From 385c7828936ef4624311af7dcc331a047533a452 Mon Sep 17 00:00:00 2001 From: Simon Hawkins Date: Wed, 19 Dec 2018 09:16:02 +0000 Subject: [PATCH 1/8] remove class --- pandas/tests/indexing/multiindex/test_xs.py | 264 ++++++++++---------- 1 file changed, 134 insertions(+), 130 deletions(-) diff --git a/pandas/tests/indexing/multiindex/test_xs.py b/pandas/tests/indexing/multiindex/test_xs.py index b8380e3a95f2a..7cb078cbcd752 100644 --- a/pandas/tests/indexing/multiindex/test_xs.py +++ b/pandas/tests/indexing/multiindex/test_xs.py @@ -8,157 +8,161 @@ from pandas.util import testing as tm -class TestMultiIndexXs(object): - - def test_xs_multiindex(self): - - # GH2903 - columns = MultiIndex.from_tuples( - [('a', 'foo'), ('a', 'bar'), ('b', 'hello'), - ('b', 'world')], names=['lvl0', 'lvl1']) - df = DataFrame(np.random.randn(4, 4), columns=columns) - df.sort_index(axis=1, inplace=True) - result = df.xs('a', level='lvl0', axis=1) - expected = df.iloc[:, 0:2].loc[:, 'a'] - tm.assert_frame_equal(result, expected) - - result = df.xs('foo', level='lvl1', axis=1) - expected = df.iloc[:, 1:2].copy() - expected.columns = expected.columns.droplevel('lvl1') - tm.assert_frame_equal(result, expected) - - def test_xs(self, multiindex_dataframe_random_data): - frame = multiindex_dataframe_random_data - xs = frame.xs(('bar', 'two')) - xs2 = frame.loc[('bar', 'two')] - - tm.assert_series_equal(xs, xs2) - tm.assert_almost_equal(xs.values, frame.values[4]) - - # GH 6574 - # missing values in returned index should be preserrved - acc = [ - ('a', 'abcde', 1), - ('b', 'bbcde', 2), - ('y', 'yzcde', 25), - ('z', 'xbcde', 24), - ('z', None, 26), - ('z', 'zbcde', 25), - ('z', 'ybcde', 26), - ] - df = DataFrame(acc, - columns=['a1', 'a2', 'cnt']).set_index(['a1', 'a2']) - expected = DataFrame({'cnt': [24, 26, 25, 26]}, index=Index( - ['xbcde', np.nan, 'zbcde', 'ybcde'], name='a2')) - - result = df.xs('z', level='a1') - tm.assert_frame_equal(result, expected) - - def test_xs_with_duplicates(self, multiindex_dataframe_random_data): - # Issue #13719 - frame = multiindex_dataframe_random_data - df_dup = concat([frame] * 2) - assert df_dup.index.is_unique is False - expected = concat([frame.xs('one', level='second')] * 2) - tm.assert_frame_equal(df_dup.xs('one', level='second'), expected) - tm.assert_frame_equal(df_dup.xs(['one'], level=['second']), expected) - - def test_xs_level(self, multiindex_dataframe_random_data): - frame = multiindex_dataframe_random_data - result = frame.xs('two', level='second') - expected = frame[frame.index.get_level_values(1) == 'two'] - expected.index = expected.index.droplevel(1) - - tm.assert_frame_equal(result, expected) - - index = MultiIndex.from_tuples([('x', 'y', 'z'), ('a', 'b', 'c'), ( - 'p', 'q', 'r')]) - df = DataFrame(np.random.randn(3, 5), index=index) - result = df.xs('c', level=2) - expected = df[1:2] - expected.index = expected.index.droplevel(2) - tm.assert_frame_equal(result, expected) - - # this is a copy in 0.14 - result = frame.xs('two', level='second') - - # setting this will give a SettingWithCopyError - # as we are trying to write a view - def f(x): - x[:] = 10 - - pytest.raises(com.SettingWithCopyError, f, result) - - def test_xs_level_multiple(self): - text = """ A B C D E +def test_xs_multiindex(): + + # GH2903 + columns = MultiIndex.from_tuples( + [('a', 'foo'), ('a', 'bar'), ('b', 'hello'), + ('b', 'world')], names=['lvl0', 'lvl1']) + df = DataFrame(np.random.randn(4, 4), columns=columns) + df.sort_index(axis=1, inplace=True) + result = df.xs('a', level='lvl0', axis=1) + expected = df.iloc[:, 0:2].loc[:, 'a'] + tm.assert_frame_equal(result, expected) + + result = df.xs('foo', level='lvl1', axis=1) + expected = df.iloc[:, 1:2].copy() + expected.columns = expected.columns.droplevel('lvl1') + tm.assert_frame_equal(result, expected) + + +def test_xs(multiindex_dataframe_random_data): + frame = multiindex_dataframe_random_data + xs = frame.xs(('bar', 'two')) + xs2 = frame.loc[('bar', 'two')] + + tm.assert_series_equal(xs, xs2) + tm.assert_almost_equal(xs.values, frame.values[4]) + + # GH 6574 + # missing values in returned index should be preserrved + acc = [ + ('a', 'abcde', 1), + ('b', 'bbcde', 2), + ('y', 'yzcde', 25), + ('z', 'xbcde', 24), + ('z', None, 26), + ('z', 'zbcde', 25), + ('z', 'ybcde', 26), + ] + df = DataFrame(acc, + columns=['a1', 'a2', 'cnt']).set_index(['a1', 'a2']) + expected = DataFrame({'cnt': [24, 26, 25, 26]}, index=Index( + ['xbcde', np.nan, 'zbcde', 'ybcde'], name='a2')) + + result = df.xs('z', level='a1') + tm.assert_frame_equal(result, expected) + + +def test_xs_with_duplicates(multiindex_dataframe_random_data): + # Issue #13719 + frame = multiindex_dataframe_random_data + df_dup = concat([frame] * 2) + assert df_dup.index.is_unique is False + expected = concat([frame.xs('one', level='second')] * 2) + tm.assert_frame_equal(df_dup.xs('one', level='second'), expected) + tm.assert_frame_equal(df_dup.xs(['one'], level=['second']), expected) + + +def test_xs_level(multiindex_dataframe_random_data): + frame = multiindex_dataframe_random_data + result = frame.xs('two', level='second') + expected = frame[frame.index.get_level_values(1) == 'two'] + expected.index = expected.index.droplevel(1) + + tm.assert_frame_equal(result, expected) + + index = MultiIndex.from_tuples([('x', 'y', 'z'), ('a', 'b', 'c'), ( + 'p', 'q', 'r')]) + df = DataFrame(np.random.randn(3, 5), index=index) + result = df.xs('c', level=2) + expected = df[1:2] + expected.index = expected.index.droplevel(2) + tm.assert_frame_equal(result, expected) + + # this is a copy in 0.14 + result = frame.xs('two', level='second') + + # setting this will give a SettingWithCopyError + # as we are trying to write a view + def f(x): + x[:] = 10 + + pytest.raises(com.SettingWithCopyError, f, result) + + +def test_xs_level_multiple(): + text = """ A B C D E one two three four a b 10.0032 5 -0.5109 -2.3358 -0.4645 0.05076 0.3640 a q 20 4 0.4473 1.4152 0.2834 1.00661 0.1744 x q 30 3 -0.6662 -0.5243 -0.3580 0.89145 2.5838""" - df = read_csv(StringIO(text), sep=r'\s+', engine='python') + df = read_csv(StringIO(text), sep=r'\s+', engine='python') - result = df.xs(('a', 4), level=['one', 'four']) - expected = df.xs('a').xs(4, level='four') - tm.assert_frame_equal(result, expected) + result = df.xs(('a', 4), level=['one', 'four']) + expected = df.xs('a').xs(4, level='four') + tm.assert_frame_equal(result, expected) - # this is a copy in 0.14 - result = df.xs(('a', 4), level=['one', 'four']) + # this is a copy in 0.14 + result = df.xs(('a', 4), level=['one', 'four']) - # setting this will give a SettingWithCopyError - # as we are trying to write a view - def f(x): - x[:] = 10 + # setting this will give a SettingWithCopyError + # as we are trying to write a view + def f(x): + x[:] = 10 - pytest.raises(com.SettingWithCopyError, f, result) + pytest.raises(com.SettingWithCopyError, f, result) - # GH2107 - dates = lrange(20111201, 20111205) - ids = 'abcde' - idx = MultiIndex.from_tuples([x for x in cart_product(dates, ids)]) - idx.names = ['date', 'secid'] - df = DataFrame(np.random.randn(len(idx), 3), idx, ['X', 'Y', 'Z']) + # GH2107 + dates = lrange(20111201, 20111205) + ids = 'abcde' + idx = MultiIndex.from_tuples([x for x in cart_product(dates, ids)]) + idx.names = ['date', 'secid'] + df = DataFrame(np.random.randn(len(idx), 3), idx, ['X', 'Y', 'Z']) - rs = df.xs(20111201, level='date') - xp = df.loc[20111201, :] - tm.assert_frame_equal(rs, xp) + rs = df.xs(20111201, level='date') + xp = df.loc[20111201, :] + tm.assert_frame_equal(rs, xp) - def test_xs_level0(self): - text = """ A B C D E + +def test_xs_level0(): + text = """ A B C D E one two three four a b 10.0032 5 -0.5109 -2.3358 -0.4645 0.05076 0.3640 a q 20 4 0.4473 1.4152 0.2834 1.00661 0.1744 x q 30 3 -0.6662 -0.5243 -0.3580 0.89145 2.5838""" - df = read_csv(StringIO(text), sep=r'\s+', engine='python') + df = read_csv(StringIO(text), sep=r'\s+', engine='python') + + result = df.xs('a', level=0) + expected = df.xs('a') + assert len(result) == 2 + tm.assert_frame_equal(result, expected) - result = df.xs('a', level=0) - expected = df.xs('a') - assert len(result) == 2 - tm.assert_frame_equal(result, expected) - def test_xs_level_series(self, multiindex_dataframe_random_data, - multiindex_year_month_day_dataframe_random_data): - frame = multiindex_dataframe_random_data - ymd = multiindex_year_month_day_dataframe_random_data - s = frame['A'] - result = s[:, 'two'] - expected = frame.xs('two', level=1)['A'] - tm.assert_series_equal(result, expected) +def test_xs_level_series(multiindex_dataframe_random_data, + multiindex_year_month_day_dataframe_random_data): + frame = multiindex_dataframe_random_data + ymd = multiindex_year_month_day_dataframe_random_data + s = frame['A'] + result = s[:, 'two'] + expected = frame.xs('two', level=1)['A'] + tm.assert_series_equal(result, expected) - s = ymd['A'] - result = s[2000, 5] - expected = ymd.loc[2000, 5]['A'] - tm.assert_series_equal(result, expected) + s = ymd['A'] + result = s[2000, 5] + expected = ymd.loc[2000, 5]['A'] + tm.assert_series_equal(result, expected) - # not implementing this for now + # not implementing this for now - pytest.raises(TypeError, s.__getitem__, (2000, slice(3, 4))) + pytest.raises(TypeError, s.__getitem__, (2000, slice(3, 4))) - # result = s[2000, 3:4] - # lv =s.index.get_level_values(1) - # expected = s[(lv == 3) | (lv == 4)] - # expected.index = expected.index.droplevel(0) - # tm.assert_series_equal(result, expected) + # result = s[2000, 3:4] + # lv =s.index.get_level_values(1) + # expected = s[(lv == 3) | (lv == 4)] + # expected.index = expected.index.droplevel(0) + # tm.assert_series_equal(result, expected) - # can do this though + # can do this though From ece270c607f3611473275f9662e44f2dcada9b41 Mon Sep 17 00:00:00 2001 From: Simon Hawkins Date: Wed, 19 Dec 2018 18:16:10 +0000 Subject: [PATCH 2/8] parametrize and rename test_xs_multiindex --- pandas/tests/indexing/multiindex/test_xs.py | 28 ++++++++++----------- 1 file changed, 13 insertions(+), 15 deletions(-) diff --git a/pandas/tests/indexing/multiindex/test_xs.py b/pandas/tests/indexing/multiindex/test_xs.py index 7cb078cbcd752..a09660f2db430 100644 --- a/pandas/tests/indexing/multiindex/test_xs.py +++ b/pandas/tests/indexing/multiindex/test_xs.py @@ -8,21 +8,19 @@ from pandas.util import testing as tm -def test_xs_multiindex(): - - # GH2903 - columns = MultiIndex.from_tuples( - [('a', 'foo'), ('a', 'bar'), ('b', 'hello'), - ('b', 'world')], names=['lvl0', 'lvl1']) - df = DataFrame(np.random.randn(4, 4), columns=columns) - df.sort_index(axis=1, inplace=True) - result = df.xs('a', level='lvl0', axis=1) - expected = df.iloc[:, 0:2].loc[:, 'a'] - tm.assert_frame_equal(result, expected) - - result = df.xs('foo', level='lvl1', axis=1) - expected = df.iloc[:, 1:2].copy() - expected.columns = expected.columns.droplevel('lvl1') +@pytest.mark.parametrize('key, level, exp_arr, exp_index', [ + ('a', 'lvl0', lambda x: x[:, 0:2], Index(['bar', 'foo'], name='lvl1')), + ('foo', 'lvl1', lambda x: x[:, 1:2], Index(['a'], name='lvl0')) +]) +def test_xs_named_levels_axis_eq_1(key, level, exp_arr, exp_index): + # see gh-2903 + arr = np.random.randn(4, 4) + index = MultiIndex(levels=[['a', 'b'], ['bar', 'foo', 'hello', 'world']], + codes=[[0, 0, 1, 1], [0, 1, 2, 3]], + names=['lvl0', 'lvl1']) + df = DataFrame(arr, columns=index) + result = df.xs(key, level=level, axis=1) + expected = DataFrame(exp_arr(arr), columns=exp_index) tm.assert_frame_equal(result, expected) From b569cfedf11dc408b6df07cb01ea61780d0692a3 Mon Sep 17 00:00:00 2001 From: Simon Hawkins Date: Wed, 19 Dec 2018 18:41:09 +0000 Subject: [PATCH 3/8] split test_xs --- pandas/tests/indexing/multiindex/test_xs.py | 17 ++++++++++++----- 1 file changed, 12 insertions(+), 5 deletions(-) diff --git a/pandas/tests/indexing/multiindex/test_xs.py b/pandas/tests/indexing/multiindex/test_xs.py index a09660f2db430..0970f4201098e 100644 --- a/pandas/tests/indexing/multiindex/test_xs.py +++ b/pandas/tests/indexing/multiindex/test_xs.py @@ -26,13 +26,20 @@ def test_xs_named_levels_axis_eq_1(key, level, exp_arr, exp_index): def test_xs(multiindex_dataframe_random_data): frame = multiindex_dataframe_random_data - xs = frame.xs(('bar', 'two')) - xs2 = frame.loc[('bar', 'two')] + result = frame.xs(('bar', 'two')).values + expected = frame.values[4] + tm.assert_almost_equal(result, expected) - tm.assert_series_equal(xs, xs2) - tm.assert_almost_equal(xs.values, frame.values[4]) - # GH 6574 +def test_xs_loc_equality(multiindex_dataframe_random_data): + frame = multiindex_dataframe_random_data + result = frame.xs(('bar', 'two')) + expected = frame.loc[('bar', 'two')] + tm.assert_series_equal(result, expected) + + +def test_xs_missing_values_in_index(): + # see gh-6574 # missing values in returned index should be preserrved acc = [ ('a', 'abcde', 1), From 51493fb57bb7a917d4d4e8b9490f927a84f20684 Mon Sep 17 00:00:00 2001 From: Simon Hawkins Date: Wed, 19 Dec 2018 19:00:21 +0000 Subject: [PATCH 4/8] parametrize test_xs_with_duplicates --- pandas/tests/indexing/multiindex/test_xs.py | 17 +++++++++++------ 1 file changed, 11 insertions(+), 6 deletions(-) diff --git a/pandas/tests/indexing/multiindex/test_xs.py b/pandas/tests/indexing/multiindex/test_xs.py index 0970f4201098e..78f0e0f47e97e 100644 --- a/pandas/tests/indexing/multiindex/test_xs.py +++ b/pandas/tests/indexing/multiindex/test_xs.py @@ -59,14 +59,19 @@ def test_xs_missing_values_in_index(): tm.assert_frame_equal(result, expected) -def test_xs_with_duplicates(multiindex_dataframe_random_data): - # Issue #13719 +@pytest.mark.parametrize('key, level', [ + ('one', 'second'), + (['one'], ['second']) +]) +def test_xs_with_duplicates(key, level, multiindex_dataframe_random_data): + # see gh-13719 frame = multiindex_dataframe_random_data - df_dup = concat([frame] * 2) - assert df_dup.index.is_unique is False + df = concat([frame] * 2) + assert df.index.is_unique is False expected = concat([frame.xs('one', level='second')] * 2) - tm.assert_frame_equal(df_dup.xs('one', level='second'), expected) - tm.assert_frame_equal(df_dup.xs(['one'], level=['second']), expected) + + result = df.xs(key, level=level) + tm.assert_frame_equal(result, expected) def test_xs_level(multiindex_dataframe_random_data): From 274f62f54b0072b7190be961b7e50f5b386ab970 Mon Sep 17 00:00:00 2001 From: Simon Hawkins Date: Wed, 19 Dec 2018 20:12:53 +0000 Subject: [PATCH 5/8] split test_xs_level --- pandas/tests/indexing/multiindex/test_xs.py | 48 +++++++++++---------- 1 file changed, 26 insertions(+), 22 deletions(-) diff --git a/pandas/tests/indexing/multiindex/test_xs.py b/pandas/tests/indexing/multiindex/test_xs.py index 78f0e0f47e97e..fa5ed52f7c08f 100644 --- a/pandas/tests/indexing/multiindex/test_xs.py +++ b/pandas/tests/indexing/multiindex/test_xs.py @@ -24,17 +24,17 @@ def test_xs_named_levels_axis_eq_1(key, level, exp_arr, exp_index): tm.assert_frame_equal(result, expected) -def test_xs(multiindex_dataframe_random_data): - frame = multiindex_dataframe_random_data - result = frame.xs(('bar', 'two')).values - expected = frame.values[4] +def test_xs_values(multiindex_dataframe_random_data): + df = multiindex_dataframe_random_data + result = df.xs(('bar', 'two')).values + expected = df.values[4] tm.assert_almost_equal(result, expected) def test_xs_loc_equality(multiindex_dataframe_random_data): - frame = multiindex_dataframe_random_data - result = frame.xs(('bar', 'two')) - expected = frame.loc[('bar', 'two')] + df = multiindex_dataframe_random_data + result = df.xs(('bar', 'two')) + expected = df.loc[('bar', 'two')] tm.assert_series_equal(result, expected) @@ -75,30 +75,34 @@ def test_xs_with_duplicates(key, level, multiindex_dataframe_random_data): def test_xs_level(multiindex_dataframe_random_data): - frame = multiindex_dataframe_random_data - result = frame.xs('two', level='second') - expected = frame[frame.index.get_level_values(1) == 'two'] - expected.index = expected.index.droplevel(1) - + df = multiindex_dataframe_random_data + result = df.xs('two', level='second') + expected = df[df.index.get_level_values(1) == 'two'] + expected.index = Index(['foo', 'bar', 'baz', 'qux'], name='first') tm.assert_frame_equal(result, expected) - index = MultiIndex.from_tuples([('x', 'y', 'z'), ('a', 'b', 'c'), ( - 'p', 'q', 'r')]) - df = DataFrame(np.random.randn(3, 5), index=index) + +def test_xs_level_eq_2(): + arr = np.random.randn(3, 5) + index = MultiIndex( + levels=[['a', 'p', 'x'], ['b', 'q', 'y'], ['c', 'r', 'z']], + codes=[[2, 0, 1], [2, 0, 1], [2, 0, 1]]) + df = DataFrame(arr, index=index) + expected = DataFrame(arr[1:2], index=[['a'], ['b']]) result = df.xs('c', level=2) - expected = df[1:2] - expected.index = expected.index.droplevel(2) tm.assert_frame_equal(result, expected) + +def test_xs_setting_with_copy_error(multiindex_dataframe_random_data): # this is a copy in 0.14 - result = frame.xs('two', level='second') + df = multiindex_dataframe_random_data + result = df.xs('two', level='second') # setting this will give a SettingWithCopyError # as we are trying to write a view - def f(x): - x[:] = 10 - - pytest.raises(com.SettingWithCopyError, f, result) + msg = 'A value is trying to be set on a copy of a slice from a DataFrame' + with pytest.raises(com.SettingWithCopyError, match=msg): + result[:] = 10 def test_xs_level_multiple(): From a827e081c01213f9346132638f862bbb903afb5f Mon Sep 17 00:00:00 2001 From: Simon Hawkins Date: Wed, 19 Dec 2018 21:28:14 +0000 Subject: [PATCH 6/8] split test_xs_level_multiple --- pandas/tests/indexing/multiindex/test_xs.py | 69 ++++++++++++++------- 1 file changed, 45 insertions(+), 24 deletions(-) diff --git a/pandas/tests/indexing/multiindex/test_xs.py b/pandas/tests/indexing/multiindex/test_xs.py index fa5ed52f7c08f..a14079e7652e6 100644 --- a/pandas/tests/indexing/multiindex/test_xs.py +++ b/pandas/tests/indexing/multiindex/test_xs.py @@ -8,6 +8,18 @@ from pandas.util import testing as tm +@pytest.fixture +def four_level_index_dataframe(): + arr = np.array([[-0.5109, -2.3358, -0.4645, 0.05076, 0.364], + [0.4473, 1.4152, 0.2834, 1.00661, 0.1744], + [-0.6662, -0.5243, -0.358, 0.89145, 2.5838]]) + index = MultiIndex( + levels=[['a', 'x'], ['b', 'q'], [10.0032, 20.0, 30.0], [3, 4, 5]], + codes=[[0, 0, 1], [0, 1, 1], [0, 1, 2], [2, 1, 0]], + names=['one', 'two', 'three', 'four']) + return DataFrame(arr, index=index, columns=list('ABCDE')) + + @pytest.mark.parametrize('key, level, exp_arr, exp_index', [ ('a', 'lvl0', lambda x: x[:, 0:2], Index(['bar', 'foo'], name='lvl1')), ('foo', 'lvl1', lambda x: x[:, 1:2], Index(['a'], name='lvl0')) @@ -93,6 +105,23 @@ def test_xs_level_eq_2(): tm.assert_frame_equal(result, expected) +@pytest.mark.parametrize('indexer', [ + lambda df: df.xs(('a', 4), level=['one', 'four']), + lambda df: df.xs('a').xs(4, level='four') +]) +def test_xs_level_multiple(indexer, four_level_index_dataframe): + df = four_level_index_dataframe + expected_values = [[0.4473, 1.4152, 0.2834, 1.00661, 0.1744]] + expected_index = MultiIndex( + levels=[['q'], [20.0]], + codes=[[0], [0]], + names=['two', 'three']) + expected = DataFrame( + expected_values, index=expected_index, columns=list('ABCDE')) + result = indexer(df) + tm.assert_frame_equal(result, expected) + + def test_xs_setting_with_copy_error(multiindex_dataframe_random_data): # this is a copy in 0.14 df = multiindex_dataframe_random_data @@ -105,39 +134,31 @@ def test_xs_setting_with_copy_error(multiindex_dataframe_random_data): result[:] = 10 -def test_xs_level_multiple(): - text = """ A B C D E -one two three four -a b 10.0032 5 -0.5109 -2.3358 -0.4645 0.05076 0.3640 -a q 20 4 0.4473 1.4152 0.2834 1.00661 0.1744 -x q 30 3 -0.6662 -0.5243 -0.3580 0.89145 2.5838""" - - df = read_csv(StringIO(text), sep=r'\s+', engine='python') - - result = df.xs(('a', 4), level=['one', 'four']) - expected = df.xs('a').xs(4, level='four') - tm.assert_frame_equal(result, expected) - +def test_xs_setting_with_copy_error_multiple(four_level_index_dataframe): # this is a copy in 0.14 + df = four_level_index_dataframe result = df.xs(('a', 4), level=['one', 'four']) # setting this will give a SettingWithCopyError # as we are trying to write a view - def f(x): - x[:] = 10 + msg = 'A value is trying to be set on a copy of a slice from a DataFrame' + with pytest.raises(com.SettingWithCopyError, match=msg): + result[:] = 10 - pytest.raises(com.SettingWithCopyError, f, result) - # GH2107 +def test_xs_integer_key(): + # see gh-2107 dates = lrange(20111201, 20111205) ids = 'abcde' - idx = MultiIndex.from_tuples([x for x in cart_product(dates, ids)]) - idx.names = ['date', 'secid'] - df = DataFrame(np.random.randn(len(idx), 3), idx, ['X', 'Y', 'Z']) - - rs = df.xs(20111201, level='date') - xp = df.loc[20111201, :] - tm.assert_frame_equal(rs, xp) + index = MultiIndex.from_tuples( + [x for x in cart_product(dates, ids)], + names=['date', 'secid']) + df = DataFrame( + np.random.randn(len(index), 3), index, ['X', 'Y', 'Z']) + + result = df.xs(20111201, level='date') + expected = df.loc[20111201, :] + tm.assert_frame_equal(result, expected) def test_xs_level0(): From 2ade6ef0b0567eb43d299f56ef02de153a5f338b Mon Sep 17 00:00:00 2001 From: Simon Hawkins Date: Thu, 20 Dec 2018 10:03:37 +0000 Subject: [PATCH 7/8] make test_xs_level0 expected explicit --- pandas/tests/indexing/multiindex/test_xs.py | 30 ++++++++++++--------- 1 file changed, 17 insertions(+), 13 deletions(-) diff --git a/pandas/tests/indexing/multiindex/test_xs.py b/pandas/tests/indexing/multiindex/test_xs.py index a14079e7652e6..2db82b26822a7 100644 --- a/pandas/tests/indexing/multiindex/test_xs.py +++ b/pandas/tests/indexing/multiindex/test_xs.py @@ -1,9 +1,9 @@ import numpy as np import pytest -from pandas.compat import StringIO, lrange, product as cart_product +from pandas.compat import lrange, product as cart_product -from pandas import DataFrame, Index, MultiIndex, concat, read_csv +from pandas import DataFrame, Index, MultiIndex, concat import pandas.core.common as com from pandas.util import testing as tm @@ -161,18 +161,22 @@ def test_xs_integer_key(): tm.assert_frame_equal(result, expected) -def test_xs_level0(): - text = """ A B C D E -one two three four -a b 10.0032 5 -0.5109 -2.3358 -0.4645 0.05076 0.3640 -a q 20 4 0.4473 1.4152 0.2834 1.00661 0.1744 -x q 30 3 -0.6662 -0.5243 -0.3580 0.89145 2.5838""" - - df = read_csv(StringIO(text), sep=r'\s+', engine='python') +@pytest.mark.parametrize('indexer', [ + lambda df: df.xs('a', level=0), + lambda df: df.xs('a') +]) +def test_xs_level0(indexer, four_level_index_dataframe): + df = four_level_index_dataframe + expected_values = [[-0.5109, -2.3358, -0.4645, 0.05076, 0.364], + [0.4473, 1.4152, 0.2834, 1.00661, 0.1744]] + expected_index = MultiIndex( + levels=[['b', 'q'], [10.0032, 20.0], [4, 5]], + codes=[[0, 1], [0, 1], [1, 0]], + names=['two', 'three', 'four']) + expected = DataFrame( + expected_values, index=expected_index, columns=list('ABCDE')) - result = df.xs('a', level=0) - expected = df.xs('a') - assert len(result) == 2 + result = indexer(df) tm.assert_frame_equal(result, expected) From 967f775c210397a5723d5c9c4d1fcbdc3bfe6e27 Mon Sep 17 00:00:00 2001 From: Simon Hawkins Date: Thu, 20 Dec 2018 10:29:12 +0000 Subject: [PATCH 8/8] split test_xs_level_series --- pandas/tests/indexing/multiindex/test_xs.py | 40 ++++++++++++--------- 1 file changed, 23 insertions(+), 17 deletions(-) diff --git a/pandas/tests/indexing/multiindex/test_xs.py b/pandas/tests/indexing/multiindex/test_xs.py index 2db82b26822a7..4a8e95b240e39 100644 --- a/pandas/tests/indexing/multiindex/test_xs.py +++ b/pandas/tests/indexing/multiindex/test_xs.py @@ -180,28 +180,34 @@ def test_xs_level0(indexer, four_level_index_dataframe): tm.assert_frame_equal(result, expected) -def test_xs_level_series(multiindex_dataframe_random_data, - multiindex_year_month_day_dataframe_random_data): - frame = multiindex_dataframe_random_data - ymd = multiindex_year_month_day_dataframe_random_data - s = frame['A'] +def test_xs_level_series(multiindex_dataframe_random_data): + # this test is not explicitly testing .xs functionality + # TODO: move to another module or refactor + df = multiindex_dataframe_random_data + s = df['A'] result = s[:, 'two'] - expected = frame.xs('two', level=1)['A'] + expected = df.xs('two', level=1)['A'] tm.assert_series_equal(result, expected) - s = ymd['A'] + +def test_xs_level_series_ymd(multiindex_year_month_day_dataframe_random_data): + # this test is not explicitly testing .xs functionality + # TODO: move to another module or refactor + df = multiindex_year_month_day_dataframe_random_data + s = df['A'] result = s[2000, 5] - expected = ymd.loc[2000, 5]['A'] + expected = df.loc[2000, 5]['A'] tm.assert_series_equal(result, expected) - # not implementing this for now - - pytest.raises(TypeError, s.__getitem__, (2000, slice(3, 4))) - # result = s[2000, 3:4] - # lv =s.index.get_level_values(1) - # expected = s[(lv == 3) | (lv == 4)] - # expected.index = expected.index.droplevel(0) - # tm.assert_series_equal(result, expected) +def test_xs_level_series_slice_not_implemented( + multiindex_year_month_day_dataframe_random_data): + # this test is not explicitly testing .xs functionality + # TODO: move to another module or refactor + # not implementing this for now + df = multiindex_year_month_day_dataframe_random_data + s = df['A'] - # can do this though + msg = r'\(2000, slice\(3, 4, None\)\)' + with pytest.raises(TypeError, match=msg): + s[2000, 3:4]