Fixed bug in outer_indexer where the special case of an empty right array resulted in bogus return data.

Author: agijsberts

doc/source/whatsnew/v0.17.0.txt

@@ -389,6 +389,7 @@ Bug Fixes
 - Bug in ``MultiIndex.get_level_values`` including ``Categorical`` raises ``AttributeError`` (:issue:`10460`)
 - Bug in ``pd.get_dummies`` with `sparse=True` not returning ``SparseDataFrame`` (:issue:`10531`)
 - Bug in ``Index`` subtypes (such as ``PeriodIndex``) not returning their own type for ``.drop`` and ``.insert`` methods (:issue:`10620`)
+- Bug in ``algos.outer_join_indexer`` when ``right`` array is empty (:issue:`10618`)
 
 
 

pandas/src/generate_code.py

@@ -2131,7 +2131,7 @@ def outer_join_indexer_%(name)s(ndarray[%(c_type)s] left,
             rindexer[j] = j
             result[j] = right[j]
     elif nright == 0:
-        for i in range(nright):
+        for i in range(nleft):
             lindexer[i] = i
             rindexer[i] = -1
             result[i] = left[i]

pandas/src/generated.pyx

@@ -10345,7 +10345,7 @@ def outer_join_indexer_float64(ndarray[float64_t] left,
             rindexer[j] = j
             result[j] = right[j]
     elif nright == 0:
-        for i in range(nright):
+        for i in range(nleft):
             lindexer[i] = i
             rindexer[i] = -1
             result[i] = left[i]
@@ -10474,7 +10474,7 @@ def outer_join_indexer_float32(ndarray[float32_t] left,
             rindexer[j] = j
             result[j] = right[j]
     elif nright == 0:
-        for i in range(nright):
+        for i in range(nleft):
             lindexer[i] = i
             rindexer[i] = -1
             result[i] = left[i]
@@ -10603,7 +10603,7 @@ def outer_join_indexer_object(ndarray[object] left,
             rindexer[j] = j
             result[j] = right[j]
     elif nright == 0:
-        for i in range(nright):
+        for i in range(nleft):
             lindexer[i] = i
             rindexer[i] = -1
             result[i] = left[i]
@@ -10732,7 +10732,7 @@ def outer_join_indexer_int32(ndarray[int32_t] left,
             rindexer[j] = j
             result[j] = right[j]
     elif nright == 0:
-        for i in range(nright):
+        for i in range(nleft):
             lindexer[i] = i
             rindexer[i] = -1
             result[i] = left[i]
@@ -10861,7 +10861,7 @@ def outer_join_indexer_int64(ndarray[int64_t] left,
             rindexer[j] = j
             result[j] = right[j]
     elif nright == 0:
-        for i in range(nright):
+        for i in range(nleft):
             lindexer[i] = i
             rindexer[i] = -1
             result[i] = left[i]

pandas/tests/test_algos.py

@@ -168,6 +168,38 @@ def _test_vector_resize(htable, uniques, dtype, nvals):
             _test_vector_resize(tbl(), vect(), dtype, 0)
             _test_vector_resize(tbl(), vect(), dtype, 10)
 
+class TestIndexer(tm.TestCase):
+    _multiprocess_can_split_ = True
+
+    def test_outer_join_indexer(self):
+        typemap = [('int32', algos.algos.outer_join_indexer_int32),
+                   ('int64', algos.algos.outer_join_indexer_int64),
+                   ('float32', algos.algos.outer_join_indexer_float32),
+                   ('float64', algos.algos.outer_join_indexer_float64),
+                   ('object', algos.algos.outer_join_indexer_object)]
+
+        for dtype, indexer in typemap:
+            left = np.arange(3, dtype = dtype)
+            right = np.arange(2,5, dtype = dtype)
+            empty = np.array([], dtype = dtype)
+
+            result, lindexer, rindexer = indexer(left, right)
+            tm.assertIsInstance(result, np.ndarray)
+            tm.assertIsInstance(lindexer, np.ndarray)
+            tm.assertIsInstance(rindexer, np.ndarray)
+            tm.assert_numpy_array_equal(result, np.arange(5, dtype = dtype))
+            tm.assert_numpy_array_equal(lindexer, np.array([0, 1, 2, -1, -1]))
+            tm.assert_numpy_array_equal(rindexer, np.array([-1, -1, 0, 1, 2]))
+
+            result, lindexer, rindexer = indexer(empty, right)
+            tm.assert_numpy_array_equal(result, right)
+            tm.assert_numpy_array_equal(lindexer, np.array([-1, -1, -1]))
+            tm.assert_numpy_array_equal(rindexer, np.array([0, 1, 2]))
+
+            result, lindexer, rindexer = indexer(left, empty)
+            tm.assert_numpy_array_equal(result, left)
+            tm.assert_numpy_array_equal(lindexer, np.array([0, 1, 2]))
+            tm.assert_numpy_array_equal(rindexer, np.array([-1, -1, -1]))
 
 class TestUnique(tm.TestCase):
     _multiprocess_can_split_ = True