ENH: Add optional argument keep_index to dataframe melt method

Setting keep_index to True will reuse the original DataFrame index +
names of melted columns as additional level. closes issue #17440

Author: NiklasKeck

pandas/core/frame.py

@@ -4367,6 +4367,10 @@ def unstack(self, level=-1, fill_value=None):
         Name to use for the 'value' column.
     col_level : int or string, optional
         If columns are a MultiIndex then use this level to melt.
+    keep_index : boolean, optional, default False
+        If True, the original index is reused.
+        In the resulting MulitIndex the names of the unpivoted columns
+        are added as an additional level to ensure uniqueness.
 
     See also
     --------
@@ -4439,11 +4443,11 @@ def unstack(self, level=-1, fill_value=None):
                    versionadded='.. versionadded:: 0.20.0\n',
                    other='melt'))
     def melt(self, id_vars=None, value_vars=None, var_name=None,
-             value_name='value', col_level=None):
+             value_name='value', col_level=None, keep_index=False):
         from pandas.core.reshape.reshape import melt
         return melt(self, id_vars=id_vars, value_vars=value_vars,
                     var_name=var_name, value_name=value_name,
-                    col_level=col_level)
+                    col_level=col_level, keep_index=keep_index)
 
     # ----------------------------------------------------------------------
     # Time series-related

pandas/core/reshape/reshape.py

@@ -720,8 +720,7 @@ def _convert_level_number(level_num, columns):
                versionadded="",
                other='DataFrame.melt'))
 def melt(frame, id_vars=None, value_vars=None, var_name=None,
-         value_name='value', col_level=None):
-    # TODO: what about the existing index?
+         value_name='value', col_level=None, keep_index=False):
     if id_vars is not None:
         if not is_list_like(id_vars):
             id_vars = [id_vars]
@@ -779,7 +778,22 @@ def melt(frame, id_vars=None, value_vars=None, var_name=None,
         mdata[col] = np.asanyarray(frame.columns
                                    ._get_level_values(i)).repeat(N)
 
-    return DataFrame(mdata, columns=mcolumns)
+    result = DataFrame(mdata, columns=mcolumns)
+
+    if keep_index:
+        orig_index_values = list(np.tile(frame.index.get_values(), K))
+
+        if len(frame.index.names) == len(set(frame.index.names)):
+            orig_index_names = frame.index.names
+        else:
+            orig_index_names = ["original_index_{i}".format(i=i)
+                                for i in range(len(frame.index.names))]
+
+        result[orig_index_names] = DataFrame(orig_index_values)
+
+        result = result.set_index(orig_index_names + list(var_name))
+
+    return result
 
 
 def lreshape(data, groups, dropna=True, label=None):