def test_except_nonnumeric():
'''
This function tests that find_bad_apples()
raises an exception
when a dataframe with non-numeric columns
is passed as an argument.
Parameter
---------
None
Returns
-------
None
Example
-------
test_find_bad_apples()
'''
df_test1 = pd.DataFrame({
'A': [
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, "test"
]
})
with pytest.raises(ValueError):
find_bad_apples(df_test1)
def test_except_min():
'''
This function tests that find_bad_apples()
raises an exception
when a dataframe with fewer than 30 rows
is passed as an argument.
Parameter
---------
None
Returns
-------
None
Example
-------
test_find_bad_apples()
'''
df_test2 = pd.DataFrame({
'A': [
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1
]
})
with pytest.raises(ValueError):
find_bad_apples(df_test2)
def test_pos_neg():
'''
This function tests that find_bad_apples()
produces the correct output
when a dataframe with positive and negative values
in the same column is passed as an argument.
Parameter
---------
None
Returns
-------
None
Example
-------
test_find_bad_apples()
'''
df_test3 = pd.DataFrame({
'A': [
100, -100, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1
]
})
output3 = find_bad_apples(df_test3)
assert (output3.query('index == 0')['Variable'].values[0] == 'A')
assert (output3.query('index == 0')['Indices'].values[0] == [0, 1])
assert (output3.query('index == 0')['Total Outliers'].values[0] == 2)
def test_one_col():
'''
This function tests that find_bad_apples()
produces the correct output
when a dataframe with only one column
is passed as an argument.
Parameter
---------
None
Returns
-------
None
Example
-------
test_find_bad_apples()
'''
df_test5 = pd.DataFrame({
'A': [
1, 1, 10, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1
]
})
output5 = find_bad_apples(df_test5)
assert (output5.query('index == 0')['Variable'].values[0] == 'A')
assert (output5.query('index == 0')['Indices'].values[0] == [2])
assert (output5.query('index == 0')['Total Outliers'].values[0] == 1)
def test_no_bad_apples():
'''
This function tests that find_bad_apples() returns
a dataframe with 'No outliers detected'
when there are no outliers present in the dataframe.
Parameter
---------
None
Returns
-------
None
Example
-------
test_find_bad_apples()
'''
df_test4 = pd.DataFrame({
'A': [
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1
]
})
output4 = find_bad_apples(df_test4)
assert (output4.query('index == 0')['Variable'].values[0] ==
'No outliers detected')
assert (output4.query('index == 0')['Indices'].values[0] == 'x')
assert (output4.query('index == 0')['Total Outliers'].values[0] == 0)