Impl Series.skew()

examples/series/series_skew.py

@@ -0,0 +1,39 @@
+# *****************************************************************************
+# Copyright (c) 2020, Intel Corporation All rights reserved.
+#
+# Redistribution and use in source and binary forms, with or without
+# modification, are permitted provided that the following conditions are met:
+#
+#     Redistributions of source code must retain the above copyright notice,
+#     this list of conditions and the following disclaimer.
+#
+#     Redistributions in binary form must reproduce the above copyright notice,
+#     this list of conditions and the following disclaimer in the documentation
+#     and/or other materials provided with the distribution.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+# AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO,
+# THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+# PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR
+# CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+# EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+# PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+# OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+# WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+# OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
+# EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+# *****************************************************************************
+
+import pandas as pd
+import numpy as np
+from numba import njit
+
+
+@njit
+def series_skew():
+    s = pd.Series([np.nan, -2., 3., 5.0])
+
+    return s.skew()  # Expect -1.1520696383139375
+
+
+print(series_skew())

sdc/datatypes/hpat_pandas_series_functions.py

@@ -5725,3 +5725,89 @@ def sdc_pandas_series_groupby_impl(self, by=None, axis=0, level=None, as_index=T
         return init_series_groupby(self, by, grouped, sort)
 
     return sdc_pandas_series_groupby_impl
+
+
+@sdc_overload_method(SeriesType, 'skew')
+def sdc_pandas_series_skew(self, axis=None, skipna=None, level=None, numeric_only=None):
+    """
+        Intel Scalable Dataframe Compiler User Guide
+        ********************************************
+
+        Pandas API: pandas.Series.skew
+
+        Limitations
+        -----------
+        - Parameters ``level`` and ``numeric_only`` are supported only with default value ``None``.
+
+        Examples
+        --------
+        .. literalinclude:: ../../../examples/series/series_skew.py
+           :language: python
+           :lines: 27-
+           :caption: Unbiased rolling skewness.
+           :name: ex_series_skew
+
+        .. command-output:: python ./series/series_skew.py
+           :cwd: ../../../examples
+
+        Intel Scalable Dataframe Compiler Developer Guide
+        *************************************************
+        Pandas Series method :meth:`pandas.Series.skew` implementation.
+
+        .. only:: developer
+            Test: python -m sdc.runtests -k sdc.tests.test_series.TestSeries.test_series_skew*
+        """
+    _func_name = 'Method Series.skew()'
+
+    ty_checker = TypeChecker(_func_name)
+    ty_checker.check(self, SeriesType)
+
+    if not isinstance(axis, (types.Integer, types.NoneType, types.Omitted)) and axis is not None:
+        ty_checker.raise_exc(axis, 'int64', 'axis')
+
+    if not isinstance(skipna, (types.Boolean, types.NoneType, types.Omitted)) and skipna is not None:
+        ty_checker.raise_exc(skipna, 'bool', 'skipna')
+
+    if not isinstance(level, (types.Omitted, types.NoneType)) and level is not None:
+        ty_checker.raise_exc(level, 'None', 'level')
+
+    if not isinstance(numeric_only, (types.Omitted, types.NoneType)) and numeric_only is not None:
+        ty_checker.raise_exc(numeric_only, 'None', 'numeric_only')
+
+    def sdc_pandas_series_skew_impl(self, axis=None, skipna=None, level=None, numeric_only=None):
+        if axis != 0 and axis is not None:
+            raise ValueError('Parameter axis must be only 0 or None.')
+
+        if skipna is None:
+            _skipna = True
+        else:
+            _skipna = skipna
+
+        infinite_mask = numpy.isfinite(self._data)
+        len_val = len(infinite_mask)
+        data = self._data[infinite_mask]
+        nfinite = len(data)
+
+        if not _skipna and nfinite < len_val:
+            return numpy.nan
+        else:
+            _sum = 0
+            square_sum = 0
+            cube_sum = 0
+
+            for i in numba.prange(nfinite):
+                _sum += data[i]
+                square_sum += data[i] ** 2
+                cube_sum += data[i] ** 3
+
+            n = nfinite
+            m2 = (square_sum - _sum * _sum / n) / n
+            m3 = (cube_sum - 3. * _sum * square_sum / n + 2. * _sum * _sum * _sum / n / n) / n
+            res = numpy.nan if m2 == 0 else m3 / m2 ** 1.5
+
+            if (n > 2) & (m2 > 0):
+                res = numpy.sqrt((n - 1.) * n) / (n - 2.) * m3 / m2 ** 1.5
+
+            return res
+
+    return sdc_pandas_series_skew_impl

sdc/tests/test_series.py

@@ -7111,6 +7111,62 @@ def test_impl(A, idx):
                 msg = 'The index of boolean indexer is not comparable to Series index.'
                 self.assertIn(msg, str(raises.exception))
 
+    def test_series_skew(self):
+        def test_impl(series, axis, skipna):
+            return series.skew(axis=axis, skipna=skipna)
+
+        hpat_func = self.jit(test_impl)
+        for data in test_global_input_data_numeric:
+            with self.subTest(data=data):
+                s = pd.Series(data)
+                for axis in [0, None]:
+                    with self.subTest(axis=axis):
+                        for skipna in [None, False, True]:
+                            with self.subTest(skipna=skipna):
+                                res1 = test_impl(s, axis, skipna)
+                                res2 = hpat_func(s, axis, skipna)
+                                np.testing.assert_allclose(res1, res2)
+
+    def test_series_skew_default(self):
+        def test_impl():
+            s = pd.Series([np.nan, -2., 3., 9.1])
+            return s.skew()
+
+        hpat_func = self.jit(test_impl)
+        np.testing.assert_allclose(test_impl(), hpat_func())
+
+    def test_series_skew_not_supported(self):
+        def test_impl(series, axis=None, skipna=None, level=None, numeric_only=None):
+            return series.skew(axis=axis, skipna=skipna, level=level, numeric_only=numeric_only)
+
+        hpat_func = self.jit(test_impl)
+        s = pd.Series([1.1, 0.3, np.nan, 1, np.inf, 0, 1.1, np.nan, 2.2, np.inf, 2, 2])
+
+        with self.assertRaises(TypingError) as raises:
+            hpat_func(s, axis=0.75)
+        msg = 'TypingError: Method Series.skew() The object axis\n given: float64\n expected: int64'
+        self.assertIn(msg, str(raises.exception))
+
+        with self.assertRaises(TypingError) as raises:
+            hpat_func(s, skipna=0)
+        msg = 'TypingError: Method Series.skew() The object skipna\n given: int64\n expected: bool'
+        self.assertIn(msg, str(raises.exception))
+
+        with self.assertRaises(TypingError) as raises:
+            hpat_func(s, level=0)
+        msg = 'TypingError: Method Series.skew() The object level\n given: int64\n expected: None'
+        self.assertIn(msg, str(raises.exception))
+
+        with self.assertRaises(TypingError) as raises:
+            hpat_func(s, numeric_only=0)
+        msg = 'TypingError: Method Series.skew() The object numeric_only\n given: int64\n expected: None'
+        self.assertIn(msg, str(raises.exception))
+
+        with self.assertRaises(ValueError) as raises:
+            hpat_func(s, axis=5)
+        msg = 'Parameter axis must be only 0 or None.'
+        self.assertIn(msg, str(raises.exception))
+
 
 if __name__ == "__main__":
     unittest.main()

sdc/tests/tests_perf/test_perf_series.py

@@ -135,6 +135,7 @@ def _test_case(self, pyfunc, name, total_data_length, input_data=None, data_num=
     TC(name='shape', size=[10 ** 7], call_expr='data.shape', usecase_params='data'),
     TC(name='shift', size=[10 ** 8]),
     TC(name='size', size=[10 ** 7], call_expr='data.size', usecase_params='data'),
+    TC(name='skew', size=[10 ** 8]),
     TC(name='sort_values', size=[10 ** 5]),
     TC(name='std', size=[10 ** 7], check_skipna=True),
     TC(name='sub', size=[10 ** 7], params='other', data_num=2),