Merge pull request #9727 from evanpw/diff_axis

ENH: Add an axis parameter to DataFrame.diff

Author: jreback

doc/source/whatsnew/v0.16.1.txt

@@ -23,6 +23,7 @@ Enhancements
 ~~~~~~~~~~~~
 
 - Added ``StringMethods.capitalize()`` and ``swapcase`` which behave as the same as standard ``str`` (:issue:`9766`)
+- ``DataFrame.diff`` now takes an ``axis`` parameter that determines the direction of differencing (:issue:`9727`)
 - Added ``StringMethods`` (.str accessor) to ``Index`` (:issue:`9068`)
 - Allow clip, clip_lower, and clip_upper to accept array-like arguments as thresholds (:issue:`6966`). These methods now have an ``axis`` parameter which determines how the Series or DataFrame will be aligned with the threshold(s).
 

pandas/core/frame.py

@@ -3586,20 +3586,22 @@ def unstack(self, level=-1):
     #----------------------------------------------------------------------
     # Time series-related
 
-    def diff(self, periods=1):
+    def diff(self, periods=1, axis=0):
         """
         1st discrete difference of object
 
         Parameters
         ----------
         periods : int, default 1
             Periods to shift for forming difference
+        axis : {0 or 'index', 1 or 'columns'}, default 0
 
         Returns
         -------
         diffed : DataFrame
         """
-        new_data = self._data.diff(n=periods)
+        bm_axis = self._get_block_manager_axis(axis)
+        new_data = self._data.diff(n=periods, axis=bm_axis)
         return self._constructor(new_data)
 
     #----------------------------------------------------------------------

pandas/core/internals.py

@@ -874,9 +874,9 @@ def take_nd(self, indexer, axis, new_mgr_locs=None, fill_tuple=None):
     def get_values(self, dtype=None):
         return self.values
 
-    def diff(self, n):
+    def diff(self, n, axis=1):
         """ return block for the diff of the values """
-        new_values = com.diff(self.values, n, axis=1)
+        new_values = com.diff(self.values, n, axis=axis)
         return [make_block(values=new_values,
                            ndim=self.ndim, fastpath=True,
                            placement=self.mgr_locs)]

pandas/tests/test_frame.py

@@ -10100,6 +10100,12 @@ def test_diff_float_n(self):
         xp = self.tsframe.diff(1)
         assert_frame_equal(rs, xp)
 
+    def test_diff_axis(self):
+        # GH 9727
+        df = DataFrame([[1., 2.], [3., 4.]])
+        assert_frame_equal(df.diff(axis=1), DataFrame([[np.nan, 1.], [np.nan, 1.]]))
+        assert_frame_equal(df.diff(axis=0), DataFrame([[np.nan, np.nan], [2., 2.]]))
+
     def test_pct_change(self):
         rs = self.tsframe.pct_change(fill_method=None)
         assert_frame_equal(rs, self.tsframe / self.tsframe.shift(1) - 1)