sort grouper if series/dataframe, sort df in key/name

mrocklin · Feb 1, 2018 · a3e5b4d · a3e5b4d
1 parent 5a90cca
commit a3e5b4d
Show file tree

Hide file tree

Showing 2 changed files with 8 additions and 1 deletion.
diff --git a/dask/dataframe/groupby.py b/dask/dataframe/groupby.py
@@ -141,7 +141,13 @@ def _groupby_raise_unaligned(df, **kwargs):
 def _groupby_slice_apply(df, grouper, key, func):
     # No need to use raise if unaligned here - this is only called after
     # shuffling, which makes everything aligned already
-    df = df.sort_values(grouper)
+    if isinstance(grouper, (pd.DataFrame, pd.Series, pd.Index)):
+        grouper = grouper.sort_values()
+    else:
+        try:
+            df = df.sort_values(grouper)
+        except KeyError:  # this fails when the grouper includes the index
+            pass
     g = df.groupby(grouper)
     if key:
         g = g[key]

diff --git a/docs/source/changelog.rst b/docs/source/changelog.rst
@@ -18,6 +18,7 @@ DataFrame
 +++++++++
 
 -  Support month timedeltas in repartition(freq=...) (:pr:`3110`) `Matthew Rocklin`_
+-  Sort grouper values prior to groupby-apply (:pr:`3118`) `Matthew Rocklin`_
 
 Bag
 +++