def test_fit_with_df_input_without_column_arg(self, example_train_df):
"""
In case we give no column argument to the initalizer, the input during fit
should be a pd.Series. Otherwise raise TypeError.
"""
transformer = PandasTfidfVectorizer()
with pytest.raises(TypeError):
transformer.fit(example_train_df)
def test_missing_values_fit(self, example_missing_values_df):
"""
Tests the case where there are missing values in the training data.
Should return a ValueError.
"""
transformer = PandasTfidfVectorizer(column="text")
with pytest.raises(ValueError):
transformer.fit(example_missing_values_df)
def test_fit_with_series_input_with_column_arg(self, example_series):
"""
In case we do give a value for the column keyword argument, the input
should be a pd.DataFrame.
Otherwise, return a TypeError.
"""
transformer = PandasTfidfVectorizer(column="text")
with pytest.raises(TypeError):
transformer.fit(example_series)
def test_missing_column(self, example_train_df,
example_test_df_diff_column):
"""
Test transformer when test set does not have the required columns.
In that case, it should return a KeyError
"""
transformer = PandasTfidfVectorizer(column="text")
transformer.fit(example_train_df)
with pytest.raises(KeyError):
transformer.transform(example_test_df_diff_column)
def test_example(self, example_train_df):
""" Tests a simple example. """
transformer = PandasTfidfVectorizer(column="text")
transformer.fit(example_train_df)
transformed = transformer.transform(example_train_df)
expected = pd.DataFrame({
"num": pd.Series([3, 4, 4]),
"animal": pd.Series([0.0, 1.0, 0.0]),
"house": pd.Series([1.0, 0.0, 1.0]),
})
# The column order shouldnt matter (therefore we sort them)
pd.testing.assert_frame_equal(transformed.sort_index(axis=1),
expected.sort_index(axis=1))
def test_series_input(self, example_series):
"""
In case we don't give a value for the column keyword argument, the input
should be a pandas series or np.ndarray.
Otherwise, return a TypeError.
"""
transformer = PandasTfidfVectorizer()
transformer.fit(example_series)
transformed = transformer.transform(example_series)
expected = pd.DataFrame({
"animal": pd.Series([0.0, 1.0, 0.0]),
"house": pd.Series([1.0, 0.0, 1.0]),
})
pd.testing.assert_frame_equal(transformed.sort_index(axis=1),
expected.sort_index(axis=1))