Merge remote-tracking branch 'upstream/master' into fix/4107

* upstream/master: (24 commits)
  Compatibility with dask 2021.02.0 (pydata#4884)
  Ensure maximum accuracy when encoding and decoding cftime.datetime values (pydata#4758)
  Fix `bounds_error=True` ignored with 1D interpolation (pydata#4855)
  add a drop_conflicts strategy for merging attrs (pydata#4827)
  update pre-commit hooks (mypy) (pydata#4883)
  ensure warnings cannot become errors in assert_ (pydata#4864)
  update pre-commit hooks (pydata#4874)
  small fixes for the docstrings of swap_dims and integrate (pydata#4867)
  Modify _encode_datetime_with_cftime for compatibility with cftime > 1.4.0 (pydata#4871)
  vélin (pydata#4872)
  don't skip the doctests CI (pydata#4869)
  fix da.pad example for numpy 1.20 (pydata#4865)
  temporarily pin dask (pydata#4873)
  Add units if "unit" is in the attrs. (pydata#4850)
  speed up the repr for big MultiIndex objects (pydata#4846)
  dim -> coord in DataArray.integrate (pydata#3993)
  WIP: backend interface, now it uses subclassing  (pydata#4836)
  weighted: small improvements (pydata#4818)
  Update related-projects.rst (pydata#4844)
  iris update doc url (pydata#4845)
  ...

Author: dcherian

.github/PULL_REQUEST_TEMPLATE.md

@@ -5,11 +5,3 @@
 - [ ] Passes `pre-commit run --all-files`
 - [ ] User visible changes (including notable bug fixes) are documented in `whats-new.rst`
 - [ ] New functions/methods are listed in `api.rst`
-
-
-<sub>
-<h3>
-  Overriding CI behaviors
-</h3>
-   By default, the upstream dev CI is disabled on pull request and push events. You can override this behavior per commit by adding a <tt>[test-upstream]</tt> tag to the first line of the commit message. For documentation-only commits, you can skip the CI per commit by adding a <tt>[skip-ci]</tt> tag to the first line of the commit message
-</sub>

.github/workflows/ci-additional.yaml

@@ -121,8 +121,6 @@ jobs:
   doctest:
     name: Doctests
     runs-on: "ubuntu-latest"
-    needs: detect-ci-trigger
-    if: needs.detect-ci-trigger.outputs.triggered == 'false'
     defaults:
       run:
         shell: bash -l {0}

.pre-commit-config.yaml

@@ -17,15 +17,20 @@ repos:
     hooks:
       - id: black
   - repo: https://github.com/keewis/blackdoc
-    rev: v0.3.2
+    rev: v0.3.3
     hooks:
       - id: blackdoc
   - repo: https://gitlab.com/pycqa/flake8
     rev: 3.8.4
     hooks:
       - id: flake8
+  # - repo: https://github.com/Carreau/velin
+  #   rev: 0.0.8
+  #   hooks:
+  #     - id: velin
+  #       args: ["--write", "--compact"]
   - repo: https://github.com/pre-commit/mirrors-mypy
-    rev: v0.790  # Must match ci/requirements/*.yml
+    rev: v0.800
     hooks:
       - id: mypy
         exclude: "properties|asv_bench"

asv_bench/benchmarks/repr.py

@@ -0,0 +1,18 @@
+import pandas as pd
+
+import xarray as xr
+
+
+class ReprMultiIndex:
+    def setup(self, key):
+        index = pd.MultiIndex.from_product(
+            [range(10000), range(10000)], names=("level_0", "level_1")
+        )
+        series = pd.Series(range(100000000), index=index)
+        self.da = xr.DataArray(series)
+
+    def time_repr(self):
+        repr(self.da)
+
+    def time_repr_html(self):
+        self.da._repr_html_()

asv_bench/benchmarks/unstacking.py

@@ -7,18 +7,23 @@
 
 class Unstacking:
     def setup(self):
-        data = np.random.RandomState(0).randn(1, 1000, 500)
-        self.ds = xr.DataArray(data).stack(flat_dim=["dim_1", "dim_2"])
+        data = np.random.RandomState(0).randn(500, 1000)
+        self.da_full = xr.DataArray(data, dims=list("ab")).stack(flat_dim=[...])
+        self.da_missing = self.da_full[:-1]
+        self.df_missing = self.da_missing.to_pandas()
 
     def time_unstack_fast(self):
-        self.ds.unstack("flat_dim")
+        self.da_full.unstack("flat_dim")
 
     def time_unstack_slow(self):
-        self.ds[:, ::-1].unstack("flat_dim")
+        self.da_missing.unstack("flat_dim")
+
+    def time_unstack_pandas_slow(self):
+        self.df_missing.unstack()
 
 
 class UnstackingDask(Unstacking):
     def setup(self, *args, **kwargs):
         requires_dask()
         super().setup(**kwargs)
-        self.ds = self.ds.chunk({"flat_dim": 50})
+        self.da_full = self.da_full.chunk({"flat_dim": 50})

ci/requirements/py37-min-all-deps.yml

@@ -8,7 +8,6 @@ dependencies:
   # When upgrading python, numpy, or pandas, must also change
   # doc/installing.rst and setup.py.
   - python=3.7
-  - black
   - boto3=1.9
   - bottleneck=1.2
   - cartopy=0.17
@@ -18,16 +17,13 @@ dependencies:
   - coveralls
   - dask=2.9
   - distributed=2.9
-  - flake8
   - h5netcdf=0.7
   - h5py=2.9  # Policy allows for 2.10, but it's a conflict-fest
   - hdf5=1.10
   - hypothesis
   - iris=2.2
-  - isort
   - lxml=4.4  # Optional dep of pydap
   - matplotlib-base=3.1
-  - mypy=0.782  # Must match .pre-commit-config.yaml
   - nc-time-axis=1.2
   - netcdf4=1.4
   - numba=0.46

ci/requirements/py38-all-but-dask.yml

@@ -4,23 +4,19 @@ channels:
   - nodefaults
 dependencies:
   - python=3.8
-  - black
   - boto3
   - bottleneck
   - cartopy
   - cdms2
   - cfgrib
   - cftime
   - coveralls
-  - flake8
   - h5netcdf
   - h5py=2
   - hdf5
   - hypothesis
-  - isort
   - lxml    # Optional dep of pydap
   - matplotlib-base
-  - mypy=0.790  # Must match .pre-commit-config.yaml
   - nc-time-axis
   - netcdf4
   - numba

doc/conf.py

@@ -411,7 +411,7 @@
 intersphinx_mapping = {
     "python": ("https://docs.python.org/3/", None),
     "pandas": ("https://pandas.pydata.org/pandas-docs/stable", None),
-    "iris": ("https://scitools.org.uk/iris/docs/latest", None),
+    "iris": ("https://scitools-iris.readthedocs.io/en/latest", None),
     "numpy": ("https://numpy.org/doc/stable", None),
     "scipy": ("https://docs.scipy.org/doc/scipy/reference", None),
     "numba": ("https://numba.pydata.org/numba-doc/latest", None),

doc/contributing.rst

@@ -836,6 +836,7 @@ PR checklist
 
     - Write new tests if needed. See `"Test-driven development/code writing" <https://xarray.pydata.org/en/stable/contributing.html#test-driven-development-code-writing>`_.
     - Test the code using `Pytest <http://doc.pytest.org/en/latest/>`_. Running all tests (type ``pytest`` in the root directory) takes a while, so feel free to only run the tests you think are needed based on your PR (example: ``pytest xarray/tests/test_dataarray.py``). CI will catch any failing tests.
+    - By default, the upstream dev CI is disabled on pull request and push events. You can override this behavior per commit by adding a <tt>[test-upstream]</tt> tag to the first line of the commit message. For documentation-only commits, you can skip the CI per commit by adding a "[skip-ci]" tag to the first line of the commit message.
 
 - **Properly format your code** and verify that it passes the formatting guidelines set by `Black <https://black.readthedocs.io/en/stable/>`_ and `Flake8 <http://flake8.pycqa.org/en/latest/>`_. See `"Code formatting" <https://xarray.pydata.org/en/stablcontributing.html#code-formatting>`_. You can use `pre-commit <https://pre-commit.com/>`_ to run these automatically on each commit.
 

doc/faq.rst

@@ -166,7 +166,7 @@ different approaches to handling metadata: Iris strictly interprets
 `CF conventions`_. Iris particularly shines at mapping, thanks to its
 integration with Cartopy_.
 
-.. _Iris: http://scitools.org.uk/iris/
+.. _Iris: https://scitools-iris.readthedocs.io/en/stable/
 .. _Cartopy: http://scitools.org.uk/cartopy/docs/latest/
 
 `UV-CDAT`__ is another Python library that implements in-memory netCDF-like