def test_happy_scale_features():
"""Happy path test for testing `scale_features`"""
test_data, columns = create_data()
columns_to_encode = ['genres', 'parent_platforms']
columns_to_drop = [
'name', 'genres', 'parent_platforms', 'stores', 'background_image',
'released'
]
test_data.genres = test_data.genres.apply(
lambda x: test_featurize_helpers.get_genres(x))
test_data.parent_platforms = test_data.parent_platforms.apply(
lambda x: test_featurize_helpers.get_platforms(x))
# encode features
output_data, intermediate_data = test_featurize_helpers.encode_features(
test_data,
columns_to_encode=columns_to_encode,
columns_to_drop=columns_to_drop)
scaled_output = test_featurize_helpers.scale_features(output_data)
# check datatype, shape and column names
test1 = (type(scaled_output).__module__ == np.__name__)
test2 = (scaled_output.shape[0] == test_data.shape[0])
assert (test1 and test2)
def test_happy_encode_features():
"""Happy path test for testing `encode_features`"""
test_data, columns = create_data()
test_data.genres = test_data.genres.apply(
lambda x: test_featurize_helpers.get_genres(x))
test_data.parent_platforms = test_data.parent_platforms.apply(
lambda x: test_featurize_helpers.get_platforms(x))
columns_to_encode = ['genres', 'parent_platforms']
columns_to_drop = [
'name', 'genres', 'parent_platforms', 'stores', 'background_image',
'released'
]
#encode features
output_data, intermediate_data = test_featurize_helpers.encode_features(
test_data,
columns_to_encode=columns_to_encode,
columns_to_drop=columns_to_drop)
# check datatype, shape and column names
test1 = isinstance(output_data, pd.DataFrame)
test2 = (output_data.shape[0] == test_data.shape[0])
test3 = np.any(test_data.columns.isin(columns_to_encode))
test4 = np.any(test_data.columns.isin(columns_to_drop))
assert (test1 and test2 and test3 and test4)
def test_unhappy_get_platfroms():
"""Unhappy path test for testing `get_platforms` with adding missing data"""
test_data, columns = create_data()
test_data = test_data.append(pd.Series(dtype='object'), ignore_index=True)
output_data = test_data.copy()
# create feature
with pytest.raises(SystemExit):
output_data.parent_platforms = test_data.parent_platforms.\
apply(lambda x: test_featurize_helpers.get_platforms(x))
assert output_data.parent_platforms is None
def test_happy_get_platfroms():
"""Happy path test for testing `get_platforms`"""
test_data, columns = create_data()
# create feature
output_platforms = test_data.parent_platforms.apply(
lambda x: test_featurize_helpers.get_platforms(x))
# check datatype, shape and column names
test1 = (output_platforms.dtype == 'object')
test2 = (output_platforms.shape[0] == test_data.shape[0])
test3 = np.any(test_data.columns.isin(['parent_platforms']))
test4 = ('parent_platforms' == output_platforms.name)
assert (test1 and test2 and test3 and test4)
def test_unhappy_scale_features():
"""UnHappy path test for testing `scale_features`,by giving dropping giving a categorical variable as well"""
test_data, columns = create_data()
test_data.genres = test_data.genres.apply(
lambda x: test_featurize_helpers.get_genres(x))
test_data.parent_platforms = test_data.parent_platforms.apply(
lambda x: test_featurize_helpers.get_platforms(x))
columns_to_encode = ['genres', 'parent_platforms']
#haven't dropped name
columns_to_drop = [
'genres', 'parent_platforms', 'stores', 'background_image', 'released'
]
# encode features
output_data, intermediate_data = test_featurize_helpers.encode_features(
test_data,
columns_to_encode=columns_to_encode,
columns_to_drop=columns_to_drop)
#name has been included to be scaled
with pytest.raises(SystemExit):
output_scaled = test_featurize_helpers.scale_features(output_data)
assert output_scaled is None
def test_unhappy_encode_features():
"""UnHappy path test for testing `encode_features`,by dropping columns to encode"""
test_data, columns = create_data()
test_data.genres = test_data.genres.apply(
lambda x: test_featurize_helpers.get_genres(x))
test_data.parent_platforms = test_data.parent_platforms.apply(
lambda x: test_featurize_helpers.get_platforms(x))
columns_to_encode = None
columns_to_drop = [
'name', 'genres', 'parent_platforms', 'stores', 'background_image',
'released'
]
# create feature
with pytest.raises(SystemExit):
output_data, intermediate_data = test_featurize_helpers.encode_features(
test_data,
columns_to_encode=columns_to_encode,
columns_to_drop=columns_to_drop)
assert output_data is None
assert intermediate_data is None