Implement load_dataset() and load_dataarray()

BUG: Fixes #2887 by adding @shoyer solution for load_dataset and load_dataarray, wrappers around open_dataset and open_dataarray which open, load, and close the file and return the Dataset/DataArray
TST: Add tests for sequentially opening and writing to files using new functions
DOC: Add to whats-new.rst. Also a tiny change to the open_dataset docstring

Update docstrings and check for cache in kwargs

Undeprecate load_dataset

Add to api.rst, fix whats-new.rst typo, raise error instead of warning

Author: dnowacki-usgs

doc/api.rst

@@ -460,6 +460,7 @@ Dataset methods
    :toctree: generated/
 
    open_dataset
+   load_dataset
    open_mfdataset
    open_rasterio
    open_zarr
@@ -487,6 +488,7 @@ DataArray methods
    :toctree: generated/
 
    open_dataarray
+   load_dataarray
    DataArray.to_dataset
    DataArray.to_netcdf
    DataArray.to_pandas

doc/whats-new.rst

@@ -29,6 +29,12 @@ Enhancements
   By `James McCreight <https://github.com/jmccreight>`_.
 - Clean up Python 2 compatibility in code (:issue:`2950`)
   By `Guido Imperiale <https://github.com/crusaderky>`_.
+- Implement ``load_dataset()`` and ``load_dataarray()`` as alternatives to
+  ``open_dataset()`` and ``open_dataarray()`` to open, load into memory,
+  and close files, returning the Dataset or DataArray. These functions are
+  helpful for avoiding file-lock errors when trying to write to files opened
+  using ``open_dataset()`` or ``open_dataarray()``. (:issue:`2887`)
+  By `Dan Nowacki <https://github.com/dnowacki-usgs>`_.
 
 Bug fixes
 ~~~~~~~~~
@@ -153,9 +159,9 @@ Other enhancements
   By `Keisuke Fujii <https://github.com/fujiisoup>`_.
 - Added :py:meth:`~xarray.Dataset.drop_dims` (:issue:`1949`).
   By `Kevin Squire <https://github.com/kmsquire>`_.
-- ``xr.open_zarr`` now accepts manually specified chunks with the ``chunks=`` 
-  parameter. ``auto_chunk=True`` is equivalent to ``chunks='auto'`` for 
-  backwards compatibility. The ``overwrite_encoded_chunks`` parameter is 
+- ``xr.open_zarr`` now accepts manually specified chunks with the ``chunks=``
+  parameter. ``auto_chunk=True`` is equivalent to ``chunks='auto'`` for
+  backwards compatibility. The ``overwrite_encoded_chunks`` parameter is
   added to remove the original zarr chunk encoding.
   By `Lily Wang <https://github.com/lilyminium>`_.
 

xarray/__init__.py

@@ -17,7 +17,7 @@
 from .core.options import set_options
 
 from .backends.api import (open_dataset, open_dataarray, open_mfdataset,
-                           save_mfdataset)
+                           save_mfdataset, load_dataset, load_dataarray)
 from .backends.rasterio_ import open_rasterio
 from .backends.zarr import open_zarr
 

xarray/backends/api.py

@@ -185,12 +185,64 @@ def _finalize_store(write, store):
     store.close()
 
 
+def load_dataset(filename_or_obj, **kwargs):
+    """Open, load into memory, and close a Dataset from a file or file-like
+    object.
+
+    This is a thin wrapper around :py:meth:`~xarray.open_dataset`. It differs
+    from `open_dataset` in that it loads the Dataset into memory, closes the
+    file, and returns the Dataset. In contrast, `open_dataset` keeps the file
+    handle open and lazy loads its contents. All parameters are passed directly
+    to `open_dataset`. See that documentation for further details.
+
+    Returns
+    -------
+    dataset : Dataset
+        The newly created Dataset.
+
+    See Also
+    --------
+    open_dataset
+    """
+    if 'cache' in kwargs:
+        raise TypeError('cache has no effect in this context')
+
+    with open_dataset(filename_or_obj, **kwargs) as ds:
+        return ds.load()
+
+
+def load_dataarray(filename_or_obj, **kwargs):
+    """Open, load into memory, and close a DataArray from a file or file-like
+    object containing a single data variable.
+
+    This is a thin wrapper around :py:meth:`~xarray.open_dataarray`. It differs
+    from `open_dataarray` in that it loads the Dataset into memory, closes the
+    file, and returns the Dataset. In contrast, `open_dataarray` keeps the file
+    handle open and lazy loads its contents. All parameters are passed directly
+    to `open_dataarray`. See that documentation for further details.
+
+    Returns
+    -------
+    datarray : DataArray
+        The newly created DataArray.
+
+    See Also
+    --------
+    open_dataarray
+    """
+    if 'cache' in kwargs:
+        raise TypeError('cache has no effect in this context')
+
+    with open_dataarray(filename_or_obj, **kwargs) as da:
+        return da.load()
+
+
 def open_dataset(filename_or_obj, group=None, decode_cf=True,
                  mask_and_scale=None, decode_times=True, autoclose=None,
                  concat_characters=True, decode_coords=True, engine=None,
                  chunks=None, lock=None, cache=None, drop_variables=None,
                  backend_kwargs=None, use_cftime=None):
-    """Load and decode a dataset from a file or file-like object.
+    """Open and decode a dataset from a file or file-like object.
 
     Parameters
     ----------
@@ -406,7 +458,8 @@ def open_dataarray(filename_or_obj, group=None, decode_cf=True,
                    concat_characters=True, decode_coords=True, engine=None,
                    chunks=None, lock=None, cache=None, drop_variables=None,
                    backend_kwargs=None, use_cftime=None):
-    """Open an DataArray from a netCDF file containing a single data variable.
+    """Open an DataArray from a file or file-like object containing a single
+    data variable.
 
     This is designed to read netCDF files with only one data variable. If
     multiple variables are present then a ValueError is raised.

xarray/tests/test_backends.py

@@ -19,7 +19,7 @@
 import xarray as xr
 from xarray import (
     DataArray, Dataset, backends, open_dataarray, open_dataset, open_mfdataset,
-    save_mfdataset)
+    save_mfdataset, load_dataset, load_dataarray)
 from xarray.backends.common import robust_getitem
 from xarray.backends.netCDF4_ import _extract_nc4_variable_encoding
 from xarray.backends.pydap_ import PydapDataStore
@@ -2641,6 +2641,23 @@ def test_save_mfdataset_compute_false_roundtrip(self):
                 with open_mfdataset([tmp1, tmp2]) as actual:
                     assert_identical(actual, original)
 
+    def test_load_dataset(self):
+        with create_tmp_file() as tmp:
+            original = Dataset({'foo': ('x', np.random.randn(10))})
+            original.to_netcdf(tmp)
+            ds = load_dataset(tmp)
+            # this would fail if we used open_dataset instead of load_dataset
+            ds.to_netcdf(tmp)
+
+    def test_load_dataarray(self):
+        with create_tmp_file() as tmp:
+            original = Dataset({'foo': ('x', np.random.randn(10))})
+            original.to_netcdf(tmp)
+            ds = load_dataarray(tmp)
+            # this would fail if we used open_dataarray instead of
+            # load_dataarray
+            ds.to_netcdf(tmp)
+
 
 @requires_scipy_or_netCDF4
 @requires_pydap

xarray/tutorial.py

@@ -27,7 +27,7 @@ def open_dataset(name, cache=True, cache_dir=_default_cache_dir,
                  github_url='https://github.com/pydata/xarray-data',
                  branch='master', **kws):
     """
-    Load a dataset from the online repository (requires internet).
+    Open a dataset from the online repository (requires internet).
 
     If a local copy is found then always use that to avoid network traffic.
 
@@ -91,17 +91,12 @@ def open_dataset(name, cache=True, cache_dir=_default_cache_dir,
 
 def load_dataset(*args, **kwargs):
     """
-    `load_dataset` will be removed a future version of xarray. The current
-    behavior of this function can be achived by using
-    `tutorial.open_dataset(...).load()`.
+    Open, load into memory, and close a dataset from the online repository
+    (requires internet).
 
     See Also
     --------
     open_dataset
     """
-    warnings.warn(
-        "load_dataset` will be removed in a future version of xarray. The "
-        "current behavior of this function can be achived by using "
-        "`tutorial.open_dataset(...).load()`.",
-        DeprecationWarning, stacklevel=2)
-    return open_dataset(*args, **kwargs).load()
+    with open_dataset(*args, **kwargs) as ds:
+        return ds.load()