Merge branch 'master' of https://github.com/pandas-dev/pandas into cln-idx

Author: jbrockmendel

pandas/core/indexes/multi.py

@@ -2058,7 +2058,7 @@ def drop(self, codes, level=None, errors="raise"):
 
         if not isinstance(codes, (np.ndarray, Index)):
             try:
-                codes = com.index_labels_to_array(codes)
+                codes = com.index_labels_to_array(codes, dtype=object)
             except ValueError:
                 pass
 

pandas/core/strings.py

@@ -79,7 +79,7 @@ def cat_core(list_of_columns: List, sep: str):
         return np.sum(arr_of_cols, axis=0)
     list_with_sep = [sep] * (2 * len(list_of_columns) - 1)
     list_with_sep[::2] = list_of_columns
-    arr_with_sep = np.asarray(list_with_sep)
+    arr_with_sep = np.asarray(list_with_sep, dtype=object)
     return np.sum(arr_with_sep, axis=0)
 
 

pandas/io/pytables.py

@@ -2472,6 +2472,7 @@ class Fixed:
         """
 
     pandas_kind: str
+    format_type: str = "fixed"  # GH#30962 needed by dask
     obj_type: Type[Union[DataFrame, Series]]
     ndim: int
     encoding: str
@@ -3129,6 +3130,7 @@ class Table(Fixed):
         """
 
     pandas_kind = "wide_table"
+    format_type: str = "table"  # GH#30962 needed by dask
     table_type: str
     levels = 1
     is_table = True

pandas/tests/arrays/categorical/test_constructors.py

@@ -605,6 +605,6 @@ def test_constructor_imaginary(self):
     @pytest.mark.skipif(_np_version_under1p16, reason="Skipping for NumPy <1.16")
     def test_constructor_string_and_tuples(self):
         # GH 21416
-        c = pd.Categorical(["c", ("a", "b"), ("b", "a"), "c"])
+        c = pd.Categorical(np.array(["c", ("a", "b"), ("b", "a"), "c"], dtype=object))
         expected_index = pd.Index([("a", "b"), ("b", "a"), "c"])
         assert c.categories.equals(expected_index)

pandas/tests/arrays/categorical/test_missing.py

@@ -77,7 +77,7 @@ def test_fillna_iterable_category(self, named):
             Point = collections.namedtuple("Point", "x y")
         else:
             Point = lambda *args: args  # tuple
-        cat = Categorical([Point(0, 0), Point(0, 1), None])
+        cat = Categorical(np.array([Point(0, 0), Point(0, 1), None], dtype=object))
         result = cat.fillna(Point(0, 0))
         expected = Categorical([Point(0, 0), Point(0, 1), Point(0, 0)])
 

pandas/tests/extension/base/getitem.py

@@ -245,7 +245,9 @@ def test_take_non_na_fill_value(self, data_missing):
         fill_value = data_missing[1]  # valid
         na = data_missing[0]
 
-        array = data_missing._from_sequence([na, fill_value, na])
+        array = data_missing._from_sequence(
+            [na, fill_value, na], dtype=data_missing.dtype
+        )
         result = array.take([-1, 1], fill_value=fill_value, allow_fill=True)
         expected = array.take([1, 1])
         self.assert_extension_array_equal(result, expected)
@@ -293,10 +295,12 @@ def test_reindex_non_na_fill_value(self, data_missing):
         valid = data_missing[1]
         na = data_missing[0]
 
-        array = data_missing._from_sequence([na, valid])
+        array = data_missing._from_sequence([na, valid], dtype=data_missing.dtype)
         ser = pd.Series(array)
         result = ser.reindex([0, 1, 2], fill_value=valid)
-        expected = pd.Series(data_missing._from_sequence([na, valid, valid]))
+        expected = pd.Series(
+            data_missing._from_sequence([na, valid, valid], dtype=data_missing.dtype)
+        )
 
         self.assert_series_equal(result, expected)
 

pandas/tests/extension/json/array.py

@@ -113,6 +113,11 @@ def __setitem__(self, key, value):
     def __len__(self) -> int:
         return len(self.data)
 
+    def __array__(self, dtype=None):
+        if dtype is None:
+            dtype = object
+        return np.asarray(self.data, dtype=dtype)
+
     @property
     def nbytes(self) -> int:
         return sys.getsizeof(self.data)

pandas/tests/extension/json/test_json.py

@@ -163,10 +163,6 @@ def test_unstack(self, data, index):
         # this matches otherwise
         return super().test_unstack(data, index)
 
-    @pytest.mark.xfail(reason="Inconsistent sizes.")
-    def test_transpose(self, data):
-        super().test_transpose(data)
-
 
 class TestGetitem(BaseJSON, base.BaseGetitemTests):
     pass

pandas/tests/indexes/conftest.py

@@ -1,7 +1,5 @@
-import numpy as np
 import pytest
 
-import pandas as pd
 import pandas._testing as tm
 from pandas.core.indexes.api import Index, MultiIndex
 
@@ -28,25 +26,3 @@
 def indices(request):
     # copy to avoid mutation, e.g. setting .name
     return indices_dict[request.param].copy()
-
-
-@pytest.fixture(params=[1, np.array(1, dtype=np.int64)])
-def one(request):
-    # zero-dim integer array behaves like an integer
-    return request.param
-
-
-zeros = [
-    box([0] * 5, dtype=dtype)
-    for box in [pd.Index, np.array]
-    for dtype in [np.int64, np.uint64, np.float64]
-]
-zeros.extend([np.array(0, dtype=dtype) for dtype in [np.int64, np.uint64, np.float64]])
-zeros.extend([0, 0.0])
-
-
-@pytest.fixture(params=zeros)
-def zero(request):
-    # For testing division by (or of) zero for Index with length 5, this
-    # gives several scalar-zeros and length-5 vector-zeros
-    return request.param

pandas/tests/indexes/datetimes/test_snap.py

@@ -0,0 +1,37 @@
+import pytest
+
+from pandas import DatetimeIndex, date_range
+import pandas._testing as tm
+
+
+@pytest.mark.filterwarnings("ignore::DeprecationWarning")
+@pytest.mark.parametrize("tz", [None, "Asia/Shanghai", "Europe/Berlin"])
+@pytest.mark.parametrize("name", [None, "my_dti"])
+def test_dti_snap(name, tz):
+    dti = DatetimeIndex(
+        [
+            "1/1/2002",
+            "1/2/2002",
+            "1/3/2002",
+            "1/4/2002",
+            "1/5/2002",
+            "1/6/2002",
+            "1/7/2002",
+        ],
+        name=name,
+        tz=tz,
+        freq="D",
+    )
+
+    result = dti.snap(freq="W-MON")
+    expected = date_range("12/31/2001", "1/7/2002", name=name, tz=tz, freq="w-mon")
+    expected = expected.repeat([3, 4])
+    tm.assert_index_equal(result, expected)
+    assert result.tz == expected.tz
+
+    result = dti.snap(freq="B")
+
+    expected = date_range("1/1/2002", "1/7/2002", name=name, tz=tz, freq="b")
+    expected = expected.repeat([1, 1, 1, 2, 2])
+    tm.assert_index_equal(result, expected)
+    assert result.tz == expected.tz

pandas/tests/indexes/multi/test_setops.py

@@ -110,6 +110,17 @@ def test_symmetric_difference(idx, sort):
         first.symmetric_difference([1, 2, 3], sort=sort)
 
 
+def test_multiindex_symmetric_difference():
+    # GH 13490
+    idx = MultiIndex.from_product([["a", "b"], ["A", "B"]], names=["a", "b"])
+    result = idx ^ idx
+    assert result.names == idx.names
+
+    idx2 = idx.copy().rename(["A", "B"])
+    result = idx ^ idx2
+    assert result.names == [None, None]
+
+
 def test_empty(idx):
     # GH 15270
     assert not idx.empty

pandas/tests/indexing/multiindex/test_set_ops.py renamed to pandas/tests/indexing/multiindex/test_insert.py

@@ -4,17 +4,7 @@
 import pandas._testing as tm
 
 
-class TestMultiIndexSetOps:
-    def test_multiindex_symmetric_difference(self):
-        # GH 13490
-        idx = MultiIndex.from_product([["a", "b"], ["A", "B"]], names=["a", "b"])
-        result = idx ^ idx
-        assert result.names == idx.names
-
-        idx2 = idx.copy().rename(["A", "B"])
-        result = idx ^ idx2
-        assert result.names == [None, None]
-
+class TestMultiIndexInsertion:
     def test_mixed_depth_insert(self):
         arrays = [
             ["a", "top", "top", "routine1", "routine1", "routine2"],

pandas/tests/io/pytables/test_store.py

@@ -64,6 +64,16 @@
 
 @pytest.mark.single
 class TestHDFStore:
+    def test_format_type(self, setup_path):
+        df = pd.DataFrame({"A": [1, 2]})
+        with ensure_clean_path(setup_path) as path:
+            with HDFStore(path) as store:
+                store.put("a", df, format="fixed")
+                store.put("b", df, format="table")
+
+                assert store.get_storer("a").format_type == "fixed"
+                assert store.get_storer("b").format_type == "table"
+
     def test_format_kwarg_in_constructor(self, setup_path):
         # GH 13291
 

pandas/tests/series/indexing/test_datetime.py

@@ -49,39 +49,6 @@ def test_fancy_setitem():
     assert (s[48:54] == -3).all()
 
 
-@pytest.mark.filterwarnings("ignore::DeprecationWarning")
-@pytest.mark.parametrize("tz", [None, "Asia/Shanghai", "Europe/Berlin"])
-@pytest.mark.parametrize("name", [None, "my_dti"])
-def test_dti_snap(name, tz):
-    dti = DatetimeIndex(
-        [
-            "1/1/2002",
-            "1/2/2002",
-            "1/3/2002",
-            "1/4/2002",
-            "1/5/2002",
-            "1/6/2002",
-            "1/7/2002",
-        ],
-        name=name,
-        tz=tz,
-        freq="D",
-    )
-
-    result = dti.snap(freq="W-MON")
-    expected = date_range("12/31/2001", "1/7/2002", name=name, tz=tz, freq="w-mon")
-    expected = expected.repeat([3, 4])
-    tm.assert_index_equal(result, expected)
-    assert result.tz == expected.tz
-
-    result = dti.snap(freq="B")
-
-    expected = date_range("1/1/2002", "1/7/2002", name=name, tz=tz, freq="b")
-    expected = expected.repeat([1, 1, 1, 2, 2])
-    tm.assert_index_equal(result, expected)
-    assert result.tz == expected.tz
-
-
 def test_dti_reset_index_round_trip():
     dti = date_range(start="1/1/2001", end="6/1/2001", freq="D")
     d1 = DataFrame({"v": np.random.rand(len(dti))}, index=dti)
@@ -751,16 +718,6 @@ def test_nat_operations():
     assert s.max() == exp
 
 
-@pytest.mark.parametrize("method", ["round", "floor", "ceil"])
-@pytest.mark.parametrize("freq", ["s", "5s", "min", "5min", "h", "5h"])
-def test_round_nat(method, freq):
-    # GH14940
-    s = Series([pd.NaT])
-    expected = Series(pd.NaT)
-    round_method = getattr(s.dt, method)
-    tm.assert_series_equal(round_method(freq), expected)
-
-
 def test_setitem_tuple_with_datetimetz():
     # GH 20441
     arr = date_range("2017", periods=4, tz="US/Eastern")

pandas/tests/series/methods/test_round.py

@@ -1,6 +1,7 @@
 import numpy as np
 import pytest
 
+import pandas as pd
 from pandas import Series
 import pandas._testing as tm
 
@@ -44,3 +45,12 @@ def test_round_builtin(self):
         expected_rounded = Series([1.12, 2.12, 3.12], index=range(3))
         result = round(ser, decimals)
         tm.assert_series_equal(result, expected_rounded)
+
+    @pytest.mark.parametrize("method", ["round", "floor", "ceil"])
+    @pytest.mark.parametrize("freq", ["s", "5s", "min", "5min", "h", "5h"])
+    def test_round_nat(self, method, freq):
+        # GH14940
+        ser = Series([pd.NaT])
+        expected = Series(pd.NaT)
+        round_method = getattr(ser.dt, method)
+        tm.assert_series_equal(round_method(freq), expected)