Support create_index: bool in to_dataframe() to skip creating MultiIndex

[dhruvak001]

• Closes Support create_index: bool in to_dataframe to skip creating MultiIndex #10912
• Tests Added

Support [create_index: bool] in [to_dataframe()] to skip creating MultiIndex.

This PR adds a new create_index parameter to both Dataset.to_dataframe() and DataArray.to_dataframe() methods, allowing users to skip the potentially expensive MultiIndex creation and use a simple RangeIndex instead.

[Copilot]

Pull request overview

This pull request adds a create_index parameter to both Dataset.to_dataframe() and DataArray.to_dataframe() methods, allowing users to bypass the potentially expensive MultiIndex creation and use a simple RangeIndex instead. This is a performance optimization feature that addresses issue #10912.

• Adds optional create_index: bool = True parameter to maintain backward compatibility
• When create_index=False, uses pd.RangeIndex instead of constructing a MultiIndex from coordinates
• Preserves data integrity and ordering while avoiding the MultiIndex overhead

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 1 comment.

File	Description
xarray/core/dataset.py	Adds create_index parameter to to_dataframe() and _to_dataframe() methods, implementing RangeIndex creation when parameter is False
xarray/core/dataarray.py	Adds create_index parameter to to_dataframe() method and passes it through to Dataset's _to_dataframe()
xarray/tests/test_dataset.py	Adds comprehensive tests for the new create_index parameter on Dataset, including default behavior, RangeIndex creation, data integrity, and interaction with dim_order
xarray/tests/test_dataarray.py	Adds comprehensive tests for the new create_index parameter on DataArray, including default behavior, RangeIndex creation, data integrity, and interaction with additional coordinates

Comments suppressed due to low confidence (1)

xarray/core/dataset.py:7277

• The docstring states "The DataFrame is indexed by the Cartesian product of this dataset's indices" but this is no longer true when create_index=False. Consider updating the first paragraph of the docstring to clarify this behavior, e.g., "When create_index=True (default), the DataFrame is indexed by the Cartesian product of this dataset's indices. When create_index=False, a simple RangeIndex is used instead."

        Non-index variables in this dataset form the columns of the
        DataFrame. The DataFrame is indexed by the Cartesian product of
        this dataset's indices.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[dcherian]

Looks great, can you add a note to whats-new.rst please

[benbovy]

Thanks @dhruvak001. I just left one small comment on the docstrings.

[benbovy]

Suggested change

	create_index : bool, default: True
	If True (default), create a MultiIndex from the Cartesian product
	of this dataset's indices. If False, use a RangeIndex instead.
	This can be useful to avoid the potentially expensive MultiIndex
	creation.
	create_index : bool, default: True
	If True (default), create a :py:class:`pandas.MultiIndex` from the Cartesian product
	of this dataset's indices. If False, use a :py:class:`pandas.RangeIndex` instead.
	This can be useful to avoid the potentially expensive MultiIndex
	creation.

To avoid any confusion with xarray.indexes.RangeIndex (float range) and xarray.indexes.PandasMultiIndex.

[dcherian]

Can you make this change for to_dask_dataframe too please?