def test__analyze_invalid_dtype(self):
"""Test _analyze when a list of dtypes containing an invalid dtype is passed."""
# Setup
hp = HyperTransformer(dtypes=['int', 'complex'])
# Run
data = pd.DataFrame({
'int': [1, 2, None],
'complex': [1.0 + 0j, 2.0 + 1j, None],
})
with pytest.raises(ValueError):
hp._analyze(data)
def test__analyze_raise_error(self):
"""Test _analyze raise error"""
# Setup
data = Mock()
data.columns = ['foo']
dtypes = [Mock()]
# Run
transformer = Mock()
transformer.dtypes = dtypes
with self.assertRaises(ValueError):
HyperTransformer._analyze(transformer, data)
def test__analyze(self):
"""Test _analyze"""
# Setup
hp = HyperTransformer(dtype_transformers={'O': 'one_hot_encoding'})
# Run
data = pd.DataFrame({
'int': [1, 2, None],
'float': [1.0, 2.0, None],
'object': ['foo', 'bar', None],
'category': [1, 2, None],
'bool': [True, False, None],
'datetime': pd.to_datetime(['1965-05-23', None, '1997-10-17']),
})
data['category'] = data['category'].astype('category')
result = hp._analyze(data)
# Asserts
assert isinstance(result, dict)
assert set(result.keys()) == {'int', 'float', 'object', 'category', 'bool', 'datetime'}
assert isinstance(result['int'], NumericalTransformer)
assert isinstance(result['float'], NumericalTransformer)
assert isinstance(result['object'], OneHotEncodingTransformer)
assert isinstance(result['category'], OneHotEncodingTransformer)
assert isinstance(result['bool'], BooleanTransformer)
assert isinstance(result['datetime'], DatetimeTransformer)
def test__analyze_bool(self):
"""Test _analyze bool dtype"""
# Setup
data = pd.DataFrame({'booleans': [True, False, None, False, True]})
dtypes = [bool]
# Run
transformer = Mock()
transformer.dtypes = dtypes
result = HyperTransformer._analyze(transformer, data)
# Asserts
expect_class = BooleanTransformer
self.assertIsInstance(result['booleans'], expect_class)
def test__analyze_object(self):
"""Test _analyze object dtype"""
# Setup
data = pd.DataFrame({'objects': ['foo', 'bar', None, 'tar']})
dtypes = [np.object]
# Run
transformer = Mock()
transformer.dtypes = dtypes
result = HyperTransformer._analyze(transformer, data)
# Asserts
expect_class = CategoricalTransformer
self.assertIsInstance(result['objects'], expect_class)
def test__analyze_int(self):
"""Test _analyze int dtype"""
# Setup
data = pd.DataFrame({'integers': [1, 2, 3, 4, 5, None, 6, 7, 8, 9, 0]})
dtypes = [int]
# Run
transformer = Mock()
transformer.dtypes = dtypes
result = HyperTransformer._analyze(transformer, data)
# Asserts
expect_class = NumericalTransformer
self.assertIsInstance(result['integers'], expect_class)
def test__analyze_float(self):
"""Test _analyze float dtype"""
# Setup
data = pd.DataFrame({
'floats': [1.1, 2.2, 3.3, 4.4, 5.5, None, 6.6, 7.7, 8.8, 9.9, 0.0]
})
dtypes = [float]
# Run
transformer = Mock()
transformer.dtypes = dtypes
result = HyperTransformer._analyze(transformer, data)
# Asserts
expect_class = NumericalTransformer
self.assertIsInstance(result['floats'], expect_class)
def test__analyze_datetime64(self):
"""Test _analyze datetime64 dtype"""
# Setup
data = pd.DataFrame({'datetimes': ['1965-05-23', None, '1997-10-17']})
data['datetimes'] = pd.to_datetime(data['datetimes'],
format='%Y-%m-%d',
errors='coerce')
dtypes = [np.datetime64]
# Run
transformer = Mock()
transformer.dtypes = dtypes
result = HyperTransformer._analyze(transformer, data)
# Asserts
expect_class = DatetimeTransformer
self.assertIsInstance(result['datetimes'], expect_class)