From de81cc1c124e268c73c082fb3cd8892dbb4d83be Mon Sep 17 00:00:00 2001
From: Wes McKinney <wesmckinn@gmail.com>
Date: Sat, 3 Mar 2012 18:37:18 -0500
Subject: [PATCH] BUG: fix int32 overflow when computing group_index, GH #850

---
 pandas/core/groupby.py       |  2 +-
 pandas/tests/test_groupby.py |  9 +++++++++
 pandas/tools/plotting.py     | 25 ++++++++++++++++---------
 setup.py                     |  4 ++--
 4 files changed, 28 insertions(+), 12 deletions(-)

diff --git a/pandas/core/groupby.py b/pandas/core/groupby.py
index 8cd0c38eec246..9fe0810aeb324 100644
--- a/pandas/core/groupby.py
+++ b/pandas/core/groupby.py
@@ -1472,7 +1472,7 @@ def get_group_index(label_list, shape):
     mask = np.zeros(n, dtype=bool)
     for i in xrange(len(shape)):
         stride = np.prod([x for x in shape[i+1:]], dtype=int)
-        group_index += label_list[i] * stride
+        group_index += com._ensure_int64(label_list[i]) * stride
         mask |= label_list[i] < 0
 
     np.putmask(group_index, mask, -1)
diff --git a/pandas/tests/test_groupby.py b/pandas/tests/test_groupby.py
index 355a8cf94aef2..153b8c5f7bb85 100644
--- a/pandas/tests/test_groupby.py
+++ b/pandas/tests/test_groupby.py
@@ -1457,6 +1457,15 @@ def test_groupby_2d_malformed(self):
         self.assert_(np.array_equal(tmp.columns, ['zeros', 'ones']))
         self.assert_(np.array_equal(tmp.values, res_values))
 
+    def test_int32_overflow(self):
+        B = np.concatenate((np.arange(100000), np.arange(100000),
+                            np.arange(50000)))
+        A = np.arange(250000)
+        df = DataFrame({'A' : A, 'B' : B, 'C' : np.random.randn(250000)})
+
+        left = df.groupby(['A', 'B']).sum()
+        right = df.groupby(['B', 'A']).sum()
+        self.assert_(len(left) == len(right))
 
 def test_decons():
     from pandas.core.groupby import decons_group_index, get_group_index
diff --git a/pandas/tools/plotting.py b/pandas/tools/plotting.py
index f6bc33ee15624..3e81643ee3490 100644
--- a/pandas/tools/plotting.py
+++ b/pandas/tools/plotting.py
@@ -22,21 +22,24 @@ def hist(data, column, by=None, ax=None, fontsize=None):
     return ax
 
 def grouped_hist(data, column=None, by=None, ax=None, bins=50, log=False,
-                 figsize=None, layout=None):
+                 figsize=None, layout=None, sharex=False, sharey=False,
+                 rot=90):
     """
 
     Returns
     -------
     fig : matplotlib.Figure
     """
-    if isinstance(data, DataFrame):
-        data = data[column]
+    # if isinstance(data, DataFrame):
+    #     data = data[column]
 
     def plot_group(group, ax):
         ax.hist(group.dropna(), bins=bins)
-    fig, axes = _grouped_plot(plot_group, data, by=by, sharex=False,
-                              sharey=False, figsize=figsize,
-                              layout=layout)
+
+    fig, axes = _grouped_plot(plot_group, data, column=column,
+                              by=by, sharex=sharex, sharey=sharey,
+                              figsize=figsize, layout=layout,
+                              rot=rot)
     fig.subplots_adjust(bottom=0.15, top=0.9, left=0.1, right=0.9,
                         hspace=0.3, wspace=0.2)
     return fig
@@ -45,7 +48,7 @@ def plot_group(group, ax):
 def boxplot(data, column=None, by=None, ax=None, fontsize=None,
             rot=0, grid=True, figsize=None):
     """
-    Make a box plot from DataFrame column optionally grouped by some columns or
+    Make a box plot from DataFrame column optionally grouped b ysome columns or
     other inputs
 
     Parameters
@@ -142,8 +145,9 @@ def plot_group(group, ax):
 
     return fig
 
-def _grouped_plot(plotf, data, by=None, numeric_only=True, figsize=None,
-                  sharex=True, sharey=True, layout=None):
+def _grouped_plot(plotf, data, column=None, by=None, numeric_only=True,
+                  figsize=None, sharex=True, sharey=True, layout=None,
+                  rot=0):
     import matplotlib.pyplot as plt
 
     # allow to specify mpl default with 'default'
@@ -151,6 +155,9 @@ def _grouped_plot(plotf, data, by=None, numeric_only=True, figsize=None,
         figsize = (10, 5)               # our default
 
     grouped = data.groupby(by)
+    if column is not None:
+        grouped = grouped[column]
+
     ngroups = len(grouped)
 
     nrows, ncols = layout or _get_layout(ngroups)
diff --git a/setup.py b/setup.py
index d27f264160e71..2868b765d49dc 100755
--- a/setup.py
+++ b/setup.py
@@ -164,8 +164,8 @@
 
 MAJOR = 0
 MINOR = 7
-MICRO = 1
-ISRELEASED = True
+MICRO = 2
+ISRELEASED = False
 VERSION = '%d.%d.%d' % (MAJOR, MINOR, MICRO)
 QUALIFIER = ''