BUG: Fix numpy boolean subtraction error in Series.diff

pandas/core/algorithms.py

@@ -1925,6 +1925,7 @@ def diff(arr, n, axis=0):
     dtype = arr.dtype
 
     is_timedelta = False
+    is_bool = False
     if needs_i8_conversion(arr):
         dtype = np.float64
         arr = arr.view("i8")
@@ -1933,6 +1934,7 @@ def diff(arr, n, axis=0):
 
     elif is_bool_dtype(dtype):
         dtype = np.object_
+        is_bool = True
 
     elif is_integer_dtype(dtype):
         dtype = np.float64
@@ -1972,6 +1974,8 @@ def diff(arr, n, axis=0):
             result = res - lag
             result[mask] = na
             out_arr[res_indexer] = result
+        elif is_bool:
+            out_arr[res_indexer] = arr[res_indexer] ^ arr[lag_indexer]
         else:
             out_arr[res_indexer] = arr[res_indexer] - arr[lag_indexer]
 

pandas/tests/series/test_diff.py

@@ -0,0 +1,24 @@
+from pandas import (
+    Series
+)
+
+from numpy import nan
+
+
+def test_diff():
+    data = Series(
+        [0, -1, -2, -3, -4, -3, -2, -1, 0, -1, -1, 0, -1, -2, -3, -2, 0]
+    )
+    filtered = data.between(-2, 0, inclusive=True)
+    diff_boolean = filtered.diff()
+    expected_boolean = Series(
+        [nan, False, False, True, False, False, True, False, False,
+         False, False, False, False, False, True, True, False]
+    )
+    assert diff_boolean.equals(expected_boolean)
+    diff_data = data.diff()
+    expected_data = Series(
+        [nan, -1.0, -1.0, -1.0, -1.0, 1.0, 1.0, 1.0, 1.0,
+         -1.0, 0.0, 1.0, -1.0, -1.0, -1.0, 1.0, 2.0]
+    )
+    assert diff_data.equals(expected_data)