Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

VectorData Refactor Expandable #1158

Merged
merged 22 commits into from
Aug 27, 2024
Merged
Show file tree
Hide file tree
Changes from 5 commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
38 changes: 30 additions & 8 deletions src/hdmf/backends/hdf5/h5tools.py
Original file line number Diff line number Diff line change
Expand Up @@ -383,7 +383,10 @@ def copy_file(self, **kwargs):
'default': True},
{'name': 'herd', 'type': 'hdmf.common.resources.HERD',
'doc': 'A HERD object to populate with references.',
'default': None})
'default': None},
{'name': 'expandable', 'type': bool, 'default': True,
'doc': ('Bool to set whether datasets are expandable by setting the max shape for all dimensions',
'of a dataset to None and enabling auto-chunking by default.')})
def write(self, **kwargs):
"""Write the container to an HDF5 file."""
if self.__mode == 'r':
Expand Down Expand Up @@ -826,10 +829,16 @@ def close_linked_files(self):
'doc': 'exhaust DataChunkIterators one at a time. If False, exhaust them concurrently',
'default': True},
{'name': 'export_source', 'type': str,
'doc': 'The source of the builders when exporting', 'default': None})
'doc': 'The source of the builders when exporting', 'default': None},
{'name': 'expandable', 'type': bool, 'default': True,
'doc': ('Bool to set whether datasets are expandable by setting the max shape for all dimensions',
'of a dataset to None and enabling auto-chunking by default.')})
def write_builder(self, **kwargs):
f_builder = popargs('builder', kwargs)
link_data, exhaust_dci, export_source = getargs('link_data', 'exhaust_dci', 'export_source', kwargs)
link_data, exhaust_dci, export_source = getargs('link_data',
'exhaust_dci',
'export_source',
kwargs)
self.logger.debug("Writing GroupBuilder '%s' to path '%s' with kwargs=%s"
% (f_builder.name, self.source, kwargs))
for name, gbldr in f_builder.groups.items():
Expand Down Expand Up @@ -1000,6 +1009,9 @@ def _filler():
'default': True},
{'name': 'export_source', 'type': str,
'doc': 'The source of the builders when exporting', 'default': None},
{'name': 'expandable', 'type': bool, 'default': True,
'doc': ('Bool to set whether datasets are expandable by setting the max shape for all dimensions',
'of a dataset to None and enabling auto-chunking by default.')},
returns='the Group that was created', rtype=Group)
def write_group(self, **kwargs):
parent, builder = popargs('parent', 'builder', kwargs)
Expand Down Expand Up @@ -1100,21 +1112,25 @@ def write_link(self, **kwargs):
'default': True},
{'name': 'export_source', 'type': str,
'doc': 'The source of the builders when exporting', 'default': None},
{'name': 'expandable', 'type': bool, 'default': True,
'doc': ('Bool to set whether datasets are expandable by setting the max shape for all dimensions',
'of a dataset to None and enabling auto-chunking by default.')},
returns='the Dataset that was created', rtype=Dataset)
def write_dataset(self, **kwargs): # noqa: C901
""" Write a dataset to HDF5

The function uses other dataset-dependent write functions, e.g,
``__scalar_fill__``, ``__list_fill__``, and ``__setup_chunked_dset__`` to write the data.
"""
parent, builder = popargs('parent', 'builder', kwargs)
parent, builder, expandable = popargs('parent', 'builder', 'expandable', kwargs)
link_data, exhaust_dci, export_source = getargs('link_data', 'exhaust_dci', 'export_source', kwargs)
self.logger.debug("Writing DatasetBuilder '%s' to parent group '%s'" % (builder.name, parent.name))
if self.get_written(builder):
self.logger.debug(" DatasetBuilder '%s' is already written" % builder.name)
return None
name = builder.name
data = builder.data
matched_spec_shape = builder.spec_shapes
dataio = None
options = dict() # dict with additional
if isinstance(data, H5DataIO):
Expand Down Expand Up @@ -1228,7 +1244,7 @@ def _filler():
return
# If the compound data type contains only regular data (i.e., no references) then we can write it as usual
else:
dset = self.__list_fill__(parent, name, data, options)
dset = self.__list_fill__(parent, name, data, matched_spec_shape, expandable, options)
# Write a dataset containing references, i.e., a region or object reference.
# NOTE: we can ignore options['io_settings'] for scalar data
elif self.__is_ref(options['dtype']):
Expand Down Expand Up @@ -1323,7 +1339,7 @@ def _filler():
self.__dci_queue.append(dataset=dset, data=data)
# Write a regular in memory array (e.g., numpy array, list etc.)
elif hasattr(data, '__len__'):
dset = self.__list_fill__(parent, name, data, options)
dset = self.__list_fill__(parent, name, data, matched_spec_shape, expandable, options)
# Write a regular scalar dataset
else:
dset = self.__scalar_fill__(parent, name, data, options)
Expand Down Expand Up @@ -1451,7 +1467,7 @@ def __chunked_iter_fill__(cls, parent, name, data, options=None):
return dset

@classmethod
def __list_fill__(cls, parent, name, data, options=None):
def __list_fill__(cls, parent, name, data, matched_spec_shape, expandable, options=None):
# define the io settings and data type if necessary
io_settings = {}
dtype = None
Expand All @@ -1473,7 +1489,13 @@ def __list_fill__(cls, parent, name, data, options=None):
data_shape = (len(data),)
else:
data_shape = get_data_shape(data)

if expandable:
# Don't override existing settings
if 'maxshape' not in io_settings:
if matched_spec_shape is not None:
io_settings['maxshape'] = matched_spec_shape
mavaylon1 marked this conversation as resolved.
Show resolved Hide resolved
else:
io_settings['maxshape'] = tuple([None]*len(data_shape))
# Create the dataset
try:
dset = parent.create_dataset(name, shape=data_shape, dtype=dtype, **io_settings)
Expand Down
16 changes: 12 additions & 4 deletions src/hdmf/build/builders.py
Original file line number Diff line number Diff line change
Expand Up @@ -330,6 +330,9 @@ class DatasetBuilder(BaseBuilder):
'doc': 'The datatype of this dataset.', 'default': None},
{'name': 'attributes', 'type': dict,
'doc': 'A dictionary of attributes to create in this dataset.', 'default': dict()},
{'name': 'spec_shapes', 'type': tuple,
'doc': ('The shape(s) defined in the spec.'),
'default': None},
{'name': 'dimension_labels', 'type': tuple,
'doc': ('A list of labels for each dimension of this dataset from the spec. Currently this is '
'supplied only on build.'),
Expand All @@ -341,22 +344,27 @@ class DatasetBuilder(BaseBuilder):
{'name': 'source', 'type': str, 'doc': 'The source of the data in this builder.', 'default': None})
def __init__(self, **kwargs):
""" Create a Builder object for a dataset """
name, data, dtype, attributes, dimension_labels, maxshape, chunks, parent, source = getargs(
'name', 'data', 'dtype', 'attributes', 'dimension_labels', 'maxshape', 'chunks', 'parent', 'source',
kwargs
)
name, data, dtype, attributes, spec_shapes, dimension_labels, maxshape, chunks, parent, source = getargs(
'name', 'data', 'dtype', 'attributes', 'spec_shapes', 'dimension_labels', 'maxshape', 'chunks',
'parent', 'source', kwargs)
super().__init__(name, attributes, parent, source)
self['data'] = data
self['attributes'] = _copy.copy(attributes)
self.__dimension_labels = dimension_labels
self.__chunks = chunks
self.__spec_shapes = spec_shapes
self.__maxshape = maxshape
if isinstance(data, BaseBuilder):
if dtype is None:
dtype = self.OBJECT_REF_TYPE
self.__dtype = dtype
self.__name = name

@property
def spec_shapes(self):
"""The shapes defined in the spec."""
return self.__spec_shapes

@property
def data(self):
"""The data stored in the dataset represented by this builder."""
Expand Down
23 changes: 14 additions & 9 deletions src/hdmf/build/objectmapper.py
Original file line number Diff line number Diff line change
Expand Up @@ -723,7 +723,7 @@ def build(self, **kwargs):
msg = "'container' must be of type Data with DatasetSpec"
raise ValueError(msg)
spec_dtype, spec_shape, spec_dims, spec = self.__check_dset_spec(self.spec, spec_ext)
dimension_labels = self.__get_dimension_labels_from_spec(container.data, spec_shape, spec_dims)
dimension_labels, matched_shape = self.__get_spec_info(container.data, spec_shape, spec_dims)
if isinstance(spec_dtype, RefSpec):
self.logger.debug("Building %s '%s' as a dataset of references (source: %s)"
% (container.__class__.__name__, container.name, repr(source)))
Expand All @@ -734,6 +734,7 @@ def build(self, **kwargs):
parent=parent,
source=source,
dtype=spec_dtype.reftype,
spec_shapes=matched_shape,
dimension_labels=dimension_labels,
)
manager.queue_ref(self.__set_dataset_to_refs(builder, spec_dtype, spec_shape, container, manager))
Expand All @@ -748,6 +749,7 @@ def build(self, **kwargs):
parent=parent,
source=source,
dtype=spec_dtype,
spec_shapes=matched_shape,
dimension_labels=dimension_labels,
)
manager.queue_ref(self.__set_compound_dataset_to_refs(builder, spec, spec_dtype, container,
Expand All @@ -766,6 +768,7 @@ def build(self, **kwargs):
parent=parent,
source=source,
dtype="object",
spec_shapes=matched_shape,
dimension_labels=dimension_labels,
)
manager.queue_ref(self.__set_untyped_dataset_to_refs(builder, container, manager))
Expand All @@ -789,6 +792,7 @@ def build(self, **kwargs):
parent=parent,
source=source,
dtype=dtype,
spec_shapes=matched_shape,
dimension_labels=dimension_labels,
)

Expand Down Expand Up @@ -820,9 +824,10 @@ def __check_dset_spec(self, orig, ext):
spec = ext
return dtype, shape, dims, spec

def __get_dimension_labels_from_spec(self, data, spec_shape, spec_dims) -> tuple:
def __get_spec_info(self, data, spec_shape, spec_dims):
"""This will return the dimension labels and shape by matching the data shape to a permissible spec shape."""
if spec_shape is None or spec_dims is None:
return None
return None, None
data_shape = get_data_shape(data)
# if shape is a list of allowed shapes, find the index of the shape that matches the data
if isinstance(spec_shape[0], list):
Expand All @@ -842,22 +847,22 @@ def __get_dimension_labels_from_spec(self, data, spec_shape, spec_dims) -> tuple
# use the most specific match -- the one with the fewest Nones
if match_shape_inds:
if len(match_shape_inds) == 1:
return tuple(spec_dims[match_shape_inds[0]])
return tuple(spec_dims[match_shape_inds[0]]), tuple(spec_shape[match_shape_inds[0]])
else:
count_nones = [len([x for x in spec_shape[k] if x is None]) for k in match_shape_inds]
index_min_count = count_nones.index(min(count_nones))
best_match_ind = match_shape_inds[index_min_count]
return tuple(spec_dims[best_match_ind])
return tuple(spec_dims[best_match_ind]), tuple(spec_shape[best_match_ind])
else:
# no matches found
msg = "Shape of data does not match any allowed shapes in spec '%s'" % self.spec.path
warnings.warn(msg, IncorrectDatasetShapeBuildWarning)
return None
return None, None
else:
if len(data_shape) != len(spec_shape):
msg = "Shape of data does not match shape in spec '%s'" % self.spec.path
warnings.warn(msg, IncorrectDatasetShapeBuildWarning)
return None
return None, None
# check each dimension. None means any length is allowed
match = True
for j, d in enumerate(data_shape):
Expand All @@ -867,9 +872,9 @@ def __get_dimension_labels_from_spec(self, data, spec_shape, spec_dims) -> tuple
if not match:
msg = "Shape of data does not match shape in spec '%s'" % self.spec.path
warnings.warn(msg, IncorrectDatasetShapeBuildWarning)
return None
return None, None
# shape is a single list of allowed dimension lengths
return tuple(spec_dims)
return tuple(spec_dims), tuple(spec_shape)

def __is_reftype(self, data):
if (isinstance(data, AbstractDataChunkIterator) or
Expand Down
2 changes: 1 addition & 1 deletion tests/unit/test_io_hdf5.py
Original file line number Diff line number Diff line change
Expand Up @@ -225,5 +225,5 @@ def test_dataset_shape(self):
io.write_builder(self.builder)
builder = io.read_builder()
dset = builder['test_bucket']['foo_holder']['foo1']['my_data'].data
self.assertEqual(get_data_shape(dset), (10,))
self.assertEqual(get_data_shape(dset), (None,))
io.close()
27 changes: 25 additions & 2 deletions tests/unit/test_io_hdf5_h5tools.py
Original file line number Diff line number Diff line change
Expand Up @@ -28,6 +28,7 @@
from hdmf.testing import TestCase, remove_test_file
from hdmf.common.resources import HERD
from hdmf.term_set import TermSet, TermSetWrapper
from hdmf.utils import get_data_shape


from tests.unit.helpers.utils import (Foo, FooBucket, FooFile, get_foo_buildmanager,
Expand Down Expand Up @@ -739,12 +740,12 @@ def test_copy_h5py_dataset_h5dataio_input(self):
self.f['test_copy'][:].tolist())

def test_list_fill_empty(self):
dset = self.io.__list_fill__(self.f, 'empty_dataset', [], options={'dtype': int, 'io_settings': {}})
dset = self.io.__list_fill__(self.f, 'empty_dataset', [], None, True, options={'dtype': int, 'io_settings': {}})
self.assertTupleEqual(dset.shape, (0,))

def test_list_fill_empty_no_dtype(self):
with self.assertRaisesRegex(Exception, r"cannot add \S+ to [/\S]+ - could not determine type"):
self.io.__list_fill__(self.f, 'empty_dataset', [])
self.io.__list_fill__(self.f, 'empty_dataset', [], None, True)

def test_read_str(self):
a = ['a', 'bb', 'ccc', 'dddd', 'e']
Expand Down Expand Up @@ -3725,3 +3726,25 @@ def test_set_data_io(self):
self.data.set_data_io(H5DataIO, dict(chunks=True))
assert isinstance(self.data.data, H5DataIO)
assert self.data.data.io_settings["chunks"]


class TestExpand(TestCase):
def setUp(self):
self.manager = get_foo_buildmanager()
self.path = get_temp_filepath()

def test_expand_false(self):
# Setup all the data we need
foo1 = Foo('foo1', [1, 2, 3, 4, 5], "I am foo1", 17, 3.14)
foobucket = FooBucket('bucket1', [foo1])
foofile = FooFile(buckets=[foobucket])

with HDF5IO(self.path, manager=self.manager, mode='w') as io:
io.write(foofile, expandable=False)

io = HDF5IO(self.path, manager=self.manager, mode='r')
read_foofile = io.read()
self.assertListEqual(foofile.buckets['bucket1'].foos['foo1'].my_data,
read_foofile.buckets['bucket1'].foos['foo1'].my_data[:].tolist())
self.assertEqual(get_data_shape(read_foofile.buckets['bucket1'].foos['foo1'].my_data),
(5,))
Loading