Squashed commit of the following:

commit 620b946
Author: Stephan Hoyer <[email protected]>
Date:   Sun Jan 27 13:02:03 2019 -0800

    Fix test failures / warnings for pandas 0.24 (#2720)

    * Fix test failures / warnings for pandas 0.24

    Fixes GH2717

    * Brief doc note

    * Comment on name order

commit 882deac
Author: Stephan Hoyer <[email protected]>
Date:   Sun Jan 27 09:11:47 2019 -0800

    DOC: refresh whats-new for 0.11.3 / 0.12.0 (#2718)

commit 8ca8efe
Author: Stephan Hoyer <[email protected]>
Date:   Sat Jan 26 10:14:49 2019 -0800

    Update environment for doc build (#2708)

    * Update environment for doc build

    We were pinning very old versions for most of these packages. This should
    fix the failures on ReadTheDocs.

    * Build fixes

commit 2e99c7d
Author: Stephan Hoyer <[email protected]>
Date:   Fri Jan 25 13:52:17 2019 -0800

    Print full environment fron conf.py (#2709)

    This should make it easier to debug the doc build environment.

commit cc5015a
Author: Stephan Hoyer <[email protected]>
Date:   Fri Jan 25 11:55:07 2019 -0800

    BUG: ensure indexes are reset when coords are modified (#2707)

    This was introduced by the recent indexes refactor, but never made it into a
    release.

commit aabda43
Author: Joe Hamman <[email protected]>
Date:   Fri Jan 25 08:38:45 2019 -0800

    Remove py2 compat (#2645)

    * strip out PY2 compat code from pycompat.py

    * isort

    * remove 2 unused imports

    * remove extra import

    * no more future

    * no unicode literals

    * no more ReprMixin

    * cleanup merge

    * remove deprecated imports from collections

    * 2 more cleanups from shoyer

Author: dcherian

.travis.yml

@@ -60,8 +60,8 @@ script:
   - python --version
   - python -OO -c "import xarray"
   - if [[ "$CONDA_ENV" == "docs" ]]; then
-      conda install -c conda-forge --override-channels sphinx sphinx_rtd_theme sphinx-gallery numpydoc "gdal>2.2.4";
-      sphinx-build -n -j auto -b html -d _build/doctrees doc _build/html;
+      cd doc;
+      sphinx-build -n -j auto -b html -d _build/doctrees . _build/html;
     elif [[ "$CONDA_ENV" == "lint" ]]; then
       pycodestyle xarray ;
     elif [[ "$CONDA_ENV" == "py36-hypothesis" ]]; then

ci/requirements-py36.yml

@@ -26,8 +26,8 @@ dependencies:
   - pseudonetcdf>=3.0.1
   - eccodes
   - cdms2
-  # - pynio  # xref #2683
-  # - iris>=1.10    # xref #2683
+# - pynio  # xref #2683
+# - iris>=1.10  # xref #2683
   - pydap
   - lxml
   - pip:

doc/conf.py

@@ -13,9 +13,10 @@
 # serve to show the default.
 from __future__ import absolute_import, division, print_function
 
+from contextlib import suppress
 import datetime
-import importlib
 import os
+import subprocess
 import sys
 
 import xarray
@@ -24,29 +25,33 @@
 
 print("python exec:", sys.executable)
 print("sys.path:", sys.path)
-for name in ('numpy scipy pandas matplotlib dask IPython seaborn '
-             'cartopy netCDF4 rasterio zarr iris flake8 '
-             'sphinx_gallery cftime').split():
-    try:
-        module = importlib.import_module(name)
-        if name == 'matplotlib':
-            module.use('Agg')
-        fname = module.__file__.rstrip('__init__.py')
-        print("%s: %s, %s" % (name, module.__version__, fname))
-    except ImportError:
-        print("no %s" % name)
-        # neither rasterio nor cartopy should be hard requirements for
-        # the doc build.
-        if name == 'rasterio':
-            allowed_failures.update(['gallery/plot_rasterio_rgb.py',
-                                     'gallery/plot_rasterio.py'])
-        elif name == 'cartopy':
-            allowed_failures.update(['gallery/plot_cartopy_facetgrid.py',
-                                     'gallery/plot_rasterio_rgb.py',
-                                     'gallery/plot_rasterio.py'])
+
+if 'conda' in sys.executable:
+    print('conda environment:')
+    subprocess.run(['conda', 'list'])
+else:
+    print('pip environment:')
+    subprocess.run(['pip', 'list'])
 
 print("xarray: %s, %s" % (xarray.__version__, xarray.__file__))
 
+with suppress(ImportError):
+    import matplotlib
+    matplotlib.use('Agg')
+
+try:
+    import rasterio
+except ImportError:
+    allowed_failures.update(['gallery/plot_rasterio_rgb.py',
+                             'gallery/plot_rasterio.py'])
+
+try:
+    import cartopy
+except ImportError:
+    allowed_failures.update(['gallery/plot_cartopy_facetgrid.py',
+                             'gallery/plot_rasterio_rgb.py',
+                             'gallery/plot_rasterio.py'])
+
 # -- General configuration ------------------------------------------------
 
 # If your documentation needs a minimal Sphinx version, state it here.

doc/environment.yml

@@ -2,22 +2,25 @@ name: xarray-docs
 channels:
   - conda-forge
 dependencies:
-  - python=3.6
-  - numpy=1.14.5
+  - python=3.7
+  - numpy=1.16.0
   - pandas=0.23.3
-  - scipy=1.1.0
-  - matplotlib=2.2.2
+  - scipy=1.2.0
+  - matplotlib=3.0.2
   - seaborn=0.9.0
-  - dask=0.18.2
-  - ipython=6.4.0
-  - netCDF4=1.4.0
-  - cartopy=0.16.0
-  - rasterio=1.0.1
+  - dask=1.1.0
+  - ipython=7.2.0
+  - netCDF4=1.4.2
+  - cartopy=0.17.0
+  - rasterio=1.0.13
   - zarr=2.2.0
-  - iris=2.1.0
-  - flake8=3.5.0
+  - iris=2.2.0
+  - flake8=3.6.0
   - cftime=1.0.3.4
-  - bottleneck=1.2
-  - sphinx=1.7.6
+  - bottleneck=1.2.1
+  - sphinx=1.8.2
   - numpydoc=0.8.0
   - sphinx-gallery=0.2.0
+  - pillow=5.4.1
+  - sphinx_rtd_theme=0.4.2
+  - mock=2.0.0

doc/examples/multidimensional-coords.rst

@@ -78,9 +78,8 @@ grid, we can take advantage of xarray's ability to apply
     ax = plt.axes(projection=ccrs.PlateCarree());
     ds.Tair[0].plot.pcolormesh(ax=ax, transform=ccrs.PlateCarree(),
                                x='xc', y='yc', add_colorbar=False);
-    ax.coastlines();
     @savefig xarray_multidimensional_coords_12_0.png width=100%
-    plt.tight_layout();
+    ax.coastlines();
 
 Multidimensional Groupby
 ------------------------

doc/pandas.rst

@@ -14,7 +14,7 @@ __ http://pandas.pydata.org/pandas-docs/stable/visualization.html
 __ http://stanford.edu/~mwaskom/software/seaborn/
 
 .. ipython:: python
-   :suppress:
+    :suppress:
 
     import numpy as np
     import pandas as pd
@@ -93,7 +93,6 @@ DataFrames:
 
     s = ds['foo'].to_series()
     s
-
     # or equivalently, with Series.to_xarray()
     xr.DataArray.from_series(s)
 
@@ -173,11 +172,10 @@ So you can represent a Panel, in two ways:
 Let's take a look:
 
 .. ipython:: python
-   :okwarning:
+    :okwarning:
 
     panel = pd.Panel(np.random.rand(2, 3, 4), items=list('ab'), major_axis=list('mno'),
                      minor_axis=pd.date_range(start='2000', periods=4, name='date'))
-
     panel
 
 As a DataArray:

doc/time-series.rst

@@ -15,7 +15,6 @@ core functionality.
     import numpy as np
     import pandas as pd
     import xarray as xr
-
     np.random.seed(123456)
 
 Creating datetime64 data
@@ -241,7 +240,6 @@ coordinate with dates from a no-leap calendar and a
 
    from itertools import product
    from cftime import DatetimeNoLeap
-   
    dates = [DatetimeNoLeap(year, month, 1) for year, month in
             product(range(1, 3), range(1, 13))]
    da = xr.DataArray(np.arange(24), coords=[dates], dims=['time'], name='foo')

doc/whats-new.rst

@@ -13,9 +13,9 @@ What's New
     import xarray as xr
     np.random.seed(123456)
 
-.. _whats-new.0.11.3:
+.. _whats-new.0.12.0:
 
-v0.11.3 (unreleased)
+v0.12.0 (unreleased)
 --------------------
 
 Breaking changes
@@ -49,14 +49,30 @@ Enhancements
 Bug fixes
 ~~~~~~~~~
 
+- Silenced warnings that appear when using pandas 0.24.
+  By `Stephan Hoyer <https://github.com/shoyer>`_
 - Interpolating via resample now internally specifies ``bounds_error=False``
   as an argument to ``scipy.interpolate.interp1d``, allowing for interpolation
   from higher frequencies to lower frequencies.  Datapoints outside the bounds
   of the original time coordinate are now filled with NaN (:issue:`2197`). By
   `Spencer Clark <https://github.com/spencerkclark>`_.
+
+.. _whats-new.0.11.3:
+
+v0.11.3 (26 January 2019)
+-------------------------
+
+Bug fixes
+~~~~~~~~~
+
 - Saving files with times encoded with reference dates with timezones
   (e.g. '2000-01-01T00:00:00-05:00') no longer raises an error
   (:issue:`2649`).  By `Spencer Clark <https://github.com/spencerkclark>`_.
+- Fixed performance regression with ``open_mfdataset`` (:issue:`2662`).
+  By `Tom Nicholas <http://github.com/TomNicholas>`_.
+- Fixed supplying an explicit dimension in the ``concat_dim`` argument to
+  to ``open_mfdataset`` (:issue:`2647`).
+  By `Ben Root <https://github.com/WeatherGod>`_.
 
 .. _whats-new.0.11.2:
 

setup.py

@@ -15,8 +15,6 @@
     'Operating System :: OS Independent',
     'Intended Audience :: Science/Research',
     'Programming Language :: Python',
-    'Programming Language :: Python :: 2',
-    'Programming Language :: Python :: 2.7',
     'Programming Language :: Python :: 3',
     'Programming Language :: Python :: 3.5',
     'Programming Language :: Python :: 3.6',

xarray/__init__.py

@@ -1,7 +1,4 @@
 # flake8: noqa
-from __future__ import absolute_import
-from __future__ import division
-from __future__ import print_function
 
 from ._version import get_versions
 __version__ = get_versions()['version']

xarray/backends/api.py

@@ -1,18 +1,16 @@
-from __future__ import absolute_import, division, print_function
-
 import os.path
 import warnings
 from glob import glob
 from io import BytesIO
 from numbers import Number
+from pathlib import Path
 
 import numpy as np
 
 from .. import Dataset, backends, conventions
 from ..core import indexing
 from ..core.combine import (
     _CONCAT_DIM_DEFAULT, _auto_combine, _infer_concat_order_from_positions)
-from ..core.pycompat import basestring, path_type
 from ..core.utils import close_on_error, is_grib_path, is_remote_uri
 from .common import ArrayWriter
 from .locks import _get_scheduler
@@ -99,7 +97,7 @@ def _normalize_path(path):
 def _validate_dataset_names(dataset):
     """DataArray.name and Dataset keys must be a string or None"""
     def check_name(name):
-        if isinstance(name, basestring):
+        if isinstance(name, str):
             if not name:
                 raise ValueError('Invalid name for DataArray or Dataset key: '
                                  'string must be length 1 or greater for '
@@ -117,7 +115,7 @@ def _validate_attrs(dataset):
     a string, an ndarray or a list/tuple of numbers/strings.
     """
     def check_attr(name, value):
-        if isinstance(name, basestring):
+        if isinstance(name, str):
             if not name:
                 raise ValueError('Invalid name for attr: string must be '
                                  'length 1 or greater for serialization to '
@@ -126,7 +124,7 @@ def check_attr(name, value):
             raise TypeError("Invalid name for attr: {} must be a string for "
                             "serialization to netCDF files".format(name))
 
-        if not isinstance(value, (basestring, Number, np.ndarray, np.number,
+        if not isinstance(value, (str, Number, np.ndarray, np.number,
                                   list, tuple)):
             raise TypeError('Invalid value for attr: {} must be a number, '
                             'a string, an ndarray or a list/tuple of '
@@ -279,7 +277,7 @@ def maybe_decode_store(store, lock=False):
             from dask.base import tokenize
             # if passed an actual file path, augment the token with
             # the file modification time
-            if (isinstance(filename_or_obj, basestring) and
+            if (isinstance(filename_or_obj, str) and
                     not is_remote_uri(filename_or_obj)):
                 mtime = os.path.getmtime(filename_or_obj)
             else:
@@ -295,13 +293,13 @@ def maybe_decode_store(store, lock=False):
 
         return ds2
 
-    if isinstance(filename_or_obj, path_type):
+    if isinstance(filename_or_obj, Path):
         filename_or_obj = str(filename_or_obj)
 
     if isinstance(filename_or_obj, backends.AbstractDataStore):
         store = filename_or_obj
         ds = maybe_decode_store(store)
-    elif isinstance(filename_or_obj, basestring):
+    elif isinstance(filename_or_obj, str):
 
         if (isinstance(filename_or_obj, bytes) and
                 filename_or_obj.startswith(b'\x89HDF')):
@@ -310,7 +308,7 @@ def maybe_decode_store(store, lock=False):
                 filename_or_obj.startswith(b'CDF')):
             # netCDF3 file images are handled by scipy
             pass
-        elif isinstance(filename_or_obj, basestring):
+        elif isinstance(filename_or_obj, str):
             filename_or_obj = _normalize_path(filename_or_obj)
 
         if engine is None:
@@ -352,7 +350,7 @@ def maybe_decode_store(store, lock=False):
 
     # Ensure source filename always stored in dataset object (GH issue #2550)
     if 'source' not in ds.encoding:
-        if isinstance(filename_or_obj, basestring):
+        if isinstance(filename_or_obj, str):
             ds.encoding['source'] = filename_or_obj
 
     return ds
@@ -588,15 +586,15 @@ def open_mfdataset(paths, chunks=None, concat_dim=_CONCAT_DIM_DEFAULT,
     .. [1] http://xarray.pydata.org/en/stable/dask.html
     .. [2] http://xarray.pydata.org/en/stable/dask.html#chunking-and-performance
     """  # noqa
-    if isinstance(paths, basestring):
+    if isinstance(paths, str):
         if is_remote_uri(paths):
             raise ValueError(
                 'cannot do wild-card matching for paths that are remote URLs: '
                 '{!r}. Instead, supply paths as an explicit list of strings.'
                 .format(paths))
         paths = sorted(glob(paths))
     else:
-        paths = [str(p) if isinstance(p, path_type) else p for p in paths]
+        paths = [str(p) if isinstance(p, Path) else p for p in paths]
 
     if not paths:
         raise IOError('no files to open')
@@ -681,7 +679,7 @@ def to_netcdf(dataset, path_or_file=None, mode='w', format=None, group=None,
 
     The ``multifile`` argument is only for the private use of save_mfdataset.
     """
-    if isinstance(path_or_file, path_type):
+    if isinstance(path_or_file, Path):
         path_or_file = str(path_or_file)
 
     if encoding is None:
@@ -698,7 +696,7 @@ def to_netcdf(dataset, path_or_file=None, mode='w', format=None, group=None,
             raise NotImplementedError(
                 'to_netcdf() with compute=False is not yet implemented when '
                 'returning bytes')
-    elif isinstance(path_or_file, basestring):
+    elif isinstance(path_or_file, str):
         if engine is None:
             engine = _get_default_engine(path_or_file)
         path_or_file = _normalize_path(path_or_file)
@@ -733,7 +731,7 @@ def to_netcdf(dataset, path_or_file=None, mode='w', format=None, group=None,
 
     if unlimited_dims is None:
         unlimited_dims = dataset.encoding.get('unlimited_dims', None)
-    if isinstance(unlimited_dims, basestring):
+    if isinstance(unlimited_dims, str):
         unlimited_dims = [unlimited_dims]
 
     writer = ArrayWriter()
@@ -896,7 +894,7 @@ def to_zarr(dataset, store=None, mode='w-', synchronizer=None, group=None,
 
     See `Dataset.to_zarr` for full API docs.
     """
-    if isinstance(store, path_type):
+    if isinstance(store, Path):
         store = str(store)
     if encoding is None:
         encoding = {}

xarray/backends/cfgrib_.py

@@ -1,5 +1,3 @@
-from __future__ import absolute_import, division, print_function
-
 import numpy as np
 
 from .. import Variable