def test_array_td(self):
a = np.array([10, 20, 30], dtype='timedelta64[Y]')
b = np.array([1, 2, 3], dtype='timedelta64[Y]')
arr = np.column_stack((a, b)).astype('timedelta64[ns]')
df1 = dx.DataFrame({'a': a, 'b': b})
assert_array_equal(arr, df1._data['m'])
assert df1._column_info['a'].values == ('m', 0, 0)
assert df1._column_info['b'].values == ('m', 1, 1)
def test_array_bool(self):
a = np.array([True, False, True])
b = np.array([False, False, False])
arr = np.column_stack((a, b)).astype('int8')
df1 = dx.DataFrame({'a': a, 'b': b})
assert_array_equal(arr, df1._data['b'])
assert df1._column_info['a'].values == ('b', 0, 0)
assert df1._column_info['b'].values == ('b', 1, 1)
def test_array_dt(self):
a = np.array([10, 20, 30], dtype='datetime64[ns]')
b = np.array([100, 200, 300], dtype='datetime64[ns]')
arr = np.column_stack((a, b))
df1 = dx.DataFrame({'a': a, 'b': b})
assert_array_equal(arr, df1._data['M'])
assert df1._column_info['a'].values == ('M', 0, 0)
assert df1._column_info['b'].values == ('M', 1, 1)
def test_array_string(self):
a = np.array(['asdf', 'wer'])
b = np.array(['wyw', 'xcvd'])
df1 = dx.DataFrame({'a': a, 'b': b})
a1 = array([[1, 1], [2, 2]], dtype='uint32')
assert_array_equal(a1, df1._data['S'])
assert df1._column_info['a'].values == ('S', 0, 0)
assert df1._column_info['b'].values == ('S', 1, 1)
def test_array_float(self):
a = np.array([1.1, 2, 3])
b = np.array([10, 20.2, 30])
arr = np.column_stack((a, b))
df1 = dx.DataFrame({'a': a, 'b': b})
assert_array_equal(arr, df1._data['f'])
assert df1._column_info['a'].values == ('f', 0, 0)
assert df1._column_info['b'].values == ('f', 1, 1)
def test_to_dict(self):
d1 = self.df1.to_dict('array')
d2 = {
'a': np.array([1, 5, 7, 11]),
'b': np.array([nan, 5.4, -1.1, .045])
}
for key, arr in d1.items():
assert_array_equal(arr, d2[key])
d1 = self.df1.to_dict('list')
d2 = {'a': [1, 5, 7, 11], 'b': [nan, 5.4, -1.1, .045]}
assert_dict_list(d1, d2)
def test_streak_group(self):
df = de.DataFrame(
{'AIRLINE': ['AA', 'AA', 'AA', 'UA', 'DL', 'DL', 'WN', 'WN', 'AA', 'AA', None],
'DAY_OF_WEEK': [2, 3, 6, 6, 6, 6, 4, 4, 1, 6, 6],
'DEPARTURE_DELAY': [nan, nan, -1.0, -1.0, -1.0, 22.0, 3.0, 3.0, 21.0,
-2.0, nan]})
arr1 = df.streak('AIRLINE', group=True)
arr2 = array([1, 1, 1, 2, 3, 3, 4, 4, 5, 5, 6])
assert_array_equal(arr1, arr2)
arr1 = df.streak('DEPARTURE_DELAY', group=True)
arr2 = array([1, 2, 3, 3, 3, 4, 5, 5, 6, 7, 8])
assert_array_equal(arr1, arr2)
def test_streak_value(self):
df = de.DataFrame(
{'AIRLINE': ['AA', 'AA', 'AA', 'UA', 'DL', 'DL', 'WN', 'WN', 'WN', 'AS', None],
'DAY_OF_WEEK': [2, 3, 6, 6, 6, 6, 4, 4, 1, 2, 2],
'DEPARTURE_DELAY': [nan, nan, -1.0, -1.0, -1.0, 22.0, 3.0, 3.0, 21.0,
-2.0, nan]})
with pytest.raises(TypeError):
df.streak('DEPARTURE_DELAY', 'AA')
arr1 = df.streak('DEPARTURE_DELAY', -1)
arr2 = array([0, 0, 1, 2, 3, 0, 0, 0, 0, 0, 0])
assert_array_equal(arr1, arr2)
arr1 = df.streak('DAY_OF_WEEK', 6)
arr2 = array([0, 0, 1, 2, 3, 4, 0, 0, 0, 0, 0])
assert_array_equal(arr1, arr2)
def test_get_values(self):
values1 = self.df1.values
values2 = np.array([[1, 5, 7, 11], [nan, 5.4, -1.1, .045]]).T
assert_array_equal(values1, values2)
a = np.random.rand(100, 5)
df = dx.DataFrame(a)
assert_array_equal(df.values, a)
values1 = self.df2.values
values2 = np.array([[1, 5, 7, 11], [nan, 5.4, -1.1, .045],
['ted', 'fred', 'ted', 'fred']],
dtype='O').T
assert_array_equal(values1, values2)
def test_streak(self):
df = de.DataFrame(
{'AIRLINE': ['AA', 'AA', 'AA', 'UA', 'DL', 'DL', 'WN', 'WN', 'WN', 'AS', None],
'DAY_OF_WEEK': [2, 3, 6, 6, 6, 6, 4, 4, 1, 2, 2],
'DEPARTURE_DELAY': [nan, nan, -1.0, -1.0, -1.0, 22.0, 3.0, 3.0, 21.0,
-2.0, nan]})
arr1 = df.streak('AIRLINE')
arr2 = array([1, 2, 3, 1, 1, 2, 1, 2, 3, 1, 1])
assert_array_equal(arr1, arr2)
arr1 = df.streak('DAY_OF_WEEK')
arr2 = array([1, 1, 1, 2, 3, 4, 1, 2, 1, 1, 2])
assert_array_equal(arr1, arr2)
arr1 = df.streak('DEPARTURE_DELAY')
arr2 = array([1, 1, 1, 2, 3, 1, 1, 2, 1, 1, 1])
assert_array_equal(arr1, arr2)
def test_single_list_string(self):
a = np.array(['a', 'b'])
df1 = dx.DataFrame({'a': a.tolist()})
a1 = array([1, 2], dtype='uint32')
assert_array_equal(a1, df1._data['S'][:, 0])
assert df1._column_info['a'].values == ('S', 0, 0)
def test_all(self):
assert_array_equal(np.array(a), df_mix._data['i'][:, 0])
assert_array_equal(np.array(b), df_mix._data['f'][:, 0])
a1 = array([1, 2, 3, 4, 5, 6, 7, 8], dtype='uint32')
assert_array_equal(a1, df_mix._data['S'][:, 0])
assert_array_equal(np.array(d).astype('int8'), df_mix._data['b'][:, 0])
assert_array_equal(np.array(e, dtype='datetime64[ns]'), df_mix._data['M'][:, 0])
assert_array_equal(np.array(f, dtype='timedelta64[ns]'), df_mix._data['m'][:, 0])
assert df_mix._column_info['a'].values == ('i', 0, 0)
assert df_mix._column_info['b'].values == ('f', 0, 1)
assert df_mix._column_info['c'].values == ('S', 0, 2)
assert df_mix._column_info['d'].values == ('b', 0, 3)
assert df_mix._column_info['e'].values == ('M', 0, 4)
assert df_mix._column_info['f'].values == ('m', 0, 5)
def test_single_array_bool(self):
a = np.array([True, False])
df1 = dx.DataFrame(a)
assert_array_equal(a.astype('int8'), df1._data['b'][:, 0])
assert df1._column_info['a0'].values == ('b', 0, 0)
def test_single_array_float(self):
a = np.array([1, 2.5, 3.2])
df1 = dx.DataFrame(a)
assert_array_equal(a, df1._data['f'][:, 0])
assert df1._column_info['a0'].values == ('f', 0, 0)
def test_single_array_int(self):
a = np.array([1, 2, 3])
df1 = dx.DataFrame({'a': a})
assert_array_equal(a, df1._data['i'][:, 0])
assert df1._column_info['a'].values == ('i', 0, 0)
def test_all(self):
assert_array_equal(a1, df_mix2._data['i'][:, 0])
assert_array_equal(b1, df_mix2._data['f'][:, 0])
arr1 = array([1, 2, 3, 4, 5, 6, 7, 8], dtype='uint32')
assert_array_equal(arr1, df_mix2._data['S'][:, 0])
assert_array_equal(d1.astype('int8'), df_mix2._data['b'][:, 0])
assert_array_equal(e1, df_mix2._data['M'][:, 0])
assert_array_equal(f1, df_mix2._data['m'][:, 0])
assert df_mix2._column_info['a0'].values == ('i', 0, 0)
assert df_mix2._column_info['a1'].values == ('f', 0, 1)
assert df_mix2._column_info['a2'].values == ('S', 0, 2)
assert df_mix2._column_info['a3'].values == ('b', 0, 3)
assert df_mix2._column_info['a4'].values == ('M', 0, 4)
assert df_mix2._column_info['a5'].values == ('m', 0, 5)
def test_single_array_dt(self):
a = np.array([10, 20, 30], dtype='datetime64[ns]')
df1 = dx.DataFrame({'a': a})
assert_array_equal(a, df1._data['M'][:, 0])
assert df1._column_info['a'].values == ('M', 0, 0)
def test_single_array_td(self):
a = np.array([10, 20, 30], dtype='timedelta64[Y]')
df1 = dx.DataFrame({'a': a})
assert_array_equal(a.astype('timedelta64[ns]'), df1._data['m'][:, 0])
assert df1._column_info['a'].values == ('m', 0, 0)
def test_factorize(self):
data = {'a': [9, 10, 9, 9, 10],
'b': [0, nan, nan, 0, 1],
'c': [''] + list('eeaz'),
'd': [False, False, True, False, True],
'e': [0, 20, 30, 4, 4],
'f': ['a', nan, 'ad', None, 'ad'],
'g': [np.nan] * 5}
df = de.DataFrame(data)
arr11, arr12 = df.factorize('a')
arr21, arr22 = (array([0, 1, 0, 0, 1]), array([9, 10]))
assert_array_equal(arr11, arr21)
assert_array_equal(arr12, arr22)
arr11, arr12 = df.factorize('b')
arr21, arr22 = (array([0, 1, 1, 0, 2]), array([0., nan, 1.]))
assert_array_equal(arr11, arr21)
assert_array_equal(arr12, arr22)
arr11, arr12 = df.factorize('c')
arr21, arr22 = (array([0, 1, 1, 2, 3]), array(['', 'e', 'a', 'z'], dtype=object))
assert_array_equal(arr11, arr21)
assert_array_equal(arr12, arr22)
arr11, arr12 = df.factorize('d')
arr21, arr22 = (array([0, 0, 1, 0, 1]), array([False, True]))
assert_array_equal(arr11, arr21)
assert_array_equal(arr12, arr22)
def test_single_list_td(self):
a = [np.timedelta64(x, 'ns') for x in [10, 20, 30]]
df1 = dx.DataFrame({'a': a})
assert_array_equal(np.array(a), df1._data['m'][:, 0])
assert df1._column_info['a'].values == ('m', 0, 0)
