Merge pull request #3 from CSCD01-team14/bugs/pandas-dev#32572

Bug pandas-dev#32572 fix

Author: letapxad

pandas/_libs/algos.pyx

@@ -1050,15 +1050,15 @@ def rank_2d(rank_t[:, :] in_arr, axis=0, ties_method='average',
             if rank_t is object:
                 nan_value = Infinity()
             elif rank_t is float64_t:
-                nan_value = np.inf
+                nan_value = NaN
             elif rank_t is int64_t:
                 nan_value = np.iinfo(np.int64).max
 
         else:
             if rank_t is object:
                 nan_value = NegInfinity()
             elif rank_t is float64_t:
-                nan_value = -np.inf
+                nan_value = NaN
             elif rank_t is int64_t:
                 nan_value = NPY_NAT
 
@@ -1120,7 +1120,7 @@ def rank_2d(rank_t[:, :] in_arr, axis=0, ties_method='average',
                 if rank_t is object:
                     skip_condition = (val is nan_value) and keep_na
                 else:
-                    skip_condition = (val == nan_value) and keep_na
+                    skip_condition = (val == nan_value or (np.isnan(val) and np.isnan(nan_value))) and keep_na
                 if skip_condition:
                     ranks[i, argsorted[i, j]] = NaN
 

pandas/tests/frame/methods/test_rank.py

@@ -329,3 +329,27 @@ def test_pct_max_many_rows(self):
         )
         result = df.rank(pct=True).max()
         assert (result == 1).all()
+
+    def test_rank_minus_inf_keep_nan(self):
+        # GH 32593
+        expected_df = DataFrame({'col': np.array([2.0,  4.0, np.nan, 3.0, 1.0])})
+        result_df = DataFrame({'col': np.array([1,  np.inf, np.nan, 10, -np.inf])}).rank(na_option='keep')
+        tm.assert_frame_equal(expected_df, result_df)
+
+    def test_rank_inf_keep_nan(self):
+        # GH 32593
+        expected_df = DataFrame({'col': np.array([1.0, 2.0, 3.0])})
+        result_df = DataFrame({'col': np.array([-np.inf, 0, np.inf])}).rank(na_option='keep')
+        tm.assert_frame_equal(expected_df, result_df)
+
+    def test_rank_inf_bottom_nan(self):
+        # GH 32593
+        expected_df = DataFrame({'col': np.array([1.0, 2.0, 4.0, 3.0])})
+        result_df = DataFrame({'col': np.array([-np.inf, 0, np.nan, np.inf])}).rank(na_option='bottom')
+        tm.assert_frame_equal(expected_df, result_df)
+
+    def test_rank_inf_decimal_nan(self):
+        # GH 32593
+        expected_df = DataFrame({'col': np.array([2.0,3.0,4.5,6.0,1.0,4.5])})
+        result_df = DataFrame({'col': np.array([5.5, 6.99, np.inf, np.nan, 0.7, np.inf])}).rank(na_option='bottom')
+        tm.assert_frame_equal(expected_df, result_df)