ENH: Add optional argument index to pd.melt to maintain index values

doc/source/user_guide/reshaping.rst

@@ -296,6 +296,7 @@ For instance,
    cheese.melt(id_vars=['first', 'last'])
    cheese.melt(id_vars=['first', 'last'], var_name='quantity')
 
+
 Another way to transform is to use the :func:`~pandas.wide_to_long` panel data
 convenience function. It is less flexible than :func:`~pandas.melt`, but more
 user-friendly.
@@ -312,6 +313,27 @@ user-friendly.
   dft
   pd.wide_to_long(dft, ["A", "B"], i="id", j="year")
 
+
+Keeping the index around
+~~~~~~~~~~~~~~~~~~~~~~~~
+When transforming a DataFrame using :func:`~pandas.melt`, original index values can be kept around by setting the ``keep_index`` parameter to True. This will however duplicate the index values.
+
+.. ipython:: python
+
+   index = pd.MultiIndex.from_tuples([('person', 'first'), ('person', 'second')])
+   cheese = pd.DataFrame({'first': ['John', 'Mary'],
+                          'last': ['Doe', 'Bo'],
+                          'height': [5.5, 6.0],
+                          'weight': [130, 150]},
+                         index=index)
+   cheese
+
+.. ipython:: python
+
+   cheese.melt(id_vars=['first', 'last'], keep_index=True)
+
+.. versionadded:: 1.1.0
+
 .. _reshaping.combine_with_groupby:
 
 Combining with stats and GroupBy

doc/source/whatsnew/v1.1.0.rst

@@ -145,11 +145,12 @@ Other enhancements
   This can be used to set a custom compression level, e.g.,
   ``df.to_csv(path, compression={'method': 'gzip', 'compresslevel': 1}``
   (:issue:`33196`)
+- :meth:`~pandas.core.groupby.GroupBy.transform` has gained ``engine`` and ``engine_kwargs`` arguments that supports executing functions with ``Numba`` (:issue:`32854`)
+- :meth:`melt` has gained an ``keep_index`` (default ``False``) argument that prevents the method from dropping the index (:issue:`17440`).
 - :meth:`Series.update` now accepts objects that can be coerced to a :class:`Series`,
   such as ``dict`` and ``list``, mirroring the behavior of :meth:`DataFrame.update` (:issue:`33215`)
 - :meth:`~pandas.core.groupby.GroupBy.transform` and :meth:`~pandas.core.groupby.GroupBy.aggregate` has gained ``engine`` and ``engine_kwargs`` arguments that supports executing functions with ``Numba`` (:issue:`32854`, :issue:`33388`)
 - :meth:`~pandas.core.resample.Resampler.interpolate` now supports SciPy interpolation method :class:`scipy.interpolate.CubicSpline` as method ``cubicspline`` (:issue:`33670`)
--
 
 .. ---------------------------------------------------------------------------
 

pandas/core/frame.py

@@ -2073,7 +2073,7 @@ def to_stata(
             from pandas.io.stata import StataWriter117 as statawriter  # type: ignore
         else:  # versions 118 and 119
             # mypy: Name 'statawriter' already defined (possibly by an import)
-            from pandas.io.stata import StataWriterUTF8 as statawriter  # type:ignore
+            from pandas.io.stata import StataWriterUTF8 as statawriter  # type: ignore
 
         kwargs: Dict[str, Any] = {}
         if version is None or version >= 117:
@@ -6795,7 +6795,7 @@ def unstack(self, level=-1, fill_value=None):
     or more columns are identifier variables (`id_vars`), while all other
     columns, considered measured variables (`value_vars`), are "unpivoted" to
     the row axis, leaving just two non-identifier columns, 'variable' and
-    'value'.
+    'value'. Optionally keeps original index values around.
     %(versionadded)s
     Parameters
     ----------
@@ -6811,6 +6811,9 @@ def unstack(self, level=-1, fill_value=None):
         Name to use for the 'value' column.
     col_level : int or str, optional
         If columns are a MultiIndex then use this level to melt.
+    keep_index : bool, optional, default False
+            If True, the original index is reused.
+            There is no uniqueness guarantee.
 
     Returns
     -------
@@ -6861,6 +6864,17 @@ def unstack(self, level=-1, fill_value=None):
     1  b         B          3
     2  c         B          5
 
+    Original index values can be kept around:
+
+    >>> %(caller)sid_vars=['A'], value_vars=['B', 'C'], keep_index=True)
+       A variable  value
+    0  a        B      1
+    1  b        B      3
+    2  c        B      5
+    0  a        C      2
+    1  b        C      4
+    2  c        C      6
+
     If you have multi-index columns:
 
     >>> df.columns = [list('ABC'), list('DEF')]
@@ -6899,6 +6913,7 @@ def melt(
         var_name=None,
         value_name="value",
         col_level=None,
+        keep_index=False,
     ) -> "DataFrame":
         from pandas.core.reshape.melt import melt
 
@@ -6909,6 +6924,7 @@ def melt(
             var_name=var_name,
             value_name=value_name,
             col_level=col_level,
+            keep_index=keep_index,
         )
 
     # ----------------------------------------------------------------------

pandas/core/reshape/melt.py

@@ -14,6 +14,7 @@
 import pandas.core.common as com
 from pandas.core.frame import DataFrame, _shared_docs
 from pandas.core.indexes.base import Index
+from pandas.core.indexes.multi import MultiIndex
 from pandas.core.reshape.concat import concat
 from pandas.core.tools.numeric import to_numeric
 
@@ -29,8 +30,8 @@ def melt(
     var_name=None,
     value_name="value",
     col_level=None,
+    keep_index=False,
 ) -> DataFrame:
-    # TODO: what about the existing index?
     # If multiindex, gather names of columns on all level for checking presence
     # of `id_vars` and `value_vars`
     if isinstance(frame.columns, ABCMultiIndex):
@@ -115,7 +116,19 @@ def melt(
         # asanyarray will keep the columns as an Index
         mdata[col] = np.asanyarray(frame.columns._get_level_values(i)).repeat(N)
 
-    return frame._constructor(mdata, columns=mcolumns)
+    result = frame._constructor(mdata, columns=mcolumns)
+
+    if keep_index:
+        new_index = np.tile(frame.index, K)
+
+        if isinstance(frame.index, MultiIndex):
+            new_index = MultiIndex.from_tuples(new_index, names=frame.index.names)
+        else:
+            new_index = Index(new_index, dtype=frame.index.dtype, name=frame.index.name)
+
+        result = result.set_index(new_index)
+
+    return result
 
 
 @deprecate_kwarg(old_arg_name="label", new_arg_name=None)

pandas/tests/reshape/test_melt.py

@@ -333,6 +333,42 @@ def test_melt_mixed_int_str_value_vars(self):
         expected = DataFrame({"variable": [0, "a"], "value": ["foo", "bar"]})
         tm.assert_frame_equal(result, expected)
 
+    def test_keep_index(self):
+        # GH 17440
+        df = DataFrame({"foo": [0], "bar": [1]}, index=["first"])
+        result = melt(df, keep_index=True)
+        expected = DataFrame(
+            {"variable": ["foo", "bar"], "value": [0, 1]}, index=["first", "first"]
+        )
+        tm.assert_frame_equal(result, expected)
+
+    def test_keep_multiindex(self):
+        # GH 17440
+        index = pd.MultiIndex.from_tuples(
+            [("first", "second"), ("first", "third")], names=["baz", "foobar"]
+        )
+        df = DataFrame({"foo": [0, 1], "bar": [2, 3]}, index=index)
+        result = melt(df, keep_index=True)
+
+        expected_index = pd.MultiIndex.from_tuples(
+            [("first", "second"), ("first", "third")] * 2, names=["baz", "foobar"]
+        )
+        expected = DataFrame(
+            {"variable": ["foo"] * 2 + ["bar"] * 2, "value": [0, 1, 2, 3]},
+            index=expected_index,
+        )
+
+        tm.assert_frame_equal(result, expected)
+
+    def test_keep_index_name_and_type(self):
+        # GH 17440
+        index = pd.Index(["foo", "bar"], dtype="category", name="baz")
+        df = DataFrame({"x": [0, 1], "y": [2, 3]}, index=index)
+        result = melt(df, keep_index=True)
+
+        expected_index = pd.Index(["foo", "bar"] * 2, dtype="category", name="baz")
+        tm.assert_index_equal(result.index, expected_index)
+
 
 class TestLreshape:
     def test_pairs(self):