Fix main build and windows build

[tomwhite]

• Fixes Main build is failing with Pandas 2.1.0 #1121
• Fixes Windows build is failing #1109

[tomwhite]

Unpinning xarray (which was pinned in #934) fixes the immediate problem of failing tests: https://github.com/pystatgen/sgkit/actions/runs/6071972311/job/16471074834?pr=1122.

However, this breaks the Sphinx build (maybe due to #934, or something else?), and the Windows build is still failing (see #1111 too). These both need more work.

[tomwhite]

Running the doc build locally I get:

/Users/tom/workspace/sgkit/docs/examples/gwas_tutorial.ipynb: WARNING: Executing notebook failed: CellTimeoutError [mystnb.exec]

Which suggests that #934 is still a problem.

Also,

/Users/tom/workspace/sgkit/docs/examples/gwas_tutorial.ipynb: WARNING: Executing notebook failed: CellTimeoutError [mystnb.exec]
/Users/tom/workspace/sgkit/docs/examples/gwas_tutorial.ipynb: WARNING: Notebook exception traceback saved in: /Users/tom/workspace/sgkit/docs/_build/html/reports/examples/gwas_tutorial.err.log [mystnb.exec]
WARNING: 
>>>-------------------------------------------------------------------------
Exception in /Users/tom/workspace/sgkit/docs/how_do_i.rst at block ending on line 139
Specify :okexcept: as an option in the ipython:: block to suppress this message
---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
Cell In[18], line 1
----> 1 ds2.sel(variants=(ds2.variant_hwe_p_value > 1e-2))

File ~/miniconda3/envs/sgkit-dev-3.10/lib/python3.10/site-packages/xarray/core/dataset.py:3034, in Dataset.sel(self, indexers, method, tolerance, drop, **indexers_kwargs)
   3031                 query_results.drop_coords.append(k)
   3032     query_results.variables = no_scalar_variables
-> 3034 result = self.isel(indexers=query_results.dim_indexers, drop=drop)
   3035 return result._overwrite_indexes(*query_results.as_tuple()[1:])

File ~/miniconda3/envs/sgkit-dev-3.10/lib/python3.10/site-packages/xarray/core/dataset.py:2874, in Dataset.isel(self, indexers, drop, missing_dims, **indexers_kwargs)
   2872 indexers = either_dict_or_kwargs(indexers, indexers_kwargs, "isel")
   2873 if any(is_fancy_indexer(idx) for idx in indexers.values()):
-> 2874     return self._isel_fancy(indexers, drop=drop, missing_dims=missing_dims)
   2876 # Much faster algorithm for when all indexers are ints, slices, one-dimensional
   2877 # lists, or zero or one-dimensional np.ndarray's
   2878 indexers = drop_dims_from_indexers(indexers, self.dims, missing_dims)

File ~/miniconda3/envs/sgkit-dev-3.10/lib/python3.10/site-packages/xarray/core/dataset.py:2930, in Dataset._isel_fancy(self, indexers, drop, missing_dims)
   2926 var_indexers = {
   2927     k: v for k, v in valid_indexers.items() if k in var.dims
   2928 }
   2929 if var_indexers:
-> 2930     new_var = var.isel(indexers=var_indexers)
   2931     # drop scalar coordinates
   2932     # https://github.com/pydata/xarray/issues/6554
   2933     if name in self.coords and drop and new_var.ndim == 0:

File ~/miniconda3/envs/sgkit-dev-3.10/lib/python3.10/site-packages/xarray/core/variable.py:1368, in Variable.isel(self, indexers, missing_dims, **indexers_kwargs)
   1365 indexers = drop_dims_from_indexers(indexers, self.dims, missing_dims)
   1367 key = tuple(indexers.get(dim, slice(None)) for dim in self.dims)
-> 1368 return self[key]

File ~/miniconda3/envs/sgkit-dev-3.10/lib/python3.10/site-packages/xarray/core/variable.py:891, in Variable.__getitem__(self, key)
    878 def __getitem__(self: T_Variable, key) -> T_Variable:
    879     """Return a new Variable object whose contents are consistent with
    880     getting the provided key from the underlying data.
    881 
   (...)
    889     array `x.values` directly.
    890     """
--> 891     dims, indexer, new_order = self._broadcast_indexes(key)
    892     data = as_indexable(self._data)[indexer]
    893     if new_order:

File ~/miniconda3/envs/sgkit-dev-3.10/lib/python3.10/site-packages/xarray/core/variable.py:724, in Variable._broadcast_indexes(self, key)
    721 if all(isinstance(k, BASIC_INDEXING_TYPES) for k in key):
    722     return self._broadcast_indexes_basic(key)
--> 724 self._validate_indexers(key)
    725 # Detect it can be mapped as an outer indexer
    726 # If all key is unlabeled, or
    727 # key can be mapped as an OuterIndexer.
    728 if all(not isinstance(k, Variable) for k in key):

File ~/miniconda3/envs/sgkit-dev-3.10/lib/python3.10/site-packages/xarray/core/variable.py:776, in Variable._validate_indexers(self, key)
    771     raise IndexError(
    772         "{}-dimensional boolean indexing is "
    773         "not supported. ".format(k.ndim)
    774     )
    775 if is_duck_dask_array(k.data):
--> 776     raise KeyError(
    777         "Indexing with a boolean dask array is not allowed. "
    778         "This will result in a dask array of unknown shape. "
    779         "Such arrays are unsupported by Xarray."
    780         "Please compute the indexer first using .compute()"
    781     )
    782 if getattr(k, "dims", (dim,)) != (dim,):
    783     raise IndexError(
    784         "Boolean indexer should be unlabeled or on the "
    785         "same dimension to the indexed array. Indexer is "
   (...)
    788         )
    789     )

KeyError: 'Indexing with a boolean dask array is not allowed. This will result in a dask array of unknown shape. Such arrays are unsupported by Xarray.Please compute the indexer first using .compute()'

[codecov-commenter]

Codecov Report

Merging #1122 (8c8ad28) into main (cc04858) will not change coverage.
The diff coverage is n/a.

❗ Current head 8c8ad28 differs from pull request most recent head 73151df. Consider uploading reports for the commit 73151df to get more accurate results

❗ Your organization is not using the GitHub App Integration. As a result you may experience degraded service beginning May 15th. Please install the GitHub App Integration for your organization. Read more.

@@            Coverage Diff            @@
##              main     #1122   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files           50        50           
  Lines         5054      5054           
=========================================
  Hits          5054      5054           

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[tomwhite]

This fixes the various build issues we've been having: #1121 and #1109. This is ready for review and will unblock the other PRs that are waiting.

Note that I have commented out the calls to produce scatterplots in the GWAS tutorial notebook as these can be fixed later (#934).

[jeromekelleher]

LGTM