def test_concat_inplace_is_true_empty_col_in_input(self):
df = sampy.DataFrameXS()
df['col1'] = [5, 6, 7, 8]
df['col2'] = None
self.df.concat(df, inplace=True)
self.assertTrue((np.nan_to_num(self.df['col2'], nan=0.) == np.array(
[2., 3., 4., 5., 0., 0., 0., 0.])).all())
self.assertTrue((self.df['col1'] == np.array([1, 2, 3, 4, 5, 6, 7,
8])).all())
self.assertEqual(str(self.df['col1'].dtype), str(df['col1'].dtype))
def test_concat_inplace_is_true_basic(self):
df = sampy.DataFrameXS()
df['col1'] = [5, 6, 7, 8]
df['col2'] = [6., 7., 8., 9.]
self.df.concat(df, inplace=True)
# concats happened as expected
self.assertTrue((self.df['col1'] == np.array([1, 2, 3, 4, 5, 6, 7,
8])).all())
self.assertTrue(
(self.df['col2'] == np.array([2., 3., 4., 5., 6., 7., 8.,
9.])).all())
def test_concat_inplace_is_true_empty_col_on_empty_col(self):
self.df['col2'] = None
df = sampy.DataFrameXS()
df['col1'] = [5, 6, 7, 8]
df['col2'] = None
self.df.concat(df, inplace=True)
self.assertIsNone(self.df['col2'])
self.assertTrue((self.df['col1'] == np.array([1, 2, 3, 4, 5, 6, 7,
8])).all())
self.assertEqual(str(self.df['col1'].dtype), str(df['col1'].dtype))
def test_concat_inplace_is_true_on_empty_df(self):
self.df['col1'] = None
self.df['col2'] = None
df = sampy.DataFrameXS()
df['col1'] = [5, 6, 7, 8]
df['col2'] = [6., 7., 8., 9.]
self.df.concat(df, inplace=True)
self.assertTrue((self.df['col1'] == np.array([5, 6, 7, 8])).all())
self.assertEqual(str(self.df['col1'].dtype), str(df['col1'].dtype))
self.assertTrue((np.nan_to_num(self.df['col2'],
nan=0.) == np.array([6., 7., 8.,
9.])).all())
self.assertEqual(str(self.df['col2'].dtype), str(df['col2'].dtype))
self.assertEqual(self.df.nb_rows, 4)
def test_concat_inplace_is_true_extra_col(self):
df = sampy.DataFrameXS()
df['col1'] = [5, 6, 7, 8]
df['col2'] = [6., 7., 8., 9.]
df['col3'] = [1, 2, 3, 4]
self.df.concat(df, inplace=True)
# check that there is no col3 in the dataframe
with self.assertRaises(KeyError):
self.df['col3']
# perform all the previous checks in this case.
# concat happened as expected
self.assertTrue((self.df['col1'] == np.array([1, 2, 3, 4, 5, 6, 7,
8])).all())
self.assertTrue(
(self.df['col2'] == np.array([2., 3., 4., 5., 6., 7., 8.,
9.])).all())
self.assertEqual(self.df.nb_rows, 8)
def test_concat_inplace_is_true_wrong_col_name(self):
df = sampy.DataFrameXS()
df['col1'] = [5, 6, 7, 8]
df['col3'] = [6., 7., 8., 9.]
with self.assertRaises(KeyError):
self.df.concat(df, inplace=True)
def test_concat_inplace_is_false(self):
with self.subTest():
df = sampy.DataFrameXS()
df['col1'] = [5, 6, 7, 8]
df['col2'] = [6., 7., 8., 9.]
c_df = self.df.concat(df, inplace=False)
# concats happened as expected
self.assertTrue((c_df['col1'] == np.array([1, 2, 3, 4, 5, 6, 7,
8])).all())
self.assertTrue(
(c_df['col2'] == np.array([2., 3., 4., 5., 6., 7., 8.,
9.])).all())
# columns are new copied columns
c_df['col1'][0] = 0
self.assertFalse(c_df['col1'][0] == self.df['col1'][0])
with self.subTest():
# same test as the previous one, but there is an extra column
df = sampy.DataFrameXS()
df['col1'] = [5, 6, 7, 8]
df['col2'] = [6., 7., 8., 9.]
df['col3'] = [1, 2, 3, 4]
c_df = self.df.concat(df, inplace=False)
# check that there is no col3 in the dataframe
with self.assertRaises(KeyError):
c_df['col3']
# perform all the previous checks in this case.
# concat happened as expected
self.assertTrue((c_df['col1'] == np.array([1, 2, 3, 4, 5, 6, 7,
8])).all())
self.assertTrue(
(c_df['col2'] == np.array([2., 3., 4., 5., 6., 7., 8.,
9.])).all())
# columns are new copied columns
c_df['col1'][0] = 0
self.assertFalse(c_df['col1'][0] == self.df['col1'][0])
# check that an error is raised when the df don't have the right col_names
with self.subTest():
df = sampy.DataFrameXS()
df['col1'] = [5, 6, 7, 8]
df['col3'] = [6., 7., 8., 9.]
with self.assertRaises(KeyError):
self.df.concat(df, inplace=False)
# check that appending with an empty dataframe returns a copy
with self.subTest():
df = sampy.DataFrameXS()
c_df = self.df.concat(df, inplace=False)
self.assertTrue((c_df['col1'] == self.df['col1']).all())
self.assertTrue((c_df['col2'] == self.df['col2']).all())
c_df['col1'][0] = 0
self.assertFalse(c_df['col1'][0] == self.df['col1'][0])
# check that having a null column in the input is correctly dealt with
with self.subTest():
df = sampy.DataFrameXS()
df['col1'] = [5, 6, 7, 8]
df['col2'] = None
c_df = self.df.concat(df, inplace=False)
self.assertTrue((c_df['col1'] == np.array([1, 2, 3, 4, 5, 6, 7,
8])).all())
self.assertTrue((np.nan_to_num(c_df['col2'], nan=0.) == np.array(
[2., 3., 4., 5., 0., 0., 0., 0.])).all())
self.assertEqual(str(c_df['col2'].dtype),
str(self.df['col2'].dtype))
# same thing but in reverse
with self.subTest():
df_in = sampy.DataFrameXS()
df_in['col1'] = [1, 2, 3, 4]
df_in['col2'] = None
df = sampy.DataFrameXS()
df['col1'] = [5, 6, 7, 8]
df['col2'] = [6., 7., 8., 9.]
c_df = df_in.concat(df, inplace=False)
self.assertTrue((c_df['col1'] == np.array([1, 2, 3, 4, 5, 6, 7,
8])).all())
self.assertTrue((np.nan_to_num(c_df['col2'], 0.) == np.array(
[0., 0., 0., 0., 6., 7., 8., 9.])).all())
self.assertEqual(str(c_df['col2'].dtype), str(df['col2'].dtype))
# When the first df is empty
with self.subTest():
df_in = sampy.DataFrameXS()
df_in['col1'] = None
df_in['col2'] = None
df = sampy.DataFrameXS()
df['col1'] = [5, 6, 7, 8]
df['col2'] = [6., 7., 8., 9.]
c_df = df_in.concat(df, inplace=False)
self.assertTrue((c_df['col1'] == np.array([5, 6, 7, 8])).all())
self.assertEqual(str(c_df['col1'].dtype), str(df['col1'].dtype))
self.assertTrue(
(np.nan_to_num(c_df['col2'],
nan=0.) == np.array([6., 7., 8., 9.])).all())
self.assertEqual(str(c_df['col2'].dtype), str(df['col2'].dtype))
# check that a copy is retrieved
c_df['col1'][0] = 0
self.assertFalse((c_df['col1'] == df['col1']).all())
# adding an empty column on an empty column
with self.subTest():
df_in = sampy.DataFrameXS()
df_in['col1'] = [1, 2, 3, 4]
df_in['col2'] = None
df = sampy.DataFrameXS()
df['col1'] = [5, 6, 7, 8]
df['col2'] = None
c_df = df_in.concat(df, inplace=False)
self.assertIsNone(c_df['col2'])
self.assertTrue((c_df['col1'] == np.array([1, 2, 3, 4, 5, 6, 7,
8])).all())
self.assertEqual(str(c_df['col1'].dtype), str(df['col1'].dtype))
def setUp(self):
self.df = sampy.DataFrameXS()
self.df['col1'] = [1, 2, 3, 4]
self.df['col2'] = [2., 3., 4., 5.]
def setUp(self):
self.df = sampy.DataFrameXS()