def test_method_check():
"""
Error check if method in mmm is not in ("mean","median","mode")
"""
# d = pd.DataFrame(data=({'col1': [1,2,None], 'col2': ['a',None,'c']}))
with pytest.raises(Exception):
mmm.mmm(['a', 'b'], method="max")
with pytest.raises(Exception):
mmm.mmm(['a', 'b'], "")
def test_numeric_col():
"""
Checks if dataframe given to mean and median imputation has numeric column
Mean and Median only work with numeric columns, dataframe with non-numeric columns should be invalid
"""
#d = pd.DataFrame(data=({'col1': [True,False,None], 'col2': ['a',None,'c']}))
#e = pd.DataFrame(data=({'col2': ['a',None,'c']}))
with pytest.raises(Exception):
mmm.mmm(pd.DataFrame(data=({
'col1': [True, False, None],
'col2': ['a', None, 'c']
})),
method="mean")
with pytest.raises(Exception):
mmm.mmm(pd.DataFrame(data=({
'col1': [True, False, None],
'col2': ['a', None, 'c']
})),
method="median")
with pytest.raises(Exception):
mmm.mmm(pd.DataFrame(data=({'col2': ['a', None, 'c']})), method="mean")
with pytest.raises(Exception):
mmm.mmm(pd.DataFrame(data=({
'col2': ['a', None, 'c']
})),
method="median")
def test_row_input():
"""
Checks row input in dataframe
dataframe need more than 1 rows to estimate missing
"""
# check if dataframe has more than 1 row
# d = pd.DataFrame(data=({'col1': [1], 'col2': ['a']}))
with pytest.raises(TypeError):
mmm.mmm(pd.DataFrame(data=({
'col1': [1],
'col2': ['a']
})),
method="mean")
with pytest.raises(TypeError):
mmm.mmm(pd.DataFrame(data=({
'col1': [1],
'col2': ['a']
})),
method="median")
with pytest.raises(TypeError):
mmm.mmm(pd.DataFrame(data=({
'col1': [1],
'col2': ['a']
})),
method="mode")
def test_more_missing():
"""
Checks if array or dataframe columns have atleast 2 non-missing for imputing
Mean, Mode, Median all need atleast two not null values to estimate missing
"""
# dataframe with less than 2 non-missing
# d = pd.DataFrame(data=({'col1': [1,None,None], 'col2': ['a',None,None]}))
with pytest.raises(ValueError):
mmm.mmm(pd.DataFrame(data=({
'col1': [1, None, None],
'col2': ['a', None, None]
})),
method="mean")
with pytest.raises(ValueError):
mmm.mmm(pd.DataFrame(data=({
'col1': [1, None, None],
'col2': ['a', None, None]
})),
method="median")
with pytest.raises(ValueError):
mmm.mmm(pd.DataFrame(data=({
'col1': [1, None, None],
'col2': ['a', None, None]
})),
method="mode")
def outputs():
"""
check if mmm return expected output
"""
# check for mmm
df = pd.DataFrame(data=({'v_num': [4,np.nan,5,6,7,np.nan,9]}))
df2 = pd.DataFrame(data=({'v_char': ['one','two',None,'two','two','two']}))
mean_out = mmm.mmm(df,method="mean")
df_mean = pd.DataFrame(data=({'v_num': [4,6.2,5,6,7,6.2,9]}))
assert np.array_equal(mean_out,df_mean)
med_out = mmm.mmm(df,method="median")
df_med = pd.DataFrame(data=({'v_num': [4,6,5,6,7,6,9]}))
assert np.array_equal(med_out,df_med)
mod_out = mmm.mmm(df2,method="mode")
df_mod = pd.DataFrame(data=({'v_char': ['one','two','two','two','two','two']}))
assert np.array_equal(mod_out,df_mod)
def test_input():
"""
Tests incorrect/unacceptable input data for mmm
These should raise a type error.
For mmm datatype should be pandas dataframe
check data type
"""
# Check MMM input other than pandas dataframe
with pytest.raises(TypeError):
mmm.mmm((0, np.nan, 2),method ="mean")
with pytest.raises(TypeError):
mmm.mmm(2,method ="mean")
with pytest.raises(TypeError):
mmm.mmm([1,2,3],method ="mean")
with pytest.raises(TypeError):
mmm.mmm(False,method ="mean")
with pytest.raises(TypeError):
mmm.mmm(['a','b'],method ="mode")
with pytest.raises(TypeError):
mmm.mmm(['a','b'],method ="median")
def test_append():
mmm.mmm(data, "mean")
mmm.mmm(data, "median")
d = pd.DataFrame(data=({'col1': [1], 'col2': ['a']}))
d2 = pd.DataFrame(data=({'col1': [1, None, None], 'col2': ['a', None, None]}))
d3 = pd.DataFrame(data=({'col1': [1, 2, None], 'col2': ['a', None, 'c']}))
d4 = pd.DataFrame(data=({
'col1': [True, False, None],
'col2': ['a', None, 'c']
}))
e = pd.DataFrame(data=({'col2': ['a', None, 'c']}))
df = pd.DataFrame(data=({'v_num': [4, np.nan, 5, 6, 7, np.nan, 9]}))
df2 = pd.DataFrame(data=({
'v_char': ['one', 'two', None, 'two', 'two', 'two']
}))
mean_out = mmm.mmm(df, method="mean")
df_mean = pd.DataFrame(data=({'v_num': [4, 6.2, 5, 6, 7, 6.2, 9]}))
med_out = mmm.mmm(df, method="median")
df_med = pd.DataFrame(data=({'v_num': [4, 6, 5, 6, 7, 6, 9]}))
mod_out = mmm.mmm(df2, method="mode")
df_mod = pd.DataFrame(data=({
'v_char': ['one', 'two', 'two', 'two', 'two', 'two']
}))
# Usage
# mmm.mmm(data,method="mode")
# em.em(matrix)