def test_pdtype2():
"""Tests given mean/median, df is all numeric"""
test_df = pd.DataFrame([["a", "b", "c"], ["d", np.nan, "f"],
["z", "x", "y"]])
with pytest.raises(TypeError):
imputer = imputation("mean")
imputer.fit(test_df)
with pytest.raises(TypeError):
imputer = imputation("median")
imputer.fit(test_df)
def test_length():
"""Tests whether columns of fit = fill"""
test_df1 = pd.DataFrame([["a", "b", "c"], ["d", np.nan, "f"],
["z", "x", "y"]])
test_df2 = pd.DataFrame([["a", "b"], ["d", np.nan], ["z", "x"]])
with pytest.raises(TypeError):
imputer = imputation("most_frequent")
imputer.fit(test_df1)
imputer.fill(test_df2)
def test_imputation():
"""Group of tests that test for the output generated from impute"""
df1 = pd.DataFrame([[np.nan, 2, 3], [4, np.nan, 6], [10, 5, 9]])
df2 = pd.DataFrame([[np.nan, 2, 3], [4, np.nan, 6], [10, 5, 9],
[3, 15, 17]])
df3 = pd.DataFrame([[np.nan, "b", "c"], ["d", np.nan, "f"],
["d", "x", np.nan]])
df_m = pd.DataFrame([[np.nan, 1, "c"], ["d", np.nan, "f"],
["d", 3, np.nan]])
df_s = pd.DataFrame([1])
df1_t = pd.DataFrame([[7.0, 2, 3], [4, 3.5, 6], [10, 5, 9]])
df2_t = pd.DataFrame([[4.0, 2.0, 3.0], [4.0, 5.0, 6.0], [10.0, 5.0, 9.0],
[3.0, 15.0, 17.0]])
df3_t = pd.DataFrame([["d", "b", "c"], ["d", "b", "f"], ["d", "x", "c"]])
df_m_t = pd.DataFrame([["d", 1.0, "c"], ["d", 1.0, "f"], ["d", 3.0, "c"]])
# test for mean imputation
imputer = imputation("mean")
imputer.fit(df1)
df1_filled = imputer.fill(df1)
assert np.array_equal(df1_filled.to_numpy(),
df1_t.to_numpy()), "Generated Output are incorrect"
# test for median imputation
imputer = imputation("median")
imputer.fit(df2)
df2_filled = imputer.fill(df2)
assert np.array_equal(df2_filled.to_numpy(),
df2_t.to_numpy()), "Generated Output are incorrect"
# test for most frequent imputation
imputer = imputation("most_frequent")
imputer.fit(df3)
df3_filled = imputer.fill(df3)
assert np.array_equal(df3_filled.to_numpy(),
df3_t.to_numpy()), "Generated Output are incorrect"
# test for dataframe with non nan values
imputer = imputation("mean")
imputer.fit(df1_t)
df4_filled = imputer.fill(df1_t)
assert np.array_equal(df4_filled.to_numpy(),
df1_t.to_numpy()), "Generated Output are incorrect"
# test for mix type dataframe for most_frequent
imputer = imputation("most_frequent")
imputer.fit(df_m)
df5_filled = imputer.fill(df_m)
assert np.array_equal(df5_filled.to_numpy(),
df_m_t.to_numpy()), "Generated Output are incorrect"
# test for dataframe where there is only a single value
imputer = imputation("mean")
imputer.fit(df_s)
df6_filled = imputer.fill(df_s)
assert np.array_equal(df6_filled.to_numpy(),
df_s.to_numpy()), "Generated Output are incorrect"
def test_pdtype():
"""Tests whether Class Method catches wrong datatype"""
with pytest.raises(TypeError):
imputer = imputation("mean")
imputer.fit(1)
def test_method():
"""Tests catches wrong method for imputation"""
with pytest.raises(KeyError):
imputer = imputation("hello") # noqa: F841
def test_empty():
"""Tests whether Class Method catches empty dataframe"""
emp = pd.DataFrame()
imputer = imputation("mean")
with pytest.raises(ValueError):
imputer.fit(emp)