issues

As per https://github.com/pydata/xarray/issues/6209#issuecomment-1025116203

On top of https://github.com/pydata/xarray/pull/5467, given I'm not sure this is correct.

With additional_dependencies, we may not need the additional CI run.

But OTOH there's nothing confirming that the additional_dependencies has the correct dependencies and there's no material cost to the additional step.

• [x] Passes pre-commit run --all-files

This should mean that running mypy . doesn't operate differently from running it in pre-commit.

It's also possible I've misunderstood something about mypy configs, please let me know if so!

I just saw https://numpy.org/neps/nep-0029-deprecation_policy.html, which suggests dropping 3.5 support earlier this year (though NEP is dated to this July)

Ofc, existing xarray versions would still work with 3.5, only upgrades would require the new version (which I know everyone knows, but reinforces we don't need to wait for everyone to move off 3.5)

• [x] Closes some of https://github.com/pydata/xarray/issues/2830

We have a DataArray unaligned to coords here:

``` python In [12]:

coords={'a': range(4), 'b': range(3)} da = xray.DataArray(np.random.rand(4,3), coords=coords) da Out[12]: <xray.DataArray (a: 4, b: 3)> array([[ 0.05126985, 0.95460352, 0.12853847], [ 0.20577943, 0.80265117, 0.46370886], [ 0.0226791 , 0.33068145, 0.55748573], [ 0.15943175, 0.20183347, 0.46907727]]) Coordinates: * a (a) int64 0 1 2 3 * b (b) int64 0 1 2 In [13]:

da_reindex = da.reindex(b=[1,2,0]) da_reindex Out[13]: <xray.DataArray (a: 4, b: 3)> array([[ 0.95460352, 0.12853847, 0.05126985], [ 0.80265117, 0.46370886, 0.20577943], [ 0.33068145, 0.55748573, 0.0226791 ], [ 0.20183347, 0.46907727, 0.15943175]]) Coordinates: * a (a) int64 0 1 2 3 * b (b) int64 1 2 0 ```

If we add this to Dataset and supply coords=coords, it raises, since there are conflicting coords:

``` python In [16]:

: ds = xray.Dataset(variables={'da':da_reindex}, coords=coords) ds

ValueError Traceback (most recent call last) <ipython-input-16-8e4afa10f781> in <module>() ----> 1 ds = xray.Dataset(variables={'da':da_reindex}, coords=coords) 2 ds

/usr/local/lib/python2.7/dist-packages/xray/core/dataset.pyc in init(self, variables, coords, attrs, compat) 317 coords = set() 318 if variables or coords: --> 319 self._set_init_vars_and_dims(variables, coords, compat) 320 if attrs is not None: 321 self.attrs = attrs

/usr/local/lib/python2.7/dist-packages/xray/core/dataset.pyc in _set_init_vars_and_dims(self, vars, coords, compat) 371 aligned = _align_variables(variables) 372 new_variables, new_coord_names = _expand_variables(aligned, --> 373 compat=compat) 374 375 new_coord_names.update(coords)

/usr/local/lib/python2.7/dist-packages/xray/core/dataset.pyc in _expand_variables(raw_variables, old_variables, compat) 142 add_variable(dim, coord.variable) 143 var = var.variable --> 144 add_variable(name, var) 145 146 return new_variables, new_coord_names

/usr/local/lib/python2.7/dist-packages/xray/core/dataset.pyc in add_variable(name, var) 130 raise ValueError('conflicting value for variable %s:\n' 131 'first value: %r\nsecond value: %r' --> 132 % (name, variables[name], var)) 133 if compat == 'broadcast_equals': 134 maybe_promote_or_replace(name, var)

ValueError: conflicting value for variable b: first value: <xray.Coordinate 'b' (b: 3)> array([1, 2, 0]) second value: <xray.Coordinate 'b' (b: 3)> array([0, 1, 2]) ```

But adding with __setitem__ aligns as expected:

``` python In [17]:

ds ds = xray.Dataset(coords=coords) ds['da']=da_reindex ds Out[17]: <xray.Dataset> Dimensions: (a: 4, b: 3) Coordinates: * a (a) int64 0 1 2 3 * b (b) int64 0 1 2 Data variables: da (a, b) float64 0.05127 0.9546 0.1285 0.2058 0.8027 0.4637 ... ```

Is this desired behavior? I could imagine aligning to coords if they are explicitly supplied, but I don't know the nuances here.

So far, only the changes that unittest2pytest could fix. Need to work out how to sort through the exisiting xarray ones

EDIT: ready to merge (and keen to get this in before there are divergences from other branches)