def test_peel(self):
run_test_peel(self.strings, self.test_strings, self.delim)
run_test_peel(self.strings, self.test_strings, np.str_(self.delim))
run_test_peel(self.strings, self.test_strings, str.encode(str(self.delim)))
# Test gremlins delimiters
g = self._get_ak_gremlins()
with self.assertRaises(ValueError):
run_test_peel(g.gremlins_strings, g.gremlins_test_strings, '')
run_test_peel(g.gremlins_strings, g.gremlins_test_strings, '"')
run_test_peel(g.gremlins_strings, g.gremlins_test_strings, ' ')
# Run a test with a specific set of strings to verify strings.bytes matches expected output
series = pd.Series(["k1:v1", "k2:v2", "k3:v3", "no_colon"])
pda = ak.from_series(series, "string")
# Convert Pandas series of strings into a byte array where each string is terminated by a null byte.
# This mimics what should be stored server-side in the strings.bytes pdarray
expected_series_dec = convert_to_ord(series.to_list())
actual_dec = pda._comp_to_ndarray("values").tolist() #pda.bytes.to_ndarray().tolist()
self.assertListEqual(expected_series_dec, actual_dec)
# Now perform the peel and verify
a, b = pda.peel(":")
expected_a = convert_to_ord(["k1", "k2", "k3", ""])
expected_b = convert_to_ord(["v1", "v2", "v3", "no_colon"])
self.assertListEqual(expected_a, a._comp_to_ndarray("values").tolist())
self.assertListEqual(expected_b, b._comp_to_ndarray("values").tolist())
def test_peel_delimiter_length_issue(self):
# See Issue 838
d = "-" * 25 # 25 dashes as delimiter
series = pd.Series([f"abc{d}xyz", f"small{d}dog", f"blue{d}hat", "last"])
pda = ak.from_series(series)
a, b = pda.peel(d)
aa = a.to_ndarray().tolist()
bb = b.to_ndarray().tolist()
self.assertListEqual(["abc", "small", "blue", ""], aa)
self.assertListEqual(["xyz", "dog", "hat", "last"], bb)
# Try a slight permutation since we were able to get both versions to fail at one point
series = pd.Series([f"abc{d}xyz", f"small{d}dog", "last"])
pda = ak.from_series(series)
a, b = pda.peel(d)
aa = a.to_ndarray().tolist()
bb = b.to_ndarray().tolist()
self.assertListEqual(["abc", "small", ""], aa)
self.assertListEqual(["xyz", "dog", "last"], bb)
def test_from_series(self):
strings = ak.from_series(
pd.Series(['a', 'b', 'c', 'd', 'e'], dtype="string"))
self.assertIsInstance(strings, ak.Strings)
self.assertEqual(5, len(strings))
objects = ak.from_series(pd.Series(['a', 'b', 'c', 'd', 'e']),
dtype=np.str)
self.assertIsInstance(objects, ak.Strings)
self.assertEqual(np.str, objects.dtype)
objects = ak.from_series(pd.Series(['a', 'b', 'c', 'd', 'e']))
self.assertIsInstance(objects, ak.Strings)
self.assertEqual(np.str, objects.dtype)
p_array = ak.from_series(pd.Series(np.random.randint(0, 10, 10)))
self.assertIsInstance(p_array, ak.pdarray)
self.assertEqual(np.int64, p_array.dtype)
p_i_objects_array = ak.from_series(pd.Series(np.random.randint(
0, 10, 10),
dtype='object'),
dtype=np.int64)
self.assertIsInstance(p_i_objects_array, ak.pdarray)
self.assertEqual(np.int64, p_i_objects_array.dtype)
p_array = ak.from_series(
pd.Series(np.random.uniform(low=0.0, high=1.0, size=10)))
self.assertIsInstance(p_array, ak.pdarray)
self.assertEqual(np.float64, p_array.dtype)
p_f_objects_array = ak.from_series(pd.Series(np.random.uniform(
low=0.0, high=1.0, size=10),
dtype='object'),
dtype=np.float64)
self.assertIsInstance(p_f_objects_array, ak.pdarray)
self.assertEqual(np.float64, p_f_objects_array.dtype)
p_array = ak.from_series(
pd.Series(np.random.choice([True, False], size=10)))
self.assertIsInstance(p_array, ak.pdarray)
self.assertEqual(bool, p_array.dtype)
p_b_objects_array = ak.from_series(pd.Series(np.random.choice(
[True, False], size=10),
dtype='object'),
dtype=np.bool)
self.assertIsInstance(p_b_objects_array, ak.pdarray)
self.assertEqual(bool, p_b_objects_array.dtype)
p_array = ak.from_series(pd.Series([dt.datetime(2016, 1, 1, 0, 0, 1)]))
self.assertIsInstance(p_array, ak.pdarray)
self.assertEqual(np.int64, p_array.dtype)
p_array = ak.from_series(pd.Series([np.datetime64('2018-01-01')]))
self.assertIsInstance(p_array, ak.pdarray)
self.assertEqual(np.int64, p_array.dtype)
p_array = ak.from_series(
pd.Series(
pd.to_datetime([
'1/1/2018',
np.datetime64('2018-01-01'),
dt.datetime(2018, 1, 1)
])))
self.assertIsInstance(p_array, ak.pdarray)
self.assertEqual(np.int64, p_array.dtype)
with self.assertRaises(TypeError) as cm:
ak.from_series(np.ones(100))
self.assertEqual(
('type of argument "series" must be pandas.core.series.Series; ' +
'got numpy.ndarray instead'), cm.exception.args[0])
with self.assertRaises(ValueError) as cm:
ak.from_series(
pd.Series(np.random.randint(0, 10, 10), dtype=np.int8))
self.assertEqual(
('dtype int8 is unsupported. Supported dtypes are bool, ' +
'float64, int64, string, datetime64[ns], and timedelta64[ns]'),
cm.exception.args[0])