Bump black to 23.1.0

.pre-commit-config.yaml

@@ -21,7 +21,7 @@ repos:
         args:
           - --py38-plus
   - repo: https://github.com/psf/black
-    rev: 22.10.0
+    rev: 23.1.0
     hooks:
       - id: black
         language_version: python3

dask/array/backends.py

@@ -152,7 +152,6 @@ def _cupy_einsum(*args, **kwargs):
 @tensordot_lookup.register_lazy("cupyx")
 @concatenate_lookup.register_lazy("cupyx")
 def register_cupyx():
-
     from cupyx.scipy.sparse import spmatrix
 
     try:

dask/array/core.py

@@ -5304,7 +5304,7 @@ def dtype(x):
 
     result = np.empty(shape=shape, dtype=dtype(deepfirst(arrays)))
 
-    for (idx, arr) in zip(
+    for idx, arr in zip(
         slices_from_chunks(chunks), core.flatten(arrays, container=(list, tuple))
     ):
         if hasattr(arr, "ndim"):

dask/array/optimization.py

@@ -340,7 +340,6 @@ def fuse_slice(a, b):
     # If given two tuples walk through both, being mindful of uneven sizes
     # and newaxes
     if isinstance(a, tuple) and isinstance(b, tuple):
-
         # Check for non-fusible cases with fancy-indexing
         a_has_lists = any(isinstance(item, list) for item in a)
         b_has_lists = any(isinstance(item, list) for item in b)

dask/array/percentile.py

@@ -45,7 +45,6 @@ def _percentile(a, q, method="linear"):
 
 
 def _tdigest_chunk(a):
-
     from crick import TDigest
 
     t = TDigest()
@@ -55,7 +54,6 @@ def _tdigest_chunk(a):
 
 
 def _percentiles_from_tdigest(qs, digests):
-
     from crick import TDigest
 
     t = TDigest()
@@ -154,7 +152,6 @@ def percentile(a, q, method="linear", internal_method="default", **kwargs):
         and method == "linear"
         and (np.issubdtype(dtype, np.floating) or np.issubdtype(dtype, np.integer))
     ):
-
         from dask.utils import import_required
 
         import_required(

dask/array/routines.py

@@ -1948,7 +1948,6 @@ def squeeze(a, axis=None):
 
 @derived_from(np)
 def compress(condition, a, axis=None):
-
     if not is_arraylike(condition):
         # Allow `condition` to be anything array-like, otherwise ensure `condition`
         # is a numpy array.

dask/array/slicing.py

@@ -1889,7 +1889,6 @@ def value_indices_from_1d_int_index(dim, vsize, loc0, loc1):
     dsk = {}
     out_name = (out_name,)
     for in_key, locations in zip(in_keys, array_locations):
-
         # Now loop round each block dimension.
         #
         # If the block overlaps the indices then set the following
@@ -1922,7 +1921,6 @@ def value_indices_from_1d_int_index(dim, vsize, loc0, loc1):
         dim_1d_int_index = None
 
         for dim, (index, (loc0, loc1)) in enumerate(zip(indices, locations)):
-
             integer_index = isinstance(index, int)
             if isinstance(index, slice):
                 # Index is a slice

dask/array/stats.py

@@ -150,7 +150,6 @@ def chisquare(f_obs, f_exp=None, ddof=0, axis=0):
 
 @derived_from(scipy.stats)
 def power_divergence(f_obs, f_exp=None, ddof=0, axis=0, lambda_=None):
-
     if isinstance(lambda_, str):
         if lambda_ not in _power_div_lambda_names:
             names = repr(list(_power_div_lambda_names.keys()))[1:-1]
@@ -405,7 +404,6 @@ def _unequal_var_ttest_denom(v1, n1, v2, n2):
 
 
 def _ttest_ind_from_stats(mean1, mean2, denom, df):
-
     d = mean1 - mean2
     with np.errstate(divide="ignore", invalid="ignore"):
         t = da.divide(d, denom)

dask/array/tests/test_array_core.py

@@ -607,7 +607,6 @@ def test_concatenate_fixlen_strings():
 
 
 def test_concatenate_zero_size():
-
     x = np.random.random(10)
     y = da.from_array(x, chunks=3)
     result_np = np.concatenate([x, x[:0]])
@@ -2650,9 +2649,7 @@ def test_from_array_scalar(type_):
     dx = da.from_array(x, chunks=-1)
     assert_eq(np.array(x), dx)
     assert isinstance(
-        dx.dask[
-            dx.name,
-        ],
+        dx.dask[dx.name,],
         np.ndarray,
     )
 
@@ -5332,7 +5329,6 @@ def test_chunk_non_array_like():
 def test_to_backend():
     # Test that `Array.to_backend` works as expected
     with dask.config.set({"array.backend": "numpy"}):
-
         # Start with numpy-backed array
         x = da.ones(10)
         assert isinstance(x._meta, np.ndarray)

dask/array/tests/test_creation.py

@@ -40,7 +40,6 @@ def test_arr_like(
 ):
     backend_lib = pytest.importorskip(backend)
     with dask.config.set({"array.backend": backend}):
-
         np_func = getattr(backend_lib, funcname)
         da_func = getattr(da, funcname)
         shape = cast_shape(shape)

dask/array/tests/test_cupy_creation.py

@@ -173,7 +173,6 @@ def test_tri_like(xp, N, M, k, dtype, chunks):
 def test_to_backend_cupy():
     # Test that `Array.to_backend` works as expected
     with config.set({"array.backend": "numpy"}):
-
         # Start with cupy-backed array
         x = da.from_array(cupy.arange(11), chunks=(4,))
         assert isinstance(x._meta, cupy.ndarray)
@@ -190,7 +189,6 @@ def test_to_backend_cupy():
 
         # Change global "array.backend" config to `cupy`
         with config.set({"array.backend": "cupy"}):
-
             # Calling `to_backend("numpy")` should
             # always move the data to `numpy`
             x_new = x.to_backend("numpy")

dask/array/tests/test_linalg.py

@@ -815,7 +815,6 @@ def test_solve_assume_a(shape, chunk):
 
 @pytest.mark.parametrize(("shape", "chunk"), [(20, 10), (12, 3), (30, 3), (30, 6)])
 def test_cholesky(shape, chunk):
-
     A = _get_symmat(shape)
     dA = da.from_array(A, (chunk, chunk))
     assert_eq(

dask/array/tests/test_masked.py

@@ -171,7 +171,7 @@ def test_creation_functions():
     dy = da.from_array(y, chunks=4)
 
     sol = np.ma.masked_greater(x, y)
-    for (a, b) in product([dx, x], [dy, y]):
+    for a, b in product([dx, x], [dy, y]):
         assert_eq(da.ma.masked_greater(a, b), sol)
 
     # These are all the same as masked_greater, just check for correct op

dask/array/tests/test_optimization.py

@@ -106,7 +106,6 @@ def _assert_getter_dsk_eq(a, b):
 
 
 def test_fuse_getitem(getter, getter_nofancy, getitem):
-
     pairs = [
         (
             (getter, (getter, "x", slice(1000, 2000)), slice(15, 20)),

dask/array/tests/test_random.py

@@ -274,7 +274,7 @@ def test_choice():
         (4, [0.2, 0.2, 0.3]),
     ]  # p must sum to 1
 
-    for (a, p) in errs:
+    for a, p in errs:
         with pytest.raises(ValueError):
             da.random.choice(a, size=size, chunks=chunks, p=p)
 

dask/array/tests/test_rechunk.py

@@ -539,7 +539,6 @@ def test_intersect_nan_single():
 
 
 def test_intersect_nan_long():
-
     old_chunks = (tuple([float("nan")] * 4), (10,))
     new_chunks = (tuple([float("nan")] * 4), (5, 5))
     result = list(intersect_chunks(old_chunks, new_chunks))

dask/array/tests/test_reductions.py

@@ -286,7 +286,6 @@ def test_nan_reduction_warnings(dfunc, func):
     ["dfunc", "func"], [(da.nanargmin, np.nanargmin), (da.nanargmax, np.nanargmax)]
 )
 def test_nanarg_reductions(dfunc, func):
-
     x = np.random.random((10, 10, 10))
     x[5] = np.nan
     a = da.from_array(x, chunks=(3, 4, 5))

dask/array/tests/test_routines.py

@@ -2178,7 +2178,6 @@ def test_coarsen_bad_chunks(chunks):
     ],
 )
 def test_aligned_coarsen_chunks(chunks, divisor):
-
     from dask.array.routines import aligned_coarsen_chunks as acc
 
     aligned_chunks = acc(chunks, divisor)

dask/array/tests/test_ufunc.py

@@ -252,7 +252,6 @@ def test_ufunc_outer():
 
 @pytest.mark.parametrize("ufunc", ["isreal", "iscomplex", "real", "imag"])
 def test_complex(ufunc):
-
     dafunc = getattr(da, ufunc)
     # Note that these functions are not NumPy ufuncs
     npfunc = getattr(np, ufunc)
@@ -282,7 +281,6 @@ def test_complex(ufunc):
 
 @pytest.mark.parametrize("ufunc", ["frexp", "modf"])
 def test_ufunc_2results(ufunc):
-
     dafunc = getattr(da, ufunc)
     npfunc = getattr(np, ufunc)
 

dask/blockwise.py

@@ -734,7 +734,7 @@ def _cull_dependencies(self, all_hlg_keys, output_blocks):
 
         # Gather constant dependencies (for all output keys)
         const_deps = set()
-        for (arg, ind) in self.indices:
+        for arg, ind in self.indices:
             if ind is None:
                 try:
                     if arg in all_hlg_keys:
@@ -1174,7 +1174,6 @@ def make_blockwise_graph(
             key_deps[out_key] = deps
 
     if return_key_deps:
-
         # Add valid-key dependencies from io_deps
         for key, io_dep in io_deps.items():
             if io_dep.produces_keys:

dask/bytes/tests/test_local.py

@@ -47,7 +47,6 @@ def to_uri(path):
 
 
 def test_unordered_urlpath_errors():
-
     # Unordered urlpath argument
     with pytest.raises(TypeError):
         read_bytes(

dask/bytes/tests/test_s3.py

@@ -492,7 +492,6 @@ def test_parquet(s3, engine, s3so, metadata_file):
     # Check that `open_file_options` arguments are
     # really passed through to fsspec
     if fsspec_parquet:
-
         # Passing `open_file_options` kwargs will fail
         # if you set an unsupported engine
         with pytest.raises(ValueError):

dask/dataframe/accessor.py

@@ -291,7 +291,6 @@ def cat(self, others=None, sep=None, na_rep=None):
         if others is None:
 
             def str_cat_none(x):
-
                 if isinstance(x, (Series, Index)):
                     x = x.compute()
 

dask/dataframe/core.py

@@ -2948,7 +2948,6 @@ def describe(
             datetime_is_numeric_kwarg = {}
 
         if self._meta.ndim == 1:
-
             meta = self._meta_nonempty.describe(
                 percentiles=percentiles,
                 include=include,
@@ -3772,7 +3771,6 @@ def rename(self, index=None, inplace=False, sorted_index=False):
             and not is_dict_like(index)
             and not isinstance(index, dd.Series)
         ):
-
             if inplace:
                 warnings.warn(
                     "'inplace' argument for dask series will be removed in future versions",
@@ -4695,7 +4693,6 @@ def empty(self):
     def __getitem__(self, key):
         name = "getitem-%s" % tokenize(self, key)
         if np.isscalar(key) or isinstance(key, (tuple, str)):
-
             if isinstance(self._meta.index, (pd.DatetimeIndex, pd.PeriodIndex)):
                 if key not in self._meta.columns:
                     if PANDAS_GT_120:
@@ -5032,7 +5029,6 @@ def set_index(
 
         # Check other can be translated to column name or column object, possibly flattening it
         if not isinstance(other, str):
-
             # It may refer to several columns
             if isinstance(other, Sequence):  # type: ignore[unreachable]
                 # Accept ["a"], but not [["a"]]
@@ -7070,7 +7066,6 @@ def quantile(df, q, method="default"):
     if internal_method == "tdigest" and (
         np.issubdtype(df.dtype, np.floating) or np.issubdtype(df.dtype, np.integer)
     ):
-
         from dask.utils import import_required
 
         import_required(
@@ -7090,7 +7085,6 @@ def quantile(df, q, method="default"):
             (name2, 0): finalize_tsk((_percentiles_from_tdigest, qs, sorted(val_dsk)))
         }
     else:
-
         from dask.array.dispatch import percentile_lookup as _percentile
         from dask.array.percentile import merge_percentiles
 

dask/dataframe/groupby.py

@@ -940,7 +940,7 @@ def _build_agg_args(spec):
 
     # a partial may contain some arguments, pass them down
     # https://github.com/dask/dask/issues/9615
-    for (result_column, func, input_column) in spec:
+    for result_column, func, input_column in spec:
         func_args = ()
         func_kwargs = {}
         if isinstance(func, partial):
@@ -1389,7 +1389,6 @@ def __init__(
         sort=True,
         observed=False,
     ):
-
         by_ = by if isinstance(by, (tuple, list)) else [by]
         if any(isinstance(key, pd.Grouper) for key in by_):
             raise NotImplementedError("pd.Grouper is currently not supported by Dask.")

dask/dataframe/indexing.py

@@ -47,7 +47,6 @@ def _meta_indexer(self):
         return self.obj._meta.iloc
 
     def __getitem__(self, key):
-
         # dataframe
         msg = (
             "'DataFrame.iloc' only supports selecting columns. "
@@ -87,7 +86,6 @@ def _meta_indexer(self):
         return self.obj._meta.loc
 
     def __getitem__(self, key):
-
         if isinstance(key, tuple):
             # multi-dimensional selection
             if len(key) > self.obj.ndim:

dask/dataframe/io/csv.py

@@ -97,7 +97,6 @@ def project_columns(self, columns):
         )
 
     def __call__(self, part):
-
         # Part will be a 3-element tuple
         block, path, is_first, is_last = part
 

dask/dataframe/io/hdf.py

@@ -470,13 +470,11 @@ def _build_parts(paths, key, start, stop, chunksize, sorted_index, mode):
     parts = []
     global_divisions = []
     for path in paths:
-
         keys, stops, divisions = _get_keys_stops_divisions(
             path, key, stop, sorted_index, chunksize, mode
         )
 
         for k, s, d in zip(keys, stops, divisions):
-
             if d and global_divisions:
                 global_divisions = global_divisions[:-1] + d
             elif d:

dask/dataframe/io/orc/arrow.py

@@ -16,7 +16,6 @@ def read_metadata(
         aggregate_files,
         **kwargs,
     ):
-
         # Convert root directory to file list.
         # TODO: Handle hive-partitioned data
         if len(paths) == 1 and not fs.isfile(paths[0]):