Add DataFrame.plot.kde() (an alias of 'density')

databricks/koalas/frame.py

@@ -544,6 +544,11 @@ def hist(self, bins=10, **kwds):
 
     hist.__doc__ = KoalasFramePlotMethods.hist.__doc__
 
+    def kde(self, bw_method=None, ind=None, **kwds):
+        return self.plot.kde(bw_method, ind, **kwds)
+
+    kde.__doc__ = KoalasFramePlotMethods.kde.__doc__
+
     add.__doc__ = _flex_doc_FRAME.format(
         desc='Addition',
         op_name='+',

databricks/koalas/plot.py

@@ -631,7 +631,6 @@ def _plot(
             cls, ax, y, style=None, bw_method=None, ind=None,
             column_num=None, stacking_id=None, **kwds):
         # 'y' is a Spark DataFrame that selects one column.
-
         # Using RDD is slow so we might have to change it to Dataset based implementation
         # once Spark has that implementation.
         sample = y.rdd.map(lambda x: float(x[0]))
@@ -919,7 +918,6 @@ def _plot(data, x=None, y=None, subplots=False,
     if kind in ('scatter', 'hexbin'):
         plot_obj = klass(data, x, y, subplots=subplots, ax=ax, kind=kind, **kwds)
     else:
-
         # check data type and do preprocess before applying plot
         if isinstance(data, DataFrame):
             if x is not None:
@@ -1246,8 +1244,30 @@ def line(self, x=None, y=None, **kwargs):
         """
         return self(kind='line', x=x, y=y, **kwargs)
 
-    def kde(self, bw_method=None, ind=None, **kwds):
-        return _unsupported_function(class_name='pd.DataFrame', method_name='kde')()
+    def kde(self, bw_method=None, ind=None, **kwargs):
+        """
+        Generate Kernel Density Estimate plot using Gaussian kernels.
+
+        Parameters
+        ----------
+        bw_method : scalar
+            The method used to calculate the estimator bandwidth.
+            See KernelDensity in PySpark for more information.
+        ind : NumPy array or integer, optional
+            Evaluation points for the estimated PDF. If None (default),
+            1000 equally spaced points are used. If `ind` is a NumPy array, the
+            KDE is evaluated at the points passed. If `ind` is an integer,
+            `ind` number of equally spaced points are used.
+        **kwargs : optional
+            Keyword arguments to pass on to :meth:`Koalas.DataFrame.plot`.
+
+        Returns
+        -------
+        matplotlib.axes.Axes or numpy.ndarray of them
+        """
+        return self(kind="kde", bw_method=bw_method, ind=ind, **kwargs)
+
+    density = kde
 
     def pie(self, y=None, **kwds):
         """
@@ -1362,9 +1382,6 @@ def barh(self, x=None, y=None, **kwargs):
     def hexbin(self, **kwds):
         return _unsupported_function(class_name='pd.DataFrame', method_name='hexbin')()
 
-    def density(self, **kwds):
-        return _unsupported_function(class_name='pd.DataFrame', method_name='density')()
-
     def box(self, **kwds):
         return _unsupported_function(class_name='pd.DataFrame', method_name='box')()
 

databricks/koalas/tests/test_frame_plot.py

@@ -335,10 +335,26 @@ def _test_hist_plot(pdf, kdf):
         kdf.columns = columns
         _test_hist_plot(pdf, kdf)
 
+    def test_kde_plot(self):
+        pdf = self.pdf1
+        kdf = self.kdf1
+
+        ax1 = pdf.plot.kde(bw_method=3)
+        ax2 = kdf.plot.kde(bw_method=3)
+        self.compare_plots(ax1, ax2)
+
+        ax1 = pdf.plot(kind='kde', bw_method=3)
+        ax2 = kdf.plot(kind='kde', bw_method=3)
+        self.compare_plots(ax1, ax2)
+
+        ax1 = pdf.plot.kde(bw_method=3, ind=[1, 2, 3])
+        ax2 = kdf.plot.kde(bw_method=3, ind=[1, 2, 3])
+        self.compare_plots(ax1, ax2)
+
     def test_missing(self):
         ks = self.kdf1
 
-        unsupported_functions = ['box', 'density', 'hexbin', 'kde']
+        unsupported_functions = ['box', 'hexbin']
 
         for name in unsupported_functions:
             with self.assertRaisesRegex(PandasNotImplementedError,

docs/source/reference/frame.rst

@@ -243,4 +243,6 @@ specific plotting methods of the form ``DataFrame.plot.<kind>``.
    DataFrame.plot.line
    DataFrame.plot.pie
    DataFrame.plot.scatter
-   DataFrame.hist
+   DataFrame.plot.hist
+   DataFrame.plot.kde
+   DataFrame.plot.density