ENH: Series.align method to leverage faster joins, faster binary ops

Author: wesm

bench/zoo_bench.py

@@ -21,3 +21,4 @@ def sample(values, k):
 
 lx = larry(np.random.randn(100000), [list(indices)])
 ly = larry(np.random.randn(subsample_size), [list(y.index)])
+

pandas/core/index.py

@@ -248,21 +248,24 @@ def union(self, other):
         if len(self) == 0:
             return _ensure_index(other)
 
-        indexer = self.get_indexer(other)
-        indexer = (indexer == -1).nonzero()[0]
-
-        if len(indexer) > 0:
-            other_diff = other.values.take(indexer)
-            result = list(self) + list(other_diff)
+        if self.is_monotonic and other.is_monotonic:
+            result = lib.outer_join_indexer_object(self, other)[0]
         else:
-            # contained in
-            result = list(self)
+            indexer = self.get_indexer(other)
+            indexer = (indexer == -1).nonzero()[0]
 
-        # timsort wins
-        try:
-            result.sort()
-        except Exception:
-            pass
+            if len(indexer) > 0:
+                other_diff = other.values.take(indexer)
+                result = list(self) + list(other_diff)
+            else:
+                # contained in
+                result = list(self)
+
+            # timsort wins
+            try:
+                result.sort()
+            except Exception:
+                pass
 
         # for subclasses
         return self._wrap_union_result(other, result)
@@ -292,9 +295,14 @@ def intersection(self, other):
         if self.equals(other):
             return self
 
-        indexer = self.get_indexer(other)
-        indexer = indexer.take((indexer != -1).nonzero()[0])
-        return self.take(indexer)
+        other = _ensure_index(other)
+
+        if self.is_monotonic and other.is_monotonic:
+            return Index(lib.inner_join_indexer_object(self, other)[0])
+        else:
+            indexer = self.get_indexer(other)
+            indexer = indexer.take((indexer != -1).nonzero()[0])
+            return self.take(indexer)
 
     def diff(self, other):
         """

pandas/core/series.py

@@ -41,11 +41,10 @@ def wrapper(self, other):
             if self.index.equals(other.index):
                 return Series(op(self.values, other.values), index=self.index)
 
-            new_index = self.index + other.index
-            this_reindexed = self.reindex(new_index)
-            other_reindexed = other.reindex(new_index)
+            this_reindexed, other_reindexed = self.align(other, join='outer',
+                                                         copy=False)
             arr = op(this_reindexed.values, other_reindexed.values)
-            return Series(arr, index=new_index)
+            return Series(arr, index=this_reindexed.index)
         elif isinstance(other, DataFrame):
             return NotImplemented
         else:
@@ -955,9 +954,8 @@ def _binop(self, other, func, fill_value=None):
         this = self
 
         if not self.index.equals(other.index):
-            new_index = self.index + other.index
-            this = self.reindex(new_index)
-            other = other.reindex(new_index)
+            this, other = self.align(other, join='outer')
+            new_index = this.index
 
         this_vals = this.values
         other_vals = other.values
@@ -1265,6 +1263,43 @@ def apply(self, func):
         except Exception:
             return Series([func(x) for x in self], index=self.index)
 
+    def align(self, other, join='outer', copy=True):
+        """
+        Align two Series object with the specified join method
+
+        Parameters
+        ----------
+        other : Series
+        join : {'outer', 'inner', 'left', 'right'}, default 'outer'
+
+        Returns
+        -------
+        (left, right) : (Series, Series)
+            Aligned Series
+        """
+        join_index, lidx, ridx = self.index.join(other.index, how=join,
+                                                 return_indexers=True)
+
+        if lidx is not None:
+            left = Series(common.take_1d(self.values, lidx), join_index)
+        else:
+            if copy:
+                new_values = self.values.copy()
+            else:
+                new_values = self.values
+            left = Series(new_values, join_index)
+
+        if ridx is not None:
+            right = Series(common.take_1d(other.values, ridx), join_index)
+        else:
+            if copy:
+                new_values = other.values.copy()
+            else:
+                new_values = other.values
+            right = Series(new_values, join_index)
+
+        return left, right
+
     def reindex(self, index=None, method=None, copy=True):
         """Conform Series to new index with optional filling logic, placing
         NA/NaN in locations having no value in the previous index. A new object