def test_unique(index_or_series_obj):
obj = index_or_series_obj
obj = np.repeat(obj, range(1, len(obj) + 1))
with tm.maybe_produces_warning(
PerformanceWarning,
pa_version_under2p0 and str(index_or_series_obj.dtype) == "string[pyarrow]",
):
result = obj.unique()
# dict.fromkeys preserves the order
unique_values = list(dict.fromkeys(obj.values))
if isinstance(obj, pd.MultiIndex):
expected = pd.MultiIndex.from_tuples(unique_values)
expected.names = obj.names
tm.assert_index_equal(result, expected, exact=True)
elif isinstance(obj, pd.Index) and obj._is_backward_compat_public_numeric_index:
expected = NumericIndex(unique_values, dtype=obj.dtype)
tm.assert_index_equal(result, expected, exact=True)
elif isinstance(obj, pd.Index):
expected = pd.Index(unique_values, dtype=obj.dtype)
if is_datetime64tz_dtype(obj.dtype):
expected = expected.normalize()
tm.assert_index_equal(result, expected, exact=True)
else:
expected = np.array(unique_values)
tm.assert_numpy_array_equal(result, expected)
def test_insert_na(self, nulls_fixture, simple_index):
# GH 18295 (test missing)
index = simple_index
na_val = nulls_fixture
if na_val is pd.NaT:
expected = Index([index[0], pd.NaT] + list(index[1:]), dtype=object)
else:
expected = Float64Index([index[0], np.nan] + list(index[1:]))
if index._is_backward_compat_public_numeric_index:
# GH#43921 we preserve NumericIndex
if index.dtype.kind == "f":
expected = NumericIndex(expected, dtype=index.dtype)
else:
expected = NumericIndex(expected)
result = index.insert(1, na_val)
tm.assert_index_equal(result, expected, exact=True)
def test_unique_null(null_obj, index_or_series_obj):
obj = index_or_series_obj
if not allow_na_ops(obj):
pytest.skip("type doesn't allow for NA operations")
elif len(obj) < 1:
pytest.skip("Test doesn't make sense on empty data")
elif isinstance(obj, pd.MultiIndex):
pytest.skip(f"MultiIndex can't hold '{null_obj}'")
values = obj.values
if needs_i8_conversion(obj.dtype):
values[0:2] = iNaT
else:
values[0:2] = null_obj
klass = type(obj)
repeated_values = np.repeat(values, range(1, len(values) + 1))
obj = klass(repeated_values, dtype=obj.dtype)
result = obj.unique()
unique_values_raw = dict.fromkeys(obj.values)
# because np.nan == np.nan is False, but None == None is True
# np.nan would be duplicated, whereas None wouldn't
unique_values_not_null = [
val for val in unique_values_raw if not pd.isnull(val)
]
unique_values = [null_obj] + unique_values_not_null
if isinstance(obj,
pd.Index) and obj._is_backward_compat_public_numeric_index:
expected = NumericIndex(unique_values, dtype=obj.dtype)
tm.assert_index_equal(result, expected, exact=True)
elif isinstance(obj, pd.Index):
expected = pd.Index(unique_values, dtype=obj.dtype)
if is_datetime64tz_dtype(obj.dtype):
result = result.normalize()
expected = expected.normalize()
tm.assert_index_equal(result, expected, exact=True)
else:
expected = np.array(unique_values, dtype=obj.dtype)
tm.assert_numpy_array_equal(result, expected)
def makeNumericIndex(k=10, name=None, *, dtype):
dtype = pandas_dtype(dtype)
assert isinstance(dtype, np.dtype)
if is_integer_dtype(dtype):
values = np.arange(k, dtype=dtype)
if is_unsigned_integer_dtype(dtype):
values += 2**(dtype.itemsize * 8 - 1)
elif is_float_dtype(dtype):
values = np.random.random_sample(k) - np.random.random_sample(1)
values.sort()
values = values * (10**np.random.randint(0, 9))
else:
raise NotImplementedError(f"wrong dtype {dtype}")
return NumericIndex(values, dtype=dtype, name=name)