Skip to content

Commit ccef23b

Browse files
R-PalazzoR-Palazzo
committed
add one indent
1 parent fc75a82 commit ccef23b

File tree

1 file changed

+118
-118
lines changed

1 file changed

+118
-118
lines changed

‎tests/integration/test_hyper_transformer.py‎

Lines changed: 118 additions & 118 deletions
Original file line numberDiff line numberDiff line change
@@ -2103,121 +2103,121 @@ def test_detect_unsigned_integer_dtypes(self):
21032103
assert config['transformers'][column_name].__class__.__name__ == 'FloatFormatter'
21042104

21052105

2106-
def test_numerical_dtype_handling():
2107-
"""Test that the HyperTransformer correctly handle all numerical dtypes."""
2108-
# Setup
2109-
original_data = pd.DataFrame({
2110-
'Int8': pd.Series([1, 2, 3, pd.NA], dtype='Int8'),
2111-
'Int16': pd.Series([1, 2, 3, pd.NA], dtype='Int16'),
2112-
'Int32': pd.Series([1, 2, 3, pd.NA], dtype='Int32'),
2113-
'Int64': pd.Series([1, 2, 3, pd.NA], dtype='Int64'),
2114-
'UInt8': pd.Series([1, 2, 3, pd.NA], dtype='UInt8'),
2115-
'UInt16': pd.Series([1, 2, 3, pd.NA], dtype='UInt16'),
2116-
'UInt32': pd.Series([1, 2, 3, pd.NA], dtype='UInt32'),
2117-
'UInt64': pd.Series([1, 2, 3, pd.NA], dtype='UInt64'),
2118-
'Float32': pd.Series([1.1, 2.2, 3.3, pd.NA], dtype='Float32'),
2119-
'Float64': pd.Series([1.1, 2.2, 3.3, pd.NA], dtype='Float64'),
2120-
'uint8': np.array([1, 2, 3, 4], dtype='uint8'),
2121-
'uint16': np.array([1, 2, 3, 4], dtype='uint16'),
2122-
'uint32': np.array([1, 2, 3, 4], dtype='uint32'),
2123-
'uint64': np.array([1, 2, 3, 4], dtype='uint64'),
2124-
'float': np.array([1.1, 2.2, 3.3, 4.4], dtype='float'),
2125-
'int8': np.array([1, 2, 3, 4], dtype='int8'),
2126-
'int16': np.array([1, 2, 3, 4], dtype='int16'),
2127-
'int32': np.array([1, 2, 3, 4], dtype='int32'),
2128-
'int64': np.array([1, 2, 3, 4], dtype='int64'),
2129-
})
2130-
2131-
ht = HyperTransformer()
2132-
2133-
# Run
2134-
ht.detect_initial_config(original_data)
2135-
ht.fit(original_data)
2136-
transformed_data = ht.transform(original_data)
2137-
reverse_transformed_data = ht.reverse_transform(transformed_data)
2138-
2139-
# Assert
2140-
assert transformed_data.dtypes.unique() == 'float'
2141-
for column in original_data.columns:
2142-
assert reverse_transformed_data[column].dtype == column
2143-
2144-
2145-
def test_numerical_handling_with_nans():
2146-
"""Test all numerical dtypes handling when there is NaN in the transformed data."""
2147-
# Setup
2148-
original_data = pd.DataFrame({
2149-
'Int8': pd.Series([1, 2, 3, pd.NA], dtype='Int8'),
2150-
'Int16': pd.Series([1, 2, 3, pd.NA], dtype='Int16'),
2151-
'Int32': pd.Series([1, 2, 3, pd.NA], dtype='Int32'),
2152-
'Int64': pd.Series([1, 2, 3, pd.NA], dtype='Int64'),
2153-
'UInt8': pd.Series([1, 2, 3, pd.NA], dtype='UInt8'),
2154-
'UInt16': pd.Series([1, 2, 3, pd.NA], dtype='UInt16'),
2155-
'UInt32': pd.Series([1, 2, 3, pd.NA], dtype='UInt32'),
2156-
'UInt64': pd.Series([1, 2, 3, pd.NA], dtype='UInt64'),
2157-
'Float32': pd.Series([1.1, 2.2, 3.3, pd.NA], dtype='Float32'),
2158-
'Float64': pd.Series([1.1, 2.2, 3.3, pd.NA], dtype='Float64'),
2159-
'uint8': np.array([1, 2, 3, 4], dtype='uint8'),
2160-
'uint16': np.array([1, 2, 3, 4], dtype='uint16'),
2161-
'uint32': np.array([1, 2, 3, 4], dtype='uint32'),
2162-
'uint64': np.array([1, 2, 3, 4], dtype='uint64'),
2163-
'float': np.array([1.1, 2.2, 3.3, 4.4], dtype='float'),
2164-
'int8': np.array([1, 2, 3, 4], dtype='int8'),
2165-
'int16': np.array([1, 2, 3, 4], dtype='int16'),
2166-
'int32': np.array([1, 2, 3, 4], dtype='int32'),
2167-
'int64': np.array([1, 2, 3, 4], dtype='int64'),
2168-
})
2169-
2170-
data_with_nans = pd.DataFrame({
2171-
'Int8': [1.1, 2.2, 3.3, np.nan],
2172-
'Int16': [1.1, 2.2, 3.3, np.nan],
2173-
'Int32': [1.1, 2.2, 3.3, np.nan],
2174-
'Int64': [1.1, 2.2, 3.3, np.nan],
2175-
'UInt8': [1.1, 2.2, 3.3, np.nan],
2176-
'UInt16': [1.1, 2.2, 3.3, np.nan],
2177-
'UInt32': [1.1, 2.2, 3.3, np.nan],
2178-
'UInt64': [1.1, 2.2, 3.3, np.nan],
2179-
'Float32': [1.1, 2.2, 3.3, np.nan],
2180-
'Float64': [1.1, 2.2, 3.3, np.nan],
2181-
'uint8': [1.1, 2.2, 3.3, np.nan],
2182-
'uint16': [1.1, 2.2, 3.3, np.nan],
2183-
'uint32': [1.1, 2.2, 3.3, np.nan],
2184-
'uint64': [1.1, 2.2, 3.3, np.nan],
2185-
'float': [1.1, 2.2, 3.3, np.nan],
2186-
'int8': [1.1, 2.2, 3.3, np.nan],
2187-
'int16': [1.1, 2.2, 3.3, np.nan],
2188-
'int32': [1.1, 2.2, 3.3, np.nan],
2189-
'int64': [1.1, 2.2, 3.3, np.nan],
2190-
})
2191-
2192-
ht = HyperTransformer()
2193-
ht.detect_initial_config(original_data)
2194-
ht.fit(original_data)
2195-
2196-
# Run
2197-
reverse_transformed_data = ht.reverse_transform(data_with_nans)
2198-
2199-
# Assert
2200-
expected_output_dtypes = {
2201-
'Int8': 'Int8',
2202-
'Int16': 'Int16',
2203-
'Int32': 'Int32',
2204-
'Int64': 'Int64',
2205-
'UInt8': 'UInt8',
2206-
'UInt16': 'UInt16',
2207-
'UInt32': 'UInt32',
2208-
'UInt64': 'UInt64',
2209-
'Float32': 'Float32',
2210-
'Float64': 'Float64',
2211-
'uint8': 'float',
2212-
'uint16': 'float',
2213-
'uint32': 'float',
2214-
'uint64': 'float',
2215-
'float': 'float',
2216-
'int8': 'float',
2217-
'int16': 'float',
2218-
'int32': 'float',
2219-
'int64': 'float',
2220-
}
2221-
assert data_with_nans.dtypes.unique() == 'float'
2222-
for column_name, expected_dtype in expected_output_dtypes.items():
2223-
assert reverse_transformed_data[column_name].dtype == expected_dtype
2106+
def test_numerical_dtype_handling(self):
2107+
"""Test that the HyperTransformer correctly handle all numerical dtypes."""
2108+
# Setup
2109+
original_data = pd.DataFrame({
2110+
'Int8': pd.Series([1, 2, 3, pd.NA], dtype='Int8'),
2111+
'Int16': pd.Series([1, 2, 3, pd.NA], dtype='Int16'),
2112+
'Int32': pd.Series([1, 2, 3, pd.NA], dtype='Int32'),
2113+
'Int64': pd.Series([1, 2, 3, pd.NA], dtype='Int64'),
2114+
'UInt8': pd.Series([1, 2, 3, pd.NA], dtype='UInt8'),
2115+
'UInt16': pd.Series([1, 2, 3, pd.NA], dtype='UInt16'),
2116+
'UInt32': pd.Series([1, 2, 3, pd.NA], dtype='UInt32'),
2117+
'UInt64': pd.Series([1, 2, 3, pd.NA], dtype='UInt64'),
2118+
'Float32': pd.Series([1.1, 2.2, 3.3, pd.NA], dtype='Float32'),
2119+
'Float64': pd.Series([1.1, 2.2, 3.3, pd.NA], dtype='Float64'),
2120+
'uint8': np.array([1, 2, 3, 4], dtype='uint8'),
2121+
'uint16': np.array([1, 2, 3, 4], dtype='uint16'),
2122+
'uint32': np.array([1, 2, 3, 4], dtype='uint32'),
2123+
'uint64': np.array([1, 2, 3, 4], dtype='uint64'),
2124+
'float': np.array([1.1, 2.2, 3.3, 4.4], dtype='float'),
2125+
'int8': np.array([1, 2, 3, 4], dtype='int8'),
2126+
'int16': np.array([1, 2, 3, 4], dtype='int16'),
2127+
'int32': np.array([1, 2, 3, 4], dtype='int32'),
2128+
'int64': np.array([1, 2, 3, 4], dtype='int64'),
2129+
})
2130+
2131+
ht = HyperTransformer()
2132+
2133+
# Run
2134+
ht.detect_initial_config(original_data)
2135+
ht.fit(original_data)
2136+
transformed_data = ht.transform(original_data)
2137+
reverse_transformed_data = ht.reverse_transform(transformed_data)
2138+
2139+
# Assert
2140+
assert transformed_data.dtypes.unique() == 'float'
2141+
for column in original_data.columns:
2142+
assert reverse_transformed_data[column].dtype == column
2143+
2144+
2145+
def test_numerical_handling_with_nans(self):
2146+
"""Test all numerical dtypes handling when there is NaN in the transformed data."""
2147+
# Setup
2148+
original_data = pd.DataFrame({
2149+
'Int8': pd.Series([1, 2, 3, pd.NA], dtype='Int8'),
2150+
'Int16': pd.Series([1, 2, 3, pd.NA], dtype='Int16'),
2151+
'Int32': pd.Series([1, 2, 3, pd.NA], dtype='Int32'),
2152+
'Int64': pd.Series([1, 2, 3, pd.NA], dtype='Int64'),
2153+
'UInt8': pd.Series([1, 2, 3, pd.NA], dtype='UInt8'),
2154+
'UInt16': pd.Series([1, 2, 3, pd.NA], dtype='UInt16'),
2155+
'UInt32': pd.Series([1, 2, 3, pd.NA], dtype='UInt32'),
2156+
'UInt64': pd.Series([1, 2, 3, pd.NA], dtype='UInt64'),
2157+
'Float32': pd.Series([1.1, 2.2, 3.3, pd.NA], dtype='Float32'),
2158+
'Float64': pd.Series([1.1, 2.2, 3.3, pd.NA], dtype='Float64'),
2159+
'uint8': np.array([1, 2, 3, 4], dtype='uint8'),
2160+
'uint16': np.array([1, 2, 3, 4], dtype='uint16'),
2161+
'uint32': np.array([1, 2, 3, 4], dtype='uint32'),
2162+
'uint64': np.array([1, 2, 3, 4], dtype='uint64'),
2163+
'float': np.array([1.1, 2.2, 3.3, 4.4], dtype='float'),
2164+
'int8': np.array([1, 2, 3, 4], dtype='int8'),
2165+
'int16': np.array([1, 2, 3, 4], dtype='int16'),
2166+
'int32': np.array([1, 2, 3, 4], dtype='int32'),
2167+
'int64': np.array([1, 2, 3, 4], dtype='int64'),
2168+
})
2169+
2170+
data_with_nans = pd.DataFrame({
2171+
'Int8': [1.1, 2.2, 3.3, np.nan],
2172+
'Int16': [1.1, 2.2, 3.3, np.nan],
2173+
'Int32': [1.1, 2.2, 3.3, np.nan],
2174+
'Int64': [1.1, 2.2, 3.3, np.nan],
2175+
'UInt8': [1.1, 2.2, 3.3, np.nan],
2176+
'UInt16': [1.1, 2.2, 3.3, np.nan],
2177+
'UInt32': [1.1, 2.2, 3.3, np.nan],
2178+
'UInt64': [1.1, 2.2, 3.3, np.nan],
2179+
'Float32': [1.1, 2.2, 3.3, np.nan],
2180+
'Float64': [1.1, 2.2, 3.3, np.nan],
2181+
'uint8': [1.1, 2.2, 3.3, np.nan],
2182+
'uint16': [1.1, 2.2, 3.3, np.nan],
2183+
'uint32': [1.1, 2.2, 3.3, np.nan],
2184+
'uint64': [1.1, 2.2, 3.3, np.nan],
2185+
'float': [1.1, 2.2, 3.3, np.nan],
2186+
'int8': [1.1, 2.2, 3.3, np.nan],
2187+
'int16': [1.1, 2.2, 3.3, np.nan],
2188+
'int32': [1.1, 2.2, 3.3, np.nan],
2189+
'int64': [1.1, 2.2, 3.3, np.nan],
2190+
})
2191+
2192+
ht = HyperTransformer()
2193+
ht.detect_initial_config(original_data)
2194+
ht.fit(original_data)
2195+
2196+
# Run
2197+
reverse_transformed_data = ht.reverse_transform(data_with_nans)
2198+
2199+
# Assert
2200+
expected_output_dtypes = {
2201+
'Int8': 'Int8',
2202+
'Int16': 'Int16',
2203+
'Int32': 'Int32',
2204+
'Int64': 'Int64',
2205+
'UInt8': 'UInt8',
2206+
'UInt16': 'UInt16',
2207+
'UInt32': 'UInt32',
2208+
'UInt64': 'UInt64',
2209+
'Float32': 'Float32',
2210+
'Float64': 'Float64',
2211+
'uint8': 'float',
2212+
'uint16': 'float',
2213+
'uint32': 'float',
2214+
'uint64': 'float',
2215+
'float': 'float',
2216+
'int8': 'float',
2217+
'int16': 'float',
2218+
'int32': 'float',
2219+
'int64': 'float',
2220+
}
2221+
assert data_with_nans.dtypes.unique() == 'float'
2222+
for column_name, expected_dtype in expected_output_dtypes.items():
2223+
assert reverse_transformed_data[column_name].dtype == expected_dtype

0 commit comments

Comments
 (0)