_path = val_path
with self.assertRaises(ValueError):
reader.val_path = train_path
self.assertEqual(reader.val_path, val_path)
def test_csv_reading(self):
_id = "test_csv_reading"
reader1 = DataReader(self.path1, _id=_id + "1")
self.assertTrue(
(reader1.train() == pd.read_csv(self.path1)).all().all())
self.assertTrue(
(reader1.train().index == pd.read_csv(self.path1).index).all())
self.assertTrue(
(reader1.train().columns == pd.read_csv(self.path1).columns).all())
reader2 = DataReader(val_path=self.path2,
_id=_id + "2",
index_col="car")
self.assertTrue(
(reader2.val() == pd.read_csv(self.path2,
index_col="car")).all().all())
self.assertTrue(
(reader2.val().index == pd.read_csv(self.path2,
index_col="car").index).all())
self.assertTrue((reader2.val().columns == pd.read_csv(
self.path2, index_col="car").columns).all())
reader_all = DataReader(self.path1,
self.path2,
self.path3,
_id=_id + "all")
self.assertTrue(
(reader_all.train() == pd.read_csv(self.path1)).all().all())
self.assertTrue(
(reader_all.test() == pd.read_csv(self.path2)).all().all())
self.assertTrue(
(reader_all.val() == pd.read_csv(self.path3)).all().all())
def get_reader_2():
return DataReader(_id="test_singleton_for_multi_file")