def test_lab_meta(self):
ns_builder = NWBNamespaceBuilder('Extension for use in my Lab', self.prefix, version='0.1.0')
test_meta_ext = NWBGroupSpec(
neurodata_type_def='MyTestMetaData',
neurodata_type_inc='LabMetaData',
doc='my test meta data',
attributes=[
NWBAttributeSpec(name='test_attr', dtype='float', doc='test_dtype')])
ns_builder.add_spec(self.ext_source, test_meta_ext)
ns_builder.export(self.ns_path, outdir=self.tempdir)
ns_abs_path = os.path.join(self.tempdir, self.ns_path)
load_namespaces(ns_abs_path)
@register_class('MyTestMetaData', self.prefix)
class MyTestMetaData(LabMetaData):
__nwbfields__ = ('test_attr',)
@docval({'name': 'name', 'type': str, 'doc': 'name'},
{'name': 'test_attr', 'type': float, 'doc': 'test attribute'})
def __init__(self, **kwargs):
test_attr = popargs('test_attr', kwargs)
super(MyTestMetaData, self).__init__(**kwargs)
self.test_attr = test_attr
nwbfile = NWBFile("a file with header data", "NB123A", datetime(2017, 5, 1, 12, 0, 0, tzinfo=tzlocal()))
nwbfile.add_lab_meta_data(MyTestMetaData(name='test_name', test_attr=5.))
def create_LabMetaData_extension_from_schemas(schema_list, prefix):
# Initializations:
outdir = os.path.abspath(os.path.dirname(__file__))
ext_source = '%s_extension.yaml' % prefix
ns_path = '%s_namespace.yaml' % prefix
neurodata_type_list_as_str = str(
[schema.neurodata_type for schema in schema_list])
extension_doc = 'LabMetaData extensions: {neurodata_type_list_as_str} ({prefix})'.format(
neurodata_type_list_as_str=neurodata_type_list_as_str, prefix=prefix)
ns_builder = NWBNamespaceBuilder(extension_doc, prefix)
for schema in schema_list:
docval_list, attributes, nwbfields_list = extract_from_schema(schema)
neurodata_type = schema.neurodata_type
# Build the spec:
ext_group_spec = NWBGroupSpec(neurodata_type_def=neurodata_type,
neurodata_type_inc='LabMetaData',
doc=extension_doc,
attributes=attributes)
# Add spec to builder:
ns_builder.add_spec(ext_source, ext_group_spec)
# Export spec
ns_builder.export(ns_path, outdir=outdir)
def test_lab_meta_auto(self):
ns_builder = NWBNamespaceBuilder('Extension for use in my Lab',
self.prefix,
version='0.1.0')
test_meta_ext = NWBGroupSpec(neurodata_type_def='MyTestMetaData',
neurodata_type_inc='LabMetaData',
doc='my test meta data',
attributes=[
NWBAttributeSpec(name='test_attr',
dtype='float',
doc='test_dtype')
])
ns_builder.add_spec(self.ext_source, test_meta_ext)
ns_builder.export(self.ns_path, outdir=self.tempdir)
ns_abs_path = os.path.join(self.tempdir, self.ns_path)
load_namespaces(ns_abs_path)
MyTestMetaData = get_class('MyTestMetaData', self.prefix)
nwbfile = NWBFile("a file with header data", "NB123A",
datetime(2017, 5, 1, 12, 0, 0, tzinfo=tzlocal()))
nwbfile.add_lab_meta_data(
MyTestMetaData(name='test_name', test_attr=5.))
def test_catch_dup_name(self):
ns_builder1 = NWBNamespaceBuilder('Extension for us in my Lab',
"pynwb_test_extension1",
version='0.1.0')
ext1 = NWBGroupSpec('A custom ElectricalSeries for my lab',
attributes=[
NWBAttributeSpec(name='trode_id',
doc='the tetrode id',
dtype='int')
],
neurodata_type_inc='ElectricalSeries',
neurodata_type_def='TetrodeSeries')
ns_builder1.add_spec(self.ext_source1, ext1)
ns_builder1.export(self.ns_path1, outdir=self.tempdir)
ns_builder2 = NWBNamespaceBuilder('Extension for us in my Lab',
"pynwb_test_extension1",
version='0.1.0')
ext2 = NWBGroupSpec('A custom ElectricalSeries for my lab',
attributes=[
NWBAttributeSpec(name='trode_id',
doc='the tetrode id',
dtype='int')
],
neurodata_type_inc='ElectricalSeries',
neurodata_type_def='TetrodeSeries')
ns_builder2.add_spec(self.ext_source2, ext2)
ns_builder2.export(self.ns_path2, outdir=self.tempdir)
type_map = get_type_map(
extensions=os.path.join(self.tempdir, self.ns_path1))
type_map.load_namespaces(os.path.join(self.tempdir, self.ns_path2))
def test_catch_dup_name(self):
ns_builder1 = NWBNamespaceBuilder('Extension for us in my Lab',
"pynwb_test_extension1")
ext1 = NWBGroupSpec('A custom ElectricalSeries for my lab',
attributes=[
NWBAttributeSpec(name='trode_id',
doc='the tetrode id',
dtype='int')
],
neurodata_type_inc='ElectricalSeries',
neurodata_type_def='TetrodeSeries')
ns_builder1.add_spec(self.ext_source1, ext1)
ns_builder1.export(self.ns_path1, outdir=self.tempdir)
ns_builder2 = NWBNamespaceBuilder('Extension for us in my Lab',
"pynwb_test_extension1")
ext2 = NWBGroupSpec('A custom ElectricalSeries for my lab',
attributes=[
NWBAttributeSpec(name='trode_id',
doc='the tetrode id',
dtype='int')
],
neurodata_type_inc='ElectricalSeries',
neurodata_type_def='TetrodeSeries')
ns_builder2.add_spec(self.ext_source2, ext2)
ns_builder2.export(self.ns_path2, outdir=self.tempdir)
type_map = get_type_map(
extensions=os.path.join(self.tempdir, self.ns_path1))
with self.assertWarnsRegex(
UserWarning,
r"ignoring namespace '\S+' because it already exists"):
type_map.load_namespaces(os.path.join(self.tempdir, self.ns_path2))
def create_pynwb_extension_from_schemas(schema_list, prefix: str):
# Initializations:
outdir = os.path.abspath(os.path.dirname(__file__))
ext_source = f'{prefix}.extension.yaml'
ns_path = f'{prefix}.namespace.yaml'
extension_doc = ("Allen Institute behavior and optical "
"physiology extensions")
ns_builder = NWBNamespaceBuilder(
doc=extension_doc,
name=prefix,
version="0.2.0",
author="Allen Institute for Brain Science",
contact="*****@*****.**")
# Loops through and create NWB custom group specs for schemas found in:
# allensdk.brain_observatory.behavior.schemas
for schema in schema_list:
docval_list, attributes, nwbfields_list = extract_from_schema(schema)
# Build the spec:
ext_group_spec = NWBGroupSpec(
neurodata_type_def=schema.neurodata_type,
neurodata_type_inc=schema.neurodata_type_inc,
doc=schema.neurodata_doc,
attributes=attributes)
# Add spec to builder:
ns_builder.add_spec(ext_source, ext_group_spec)
# Export spec
ns_builder.export(ns_path, outdir=outdir)
def test_load_namespace_with_reftype_attribute_check_autoclass_const(self):
ns_builder = NWBNamespaceBuilder('Extension for use in my Lab',
self.prefix)
test_ds_ext = NWBDatasetSpec(
doc='test dataset to add an attr',
name='test_data',
shape=(None, ),
attributes=[
NWBAttributeSpec(name='target_ds',
doc='the target the dataset applies to',
dtype=RefSpec('TimeSeries', 'object'))
],
neurodata_type_def='my_new_type')
ns_builder.add_spec(self.ext_source, test_ds_ext)
ns_builder.export(self.ns_path, outdir=self.tempdir)
type_map = get_type_map(
extensions=os.path.join(self.tempdir, self.ns_path))
my_new_type = type_map.get_container_cls(self.prefix, 'my_new_type')
docval = None
for tmp in get_docval(my_new_type.__init__):
if tmp['name'] == 'target_ds':
docval = tmp
break
self.assertIsNotNone(docval)
self.assertEqual(docval['type'], TimeSeries)
def test_export(self):
ns_builder = NWBNamespaceBuilder('Extension for use in my Lab', self.prefix, version='0.1.0')
ext1 = NWBGroupSpec('A custom ElectricalSeries for my lab',
attributes=[NWBAttributeSpec(name='trode_id', doc='the tetrode id', dtype='int')],
neurodata_type_inc='ElectricalSeries',
neurodata_type_def='TetrodeSeries')
ns_builder.add_spec(self.ext_source, ext1)
ns_builder.export(self.ns_path, outdir=self.tempdir)
def test_export(self):
ns_builder = NWBNamespaceBuilder('Extension for us in my Lab', "pynwb_test_extension")
ext1 = NWBGroupSpec('A custom ElectricalSeries for my lab',
attributes=[NWBAttributeSpec('trode_id', 'int', 'the tetrode id')],
neurodata_type_inc='ElectricalSeries',
neurodata_type_def='TetrodeSeries')
ns_builder.add_spec(self.ext_source, ext1)
ns_builder.export(self.ns_path, outdir=self.tempdir)
def build_settings(dict):
""" WIP builds metadata based on passed in dict Does not support dict(s) deeper than 1 ATM"""
# Settings:
neurodata_type = 'MetaData'
prefix = 'NHP'
outdir = './'
extension_doc = 'lab metadata extension'
metadata_ext_group_spec = NWBGroupSpec(neurodata_type_def=neurodata_type,
neurodata_type_inc='LabMetaData',
doc=extension_doc,
attributes=[
NWBAttributeSpec(
name='experiment_id',
dtype='int',
doc='HW'),
NWBAttributeSpec(name='test',
dtype='text',
doc='HW')
])
#Export spec:
ext_source = '%s_extension.yaml' % prefix
ns_path = '%s_namespace.yaml' % prefix
ns_builder = NWBNamespaceBuilder(extension_doc, prefix, version=str(1))
ns_builder.add_spec(ext_source, metadata_ext_group_spec)
ns_builder.export(ns_path, outdir=outdir)
#Read spec and load namespace:
ns_abs_path = os.path.join(outdir, ns_path)
load_namespaces(ns_abs_path)
class MetaData(LabMetaData):
__nwbfields__ = ('experiment_id', 'test')
@docval({
'name': 'name',
'type': str,
'doc': 'name'
}, {
'name': 'experiment_id',
'type': int,
'doc': 'HW'
}, {
'name': 'test',
'type': str,
'doc': 'HW'
})
def __init__(self, **kwargs):
name, ophys_experiment_id, test = popargs('name', 'experiment_id',
'test', kwargs)
super(OphysBehaviorMetaData, self).__init__(name=name)
self.experiment_id = experiment_id
self.test = test
register_class('MetaData', prefix, MetaData)
return MetaData
def test_load_namespace_with_reftype_attribute(self):
ns_builder = NWBNamespaceBuilder('Extension for use in my Lab', self.prefix, version='0.1.0')
test_ds_ext = NWBDatasetSpec(doc='test dataset to add an attr',
name='test_data', shape=(None,),
attributes=[NWBAttributeSpec(name='target_ds',
doc='the target the dataset applies to',
dtype=RefSpec('TimeSeries', 'object'))],
neurodata_type_def='my_new_type')
ns_builder.add_spec(self.ext_source, test_ds_ext)
ns_builder.export(self.ns_path, outdir=self.tempdir)
get_type_map(extensions=os.path.join(self.tempdir, self.ns_path))
def test_catch_duplicate_spec(self):
spec1 = NWBGroupSpec("This is my new group 1",
"Group1",
neurodata_type_inc="NWBDataInterface",
neurodata_type_def="Group1")
spec2 = NWBGroupSpec("This is my new group 2",
"Group2",
groups=[spec1],
neurodata_type_inc="NWBDataInterface",
neurodata_type_def="Group2")
ns_builder = NWBNamespaceBuilder("Example namespace", "pynwb_test_ext")
ns_builder.add_spec(self.ext_source, spec1)
with self.assertRaises(ValueError):
ns_builder.add_spec(self.ext_source, spec2)
def test_catch_dup_name(self):
ns_builder1 = NWBNamespaceBuilder('Extension for us in my Lab', "pynwb_test_extension1")
ext1 = NWBGroupSpec('A custom ElectricalSeries for my lab',
attributes=[NWBAttributeSpec('trode_id', 'int', 'the tetrode id')],
neurodata_type_inc='ElectricalSeries',
neurodata_type_def='TetrodeSeries')
ns_builder1.add_spec(self.ext_source1, ext1)
ns_builder1.export(self.ns_path1, outdir=self.tempdir)
ns_builder2 = NWBNamespaceBuilder('Extension for us in my Lab', "pynwb_test_extension1")
ext2 = NWBGroupSpec('A custom ElectricalSeries for my lab',
attributes=[NWBAttributeSpec('trode_id', 'int', 'the tetrode id')],
neurodata_type_inc='ElectricalSeries',
neurodata_type_def='TetrodeSeries')
ns_builder2.add_spec(self.ext_source2, ext2)
ns_builder2.export(self.ns_path2, outdir=self.tempdir)
load_namespaces(os.path.join(self.tempdir, self.ns_path1))
with self.assertRaises(KeyError):
load_namespaces(os.path.join(self.tempdir, self.ns_path2))
name='vertices', dtype='float', dims=('vertex_number', 'xyz'))],
attributes=[
NWBAttributeSpec(
name='help', dtype='text', doc='help',
value='This holds Surface objects')])
surfaces = NWBGroupSpec(
neurodata_type_def='CorticalSurfaces',
neurodata_type_inc='NWBDataInterface',
name='cortical_surfaces',
doc='triverts for cortical surfaces', quantity='?',
groups=[surface],
attributes=[NWBAttributeSpec(
name='help', dtype='text', doc='help',
value='This holds the vertices and faces for the cortical surface '
'meshes')])
ecog_subject = NWBGroupSpec(
neurodata_type_def='ECoGSubject',
neurodata_type_inc='Subject',
name='subject',
doc='extension of subject that holds cortical surface data',
groups=[surfaces]
)
ns_builder = NWBNamespaceBuilder(doc=namespace + ' extensions', name=namespace,
version='1.1', author='Ben Dichter',
contact='*****@*****.**')
ns_builder.add_spec(ext_source, ecog_subject)
ns_builder.export(ns_path)
dtype='int'),
NWBDatasetSpec('scanimage auto-generated notes',
name='scanimage_notes',
dtype='text')
]
metadata_attr = [
NWBAttributeSpec('help',
'Metadata from Bernardo-Sabatini ScanImage',
'text',
value='Metadata from Bernardo-Sabatini ScanImage')
]
metadata_spec = NWBGroupSpec('ScanImage-specific metadata',
name='scanimage_metadata',
datasets=metadata_datasets,
attributes=metadata_attr,
neurodata_type_inc='LabMetaData',
neurodata_type_def='ScanImageMetaData')
# Export namespace
ext_source = 'sb_scanimage.specs.yaml'
ns_builder = NWBNamespaceBuilder(
'Extension for use with Bernardo-Sabatini ScanImage',
'sb_scanimage',
version='0.1',
author='Lawrence Niu',
contact='*****@*****.**')
ns_builder.add_spec(ext_source, metadata_spec)
ns_path = 'sb_scanimage.namespace.yaml'
ns_builder.export(ns_path)
def generate_extended_schema():
# set up silverlab namespace
ns_builder = NWBNamespaceBuilder(
'Extensions for acousto-optic lens data',
'silverlab_extended_schema',
'Silver lab data extension to NWB format for acousto-optic lens experiments',
version='0.4')
ns_builder.include_type('LabMetaData', namespace='core')
ns_builder.include_type('TwoPhotonSeries', namespace='core')
# define attributes Silver lab extension
cycle_time_attr = NWBAttributeSpec(
name='cycle_time',
doc='time in seconds for the microscope to acquire all ROIs once '
'and return to its starting position',
dtype='float')
cycles_per_trial_attr = NWBAttributeSpec(
name='cycles_per_trial',
doc='how many microscope cycles occur in each experimental trial',
dtype="int")
imaging_mode_attr = NWBAttributeSpec(
name='imaging_mode',
doc='the acquisition mode for the experiment; '
'pointing = single-voxel ROIs, '
'miniscan = 2d rectangular ROIs, '
'volume = 3d cuboid ROIs',
dtype='text')
frame_size_attr = NWBAttributeSpec(
name='frame_size',
doc='the 2d imaging frame size in voxels',
shape=(2, ),
dtype='int')
silverlab_api_version_attr = NWBAttributeSpec(
name='silverlab_api_version',
doc='For potential future backwards compatibility, '
'store the \'version\' of this API that created the file.',
dtype='text')
labview_version_attr = NWBAttributeSpec(
name='labview_version',
doc='The version of LabVIEW the data came from',
dtype='text',
required=False)
pockels_column_names_attr = NWBAttributeSpec(
name='columns',
doc='column names for the zplane pockels dataset',
shape=(4, ),
dtype='text')
# define datasets for Silver lab extensions
zplane_pockels_ds = NWBDatasetSpec(
doc='pockels data set, recording calibration data '
'for focusing at different z-planes in four columns: '
'Z offset from focal plane (micrometres), '
'normalised Z, '
'\'Pockels\' i.e. laser power in %, '
'and z offset for drive motors',
name='pockels',
shape=(None, 4),
attributes=[pockels_column_names_attr],
neurodata_type_def='ZplanePockelsDataset')
# define groups for Silver lab extensions
silverlab_optophys_specs = NWBGroupSpec(
doc='A place to store Silver lab specific optophysiology data',
attributes=[
cycle_time_attr, cycles_per_trial_attr, frame_size_attr,
imaging_mode_attr
],
datasets=[zplane_pockels_ds],
neurodata_type_def='SilverLabOptophysiology',
neurodata_type_inc='LabMetaData')
silverlab_metadata_specs = NWBGroupSpec(
doc='A place to store Silver lab specific metadata',
attributes=[silverlab_api_version_attr, labview_version_attr],
neurodata_type_def='SilverLabMetaData',
neurodata_type_inc='LabMetaData',
)
# dimensions ordered as t, x, y [, z], like the TimeSeries data itself
silverlab_pixel_time_offset_data = NWBDatasetSpec(
doc='A datastructure to hold time offsets for pixels. The'
'time offsets are the acquisition time of each pixel '
'relative to a starting time. The starting time is the '
'start of the cycle for pre-2018 LabView versions, '
'and the start of the trial for new versions.',
name='pixel_time_offsets',
shape=[(None, None), (None, None, None), (None, None, None, None)],
neurodata_type_def='PixelTimeOffsets')
silverlab_roi_image_specs = NWBGroupSpec(
doc='An extension to PyNWB\'s TwoPhotonSeries class, designed to hold '
'pixels from an ROI as well as the PixelTimeOffsets for them.',
datasets=[silverlab_pixel_time_offset_data],
neurodata_type_def='ROISeriesWithPixelTimeOffsets',
neurodata_type_inc='TwoPhotonSeries')
# export as schema extension
ext_source = 'silverlab.ophys.yaml'
ns_builder.add_spec(ext_source, silverlab_optophys_specs)
ext_source = 'silverlab.metadata.yaml'
ns_builder.add_spec(ext_source, silverlab_metadata_specs)
ext_source = 'silverlab.roi.yaml'
ns_builder.add_spec(ext_source, silverlab_roi_image_specs)
ns_builder.export('silverlab.namespace.yaml')
def main():
'''
Create the specification using PyNWB helpers.
Now we will create the specification ('draw up the blueprints') for the
Frank Lab extension. The main entries in a spec file are Groups, Attributes
and Datasets; PyNWB provides helpers that allow us to generate our spec
files using these components.
We primarily use the [NWBGroupSpec](https://pynwb.readthedocs.io/en/stable/pynwb.spec.html#pynwb.spec.NWBGroupSpec) # noqa
class to help us create valid NWB groups. An NWB group is basically just a
container that can have things like a name, attributes, datasets, and even
nested groups. In fl_extension.py (where we implement this extension in
Python), each of these groups will get its own Python class.
We add items within a group using [NWBAttributeSpec]() and
[NWBDatasetSpec](). An NWB attribute is just what it sounds like: a short
piece of metadata defining some attribute, such as a "help" text. An NWB
dataset is also pretty self-explanatory: it's just some data
(numbers, text, etc.).
As an example, the cell below describes the representation of a behavioral
task, and will generate the following lines in the
franklab.extensions.yaml file:
```
- neurodata_type_def: Task
neurodata_type_inc: NWBDataInterface
doc: a behavioral task
attributes:
- name: name
dtype: text
doc: the name of this task
- name: description
dtype: text
doc: description of this task
- name: help
dtype: text
doc: help doc
value: help value
```
---------------------------
Task (i.e. free exploration, W-alternation, sleep)
------
A Task consists simply of two attributes:
- a name (i.e. W-alternation, sleep, free exploration)
- a description of the task
Note that Task inherits from something called NWBDataInterface.
NWBDataInterface is a group in PyNWB representing basically any kind of
data, and which we can store in the NWB file in a ProcessingModule.
See create_franklab_nwbfile.ipynb for a discussion of Processing Modules.
---------------------------
'''
task = NWBGroupSpec(neurodata_type_def='Task',
neurodata_type_inc='NWBDataInterface',
doc='a behavioral task',
attributes=[
NWBAttributeSpec(
name='name',
doc='the name of this task',
dtype='text'),
NWBAttributeSpec(
name='description',
doc='description of this task',
dtype='text'),
NWBAttributeSpec(
name='help',
doc='help doc',
dtype='text',
value='Behavioral Task')])
# ---------------------------
# Apparatus (i.e. tracks, mazes, sleep boxes, arenas)
# --------
# We represent the topology of apparatuses using a graph representation
# (i.e. nodes and edges)
# - Nodes represent a component of an apparatus that you'd like the ability
# to refer
# (e.g. W-track arms, reward wells, novel object, open field components)
# - Edges represent the topological connectivity of the nodes
# (i.e. there should be an edge between the left track arm and the left
# reward well)
# In addition, all nodes will contain x/y coordinates that allow us to
# reconstruct not just
# the topology, but also the spatial geometry, of the apparatuses.
#
# Below, we will first define the nodes and edges. Finally we will define
# the Apparatus itself
# as a container that holds the nodes and edges as sub-groups.
# ---------------------------
# Node
# -----
# Abstract represention for any kind of node in the topological graph
# We won't actually implement abstract nodes. Rather this is a parent group
# from which our more specific types of nodes will inherit. Note that NWB
# specifications have inheritance.
# The quantity '*' means that we can have any number (0 or more) nodes.
node = NWBGroupSpec(
neurodata_type_def='Node',
neurodata_type_inc='NWBDataInterface',
doc='nodes in the graph',
quantity='*',
attributes=[NWBAttributeSpec(name='name',
doc='the name of this node',
dtype='text'),
NWBAttributeSpec(name='help',
doc='help doc',
dtype='text',
value='Apparatus Node')])
# Edge
# -------
# Edges between any two nodes in the graph.
# An edge's only dataset is the name (string) of the two nodes that the
# edge connects
# Note that we don't actually include the nodes themselves, just their
# names, in an edge.
edge = NWBGroupSpec(
neurodata_type_def='Edge',
neurodata_type_inc='NWBDataInterface',
doc='edges in the graph',
quantity='*',
datasets=[
NWBDatasetSpec(
doc='names of the nodes this edge connects',
name='edge_nodes',
dtype='text',
dims=['first_node_name|second_node_name'],
shape=[2])],
attributes=[
NWBAttributeSpec(
name='help',
doc='help doc',
dtype='text',
value='Apparatus Edge')])
# Point Node
# -----------
# A node that represents a single 2D point in space (e.g. reward well,
# novel object location)
point_node = NWBGroupSpec(
neurodata_type_def='PointNode',
neurodata_type_inc='Node',
doc='node representing a point in 2D space',
quantity='*',
datasets=[NWBDatasetSpec(doc='x/y coordinate of this 2D point',
name='coords',
dtype='float',
dims=['num_coords',
'x_vals|y_vals'],
shape=[1, 2])],
attributes=[NWBAttributeSpec(name='help',
doc='help doc',
dtype='text',
value='Apparatus Point')])
# Segment Node
# -------------
# A node that represents a linear segement in 2D space, defined by its
# start and end points
# (e.g. a single arm of W-track maze)
segment_node = NWBGroupSpec(
neurodata_type_def='SegmentNode',
neurodata_type_inc='Node',
doc=('node representing a 2D linear segment defined by its start and'
'end points'),
quantity='*',
datasets=[
NWBDatasetSpec(doc=('x/y coordinates of the start and end points '
'of this segment'),
name='coords',
dtype='float',
dims=['num_coords', 'x_vals|y_vals'],
shape=[2, 2])],
attributes=[
NWBAttributeSpec(name='help',
doc='help doc',
dtype='text', value='Apparatus Segment')])
# Polygon Node
# -------------
# A node that represents a polygon area (e.g. open field, sleep box)
# A polygon is defined by its external vertices and, optionally, by
# any interior points of interest (e.g. interior wells, objects)
polygon_node = NWBGroupSpec(
neurodata_type_def='PolygonNode',
neurodata_type_inc='Node',
doc='node representing a 2D polygon area',
quantity='*',
datasets=[
NWBDatasetSpec(
doc='x/y coordinates of the exterior points of this polygon',
name='coords',
dtype='float',
dims=['num_coords', 'x_vals|y_vals'],
shape=['null', 2]),
NWBDatasetSpec(
doc='x/y coordinates of interior points inside this polygon',
name='interior_coords',
dtype='float',
quantity='?',
dims=['num_coords',
'x_vals|y_vals'],
shape=['null', 2])],
attributes=[
NWBAttributeSpec(name='help',
doc='help doc',
dtype='text',
value='Apparatus Polygon')])
# Apparatus
# -------------
# Finally, we define the apparatus itself.
# It is has two sub-groups: nodes and edges.
apparatus = NWBGroupSpec(
neurodata_type_def='Apparatus',
neurodata_type_inc='NWBDataInterface',
doc='a graph of nodes and edges',
quantity='*',
groups=[node, edge],
attributes=[
NWBAttributeSpec(name='name',
doc='the name of this apparatus',
dtype='text'),
NWBAttributeSpec(name='help',
doc='help doc',
dtype='text',
value='Behavioral Apparatus')])
# ### Save the extension specification
# The specification consists of two YAML (.yaml) files: one for the actual
# blueprint and one for the namespace. In a world with many blueprints, we
# need namespaces to effectively categorize/store them, like the drawers of
# an architect's filing cabinet.
namespace_builder = NWBNamespaceBuilder(
f'{namespace} extensions', namespace, version='0.1.0')
namespace_builder.add_spec(extension_filename, apparatus)
namespace_builder.add_spec(extension_filename, task)
namespace_builder.add_spec(extension_filename, point_node)
namespace_builder.add_spec(extension_filename, segment_node)
namespace_builder.add_spec(extension_filename, polygon_node)
# Bug: NamespaceBuilder.add_spec creates the .extensions.yaml file in the
# current directory (it errors if you pass in a file path containing '/'
# to add_spec, above.)
old_cwd = os.getcwd()
os.chdir(yaml_dir)
namespace_builder.export(namespace_path)
os.chdir(old_cwd)
LabMetaData_ext.add_attribute(
name='auditory_stim_association',
doc='Descriptor of auditory stimulus-lick port association.',
dtype='text',
shape=None,
)
LabMetaData_ext.add_attribute(
name='visual_stim_association',
doc='Descriptor of auditory stimulus-lick port association.',
dtype='text',
shape=None,
)
LabMetaData_ext.add_attribute(
name='behavior_version',
doc='Version of schema used to translate LabView data to NWB.',
dtype='text',
shape=None,
)
# Add the extension
ext_source = 'uobrainflex.specs.yaml'
ns_builder.include_type('LabMetaData', namespace='core')
ns_builder.add_spec(ext_source, LabMetaData_ext)
# Save the namespace and extensions
ns_path = 'uobrainflex.namespace.yaml'
ns_builder.export(ns_path)
# to this namespace. Finally,
# it calls :py:meth:`~pynwb.form.spec.write.NamespaceBuilder.export` to save the extensions to disk for downstream use.
from pynwb.spec import NWBNamespaceBuilder, NWBGroupSpec, NWBAttributeSpec
ns_path = "mylab.namespace.yaml"
ext_source = "mylab.extensions.yaml"
ns_builder = NWBNamespaceBuilder('Extension for use in my Lab', "mylab")
ext = NWBGroupSpec(
'A custom ElectricalSeries for my lab',
attributes=[NWBAttributeSpec('trode_id', 'the tetrode id', 'int')],
neurodata_type_inc='ElectricalSeries',
neurodata_type_def='TetrodeSeries')
ns_builder.add_spec(ext_source, ext)
ns_builder.export(ns_path)
####################
# Running this block will produce two YAML files.
#
# The first file contains the specification of the namespace.
#
# .. code-block:: yaml
#
# # mylab.namespace.yaml
# namespaces:
# - doc: Extension for use in my Lab
# name: mylab
# schema:
# - namespace: core
dtype='float',
required=False),
NWBAttributeSpec(
name='microdrive',
doc='whether a microdrive was used (0: not used, 1: used)',
dtype='int'),
NWBAttributeSpec(name='microdrive_lead',
doc='um/turn',
dtype='float',
required=False),
NWBAttributeSpec(name='microdrive_id',
doc='id of microdrive',
dtype='int',
required=False),
NWBAttributeSpec(name='help',
doc='help',
dtype='text',
value='Information about optical fiber')
])
ns_builder = NWBNamespaceBuilder(doc=namespace + ' extensions',
name=namespace,
version='1.0',
author='Ben Dichter',
contact='*****@*****.**')
specs = (subject, optical_fiber)
for spec in specs:
ns_builder.add_spec(ext_source, spec)
ns_builder.export(ns_path)
quantity='?'),
NWBDatasetSpec(name='label_index',
neurodata_type_inc='VectorIndex',
doc='indexes label',
quantity='?')
])
CompartmentsSeries = NWBGroupSpec(
neurodata_type_def='CompartmentSeries',
neurodata_type_inc='TimeSeries',
doc='Stores continuous data from cell compartments',
links=[
NWBLinkSpec(
name='compartments',
target_type='Compartments',
doc='meta-data about compartments in this CompartmentSeries',
quantity='?')
])
# Export
doc = 'NWB:N extension for storing large-scale simulation output ' \
'with multi-cell multi-compartment recordings'
ns_builder = NWBNamespaceBuilder(doc=doc,
name=name,
version='0.2.0',
author=['Ben Dichter', 'Kael Dai'],
contact='*****@*****.**')
for neurodata_type in [Compartments, CompartmentsSeries]:
ns_builder.add_spec(ext_source, neurodata_type)
ns_builder.export(ns_path)
csd_virtual_electrode_vertical_positions = NWBDatasetSpec(name="virtual_electrode_y_positions",
doc="Virtual vertical positions of electrodes from which CSD was calculated",
attributes=[unit_attr],
dtype='float32',
shape=(None,))
csd_virtual_electrode_horizontal_positions = NWBDatasetSpec(name="virtual_electrode_x_positions",
doc="Virtual horizontal positions of electrodes from which CSD was calculated",
attributes=[unit_attr],
dtype='float32',
shape=(None,))
ecephys_csd_ext = NWBGroupSpec(
doc="A group containing current source density (CSD) data and virtual electrode locations",
groups=[csd_timeseries_group],
datasets=[csd_virtual_electrode_horizontal_positions,
csd_virtual_electrode_vertical_positions],
neurodata_type_def="EcephysCSD",
neurodata_type_inc="NWBDataInterface"
)
ext_source = "ndx-aibs-ecephys.extension.yaml"
ns_builder.add_spec(ext_source, ecephys_probe_ext)
ns_builder.add_spec(ext_source, ecephys_egroup_ext)
ns_builder.add_spec(ext_source, ecephys_specimen_ext)
ns_builder.add_spec(ext_source, ecephys_eye_tracking_rig_metadata_ext)
ns_builder.add_spec(ext_source, ecephys_csd_ext)
namespace_path = "ndx-aibs-ecephys.namespace.yaml"
ns_builder.export(namespace_path)
# The following block of code demonstrates how to create a new namespace, and then add a new `neurodata_type`
# to this namespace. Finally,
# it calls :py:meth:`~pynwb.form.spec.write.NamespaceBuilder.export` to save the extensions to disk for downstream use.
from pynwb.spec import NWBNamespaceBuilder, NWBGroupSpec, NWBAttributeSpec
ns_path = "mylab.namespace.yaml"
ext_source = "mylab.extensions.yaml"
ns_builder = NWBNamespaceBuilder('Extension for use in my Lab', "mylab")
ext = NWBGroupSpec('A custom ElectricalSeries for my lab',
attributes=[NWBAttributeSpec('trode_id', 'int', 'the tetrode id')],
neurodata_type_inc='ElectricalSeries',
neurodata_type_def='TetrodeSeries')
ns_builder.add_spec(ext_source, ext)
ns_builder.export(ns_path)
####################
# Running this block will produce two YAML files.
#
# The first file contains the specification of the namespace.
#
# .. code-block:: yaml
#
# # mylab.namespace.yaml
# namespaces:
# - doc: Extension for use in my Lab
# name: mylab
# schema:
# - namespace: core
# {'name': 'vertices', 'type': ('array_data', 'data'),'doc': 'faces for this surface', 'default': None})
# def __init__(self, **kwargs):
# call_docval_func(super((BlenderSurface, self).__init__, kwargs))
# self.faces = getargs('faces', kwargs)
# self.vertices = getargs('vertices', kwargs)
#ns_builder.add_spec(ext_source, blender_surface)
blender_plane_segmentation = NWBGroupSpec(
'A plane to store data from blender',
neurodata_type_inc='PlaneSegmentation',
neurodata_type_def='BlenderPlaneSegmentation',
groups=[blender_surface])
ns_builder.add_spec(ext_source, blender_plane_segmentation)
#Writes YAML files
ns_builder.export(ns_path)
load_namespaces('blenderbits.namespace.yaml')
BlenderSurface = get_class('BlenderSurface', 'TanLab')
BlenderPlaneSegmentation = get_class('BlenderPlaneSegmentation', 'TanLab')
#Read in OBJ
os.chdir(
'C:/Users/Mrika/OneDrive/TanLab/NWBHackathonFiles/HackthonFiles/ObjectModels'
)
soma_triangles = o3d.io.read_triangle_mesh("soma.obj")
soma_triangles = np.asarray(soma_triangles.triangles)
