TST/REF: share method tests between DataFrame and Series (#37596)

Author: jbrockmendel

pandas/conftest.py

@@ -290,6 +290,16 @@ def unique_nulls_fixture(request):
 # ----------------------------------------------------------------
 # Classes
 # ----------------------------------------------------------------
+
+
+@pytest.fixture(params=[pd.DataFrame, pd.Series])
+def frame_or_series(request):
+    """
+    Fixture to parametrize over DataFrame and Series.
+    """
+    return request.param
+
+
 @pytest.fixture(
     params=[pd.Index, pd.Series], ids=["index", "series"]  # type: ignore[list-item]
 )

pandas/core/frame.py

@@ -131,7 +131,13 @@
 from pandas.core.construction import extract_array
 from pandas.core.generic import NDFrame, _shared_docs
 from pandas.core.indexes import base as ibase
-from pandas.core.indexes.api import Index, ensure_index, ensure_index_from_sequences
+from pandas.core.indexes.api import (
+    DatetimeIndex,
+    Index,
+    PeriodIndex,
+    ensure_index,
+    ensure_index_from_sequences,
+)
 from pandas.core.indexes.multi import MultiIndex, maybe_droplevels
 from pandas.core.indexing import check_bool_indexer, convert_to_index_sliceable
 from pandas.core.internals import BlockManager
@@ -9253,6 +9259,9 @@ def to_timestamp(
 
         axis_name = self._get_axis_name(axis)
         old_ax = getattr(self, axis_name)
+        if not isinstance(old_ax, PeriodIndex):
+            raise TypeError(f"unsupported Type {type(old_ax).__name__}")
+
         new_ax = old_ax.to_timestamp(freq=freq, how=how)
 
         setattr(new_obj, axis_name, new_ax)
@@ -9282,6 +9291,9 @@ def to_period(self, freq=None, axis: Axis = 0, copy: bool = True) -> DataFrame:
 
         axis_name = self._get_axis_name(axis)
         old_ax = getattr(self, axis_name)
+        if not isinstance(old_ax, DatetimeIndex):
+            raise TypeError(f"unsupported Type {type(old_ax).__name__}")
+
         new_ax = old_ax.to_period(freq=freq)
 
         setattr(new_obj, axis_name, new_ax)

pandas/tests/frame/methods/test_at_time.py

@@ -4,38 +4,57 @@
 import pytest
 import pytz
 
+from pandas._libs.tslibs import timezones
+
 from pandas import DataFrame, date_range
 import pandas._testing as tm
 
 
 class TestAtTime:
-    def test_at_time(self):
+    @pytest.mark.parametrize("tzstr", ["US/Eastern", "dateutil/US/Eastern"])
+    def test_localized_at_time(self, tzstr, frame_or_series):
+        tz = timezones.maybe_get_tz(tzstr)
+
+        rng = date_range("4/16/2012", "5/1/2012", freq="H")
+        ts = frame_or_series(np.random.randn(len(rng)), index=rng)
+
+        ts_local = ts.tz_localize(tzstr)
+
+        result = ts_local.at_time(time(10, 0))
+        expected = ts.at_time(time(10, 0)).tz_localize(tzstr)
+        tm.assert_equal(result, expected)
+        assert timezones.tz_compare(result.index.tz, tz)
+
+    def test_at_time(self, frame_or_series):
         rng = date_range("1/1/2000", "1/5/2000", freq="5min")
         ts = DataFrame(np.random.randn(len(rng), 2), index=rng)
+        if frame_or_series is not DataFrame:
+            ts = ts[0]
         rs = ts.at_time(rng[1])
         assert (rs.index.hour == rng[1].hour).all()
         assert (rs.index.minute == rng[1].minute).all()
         assert (rs.index.second == rng[1].second).all()
 
         result = ts.at_time("9:30")
         expected = ts.at_time(time(9, 30))
-        tm.assert_frame_equal(result, expected)
-
-        result = ts.loc[time(9, 30)]
-        expected = ts.loc[(rng.hour == 9) & (rng.minute == 30)]
-
-        tm.assert_frame_equal(result, expected)
+        tm.assert_equal(result, expected)
 
+    def test_at_time_midnight(self, frame_or_series):
         # midnight, everything
         rng = date_range("1/1/2000", "1/31/2000")
         ts = DataFrame(np.random.randn(len(rng), 3), index=rng)
+        if frame_or_series is not DataFrame:
+            ts = ts[0]
 
         result = ts.at_time(time(0, 0))
-        tm.assert_frame_equal(result, ts)
+        tm.assert_equal(result, ts)
 
+    def test_at_time_nonexistent(self, frame_or_series):
         # time doesn't exist
         rng = date_range("1/1/2012", freq="23Min", periods=384)
-        ts = DataFrame(np.random.randn(len(rng), 2), rng)
+        ts = DataFrame(np.random.randn(len(rng)), rng)
+        if frame_or_series is not DataFrame:
+            ts = ts[0]
         rs = ts.at_time("16:00")
         assert len(rs) == 0
 
@@ -62,12 +81,14 @@ def test_at_time_tz(self):
         expected = df.iloc[1:2]
         tm.assert_frame_equal(result, expected)
 
-    def test_at_time_raises(self):
+    def test_at_time_raises(self, frame_or_series):
         # GH#20725
-        df = DataFrame([[1, 2, 3], [4, 5, 6]])
+        obj = DataFrame([[1, 2, 3], [4, 5, 6]])
+        if frame_or_series is not DataFrame:
+            obj = obj[0]
         msg = "Index must be DatetimeIndex"
         with pytest.raises(TypeError, match=msg):  # index is not a DatetimeIndex
-            df.at_time("00:00")
+            obj.at_time("00:00")
 
     @pytest.mark.parametrize("axis", ["index", "columns", 0, 1])
     def test_at_time_axis(self, axis):

pandas/tests/frame/methods/test_between_time.py

@@ -1,16 +1,73 @@
-from datetime import time
+from datetime import datetime, time
 
 import numpy as np
 import pytest
 
-from pandas import DataFrame, date_range
+from pandas._libs.tslibs import timezones
+import pandas.util._test_decorators as td
+
+from pandas import DataFrame, Series, date_range
 import pandas._testing as tm
 
 
 class TestBetweenTime:
-    def test_between_time(self, close_open_fixture):
+    @td.skip_if_has_locale
+    def test_between_time_formats(self, frame_or_series):
+        # GH#11818
         rng = date_range("1/1/2000", "1/5/2000", freq="5min")
         ts = DataFrame(np.random.randn(len(rng), 2), index=rng)
+        if frame_or_series is Series:
+            ts = ts[0]
+
+        strings = [
+            ("2:00", "2:30"),
+            ("0200", "0230"),
+            ("2:00am", "2:30am"),
+            ("0200am", "0230am"),
+            ("2:00:00", "2:30:00"),
+            ("020000", "023000"),
+            ("2:00:00am", "2:30:00am"),
+            ("020000am", "023000am"),
+        ]
+        expected_length = 28
+
+        for time_string in strings:
+            assert len(ts.between_time(*time_string)) == expected_length
+
+    @pytest.mark.parametrize("tzstr", ["US/Eastern", "dateutil/US/Eastern"])
+    def test_localized_between_time(self, tzstr, frame_or_series):
+        tz = timezones.maybe_get_tz(tzstr)
+
+        rng = date_range("4/16/2012", "5/1/2012", freq="H")
+        ts = Series(np.random.randn(len(rng)), index=rng)
+        if frame_or_series is DataFrame:
+            ts = ts.to_frame()
+
+        ts_local = ts.tz_localize(tzstr)
+
+        t1, t2 = time(10, 0), time(11, 0)
+        result = ts_local.between_time(t1, t2)
+        expected = ts.between_time(t1, t2).tz_localize(tzstr)
+        tm.assert_equal(result, expected)
+        assert timezones.tz_compare(result.index.tz, tz)
+
+    def test_between_time_types(self, frame_or_series):
+        # GH11818
+        rng = date_range("1/1/2000", "1/5/2000", freq="5min")
+        obj = DataFrame({"A": 0}, index=rng)
+        if frame_or_series is Series:
+            obj = obj["A"]
+
+        msg = r"Cannot convert arg \[datetime\.datetime\(2010, 1, 2, 1, 0\)\] to a time"
+        with pytest.raises(ValueError, match=msg):
+            obj.between_time(datetime(2010, 1, 2, 1), datetime(2010, 1, 2, 5))
+
+    def test_between_time(self, close_open_fixture, frame_or_series):
+        rng = date_range("1/1/2000", "1/5/2000", freq="5min")
+        ts = DataFrame(np.random.randn(len(rng), 2), index=rng)
+        if frame_or_series is not DataFrame:
+            ts = ts[0]
+
         stime = time(0, 0)
         etime = time(1, 0)
         inc_start, inc_end = close_open_fixture
@@ -37,11 +94,13 @@ def test_between_time(self, close_open_fixture):
 
         result = ts.between_time("00:00", "01:00")
         expected = ts.between_time(stime, etime)
-        tm.assert_frame_equal(result, expected)
+        tm.assert_equal(result, expected)
 
         # across midnight
         rng = date_range("1/1/2000", "1/5/2000", freq="5min")
         ts = DataFrame(np.random.randn(len(rng), 2), index=rng)
+        if frame_or_series is not DataFrame:
+            ts = ts[0]
         stime = time(22, 0)
         etime = time(9, 0)
 
@@ -65,14 +124,33 @@ def test_between_time(self, close_open_fixture):
             else:
                 assert (t < etime) or (t >= stime)
 
-    def test_between_time_raises(self):
+    def test_between_time_raises(self, frame_or_series):
         # GH#20725
-        df = DataFrame([[1, 2, 3], [4, 5, 6]])
+        obj = DataFrame([[1, 2, 3], [4, 5, 6]])
+        if frame_or_series is not DataFrame:
+            obj = obj[0]
+
         msg = "Index must be DatetimeIndex"
         with pytest.raises(TypeError, match=msg):  # index is not a DatetimeIndex
-            df.between_time(start_time="00:00", end_time="12:00")
+            obj.between_time(start_time="00:00", end_time="12:00")
+
+    def test_between_time_axis(self, frame_or_series):
+        # GH#8839
+        rng = date_range("1/1/2000", periods=100, freq="10min")
+        ts = Series(np.random.randn(len(rng)), index=rng)
+        if frame_or_series is DataFrame:
+            ts = ts.to_frame()
+
+        stime, etime = ("08:00:00", "09:00:00")
+        expected_length = 7
+
+        assert len(ts.between_time(stime, etime)) == expected_length
+        assert len(ts.between_time(stime, etime, axis=0)) == expected_length
+        msg = f"No axis named {ts.ndim} for object type {type(ts).__name__}"
+        with pytest.raises(ValueError, match=msg):
+            ts.between_time(stime, etime, axis=ts.ndim)
 
-    def test_between_time_axis(self, axis):
+    def test_between_time_axis_aliases(self, axis):
         # GH#8839
         rng = date_range("1/1/2000", periods=100, freq="10min")
         ts = DataFrame(np.random.randn(len(rng), len(rng)))

pandas/tests/frame/methods/test_to_period.py

@@ -1,36 +1,87 @@
 import numpy as np
 import pytest
 
-from pandas import DataFrame, date_range, period_range
+from pandas import (
+    DataFrame,
+    DatetimeIndex,
+    PeriodIndex,
+    Series,
+    date_range,
+    period_range,
+)
 import pandas._testing as tm
 
 
 class TestToPeriod:
-    def test_frame_to_period(self):
+    def test_to_period(self, frame_or_series):
         K = 5
 
-        dr = date_range("1/1/2000", "1/1/2001")
-        pr = period_range("1/1/2000", "1/1/2001")
-        df = DataFrame(np.random.randn(len(dr), K), index=dr)
-        df["mix"] = "a"
+        dr = date_range("1/1/2000", "1/1/2001", freq="D")
+        obj = DataFrame(
+            np.random.randn(len(dr), K), index=dr, columns=["A", "B", "C", "D", "E"]
+        )
+        obj["mix"] = "a"
+        if frame_or_series is Series:
+            obj = obj["A"]
 
-        pts = df.to_period()
-        exp = df.copy()
-        exp.index = pr
-        tm.assert_frame_equal(pts, exp)
+        pts = obj.to_period()
+        exp = obj.copy()
+        exp.index = period_range("1/1/2000", "1/1/2001")
+        tm.assert_equal(pts, exp)
 
-        pts = df.to_period("M")
-        tm.assert_index_equal(pts.index, exp.index.asfreq("M"))
+        pts = obj.to_period("M")
+        exp.index = exp.index.asfreq("M")
+        tm.assert_equal(pts, exp)
+
+    def test_to_period_without_freq(self, frame_or_series):
+        # GH#7606 without freq
+        idx = DatetimeIndex(["2011-01-01", "2011-01-02", "2011-01-03", "2011-01-04"])
+        exp_idx = PeriodIndex(
+            ["2011-01-01", "2011-01-02", "2011-01-03", "2011-01-04"], freq="D"
+        )
+
+        obj = DataFrame(np.random.randn(4, 4), index=idx, columns=idx)
+        if frame_or_series is Series:
+            obj = obj[idx[0]]
+        expected = obj.copy()
+        expected.index = exp_idx
+        tm.assert_equal(obj.to_period(), expected)
+
+        if frame_or_series is DataFrame:
+            expected = obj.copy()
+            expected.columns = exp_idx
+            tm.assert_frame_equal(obj.to_period(axis=1), expected)
+
+    def test_to_period_columns(self):
+        dr = date_range("1/1/2000", "1/1/2001")
+        df = DataFrame(np.random.randn(len(dr), 5), index=dr)
+        df["mix"] = "a"
 
         df = df.T
         pts = df.to_period(axis=1)
         exp = df.copy()
-        exp.columns = pr
+        exp.columns = period_range("1/1/2000", "1/1/2001")
         tm.assert_frame_equal(pts, exp)
 
         pts = df.to_period("M", axis=1)
         tm.assert_index_equal(pts.columns, exp.columns.asfreq("M"))
 
+    def test_to_period_invalid_axis(self):
+        dr = date_range("1/1/2000", "1/1/2001")
+        df = DataFrame(np.random.randn(len(dr), 5), index=dr)
+        df["mix"] = "a"
+
         msg = "No axis named 2 for object type DataFrame"
         with pytest.raises(ValueError, match=msg):
             df.to_period(axis=2)
+
+    def test_to_period_raises(self, index, frame_or_series):
+        # https://github.com/pandas-dev/pandas/issues/33327
+        obj = Series(index=index, dtype=object)
+        if frame_or_series is DataFrame:
+            obj = obj.to_frame()
+
+        if not isinstance(index, DatetimeIndex):
+            msg = f"unsupported Type {type(index).__name__}"
+            with pytest.raises(TypeError, match=msg):
+                obj.to_period()