BUG: boolean groupby performance issue. close #2692

Author: wesm

RELEASE.rst

@@ -78,6 +78,7 @@ pandas 0.10.1
   - Exclude non-numeric data from DataFrame.quantile by default (GH2625_)
   - Fix a Cython C int64 boxing issue causing read_csv to return incorrect
     results (GH2599_)
+  - Fix groupby summing performance issue on boolean data (GH2692_)
 
 **API Changes**
 
@@ -98,6 +99,7 @@ pandas 0.10.1
 .. _GH2625: https://github.com/pydata/pandas/issues/2625
 .. _GH2643: https://github.com/pydata/pandas/issues/2643
 .. _GH2637: https://github.com/pydata/pandas/issues/2637
+.. _GH2692: https://github.com/pydata/pandas/issues/2692
 
 pandas 0.10.0
 =============

pandas/core/common.py

@@ -708,7 +708,7 @@ def _default_index(n):
 
 
 def ensure_float(arr):
-    if issubclass(arr.dtype.type, np.integer):
+    if issubclass(arr.dtype.type, (np.integer, np.bool_)):
         arr = arr.astype(float)
 
     return arr

vb_suite/groupby.py

@@ -254,3 +254,13 @@ def f(g):
 
 groupbym_frame_apply = Benchmark("df.groupby(['key', 'key2']).apply(f)", setup,
                                  start_date=datetime(2011, 10, 1))
+
+#----------------------------------------------------------------------
+# Sum booleans #2692
+
+setup = common_setup + """
+N = 500
+df = DataFrame({'ii':range(N),'bb':[True for x in range(N)]})
+"""
+
+groupby_sum_booleans = Benchmark("df.groupby('ii').sum()", setup)