Merge pull request #4316 from khgitting/kde_kwargs

Added ind and bw_method kwargs to KdePlot

Author: jreback

doc/source/release.rst

@@ -30,6 +30,10 @@ pandas 0.13
 **Release date:** not-yet-released
 
 **New features**
+  - ``plot(kind='kde')`` now accepts the optional parameters ``bw_method`` and
+    ``ind``, passed to scipy.stats.gaussian_kde() (for scipy >= 0.11.0) to set
+    the bandwidth, and to gkde.evaluate() to specify the indicies at which it
+    is evaluated, respecttively. See scipy docs.
 
   - Added ``isin`` method to DataFrame (:issue:`4211`)
   - Clipboard functionality now works with PySide (:issue:`4282`)

doc/source/v0.13.0.txt

@@ -163,6 +163,11 @@ Enhancements
          from pandas import offsets
          td + offsets.Minute(5) + offsets.Milli(5)
 
+    - ``plot(kind='kde')`` now accepts the optional parameters ``bw_method`` and
+      ``ind``, passed to scipy.stats.gaussian_kde() (for scipy >= 0.11.0) to set
+      the bandwidth, and to gkde.evaluate() to specify the indicies at which it
+      is evaluated, respecttively. See scipy docs.
+
 .. _whatsnew_0130.refactoring:
 
 Internal Refactoring

pandas/tests/test_graphics.py

@@ -257,8 +257,18 @@ def test_kde(self):
         _check_plot_works(self.ts.plot, kind='kde')
         _check_plot_works(self.ts.plot, kind='density')
         ax = self.ts.plot(kind='kde', logy=True)
-        self.assert_(ax.get_yscale() == 'log')
+        self.assertEqual(ax.get_yscale(), 'log')
 
+    @slow
+    def test_kde_kwargs(self):
+        _skip_if_no_scipy()
+        from numpy import linspace
+        _check_plot_works(self.ts.plot, kind='kde', bw_method=.5, ind=linspace(-100,100,20))
+        _check_plot_works(self.ts.plot, kind='density', bw_method=.5, ind=linspace(-100,100,20))
+        ax = self.ts.plot(kind='kde', logy=True, bw_method=.5, ind=linspace(-100,100,20))
+        self.assertEqual(ax.get_yscale(), 'log')
+
+    @slow
     def test_kde_color(self):
         _skip_if_no_scipy()
         ax = self.ts.plot(kind='kde', logy=True, color='r')

pandas/tools/plotting.py

@@ -1136,11 +1136,15 @@ def _get_marked_label(self, label, col_num):
 
 
 class KdePlot(MPLPlot):
-    def __init__(self, data, **kwargs):
+    def __init__(self, data, bw_method=None, ind=None, **kwargs):
         MPLPlot.__init__(self, data, **kwargs)
+        self.bw_method=bw_method
+        self.ind=ind
 
     def _make_plot(self):
         from scipy.stats import gaussian_kde
+        from scipy import __version__ as spv
+        from distutils.version import LooseVersion
         plotf = self._get_plot_function()
         colors = self._get_colors()
         for i, (label, y) in enumerate(self._iter_data()):
@@ -1149,10 +1153,23 @@ def _make_plot(self):
 
             label = com.pprint_thing(label)
 
-            gkde = gaussian_kde(y)
+            if LooseVersion(spv) >= '0.11.0':
+                gkde = gaussian_kde(y, bw_method=self.bw_method)
+            else:
+                gkde = gaussian_kde(y)
+                if self.bw_method is not None:
+                    msg = ('bw_method was added in Scipy 0.11.0.' +
+                           ' Scipy version in use is %s.' % spv)
+                    warnings.warn(msg)
+
             sample_range = max(y) - min(y)
-            ind = np.linspace(min(y) - 0.5 * sample_range,
-                              max(y) + 0.5 * sample_range, 1000)
+
+            if self.ind is None:
+                ind = np.linspace(min(y) - 0.5 * sample_range,
+                                  max(y) + 0.5 * sample_range, 1000)
+            else:
+                ind = self.ind
+
             ax.set_ylabel("Density")
 
             y = gkde.evaluate(ind)