REGR: Fixes first_valid_index when DataFrame or Series has duplicate row index (GH21441)

doc/source/whatsnew/v0.23.2.txt

@@ -16,8 +16,12 @@ and bug fixes. We recommend that all users upgrade to this version.
 Fixed Regressions
 ~~~~~~~~~~~~~~~~~
 
+- 
 -
--
+
+**Other Fixes**
+
+- Bug in :meth:`first_valid_index` that raised for row index with duplicate values (:issue:`21441`)
 
 .. _whatsnew_0232.performance:
 

pandas/core/generic.py

@@ -8970,17 +8970,17 @@ def _find_valid_index(self, how):
 
         if how == 'first':
             # First valid value case
-            i = is_valid.idxmax()
-            if not is_valid[i]:
-                return None
-            return i
+            i = is_valid.values[::].argmin()
+            idxpos = i
 
         elif how == 'last':
             # Last valid value case
             i = is_valid.values[::-1].argmax()
-            if not is_valid.iat[len(self) - i - 1]:
-                return None
-            return self.index[len(self) - i - 1]
+            idxpos = len(self) - i - 1
+
+        if not is_valid.iat[idxpos]:
+            return None
+        return self.index[idxpos]
 
     @Appender(_shared_docs['valid_index'] % {'position': 'first',
                                              'klass': 'NDFrame'})

pandas/tests/generic/test_generic.py

@@ -612,6 +612,16 @@ def test_pct_change(self, periods, fill_method, limit, exp):
         else:
             tm.assert_series_equal(res, Series(exp))
 
+    @pytest.mark.parametrize("DF,idx,first_idx,last_idx", [
+        ({'A': [1, 2, 3]}, [1, 1, 2], 1, 2),
+        ({'A': [1, 2, 3]}, [1, 2, 2], 1, 2),
+        ({'A': [1, 2, 3, 4]}, ['d', 'd', 'd', 'd'], 'd', 'd')])
+    def test_valid_index(self, DF, idx, first_idx, last_idx):
+        # GH 21441
+        df1 = pd.DataFrame(DF, index=idx)
+        assert first_idx == df1.first_valid_index()
+        assert last_idx == df1.last_valid_index()
+
 
 class TestNDFrame(object):
     # tests that don't fit elsewhere