Add check_skipna for perf tests (#621)

* Add check_skipna

* Add check_skipna for DataFrame perf tests

Author: PokhodenkoSA

sdc/tests/tests_perf/generator.py

@@ -26,6 +26,7 @@ class TestCase(NamedTuple):
     data_num: total number of generated data, e.g. 2 (data, other)
     input_data: input data for generating test data
     skip: flag for skipping a test
+    check_skipna: flag for checking a function with both parameters skipna=True and skipna=False
     """
     name: str
     size: list
@@ -35,15 +36,32 @@ class TestCase(NamedTuple):
     data_num: int = 1
     input_data: list = None
     skip: bool = False
+    check_skipna: bool = False
 
 
 def to_varname_without_excess_underscores(string):
     """Removing excess underscores from the string."""
     return '_'.join(i for i in to_varname(string).split('_') if i)
 
 
+def skipna_cases(cases):
+    """Generator. Replaces a test case containing check_skipna=True
+    with two cases containing parameters skipna=True and skipna=False
+    """
+    for case in cases:
+        if case.check_skipna:
+            for skipna in [True, False]:
+                params = case.params
+                if params:
+                    params += ', '
+                params += f'skipna={skipna}'
+                yield case._replace(params=params)
+        else:
+            yield case
+
+
 def generate_test_cases(cases, class_add, typ, prefix=''):
-    for test_case in cases:
+    for test_case in skipna_cases(cases):
         test_name_parts = ['test', typ, prefix, test_case.name, gen_params_wo_data(test_case)]
         test_name = to_varname_without_excess_underscores('_'.join(test_name_parts))
 

sdc/tests/tests_perf/test_perf_df.py

@@ -64,15 +64,15 @@ def _test_case(self, pyfunc, name, total_data_length, input_data, data_num=1):
     TC(name='append', size=[10 ** 7], params='other', data_num=2),
     TC(name='count', size=[10 ** 7]),
     TC(name='drop', size=[10 ** 8], params='columns="f0"'),
-    TC(name='max', size=[10 ** 7]),
-    TC(name='mean', size=[10 ** 7]),
-    TC(name='median', size=[10 ** 7]),
-    TC(name='min', size=[10 ** 7]),
+    TC(name='max', size=[10 ** 7], check_skipna=True),
+    TC(name='mean', size=[10 ** 7], check_skipna=True),
+    TC(name='median', size=[10 ** 7], check_skipna=True),
+    TC(name='min', size=[10 ** 7], check_skipna=True),
     TC(name='pct_change', size=[10 ** 7]),
-    TC(name='prod', size=[10 ** 7]),
-    TC(name='std', size=[10 ** 7]),
-    TC(name='sum', size=[10 ** 7]),
-    TC(name='var', size=[10 ** 7]),
+    TC(name='prod', size=[10 ** 7], check_skipna=True),
+    TC(name='std', size=[10 ** 7], check_skipna=True),
+    TC(name='sum', size=[10 ** 7], check_skipna=True),
+    TC(name='var', size=[10 ** 7], check_skipna=True),
 ]
 
 generate_test_cases(cases, TestDataFrameMethods, 'df')

sdc/tests/tests_perf/test_perf_series.py

@@ -77,7 +77,7 @@ def _test_case(self, pyfunc, name, total_data_length, data_num=1, input_data=tes
     TC(name='corr',  size=[10 ** 7],params='other', data_num=2),
     TC(name='count', size=[10 ** 8]),
     TC(name='cov', size=[10 ** 8], params='other', data_num=2),
-    TC(name='cumsum', size=[10 ** 8]),
+    TC(name='cumsum', size=[10 ** 8], check_skipna=True),
     TC(name='describe', size=[10 ** 7]),
     TC(name='div', size=[10 ** 7], params='other', data_num=2),
     TC(name='dropna', size=[10 ** 7]),
@@ -89,8 +89,8 @@ def _test_case(self, pyfunc, name, total_data_length, data_num=1, input_data=tes
     TC(name='gt',  size=[10 ** 7],params='other', data_num=2),
     TC(name='head', size=[10 ** 8]),
     TC(name='iat', size=[10 ** 7], call_expr='data.iat[100000]', usecase_params='data'),
-    TC(name='idxmax', size=[10 ** 8]),
-    TC(name='idxmin', size=[10 ** 8]),
+    TC(name='idxmax', size=[10 ** 8], check_skipna=True),
+    TC(name='idxmin', size=[10 ** 8], check_skipna=True),
     TC(name='iloc', size=[10 ** 7], call_expr='data.iloc[100000]', usecase_params='data'),
     TC(name='index', size=[10 ** 7], call_expr='data.index', usecase_params='data'),
     TC(name='isin', size=[10 ** 7], call_expr='data.isin([0])', usecase_params='data'),
@@ -101,13 +101,10 @@ def _test_case(self, pyfunc, name, total_data_length, data_num=1, input_data=tes
     TC(name='lt', size=[10 ** 7], params='other', data_num=2),
     TC(name='map', size=[10 ** 7], params='lambda x: x * 2'),
     TC(name='map', size=[10 ** 7], params='{2.: 42., 4.: 3.14}'),
-    TC(name='max', size=[10 ** 8], params='skipna=True'),
-    TC(name='max', size=[10 ** 8], params='skipna=False'),
-    TC(name='mean', size=[10 ** 8], params='skipna=True'),
-    TC(name='mean', size=[10 ** 8], params='skipna=False'),
-    TC(name='median', size=[10 ** 8]),
-    TC(name='min', size=[10 ** 8], params='skipna=True'),
-    TC(name='min', size=[10 ** 8], params='skipna=False'),
+    TC(name='max', size=[10 ** 8], check_skipna=True),
+    TC(name='mean', size=[10 ** 8], check_skipna=True),
+    TC(name='median', size=[10 ** 8], check_skipna=True),
+    TC(name='min', size=[10 ** 8], check_skipna=True),
     TC(name='mod', size=[10 ** 7], params='other', data_num=2),
     TC(name='mul', size=[10 ** 7], params='other', data_num=2),
     TC(name='ndim', size=[10 ** 7], call_expr='data.ndim', usecase_params='data'),
@@ -116,8 +113,7 @@ def _test_case(self, pyfunc, name, total_data_length, data_num=1, input_data=tes
     TC(name='notna', size=[10 ** 7]),
     TC(name='nsmallest', size=[10 ** 6]),
     TC(name='nunique', size=[10 ** 7]),
-    TC(name='prod', size=[10 ** 8], params='skipna=True'),
-    TC(name='prod', size=[10 ** 8], params='skipna=False'),
+    TC(name='prod', size=[10 ** 8], check_skipna=True),
     TC(name='pct_change', size=[10 ** 7], params='periods=1, limit=None, freq=None'),
     TC(name='pow', size=[10 ** 7], params='other', data_num=2),
     TC(name='quantile', size=[10 ** 8]),
@@ -127,16 +123,14 @@ def _test_case(self, pyfunc, name, total_data_length, data_num=1, input_data=tes
     TC(name='shift', size=[10 ** 8]),
     TC(name='size', size=[10 ** 7], call_expr='data.size', usecase_params='data'),
     TC(name='sort_values', size=[10 ** 5]),
-    TC(name='std', size=[10 ** 7], params='skipna=True'),
-    TC(name='std', size=[10 ** 7], params='skipna=False'),
+    TC(name='std', size=[10 ** 7], check_skipna=True),
     TC(name='sub', size=[10 ** 7], params='other', data_num=2),
-    TC(name='sum', size=[10 ** 8]),
+    TC(name='sum', size=[10 ** 8], check_skipna=True),
     TC(name='take', size=[10 ** 7], call_expr='data.take([0])', usecase_params='data'),
     TC(name='truediv', size=[10 ** 7], params='other', data_num=2),
     TC(name='values', size=[10 ** 7], call_expr='data.values', usecase_params='data'),
     TC(name='value_counts', size=[10 ** 6]),
-    TC(name='var', size=[10 ** 8], params='skipna=True'),
-    TC(name='var', size=[10 ** 8], params='skipna=False'),
+    TC(name='var', size=[10 ** 8], check_skipna=True),
     TC(name='unique', size=[10 ** 5]),
 ]