def st_valid_inds_of_same_length(draw, v1, v2):
len1, len2 = len(v1), len(v2)
ret = hyst.just(([], []))
# TODO we should include integer arrays here by chaining `| hynp.integer_array_indices(shape=(LEN_X,))`
val1 = hynp.basic_indices(shape=(len1,), allow_ellipsis=False)
if len1 == len2:
ret = ret | hyst.tuples(hyst.shared(val1, key="st_valid_inds_of_same_length"), hyst.shared(val1, key="st_valid_inds_of_same_length"))
if len1 > 0 and len2 > 0:
val2 = hynp.basic_indices(shape=(len2,), allow_ellipsis=False)
ret = ret | hyst.tuples(val1, val2)
return draw(ret)
def get_vector_index_axis(draw, max_shape: int = 10):
N = draw(integers(min_value=2, max_value=max_shape))
index = draw(
npst.basic_indices(shape=(N, ),
min_dims=1,
max_dims=1,
allow_ellipsis=False))
axis = draw(integers(min_value=0, max_value=1))
return (N, *index, axis)
def test_basic_indices_can_generate_indices_not_covering_all_dims():
# These "flat indices" are skippable in the underlying BasicIndexStrategy,
# so we ensure we're definitely generating them for nps.basic_indices().
find_any(
nps.basic_indices(shape=(3, 3, 3)),
lambda ix: (
(not isinstance(ix, tuple) and ix != Ellipsis)
or (isinstance(ix, tuple) and Ellipsis not in ix and len(ix) < 3)
),
)
def array_and_basic_indices(
draw,
array_min_dims: int = 0,
array_max_dims: int = 2,
dtype: np.dtype = None,
):
arr = draw(
anyarray(min_dims=array_min_dims, max_dims=array_max_dims, dtype=dtype)
)
ind = draw(basic_indices(arr.shape))
return arr, ind
def st_valid_inds_of_different_length(draw, v1, v2):
len1, len2 = len(v1), len(v2)
ret = nothing()
# TODO we should include integer arrays here
val = hynp.basic_indices(shape=(len1,), allow_ellipsis=False) # | hynp.integer_array_indices(shape=(len1,))
if len1 != len2:
ret = ret | hyst.just((slice(None), slice(None))) \
| hyst.tuples(hyst.shared(val, key="indfl"), hyst.shared(val, key="indfl"))
if len1 > 0 and len2 > 0:
ret = ret | hyst.tuples(val, val).filter(lambda x: len(x[0])!=len(x[1]))
return draw(ret)
def get_chainable_arrays_shapes_and_indices(draw,
base_min_dims: int = 1,
base_max_dims: int = 2,
max_chain: int = 8,
chain_min_dims: int = 1):
shape_strategy = get_chainable_array_shapes(base_min_dims=base_min_dims,
base_max_dims=base_max_dims,
max_chain=max_chain,
chain_min_dims=chain_min_dims)
shapes = draw(shape_strategy)
shape = (shapes[0][0], shapes[-1][-1])
indices = draw(npst.basic_indices(shape, allow_ellipsis=False, min_dims=2))
return shapes, indices, shape
def test_basic_indices_generate_valid_indexers(
shape, allow_newaxis, allow_ellipsis, data
):
min_dims = data.draw(
st.integers(0, 5 if allow_newaxis else len(shape)), label="min_dims"
)
max_dims = data.draw(
st.none() | st.integers(min_dims, 32 if allow_newaxis else len(shape)),
label="max_dims",
)
indexer = data.draw(
nps.basic_indices(
shape,
min_dims=min_dims,
max_dims=max_dims,
allow_ellipsis=allow_ellipsis,
allow_newaxis=allow_newaxis,
),
label="indexer",
)
# Check that disallowed things are indeed absent
if not allow_newaxis:
if isinstance(indexer, tuple):
assert 0 <= len(indexer) <= len(shape) + int(allow_ellipsis)
else:
assert 1 <= len(shape) + int(allow_ellipsis)
assert np.newaxis not in shape
if not allow_ellipsis:
assert Ellipsis not in shape
if 0 in shape:
# If there's a zero in the shape, the array will have no elements.
array = np.zeros(shape)
assert array.size == 0
elif np.prod(shape) <= 10**5:
# If it's small enough to instantiate, do so with distinct elements.
array = np.arange(np.prod(shape)).reshape(shape)
else:
# We can't cheat on this one, so just try another.
assume(False)
view = array[indexer]
if not np.isscalar(view):
assert min_dims <= view.ndim <= (32 if max_dims is None else max_dims)
if view.size:
assert np.shares_memory(view, array)
def get_boardcastable_arrays_shapes_and_indices(draw,
base_min_dims: int = 2,
base_max_dims: int = 3,
broadcast_min_dims: int = 1,
broadcast_max_dims: int = 5):
shape_strategy = get_boardcastable_arrays_shapes(
base_min_dims=base_min_dims,
base_max_dims=base_max_dims,
broadcast_min_dims=broadcast_min_dims,
broadcast_max_dims=broadcast_max_dims)
shapes = draw(shape_strategy)
# What is final shape?
ndim = max(len(shape) for shape in shapes)
padded_shapes = np.array([[1] * (ndim - len(shape)) + list(shape)
for shape in shapes])
shape = tuple(np.max(padded_shapes, axis=0).astype(int))
# TODO: raise Issue with hypothesis. Numpy types aren't allowed. uncomment next line and it fails
shape = tuple(int(s) for s in shape)
indices = draw(npst.basic_indices(shape, allow_ellipsis=False, min_dims=2))
return shapes, indices, shape
def test_basic_indices_can_generate_empty_tuple():
find_any(nps.basic_indices(shape=(0, 0), allow_ellipsis=True),
lambda ix: ix == ())
def test_basic_indices_can_generate_long_ellipsis():
# Runs of slice(None) - such as [0,:,:,:,0] - can be replaced by e.g. [0,...,0]
find_any(
nps.basic_indices(shape=(1, 0, 0, 0, 1), allow_ellipsis=True),
lambda ix: len(ix) == 3 and ix[1] == Ellipsis,
)
@given(
nps.basic_indices(shape=(0, 0, 0, 0,
0)).filter(lambda idx: Ellipsis in idx))
def test_basic_indices_replaces_whole_axis_slices_with_ellipsis(idx):
# If ... is in the slice, it replaces all ,:, entries for this shape.
assert slice(None) not in idx
@given(
shape=nps.array_shapes(min_dims=0, max_side=4)
| nps.array_shapes(min_dims=0, min_side=0, max_side=10),
min_dims=st.integers(0, 5),
allow_ellipsis=st.booleans(),
allow_newaxis=st.booleans(),
data=st.data(),
)
def test_basic_indices_generate_valid_indexers(shape, min_dims, allow_ellipsis,
allow_newaxis, data):
def arbitrary_indices(draw, shape: Tuple[int]):
"""
Hypothesis search strategy: Generate a valid index
for an array of a given shape. The index can contain
any type of valid object used for indexing, including
integers, slices, Ellipsis's, newaxis's, boolean arrays,
and integer arrays.
Parameters
----------
shape : Tuple[int]
The shape of the array to be indexed into
Notes
-----
`draw` is a parameter reserved by hypothesis, and should not be specified
by the user.
When given a shape with a 0-dimensional axis, only a basic index will be returned.
Returns
-------
hypothesis.searchstrategy.SearchStrategy[Tuple[Union[int, slice, Ellipsis, NoneType, numpy.ndarray], ...]]
"""
def group_continuous_integers(ls):
"""
Given a list of integers, find and group continuous sequences
Parameters
----------
ls: List[int]
Returns
-------
List[Tuple[int]]
Examples
--------
>>> group_continuous_integers([1, 3, 4, 5, 7, 8])
[(1,), (3, 4, 5), (7, 8)]
"""
return [
tuple(map(itemgetter(1), g))
for k, g in groupby(enumerate(ls), lambda x: x[0] - x[1])
]
if not shape or 0 in shape:
return draw(hnp.basic_indices(shape=shape, allow_newaxis=True))
shape_inds = list(range(len(shape)))
index = [] # stores tuples of (axis, indexing object)
# add integers, slices
basic_inds = sorted(
draw(st.lists(st.sampled_from(shape_inds), unique=True)))
if len(basic_inds) > 0:
basic_dims = tuple(shape[i] for i in basic_inds)
# only draw ints and slices
# will handle possible ellipsis and newaxis objects later
# as these can make array indices difficult to handle
basics = draw(hnp.basic_indices(shape=basic_dims,
allow_ellipsis=False))
if not isinstance(basics, tuple):
basics = (basics, )
index += [tup for tup in zip(basic_inds, basics)]
# will not necessarily index all axes from basic_inds as
# `basic_indices` can return indices with omitted trailing slices
# so only remove dimensions directly indexed into
for i in basic_inds[:len(basics)]:
shape_inds.pop(shape_inds.index(i))
if len(shape_inds) > 0:
# add integer arrays to index
int_arr_inds = sorted(
draw(st.lists(st.sampled_from(shape_inds), unique=True)))
if len(int_arr_inds) > 0:
int_arr_dims = tuple(shape[i] for i in int_arr_inds)
int_arrs = draw(hnp.integer_array_indices(shape=int_arr_dims))
index += [tup for tup in zip(int_arr_inds, int_arrs)]
for i in int_arr_inds:
shape_inds.pop(shape_inds.index(i))
if len(shape_inds) > 0:
# add boolean arrays to index
bool_inds = sorted(
draw(st.lists(st.sampled_from(shape_inds), unique=True)))
if len(bool_inds) > 0:
# boolean arrays can be multi-dimensional, so by grouping all
# adjacent axes to make a single boolean array, this can be tested for
grouped_bool_inds = group_continuous_integers(bool_inds)
bool_dims = [
tuple(shape[i] for i in ind) for ind in grouped_bool_inds
]
# if multiple boolean array indices, the number of trues must be such that
# the output of ind.nonzero() for each index are broadcast compatible
# this must also be the same as the trailing dim of each integer array, if any used
if len(int_arr_inds):
max_trues = max(i.shape[-1] for i in int_arrs)
else:
max_trues = st.integers(min_value=0,
max_value=min(
bool_dims,
key=lambda x: np.prod(x)))
index += [(
i[0],
draw(
hnp.arrays(shape=sh, dtype=bool).filter(
lambda x: x.sum() in (1, max_trues))),
) for i, sh in zip(grouped_bool_inds, bool_dims)]
for i in bool_inds:
shape_inds.pop(shape_inds.index(i))
grouped_shape_inds = group_continuous_integers(sorted(shape_inds))
if len(grouped_shape_inds) == 1:
# unused indices form a continuous stretch of dimensions
# so can replace with an ellipsis
# to test ellipsis vs omitted slices, randomly
# add ellipsis when the unused axes are trailing
if max(shape_inds) + 1 == len(shape):
if draw(st.booleans()):
index += [(min(shape_inds), Ellipsis)]
else:
index += [(min(shape_inds), Ellipsis)]
elif len(grouped_shape_inds) == 0 and draw(st.booleans()):
# all indices filled already
# can randomly add ellipsis that expands to 0-d tuple
# this can have counter-intuitive behavior
# (particularly in conjunction with array indices)
i = draw(st.integers(min_value=0, max_value=len(index)))
index.insert(i, (i, Ellipsis))
else:
# so that current chosen index's work,
# fill in remaining any gaps with empty slices
index += [(i, slice(None)) for i in shape_inds]
index = sorted(index, key=lambda x: x[0])
# can now randomly add in newaxis objects
newaxis_pos = sorted(
draw(
st.lists(st.integers(min_value=0, max_value=len(index)),
unique=True)),
reverse=True,
)
for i in newaxis_pos:
index.insert(i, (-1, np.newaxis))
out_ind = tuple(i[1] for i in index)
return out_ind
def test_basic_indices_can_generate_long_ellipsis():
# Runs of slice(None) - such as [0,:,:,:,0] - can be replaced by e.g. [0,...,0]
find_any(
nps.basic_indices(shape=(1, 0, 0, 0, 1), allow_ellipsis=True),
lambda ix: len(ix) == 3 and ix[1] == Ellipsis,
)
def test_basic_indices_can_generate_non_tuples():
find_any(
nps.basic_indices(shape=(0, 0), allow_ellipsis=True),
lambda ix: not isinstance(ix, tuple),
)
def test_basic_indices_can_generate_empty_tuple():
find_any(nps.basic_indices(shape=(0, 0), allow_ellipsis=True), lambda ix: ix == ())
def test_basic_indices_options(condition):
indexers = nps.array_shapes(min_dims=0, max_dims=32).flatmap(
lambda shape: nps.basic_indices(shape, allow_newaxis=True)
)
find_any(indexers, condition)
nps.basic_indices(shape=(0, 0), allow_ellipsis=True),
lambda ix: not isinstance(ix, tuple),
)
def test_basic_indices_can_generate_long_ellipsis():
# Runs of slice(None) - such as [0,:,:,:,0] - can be replaced by e.g. [0,...,0]
find_any(
nps.basic_indices(shape=(1, 0, 0, 0, 1), allow_ellipsis=True),
lambda ix: len(ix) == 3 and ix[1] == Ellipsis,
)
@given(
nps.basic_indices(shape=(0, 0, 0, 0, 0)).filter(
lambda idx: isinstance(idx, tuple) and Ellipsis in idx
)
)
def test_basic_indices_replaces_whole_axis_slices_with_ellipsis(idx):
# `slice(None)` (aka `:`) is the only valid index for an axis of size
# zero, so if all dimensions are 0 then a `...` will replace all the
# slices because we generate `...` for entire contiguous runs of `:`
assert slice(None) not in idx
def test_basic_indices_can_generate_indices_not_covering_all_dims():
# These "flat indices" are skippable in the underlying BasicIndexStrategy,
# so we ensure we're definitely generating them for nps.basic_indices().
find_any(
nps.basic_indices(shape=(3, 3, 3)),
lambda ix: (
def test_test_basic_indices_kwonly_emulation():
with pytest.raises(InvalidArgument):
nps.basic_indices((), 0, 1).validate()
with pytest.raises(InvalidArgument):
nps.basic_indices((), __reserved=None).validate()
def test_basic_indices_bad_min_dims_warns():
with pytest.warns(HypothesisDeprecationWarning):
with pytest.raises(InvalidArgument):
nps.basic_indices((3, 3, 3), min_dims=4).example()
def test_basic_indices_default_max_dims_does_not_warn():
with catch_warnings(record=True) as record:
nps.basic_indices((3, 3, 3)).example()
nps.basic_indices((3, 3, 3), allow_newaxis=True).example()
assert len(record) == 0
def test_basic_indices_bad_max_dims_warns():
with pytest.warns(HypothesisDeprecationWarning):
nps.basic_indices((3, 3, 3), max_dims=4).example()
