Merge remote-tracking branch 'upstream/master' into coarsen-reduce

* upstream/master: (22 commits)
  Resolve the version issues on RTD (pydata#3589)
  Add bottleneck & rasterio git tip to upstream-dev CI (pydata#3585)
  update whats-new.rst (pydata#3581)
  Examples for quantile (pydata#3576)
  add cftime intersphinx entries (pydata#3577)
  Add pyXpcm to Related Projects doc page (pydata#3578)
  Reimplement quantile with apply_ufunc (pydata#3559)
  add environment file for binderized examples (pydata#3568)
  Add drop to api.rst under pending deprecations (pydata#3561)
  replace duplicate method _from_vars_and_coord_names (pydata#3565)
  propagate indexes in to_dataset, from_dataset (pydata#3519)
  Switch examples to notebooks + scipy19 docs improvements (pydata#3557)
  fix whats-new.rst (pydata#3554)
  Tweaks to release instructions (pydata#3555)
  Clarify conda environments for new contributors (pydata#3551)
  Revert to dev version
  0.14.1 whatsnew (pydata#3547)
  sparse option to reindex and unstack (pydata#3542)
  Silence sphinx warnings (pydata#3516)
  Numpy 1.18 support (pydata#3537)
  ...

.binder/environment.yml

@@ -0,0 +1,39 @@
+name: xarray-examples
+channels:
+  - conda-forge
+dependencies:
+  - python=3.7
+  - boto3
+  - bottleneck
+  - cartopy
+  - cdms2
+  - cfgrib
+  - cftime
+  - coveralls
+  - dask
+  - distributed
+  - dask_labextension
+  - h5netcdf
+  - h5py
+  - hdf5
+  - iris
+  - lxml    # Optional dep of pydap
+  - matplotlib
+  - nc-time-axis
+  - netcdf4
+  - numba
+  - numpy
+  - pandas
+  - pint
+  - pip
+  - pydap
+  - pynio
+  - rasterio
+  - scipy
+  - seaborn
+  - sparse
+  - toolz
+  - xarray
+  - zarr
+  - pip:
+    - numbagg

HOW_TO_RELEASE → HOW_TO_RELEASE.md

@@ -1,9 +1,11 @@
-How to issue an xarray release in 15 easy steps
+How to issue an xarray release in 14 easy steps
 
 Time required: about an hour.
 
  1. Ensure your master branch is synced to upstream:
-       git pull upstream master
+      ```
+      git pull upstream master
+      ```
  2. Look over whats-new.rst and the docs. Make sure "What's New" is complete
     (check the date!) and consider adding a brief summary note describing the
     release at the top.
@@ -12,37 +14,53 @@ Time required: about an hour.
     - Function/method references should include links to the API docs.
     - Sometimes notes get added in the wrong section of whats-new, typically
       due to a bad merge. Check for these before a release by using git diff,
-      e.g., ``git diff v0.X.Y whats-new.rst`` where 0.X.Y is the previous
+      e.g., `git diff v0.X.Y whats-new.rst` where 0.X.Y is the previous
       release.
  3. If you have any doubts, run the full test suite one final time!
-      py.test
+      ```
+      pytest
+      ```
  4. On the master branch, commit the release in git:
+      ```
       git commit -a -m 'Release v0.X.Y'
+      ```
  5. Tag the release:
+      ```
       git tag -a v0.X.Y -m 'v0.X.Y'
+      ```
  6. Build source and binary wheels for pypi:
+      ```
       git clean -xdf  # this deletes all uncommited changes!
       python setup.py bdist_wheel sdist
+      ```
  7. Use twine to register and upload the release on pypi. Be careful, you can't
     take this back!
+      ```
       twine upload dist/xarray-0.X.Y*
+      ```
     You will need to be listed as a package owner at
     https://pypi.python.org/pypi/xarray for this to work.
  8. Push your changes to master:
+      ```
       git push upstream master
       git push upstream --tags
+      ```
  9. Update the stable branch (used by ReadTheDocs) and switch back to master:
+     ```
       git checkout stable
       git rebase master
       git push upstream stable
       git checkout master
-    It's OK to force push to 'stable' if necessary.
-    We also update the stable branch with `git cherrypick` for documentation
-    only fixes that apply the current released version.
+     ```
+    It's OK to force push to 'stable' if necessary. (We also update the stable 
+    branch with `git cherrypick` for documentation only fixes that apply the 
+    current released version.)
 10. Add a section for the next release (v.X.(Y+1)) to doc/whats-new.rst.
 11. Commit your changes and push to master again:
-      git commit -a -m 'Revert to dev version'
+      ```
+      git commit -a -m 'New whatsnew section'
       git push upstream master
+      ```
     You're done pushing to master!
 12. Issue the release on GitHub. Click on "Draft a new release" at
     https://github.com/pydata/xarray/releases. Type in the version number, but
@@ -53,11 +71,22 @@ Time required: about an hour.
 14. Issue the release announcement! For bug fix releases, I usually only email
     xarray@googlegroups.com. For major/feature releases, I will email a broader
     list (no more than once every 3-6 months):
-      pydata@googlegroups.com, xarray@googlegroups.com,
-      numpy-discussion@scipy.org, scipy-user@scipy.org,
-      pyaos@lists.johnny-lin.com
+      - pydata@googlegroups.com
+      - xarray@googlegroups.com
+      - numpy-discussion@scipy.org
+      - scipy-user@scipy.org
+      - pyaos@lists.johnny-lin.com
+
     Google search will turn up examples of prior release announcements (look for
     "ANN xarray").
+    You can get a list of contributors with:
+    ```
+    git log "$(git tag --sort="v:refname" | sed -n 'x;$p').." --format="%aN" | sort -u
+    ```
+    or by replacing `v0.X.Y` with the _previous_ release in:
+    ```
+    git log v0.X.Y.. --format="%aN" | sort -u
+    ```
 
 Note on version numbering:
 

ci/azure/install.yml

@@ -16,16 +16,18 @@ steps:
         --pre \
         --upgrade \
         matplotlib \
+        numpy \
         pandas \
         scipy
-        # numpy \  # FIXME https://github.com/pydata/xarray/issues/3409
     pip install \
         --no-deps \
         --upgrade \
         git+https://github.com/dask/dask \
         git+https://github.com/dask/distributed \
         git+https://github.com/zarr-developers/zarr \
-        git+https://github.com/Unidata/cftime
+        git+https://github.com/Unidata/cftime \
+        git+https://github.com/mapbox/rasterio \
+        git+https://github.com/pydata/bottleneck
   condition: eq(variables['UPSTREAM_DEV'], 'true')
   displayName: Install upstream dev dependencies
 

ci/requirements/doc.yml

@@ -6,16 +6,20 @@ dependencies:
   - python=3.7
   - bottleneck
   - cartopy
+  - cfgrib
   - h5netcdf
+  - ipykernel
   - ipython
   - iris
+  - jupyter_client
+  - nbsphinx
   - netcdf4
   - numpy
   - numpydoc
   - pandas<0.25  # Hack around https://github.com/pydata/xarray/issues/3369
   - rasterio
   - seaborn
   - sphinx
-  - sphinx-gallery
   - sphinx_rtd_theme
+  - xarray
   - zarr

ci/requirements/py36.yml

@@ -25,7 +25,7 @@ dependencies:
   - nc-time-axis
   - netcdf4
   - numba
-  - numpy<1.18  # FIXME https://github.com/pydata/xarray/issues/3409
+  - numpy
   - pandas
   - pint
   - pip

ci/requirements/py37.yml

@@ -25,7 +25,7 @@ dependencies:
   - nc-time-axis
   - netcdf4
   - numba
-  - numpy<1.18  # FIXME https://github.com/pydata/xarray/issues/3409
+  - numpy
   - pandas
   - pint
   - pip

doc/README.rst

@@ -1,3 +1,5 @@
+:orphan:
+
 xarray
 ------
 

doc/api-hidden.rst

@@ -2,6 +2,8 @@
 .. This extra page is a work around for sphinx not having any support for
 .. hiding an autosummary table.
 
+:orphan:
+
 .. currentmodule:: xarray
 
 .. autosummary::
@@ -30,9 +32,11 @@
    core.groupby.DatasetGroupBy.first
    core.groupby.DatasetGroupBy.last
    core.groupby.DatasetGroupBy.fillna
+   core.groupby.DatasetGroupBy.quantile
    core.groupby.DatasetGroupBy.where
 
    Dataset.argsort
+   Dataset.astype
    Dataset.clip
    Dataset.conj
    Dataset.conjugate
@@ -71,6 +75,7 @@
    core.groupby.DataArrayGroupBy.first
    core.groupby.DataArrayGroupBy.last
    core.groupby.DataArrayGroupBy.fillna
+   core.groupby.DataArrayGroupBy.quantile
    core.groupby.DataArrayGroupBy.where
 
    DataArray.argsort

doc/api.rst

@@ -675,3 +675,12 @@ arguments for the ``from_store`` and ``dump_to_store`` Dataset methods:
    backends.FileManager
    backends.CachingFileManager
    backends.DummyFileManager
+
+Deprecated / Pending Deprecation
+================================
+
+   Dataset.drop
+   DataArray.drop
+   Dataset.apply
+   core.groupby.DataArrayGroupBy.apply
+   core.groupby.DatasetGroupBy.apply

doc/combining.rst

@@ -255,11 +255,11 @@ Combining along multiple dimensions
   ``combine_nested``.
 
 For combining many objects along multiple dimensions xarray provides
-:py:func:`~xarray.combine_nested`` and :py:func:`~xarray.combine_by_coords`. These
+:py:func:`~xarray.combine_nested` and :py:func:`~xarray.combine_by_coords`. These
 functions use a combination of ``concat`` and ``merge`` across different
 variables to combine many objects into one.
 
-:py:func:`~xarray.combine_nested`` requires specifying the order in which the
+:py:func:`~xarray.combine_nested` requires specifying the order in which the
 objects should be combined, while :py:func:`~xarray.combine_by_coords` attempts to
 infer this ordering automatically from the coordinates in the data.
 
@@ -310,4 +310,4 @@ These functions can be used by :py:func:`~xarray.open_mfdataset` to open many
 files as one dataset. The particular function used is specified by setting the
 argument ``'combine'`` to ``'by_coords'`` or ``'nested'``. This is useful for
 situations where your data is split across many files in multiple locations,
-which have some known relationship between one another.
+which have some known relationship between one another.

doc/computation.rst

@@ -325,8 +325,8 @@ Broadcasting by dimension name
 ``DataArray`` objects are automatically align themselves ("broadcasting" in
 the numpy parlance) by dimension name instead of axis order. With xarray, you
 do not need to transpose arrays or insert dimensions of length 1 to get array
-operations to work, as commonly done in numpy with :py:func:`np.reshape` or
-:py:const:`np.newaxis`.
+operations to work, as commonly done in numpy with :py:func:`numpy.reshape` or
+:py:data:`numpy.newaxis`.
 
 This is best illustrated by a few examples. Consider two one-dimensional
 arrays with different sizes aligned along different dimensions:
@@ -566,7 +566,7 @@ to set ``axis=-1``. As an example, here is how we would wrap
 
 Because ``apply_ufunc`` follows a standard convention for ufuncs, it plays
 nicely with tools for building vectorized functions, like
-:func:`numpy.broadcast_arrays` and :func:`numpy.vectorize`. For high performance
+:py:func:`numpy.broadcast_arrays` and :py:class:`numpy.vectorize`. For high performance
 needs, consider using Numba's :doc:`vectorize and guvectorize <numba:user/vectorize>`.
 
 In addition to wrapping functions, ``apply_ufunc`` can automatically parallelize

doc/conf.py

@@ -15,10 +15,16 @@
 
 import datetime
 import os
+import pathlib
 import subprocess
 import sys
 from contextlib import suppress
 
+# make sure the source version is preferred (#3567)
+root = pathlib.Path(__file__).absolute().parent.parent
+os.environ["PYTHONPATH"] = str(root)
+sys.path.insert(0, str(root))
+
 import xarray
 
 allowed_failures = set()
@@ -76,20 +82,24 @@
     "numpydoc",
     "IPython.sphinxext.ipython_directive",
     "IPython.sphinxext.ipython_console_highlighting",
-    "sphinx_gallery.gen_gallery",
+    "nbsphinx",
 ]
 
 extlinks = {
     "issue": ("https://github.com/pydata/xarray/issues/%s", "GH"),
     "pull": ("https://github.com/pydata/xarray/pull/%s", "PR"),
 }
 
-sphinx_gallery_conf = {
-    "examples_dirs": "gallery",
-    "gallery_dirs": "auto_gallery",
-    "backreferences_dir": False,
-    "expected_failing_examples": list(allowed_failures),
-}
+nbsphinx_timeout = 600
+nbsphinx_execute = "always"
+nbsphinx_prolog = """
+{% set docname = env.doc2path(env.docname, base=None) %}
+
+You can run this notebook in a `live session <https://mybinder.org/v2/gh/pydata/xarray/doc/examples/master?urlpath=lab/tree/doc/{{ docname }}>`_ |Binder| or view it `on Github <https://github.com/pydata/xarray/blob/master/doc/{{ docname }}>`_.
+
+.. |Binder| image:: https://mybinder.org/badge.svg
+   :target: https://mybinder.org/v2/gh/pydata/xarray/master?urlpath=lab/tree/doc/{{ docname }}
+"""
 
 autosummary_generate = True
 autodoc_typehints = "none"
@@ -137,7 +147,7 @@
 
 # List of patterns, relative to source directory, that match files and
 # directories to ignore when looking for source files.
-exclude_patterns = ["_build"]
+exclude_patterns = ["_build", "**.ipynb_checkpoints"]
 
 # The reST default role (used for this markup: `text`) to use for all
 # documents.
@@ -346,4 +356,6 @@
     "scipy": ("https://docs.scipy.org/doc/scipy/reference", None),
     "numba": ("https://numba.pydata.org/numba-doc/latest", None),
     "matplotlib": ("https://matplotlib.org", None),
+    "dask": ("https://docs.dask.org/en/latest", None),
+    "cftime": ("https://unidata.github.io/cftime", None),
 }

doc/contributing.rst

@@ -151,7 +151,9 @@ We'll now kick off a two-step process:
 .. code-block:: none
 
    # Create and activate the build environment
-   conda env create -f ci/requirements/py36.yml
+   # This is for Linux and MacOS. On Windows, use py37-windows.yml instead.
+   conda env create -f ci/requirements/py37.yml
+
    conda activate xarray-tests
 
    # or with older versions of Anaconda:

doc/dask.rst

@@ -285,7 +285,7 @@ automate `embarrassingly parallel
 <https://en.wikipedia.org/wiki/Embarrassingly_parallel>`__ "map" type operations
 where a function written for processing NumPy arrays should be repeatedly
 applied to xarray objects containing Dask arrays. It works similarly to
-:py:func:`dask.array.map_blocks` and :py:func:`dask.array.atop`, but without
+:py:func:`dask.array.map_blocks` and :py:func:`dask.array.blockwise`, but without
 requiring an intermediate layer of abstraction.
 
 For the best performance when using Dask's multi-threaded scheduler, wrap a