Lag plot

doc/source/visualization.rst

@@ -283,3 +283,22 @@ of the same class will usually be closer together and form larger structures.
 
    @savefig andrews_curves.png width=6in
    andrews_curves(data, 'Name')
+
+Lag Plot
+~~~~~~~~
+
+Lag plots are used to check if a data set or time series is random. Random
+data should not exhibit any structure in the lag plot. Non-random structure
+implies that the underlying data are not random.
+
+.. ipython:: python
+
+   from pandas.tools.plotting import lag_plot
+
+   plt.figure()
+
+   data = Series(0.1 * np.random.random(1000) + 
+      0.9 * np.sin(np.linspace(-99 * np.pi, 99 * np.pi, num=1000)))
+
+   @savefig lag_plot.png width=6in
+   lag_plot(data)

pandas/tests/test_graphics.py

@@ -63,6 +63,11 @@ def test_kde(self):
         ax = self.ts.plot(kind='kde', logy=True)
         self.assert_(ax.get_yscale() == 'log')
 
+    @slow
+    def test_lag_plot(self):
+        from pandas.tools.plotting import lag_plot
+        _check_plot_works(lag_plot, self.ts)
+
 class TestDataFramePlots(unittest.TestCase):
 
     @classmethod

pandas/tools/plotting.py

@@ -177,6 +177,30 @@ def random_color(column):
     ax.grid()
     return ax
 
+def lag_plot(series, ax=None, **kwds):
+    """Lag plot for time series.
+
+    Parameters:
+    -----------
+    series: Time series
+    ax: Matplotlib axis object, optional
+    kwds: Matplotlib scatter method keyword arguments, optional
+
+    Returns:
+    --------
+    ax: Matplotlib axis object
+    """
+    import matplotlib.pyplot as plt
+    data = series.values
+    y1 = data[:-1]
+    y2 = data[1:]
+    if ax == None:
+        ax = plt.gca()
+    ax.set_xlabel("y(t)")
+    ax.set_ylabel("y(t + 1)")
+    ax.scatter(y1, y2, **kwds)
+    return ax
+
 def grouped_hist(data, column=None, by=None, ax=None, bins=50, log=False,
                  figsize=None, layout=None, sharex=False, sharey=False,
                  rot=90):