def test_init_delayed_bands(self):
timeseries = TimeSeries(name='dummy timeseries',
description='desc',
data=np.ones((3, 3)),
unit='Volts',
timestamps=np.ones((3, )))
spec_anal = DecompositionSeries(name='LFPSpectralAnalysis',
description='my description',
data=np.ones((3, 3, 3)),
timestamps=np.ones((3, )),
source_timeseries=timeseries,
metric='amplitude')
for band_name in ['alpha', 'beta', 'gamma']:
spec_anal.add_band(band_name=band_name,
band_limits=(1., 1.),
band_mean=1.,
band_stdev=1.)
self.assertEqual(spec_anal.name, 'LFPSpectralAnalysis')
self.assertEqual(spec_anal.description, 'my description')
np.testing.assert_equal(spec_anal.data, np.ones((3, 3, 3)))
np.testing.assert_equal(spec_anal.timestamps, np.ones((3, )))
self.assertEqual(spec_anal.bands['band_name'].data,
['alpha', 'beta', 'gamma'])
np.testing.assert_equal(spec_anal.bands['band_limits'].data,
np.ones((3, 2)))
self.assertEqual(spec_anal.source_timeseries, timeseries)
self.assertEqual(spec_anal.metric, 'amplitude')
def yuta2nwb(session_path='/Users/bendichter/Desktop/Buzsaki/SenzaiBuzsaki2017/YutaMouse41/YutaMouse41-150903',
subject_xls=None, include_spike_waveforms=True, stub=True):
subject_path, session_id = os.path.split(session_path)
fpath_base = os.path.split(subject_path)[0]
identifier = session_id
mouse_number = session_id[9:11]
if '-' in session_id:
subject_id, date_text = session_id.split('-')
b = False
else:
subject_id, date_text = session_id.split('b')
b = True
if subject_xls is None:
subject_xls = os.path.join(subject_path, 'YM' + mouse_number + ' exp_sheet.xlsx')
else:
if not subject_xls[-4:] == 'xlsx':
subject_xls = os.path.join(subject_xls, 'YM' + mouse_number + ' exp_sheet.xlsx')
session_start_time = dateparse(date_text, yearfirst=True)
df = pd.read_excel(subject_xls)
subject_data = {}
for key in ['genotype', 'DOB', 'implantation', 'Probe', 'Surgery', 'virus injection', 'mouseID']:
names = df.iloc[:, 0]
if key in names.values:
subject_data[key] = df.iloc[np.argmax(names == key), 1]
if isinstance(subject_data['DOB'], datetime):
age = session_start_time - subject_data['DOB']
else:
age = None
subject = Subject(subject_id=subject_id, age=str(age),
genotype=subject_data['genotype'],
species='mouse')
nwbfile = NWBFile(session_description='mouse in open exploration and theta maze',
identifier=identifier,
session_start_time=session_start_time.astimezone(),
file_create_date=datetime.now().astimezone(),
experimenter='Yuta Senzai',
session_id=session_id,
institution='NYU',
lab='Buzsaki',
subject=subject,
related_publications='DOI:10.1016/j.neuron.2016.12.011')
print('reading and writing raw position data...', end='', flush=True)
ns.add_position_data(nwbfile, session_path)
shank_channels = ns.get_shank_channels(session_path)[:8]
all_shank_channels = np.concatenate(shank_channels)
print('setting up electrodes...', end='', flush=True)
hilus_csv_path = os.path.join(fpath_base, 'early_session_hilus_chans.csv')
lfp_channel = get_reference_elec(subject_xls, hilus_csv_path, session_start_time, session_id, b=b)
print(lfp_channel)
custom_column = [{'name': 'theta_reference',
'description': 'this electrode was used to calculate LFP canonical bands',
'data': all_shank_channels == lfp_channel}]
ns.write_electrode_table(nwbfile, session_path, custom_columns=custom_column, max_shanks=max_shanks)
print('reading LFPs...', end='', flush=True)
lfp_fs, all_channels_data = ns.read_lfp(session_path, stub=stub)
lfp_data = all_channels_data[:, all_shank_channels]
print('writing LFPs...', flush=True)
# lfp_data[:int(len(lfp_data)/4)]
lfp_ts = ns.write_lfp(nwbfile, lfp_data, lfp_fs, name='lfp',
description='lfp signal for all shank electrodes')
for name, channel in special_electrode_dict.items():
ts = TimeSeries(name=name, description='environmental electrode recorded inline with neural data',
data=all_channels_data[channel], rate=lfp_fs, unit='V', conversion=np.nan, resolution=np.nan)
nwbfile.add_acquisition(ts)
# compute filtered LFP
print('filtering LFP...', end='', flush=True)
all_lfp_phases = []
for passband in ('theta', 'gamma'):
lfp_fft = filter_lfp(lfp_data[:, all_shank_channels == lfp_channel].ravel(), lfp_fs, passband=passband)
lfp_phase, _ = hilbert_lfp(lfp_fft)
all_lfp_phases.append(lfp_phase[:, np.newaxis])
data = np.dstack(all_lfp_phases)
print('done.', flush=True)
if include_spike_waveforms:
print('writing waveforms...', end='', flush=True)
for shankn in np.arange(1, 9, dtype=int):
ns.write_spike_waveforms(nwbfile, session_path, shankn, stub=stub)
print('done.', flush=True)
decomp_series = DecompositionSeries(name='LFPDecompositionSeries',
description='Theta and Gamma phase for reference LFP',
data=data, rate=lfp_fs,
source_timeseries=lfp_ts,
metric='phase', unit='radians')
decomp_series.add_band(band_name='theta', band_limits=(4, 10))
decomp_series.add_band(band_name='gamma', band_limits=(30, 80))
check_module(nwbfile, 'ecephys', 'contains processed extracellular electrophysiology data').add_data_interface(decomp_series)
[nwbfile.add_stimulus(x) for x in ns.get_events(session_path)]
# create epochs corresponding to experiments/environments for the mouse
sleep_state_fpath = os.path.join(session_path, '{}--StatePeriod.mat'.format(session_id))
exist_pos_data = any(os.path.isfile(os.path.join(session_path, '{}__{}.mat'.format(session_id, task_type['name'])))
for task_type in task_types)
if exist_pos_data:
nwbfile.add_epoch_column('label', 'name of epoch')
for task_type in task_types:
label = task_type['name']
file = os.path.join(session_path, session_id + '__' + label + '.mat')
if os.path.isfile(file):
print('loading position for ' + label + '...', end='', flush=True)
pos_obj = Position(name=label + '_position')
matin = loadmat(file)
tt = matin['twhl_norm'][:, 0]
exp_times = find_discontinuities(tt)
if 'conversion' in task_type:
conversion = task_type['conversion']
else:
conversion = np.nan
for pos_type in ('twhl_norm', 'twhl_linearized'):
if pos_type in matin:
pos_data_norm = matin[pos_type][:, 1:]
spatial_series_object = SpatialSeries(
name=label + '_{}_spatial_series'.format(pos_type),
data=H5DataIO(pos_data_norm, compression='gzip'),
reference_frame='unknown', conversion=conversion,
resolution=np.nan,
timestamps=H5DataIO(tt, compression='gzip'))
pos_obj.add_spatial_series(spatial_series_object)
check_module(nwbfile, 'behavior', 'contains processed behavioral data').add_data_interface(pos_obj)
for i, window in enumerate(exp_times):
nwbfile.add_epoch(start_time=window[0], stop_time=window[1],
label=label + '_' + str(i))
print('done.')
# there are occasional mismatches between the matlab struct and the neuroscope files
# regions: 3: 'CA3', 4: 'DG'
df_unit_features = get_UnitFeatureCell_features(fpath_base, session_id, session_path)
celltype_names = []
for celltype_id, region_id in zip(df_unit_features['fineCellType'].values,
df_unit_features['region'].values):
if celltype_id == 1:
if region_id == 3:
celltype_names.append('pyramidal cell')
elif region_id == 4:
celltype_names.append('granule cell')
else:
raise Exception('unknown type')
elif not np.isfinite(celltype_id):
celltype_names.append('missing')
else:
celltype_names.append(celltype_dict[celltype_id])
custom_unit_columns = [
{
'name': 'cell_type',
'description': 'name of cell type',
'data': celltype_names},
{
'name': 'global_id',
'description': 'global id for cell for entire experiment',
'data': df_unit_features['unitID'].values},
{
'name': 'max_electrode',
'description': 'electrode that has the maximum amplitude of the waveform',
'data': get_max_electrodes(nwbfile, session_path),
'table': nwbfile.electrodes
}]
ns.add_units(nwbfile, session_path, custom_unit_columns, max_shanks=max_shanks)
trialdata_path = os.path.join(session_path, session_id + '__EightMazeRun.mat')
if os.path.isfile(trialdata_path):
trials_data = loadmat(trialdata_path)['EightMazeRun']
trialdatainfo_path = os.path.join(fpath_base, 'EightMazeRunInfo.mat')
trialdatainfo = [x[0] for x in loadmat(trialdatainfo_path)['EightMazeRunInfo'][0]]
features = trialdatainfo[:7]
features[:2] = 'start_time', 'stop_time',
[nwbfile.add_trial_column(x, 'description') for x in features[4:] + ['condition']]
for trial_data in trials_data:
if trial_data[3]:
cond = 'run_left'
else:
cond = 'run_right'
nwbfile.add_trial(start_time=trial_data[0], stop_time=trial_data[1], condition=cond,
error_run=trial_data[4], stim_run=trial_data[5], both_visit=trial_data[6])
"""
mono_syn_fpath = os.path.join(session_path, session_id+'-MonoSynConvClick.mat')
matin = loadmat(mono_syn_fpath)
exc = matin['FinalExcMonoSynID']
inh = matin['FinalInhMonoSynID']
#exc_obj = CatCellInfo(name='excitatory_connections',
# indices_values=[], cell_index=exc[:, 0] - 1, indices=exc[:, 1] - 1)
#module_cellular.add_container(exc_obj)
#inh_obj = CatCellInfo(name='inhibitory_connections',
# indices_values=[], cell_index=inh[:, 0] - 1, indices=inh[:, 1] - 1)
#module_cellular.add_container(inh_obj)
"""
if os.path.isfile(sleep_state_fpath):
matin = loadmat(sleep_state_fpath)['StatePeriod']
table = TimeIntervals(name='states', description='sleep states of animal')
table.add_column(name='label', description='sleep state')
data = []
for name in matin.dtype.names:
for row in matin[name][0][0]:
data.append({'start_time': row[0], 'stop_time': row[1], 'label': name})
[table.add_row(**row) for row in sorted(data, key=lambda x: x['start_time'])]
check_module(nwbfile, 'behavior', 'contains behavioral data').add_data_interface(table)
if stub:
out_fname = session_path + '_stub.nwb'
else:
out_fname = session_path + '.nwb'
print('writing NWB file...', end='', flush=True)
with NWBHDF5IO(out_fname, mode='w') as io:
io.write(nwbfile)
print('done.')
print('testing read...', end='', flush=True)
# test read
with NWBHDF5IO(out_fname, mode='r') as io:
io.read()
print('done.')
def convert_data(self,
nwbfile: NWBFile,
metadata_dict: dict,
stub_test: bool = False):
session_path = self.input_args['folder_path']
# TODO: check/enforce format?
all_shank_channels = metadata_dict['all_shank_channels']
special_electrode_dict = metadata_dict['special_electrodes']
lfp_channel = metadata_dict['lfp_channel']
lfp_sampling_rate = metadata_dict['lfp_sampling_rate']
spikes_nsamples = metadata_dict['spikes_nsamples']
shank_channels = metadata_dict['shank_channels']
subject_path, session_id = os.path.split(session_path)
_, all_channels_lfp_data = read_lfp(session_path, stub=stub_test)
lfp_data = all_channels_lfp_data[:, all_shank_channels]
lfp_ts = write_lfp(nwbfile,
lfp_data,
lfp_sampling_rate,
name=metadata_dict['lfp']['name'],
description=metadata_dict['lfp']['description'],
electrode_inds=None)
# TODO: error checking on format?
for special_electrode in special_electrode_dict:
ts = TimeSeries(
name=special_electrode['name'],
description=special_electrode['description'],
data=all_channels_lfp_data[:, special_electrode['channel']],
rate=lfp_sampling_rate,
unit='V',
resolution=np.nan)
nwbfile.add_acquisition(ts)
# TODO: discuss/consider more robust checking well prior to this
# when missing experimental sheets for a subject, the lfp_channel cannot be determined(?)
# which causes uninformative downstream errors at this step because lfp_channel is None
# (get_reference_electrode does throw a warning, though)
if lfp_channel is not None:
all_lfp_phases = []
for passband in ('theta', 'gamma'):
lfp_fft = filter_lfp(
lfp_data[:, all_shank_channels == lfp_channel].ravel(),
lfp_sampling_rate,
passband=passband)
lfp_phase, _ = hilbert_lfp(lfp_fft)
all_lfp_phases.append(lfp_phase[:, np.newaxis])
decomp_series_data = np.dstack(all_lfp_phases)
# TODO: should units or metrics be metadata?
decomp_series = DecompositionSeries(
name=metadata_dict['lfp_decomposition']['name'],
description=metadata_dict['lfp_decomposition']['description'],
data=decomp_series_data,
rate=lfp_sampling_rate,
source_timeseries=lfp_ts,
metric='phase',
unit='radians')
# TODO: the band limits should be extracted from parse_passband in band_analysis?
decomp_series.add_band(band_name='theta', band_limits=(4, 10))
decomp_series.add_band(band_name='gamma', band_limits=(30, 80))
check_module(
nwbfile, 'ecephys',
'contains processed extracellular electrophysiology data'
).add_data_interface(decomp_series)
write_spike_waveforms(nwbfile,
session_path,
spikes_nsamples=spikes_nsamples,
shank_channels=shank_channels,
stub_test=stub_test)
def run_conversion(self, nwbfile: NWBFile, metadata: dict, stub_test: bool = False):
super().run_conversion(nwbfile=nwbfile, metadata=metadata, stub_test=stub_test)
session_path = Path(self.source_data["file_path"]).parent
session_id = session_path.name
subject_path = session_path.parent
xml_filepath = session_path / f"{session_id}.xml"
root = et.parse(str(xml_filepath)).getroot()
n_total_channels = int(root.find("acquisitionSystem").find("nChannels").text)
lfp_sampling_rate = float(root.find("fieldPotentials").find("lfpSamplingRate").text)
shank_channels = [
[int(channel.text) for channel in group.find("channels")]
for group in root.find("spikeDetection").find("channelGroups").findall("group")
]
all_shank_channels = np.concatenate(shank_channels) # Flattened
# Special electrodes
special_electrode_mapping = dict(
ch_wait=79,
ch_arm=78,
ch_solL=76,
ch_solR=77,
ch_dig1=65,
ch_dig2=68,
ch_entL=72,
ch_entR=71,
ch_SsolL=73,
ch_SsolR=70,
)
special_electrodes = []
for special_electrode_name, channel in special_electrode_mapping.items():
if channel <= n_total_channels - 1:
special_electrodes.append(
dict(
name=special_electrode_name,
channel=channel,
description="Environmental electrode recorded inline with neural data.",
)
)
_, all_channels_lfp_data = read_lfp(session_path, stub=stub_test)
for special_electrode in special_electrodes:
ts = TimeSeries(
name=special_electrode["name"],
description=special_electrode["description"],
data=all_channels_lfp_data[:, special_electrode["channel"]],
rate=lfp_sampling_rate,
unit="V",
resolution=np.nan,
)
nwbfile.add_acquisition(ts)
# DecompositionSeries
mouse_number = session_id[-9:-7]
subject_xls = str(subject_path / f"DGProject/YM{mouse_number} exp_sheet.xlsx")
hilus_csv_path = str(subject_path / "DGProject/early_session_hilus_chans.csv")
session_start = metadata["NWBFile"]["session_start_time"]
if "-" in session_id:
b = False
else:
b = True
lfp_channel = get_reference_elec(subject_xls, hilus_csv_path, session_start, session_id, b=b)
if lfp_channel is not None:
lfp_data = all_channels_lfp_data[:, all_shank_channels]
all_lfp_phases = []
for passband in ("theta", "gamma"):
lfp_fft = filter_lfp(
lfp_data[:, all_shank_channels == lfp_channel].ravel(),
lfp_sampling_rate,
passband=passband,
)
lfp_phase, _ = hilbert_lfp(lfp_fft)
all_lfp_phases.append(lfp_phase[:, np.newaxis])
decomp_series_data = np.dstack(all_lfp_phases)
ecephys_mod = check_module(
nwbfile,
"ecephys",
"Intermediate data from extracellular electrophysiology recordings, e.g., LFP.",
)
lfp_ts = ecephys_mod.data_interfaces["LFP"]["LFP"]
decomp_series = DecompositionSeries(
name="LFPDecompositionSeries",
description="Theta and Gamma phase for reference LFP",
data=decomp_series_data,
rate=lfp_sampling_rate,
source_timeseries=lfp_ts,
metric="phase",
unit="radians",
)
# TODO: the band limits should be extracted from parse_passband in band_analysis?
decomp_series.add_band(band_name="theta", band_limits=(4, 10))
decomp_series.add_band(band_name="gamma", band_limits=(30, 80))
check_module(
nwbfile,
"ecephys",
"Contains processed extracellular electrophysiology data.",
).add(decomp_series)
def chang2nwb(blockpath,
outpath=None,
session_start_time=None,
session_description=None,
identifier=None,
anin4=False,
ecog_format='auto',
external_subject=True,
include_pitch=False,
include_intensity=False,
speakers=True,
mic=False,
mini=False,
hilb=False,
verbose=False,
imaging_path=None,
parse_transcript=False,
include_cortical_surfaces=True,
include_electrodes=True,
include_ekg=True,
subject_image_list=None,
rest_period=None,
load_warped=False,
**kwargs):
"""
Parameters
----------
blockpath: str
outpath: None | str
if None, output = [blockpath]/[blockname].nwb
session_start_time: datetime.datetime
default: datetime(1900, 1, 1)
session_description: str
default: blockname
identifier: str
default: blockname
anin4: False | str
Whether or not to convert ANIN4. ANIN4 is used as an extra channel for
things like button presses, and is usually unused. If a string is
supplied, that is used as the name of the timeseries.
ecog_format: str
({'htk'}, 'mat', 'raw')
external_subject: bool (optional)
True: (default) cortical mesh is saved in an external file and a link is
provided to that file. This is useful if you have multiple sessions for a single subject.
False: cortical mesh is saved normally
include_pitch: bool (optional)
add pitch data. Default: False
include_intensity: bool (optional)
add intensity data. Default: False
speakers: bool (optional)
Default: False
mic: bool (optional)
default: False
mini: only save data stub. Used for testing
hilb: bool
include Hilbert Transform data. Default: False
verbose: bool (optional)
imaging_path: str (optional)
None: use IMAGING_DIR
'local': use subject_dir/Imaging/
else: use supplied string
parse_transcript: str (optional)
include_cortical_surfaces: bool (optional)
include_electrodes: bool (optional)
include_ekg: bool (optional)
subject_image_list: list (optional)
List of paths of images to include
rest_period: None | array-like
kwargs: dict
passed to pynwb.NWBFile
Returns
-------
"""
behav_module = None
basepath, blockname = os.path.split(blockpath)
subject_id = get_subject_id(blockname)
if identifier is None:
identifier = blockname
if session_description is None:
session_description = blockname
if outpath is None:
outpath = blockpath + '.nwb'
out_base_path = os.path.split(outpath)[0]
if session_start_time is None:
session_start_time = datetime(1900, 1, 1).astimezone(timezone('UTC'))
if imaging_path is None:
subj_imaging_path = path.join(IMAGING_PATH, subject_id)
elif imaging_path == 'local':
subj_imaging_path = path.join(basepath, 'imaging')
else:
subj_imaging_path = os.path.join(imaging_path, subject_id)
# file paths
bad_time_file = path.join(blockpath, 'Artifacts', 'badTimeSegments.mat')
ecog_path = path.join(blockpath, 'RawHTK')
ecog400_path = path.join(blockpath, 'ecog400', 'ecog.mat')
elec_metadata_file = path.join(subj_imaging_path, 'elecs',
'TDT_elecs_all.mat')
mesh_path = path.join(subj_imaging_path, 'Meshes')
pial_files = glob.glob(path.join(mesh_path, '*pial.mat'))
# Create the NWB file object
nwbfile = NWBFile(session_description,
identifier,
session_start_time,
datetime.now().astimezone(),
session_id=identifier,
institution='University of California, San Francisco',
lab='Chang Lab',
**kwargs)
nwbfile.add_electrode_column('bad', 'electrode identified as too noisy')
bad_elecs_inds = get_bad_elecs(blockpath)
if include_electrodes:
add_electrodes(nwbfile,
elec_metadata_file,
bad_elecs_inds,
load_warped=load_warped)
else:
device = nwbfile.create_device('256Grid')
electrode_group = nwbfile.create_electrode_group(
name='256Grid electrodes',
description='auto_group',
location='location',
device=device)
for elec_counter in range(256):
bad = elec_counter in bad_elecs_inds
nwbfile.add_electrode(id=elec_counter + 1,
x=np.nan,
y=np.nan,
z=np.nan,
imp=np.nan,
location=' ',
filtering='none',
group=electrode_group,
bad=bad)
ecog_elecs = list(range(len(nwbfile.electrodes)))
ecog_elecs_region = nwbfile.create_electrode_table_region(
ecog_elecs, 'ECoG electrodes on brain')
# Read electrophysiology data from HTK files and add them to NWB file
if ecog_format == 'auto':
ecog_rate, data, ecog_path = auto_ecog(blockpath,
ecog_elecs,
verbose=False)
elif ecog_format == 'htk':
if verbose:
print('reading htk acquisition...', flush=True)
ecog_rate, data = readhtks(ecog_path, ecog_elecs)
data = data.squeeze()
if verbose:
print('done', flush=True)
elif ecog_format == 'mat':
with File(ecog400_path, 'r') as f:
data = f['ecogDS']['data'][:, ecog_elecs]
ecog_rate = f['ecogDS']['sampFreq'][:].ravel()[0]
ecog_path = ecog400_path
elif ecog_format == 'raw':
ecog_path = os.path.join(tdt_data_path, subject_id, blockname,
'raw.mat')
ecog_rate, data = load_wavs(ecog_path)
else:
raise ValueError('unrecognized argument: ecog_format')
ts_desc = "all Wav data"
if mini:
data = data[:2000]
ecog_ts = ElectricalSeries(name='ElectricalSeries',
data=H5DataIO(data, compression='gzip'),
electrodes=ecog_elecs_region,
rate=ecog_rate,
description=ts_desc,
conversion=0.001)
nwbfile.add_acquisition(ecog_ts)
if include_ekg:
ekg_elecs = find_ekg_elecs(elec_metadata_file)
if len(ekg_elecs):
add_ekg(nwbfile, ecog_path, ekg_elecs)
if mic:
# Add microphone recording from room
fs, data = get_analog(blockpath, 1)
nwbfile.add_acquisition(
TimeSeries('microphone',
data,
'audio unit',
rate=fs,
description="audio recording from microphone in room"))
if speakers:
fs, data = get_analog(blockpath, 2)
# Add audio stimulus 1
nwbfile.add_stimulus(
TimeSeries('speaker 1',
data,
'NA',
rate=fs,
description="audio stimulus 1"))
# Add audio stimulus 2
fs, data = get_analog(blockpath, 3)
if fs is not None:
nwbfile.add_stimulus(
TimeSeries('speaker 2',
data,
'NA',
rate=fs,
description='the second stimulus source'))
if anin4:
fs, data = get_analog(blockpath, 4)
nwbfile.add_acquisition(
TimeSeries(anin4,
data,
'aux unit',
rate=fs,
description="aux analog recording"))
# Add bad time segments
if os.path.exists(bad_time_file) and os.stat(bad_time_file).st_size:
bad_time = sio.loadmat(bad_time_file)['badTimeSegments']
for row in bad_time:
nwbfile.add_invalid_time_interval(start_time=row[0],
stop_time=row[1],
tags=('ECoG artifact', ),
timeseries=ecog_ts)
if rest_period is not None:
nwbfile.add_epoch_column(name='label', description='label')
nwbfile.add_epoch(start_time=rest_period[0],
stop_time=rest_period[1],
label='rest_period')
if hilb:
block_hilb_path = os.path.join(hilb_dir, subject_id, blockname,
blockname + '_AA.h5')
file = File(block_hilb_path, 'r')
data = transpose_iter(
file['X']) # transposes data during iterative write
filter_center = file['filter_center'][:]
filter_sigma = file['filter_sigma'][:]
data = H5DataIO(DataChunkIterator(tqdm(data,
desc='writing hilbert data'),
buffer_size=400 * 20),
compression='gzip')
decomp_series = DecompositionSeries(
name='LFPDecompositionSeries',
description='Gaussian band Hilbert transform',
data=data,
rate=400.,
source_timeseries=ecog_ts,
metric='amplitude')
for band_mean, band_stdev in zip(filter_center, filter_sigma):
decomp_series.add_band(band_mean=band_mean, band_stdev=band_stdev)
hilb_mod = nwbfile.create_processing_module(
name='ecephys', description='holds hilbert analysis results')
hilb_mod.add_container(decomp_series)
if include_cortical_surfaces:
subject = ECoGSubject(subject_id=subject_id)
subject.cortical_surfaces = create_cortical_surfaces(
pial_files, subject_id)
else:
subject = Subject(subject_id=subject_id, species='H**o sapiens')
if subject_image_list is not None:
subject = add_images_to_subject(subject, subject_image_list)
if external_subject:
subj_fpath = path.join(out_base_path, subject_id + '.nwb')
if not os.path.isfile(subj_fpath):
subj_nwbfile = NWBFile(session_description=subject_id,
identifier=subject_id,
subject=subject,
session_start_time=datetime(
1900, 1, 1).astimezone(timezone('UTC')))
with NWBHDF5IO(subj_fpath, manager=manager, mode='w') as subj_io:
subj_io.write(subj_nwbfile)
subj_read_io = NWBHDF5IO(subj_fpath, manager=manager, mode='r')
subj_nwbfile = subj_read_io.read()
subject = subj_nwbfile.subject
nwbfile.subject = subject
if parse_transcript:
if parse_transcript == 'CV':
parseout = parse(blockpath, blockname)
df = make_df(parseout, 0, subject_id, align_pos=1)
nwbfile.add_trial_column('cv_transition_time',
'time of CV transition in seconds')
nwbfile.add_trial_column(
'speak',
'if True, subject is speaking. If False, subject is listening')
nwbfile.add_trial_column('condition', 'syllable spoken')
for _, row in df.iterrows():
nwbfile.add_trial(start_time=row['start'],
stop_time=row['stop'],
cv_transition_time=row['align'],
speak=row['mode'] == 'speak',
condition=row['label'])
elif parse_transcript == 'singing':
parseout = parse(blockpath, blockname)
df = make_df(parseout, 0, subject_id, align_pos=0)
if not len(df):
df = pd.DataFrame(parseout)
df['mode'] = 'speak'
df = df.loc[df['label'].astype('bool'), :] # handle empty labels
nwbfile.add_trial_column(
'speak',
'if True, subject is speaking. If False, subject is listening')
nwbfile.add_trial_column('condition', 'syllable spoken')
for _, row in df.iterrows():
nwbfile.add_trial(start_time=row['start'],
stop_time=row['stop'],
speak=row['mode'] == 'speak',
condition=row['label'])
elif parse_transcript == 'emphasis':
parseout = parse(blockpath, blockname)
try:
df = make_df(parseout, 0, subject_id, align_pos=0)
except:
df = pd.DataFrame(parseout)
if not len(df):
df = pd.DataFrame(parseout)
df = df.loc[df['label'].astype('bool'), :] # handle empty labels
nwbfile.add_trial_column('condition', 'word emphasized')
nwbfile.add_trial_column(
'speak',
'if True, subject is speaking. If False, subject is listening')
for _, row in df.iterrows():
nwbfile.add_trial(start_time=row['start'],
stop_time=row['stop'],
speak=True,
condition=row['label'])
elif parse_transcript == 'MOCHA':
nwbfile = create_transcription(nwbfile, transcript_path, blockname)
# behavior
if include_pitch:
if behav_module is None:
behav_module = nwbfile.create_processing_module(
'behavior', 'processing about behavior')
if os.path.isfile(
os.path.join(blockpath, 'pitch_' + blockname + '.mat')):
fs, data = load_pitch(blockpath)
pitch_ts = TimeSeries(
data=data,
rate=fs,
unit='Hz',
name='pitch',
description=
'Pitch as extracted from Praat. NaNs mark unvoiced regions.')
behav_module.add_container(
BehavioralTimeSeries(name='pitch', time_series=pitch_ts))
else:
print('No pitch file for ' + blockname)
if include_intensity:
if behav_module is None:
behav_module = nwbfile.create_processing_module(
'behavior', 'processing about behavior')
if os.path.isfile(
os.path.join(blockpath, 'intensity_' + blockname + '.mat')):
fs, data = load_pitch(blockpath)
intensity_ts = TimeSeries(
data=data,
rate=fs,
unit='dB',
name='intensity',
description='Intensity of speech in dB extracted from Praat.')
behav_module.add_container(
BehavioralTimeSeries(name='intensity',
time_series=intensity_ts))
else:
print('No intensity file for ' + blockname)
# Export the NWB file
with NWBHDF5IO(outpath, manager=manager, mode='w') as io:
io.write(nwbfile)
if external_subject:
subj_read_io.close()
if hilb:
file.close()
# read check
with NWBHDF5IO(outpath, manager=manager, mode='r') as io:
io.read()
def convert_data(self,
nwbfile: NWBFile,
metadata: dict,
stub_test: bool = False):
session_path = self.input_args["folder_path"]
# TODO: check/enforce format?
all_shank_channels = metadata["all_shank_channels"]
special_electrode_dict = metadata.get("special_electrodes", [])
lfp_channels = metadata["lfp_channels"]
lfp_sampling_rate = metadata["lfp_sampling_rate"]
spikes_nsamples = metadata["spikes_nsamples"]
shank_channels = metadata["shank_channels"]
n_total_channels = metadata["n_total_channels"]
subject_path, session_id = os.path.split(session_path)
_, all_channels_lfp_data = read_lfp(session_path,
stub=stub_test,
n_channels=n_total_channels)
try:
lfp_data = all_channels_lfp_data[:, all_shank_channels]
except IndexError:
lfp_data = all_channels_lfp_data
lfp_ts = write_lfp(
nwbfile,
lfp_data,
lfp_sampling_rate,
name=metadata["lfp"]["name"],
description=metadata["lfp"]["description"],
electrode_inds=None,
)
# TODO: error checking on format?
for special_electrode in special_electrode_dict:
ts = TimeSeries(
name=special_electrode["name"],
description=special_electrode["description"],
data=all_channels_lfp_data[:, special_electrode["channel"]],
rate=lfp_sampling_rate,
unit="V",
resolution=np.nan,
)
nwbfile.add_acquisition(ts)
for ref_name, lfp_channel in lfp_channels.items():
try:
all_lfp_phases = []
for passband in ("theta", "gamma"):
lfp_fft = filter_lfp(
lfp_data[:, all_shank_channels == lfp_channel].ravel(),
lfp_sampling_rate,
passband=passband)
lfp_phase, _ = hilbert_lfp(lfp_fft)
all_lfp_phases.append(lfp_phase[:, np.newaxis])
decomp_series_data = np.dstack(all_lfp_phases)
# TODO: should units or metrics be metadata?
decomp_series = DecompositionSeries(
name=metadata["lfp_decomposition"][ref_name]["name"],
description=metadata["lfp_decomposition"][ref_name]
["description"],
data=decomp_series_data,
rate=lfp_sampling_rate,
source_timeseries=lfp_ts,
metric="phase",
unit="radians",
)
# TODO: the band limits should be extracted from parse_passband in band_analysis?
decomp_series.add_band(band_name="theta", band_limits=(4, 10))
decomp_series.add_band(band_name="gamma", band_limits=(30, 80))
check_module(
nwbfile, "ecephys",
"contains processed extracellular electrophysiology data"
).add_data_interface(decomp_series)
except IndexError:
print(
"Unable to index lfp data for decomposition series - skipping"
)
write_spike_waveforms(nwbfile,
session_path,
spikes_nsamples=spikes_nsamples,
shank_channels=shank_channels,
stub_test=stub_test)
def yuta2nwb(
session_path='D:/BuzsakiData/SenzaiY/YutaMouse41/YutaMouse41-150903',
# '/Users/bendichter/Desktop/Buzsaki/SenzaiBuzsaki2017/YutaMouse41/YutaMouse41-150903',
subject_xls=None,
include_spike_waveforms=True,
stub=True,
cache_spec=True):
subject_path, session_id = os.path.split(session_path)
fpath_base = os.path.split(subject_path)[0]
identifier = session_id
mouse_number = session_id[9:11]
if '-' in session_id:
subject_id, date_text = session_id.split('-')
b = False
else:
subject_id, date_text = session_id.split('b')
b = True
if subject_xls is None:
subject_xls = os.path.join(subject_path,
'YM' + mouse_number + ' exp_sheet.xlsx')
else:
if not subject_xls[-4:] == 'xlsx':
subject_xls = os.path.join(subject_xls,
'YM' + mouse_number + ' exp_sheet.xlsx')
session_start_time = dateparse(date_text, yearfirst=True)
df = pd.read_excel(subject_xls)
subject_data = {}
for key in [
'genotype', 'DOB', 'implantation', 'Probe', 'Surgery',
'virus injection', 'mouseID'
]:
names = df.iloc[:, 0]
if key in names.values:
subject_data[key] = df.iloc[np.argmax(names == key), 1]
if isinstance(subject_data['DOB'], datetime):
age = session_start_time - subject_data['DOB']
else:
age = None
subject = Subject(subject_id=subject_id,
age=str(age),
genotype=subject_data['genotype'],
species='mouse')
nwbfile = NWBFile(
session_description='mouse in open exploration and theta maze',
identifier=identifier,
session_start_time=session_start_time.astimezone(),
file_create_date=datetime.now().astimezone(),
experimenter='Yuta Senzai',
session_id=session_id,
institution='NYU',
lab='Buzsaki',
subject=subject,
related_publications='DOI:10.1016/j.neuron.2016.12.011')
print('reading and writing raw position data...', end='', flush=True)
ns.add_position_data(nwbfile, session_path)
shank_channels = ns.get_shank_channels(session_path)[:8]
nshanks = len(shank_channels)
all_shank_channels = np.concatenate(shank_channels)
print('setting up electrodes...', end='', flush=True)
hilus_csv_path = os.path.join(fpath_base, 'early_session_hilus_chans.csv')
lfp_channel = get_reference_elec(subject_xls,
hilus_csv_path,
session_start_time,
session_id,
b=b)
custom_column = [{
'name': 'theta_reference',
'description':
'this electrode was used to calculate LFP canonical bands',
'data': all_shank_channels == lfp_channel
}]
ns.write_electrode_table(nwbfile,
session_path,
custom_columns=custom_column,
max_shanks=max_shanks)
print('reading raw electrode data...', end='', flush=True)
if stub:
# example recording extractor for fast testing
xml_filepath = os.path.join(session_path, session_id + '.xml')
xml_root = et.parse(xml_filepath).getroot()
acq_sampling_frequency = float(
xml_root.find('acquisitionSystem').find('samplingRate').text)
num_channels = 4
num_frames = 10000
X = np.random.normal(0, 1, (num_channels, num_frames))
geom = np.random.normal(0, 1, (num_channels, 2))
X = (X * 100).astype(int)
sre = se.NumpyRecordingExtractor(
timeseries=X, sampling_frequency=acq_sampling_frequency, geom=geom)
else:
nre = se.NeuroscopeRecordingExtractor('{}/{}.dat'.format(
session_path, session_id))
sre = se.SubRecordingExtractor(nre, channel_ids=all_shank_channels)
print('writing raw electrode data...', end='', flush=True)
se.NwbRecordingExtractor.add_electrical_series(sre, nwbfile)
print('done.')
print('reading spiking units...', end='', flush=True)
if stub:
spike_times = [200, 300, 400]
num_frames = 10000
allshanks = []
for k in range(nshanks):
SX = se.NumpySortingExtractor()
for j in range(len(spike_times)):
SX.add_unit(unit_id=j + 1,
times=np.sort(
np.random.uniform(0, num_frames,
spike_times[j])))
allshanks.append(SX)
se_allshanks = se.MultiSortingExtractor(allshanks)
se_allshanks.set_sampling_frequency(acq_sampling_frequency)
else:
se_allshanks = se.NeuroscopeMultiSortingExtractor(session_path,
keep_mua_units=False)
electrode_group = []
for shankn in np.arange(1, nshanks + 1, dtype=int):
for id in se_allshanks.sortings[shankn - 1].get_unit_ids():
electrode_group.append(nwbfile.electrode_groups['shank' +
str(shankn)])
df_unit_features = get_UnitFeatureCell_features(fpath_base, session_id,
session_path)
celltype_names = []
for celltype_id, region_id in zip(df_unit_features['fineCellType'].values,
df_unit_features['region'].values):
if celltype_id == 1:
if region_id == 3:
celltype_names.append('pyramidal cell')
elif region_id == 4:
celltype_names.append('granule cell')
else:
raise Exception('unknown type')
elif not np.isfinite(celltype_id):
celltype_names.append('missing')
else:
celltype_names.append(celltype_dict[celltype_id])
# Add custom column data into the SortingExtractor so it can be written by the converter
# Note there is currently a hidden assumption that the way in which the NeuroscopeSortingExtractor
# merges the cluster IDs matches one-to-one with the get_UnitFeatureCell_features extraction
property_descriptions = {
'cell_type': 'name of cell type',
'global_id': 'global id for cell for entire experiment',
'shank_id': '0-indexed id of cluster of shank',
'electrode_group': 'the electrode group that each spike unit came from'
}
property_values = {
'cell_type': celltype_names,
'global_id': df_unit_features['unitID'].values,
'shank_id': [x - 2 for x in df_unit_features['unitIDshank'].values],
# - 2 b/c the get_UnitFeatureCell_features removes 0 and 1 IDs from each shank
'electrode_group': electrode_group
}
for unit_id in se_allshanks.get_unit_ids():
for property_name in property_descriptions.keys():
se_allshanks.set_unit_property(
unit_id, property_name,
property_values[property_name][unit_id])
se.NwbSortingExtractor.write_sorting(
se_allshanks,
nwbfile=nwbfile,
property_descriptions=property_descriptions)
print('done.')
# Read and write LFP's
print('reading LFPs...', end='', flush=True)
lfp_fs, all_channels_lfp_data = ns.read_lfp(session_path, stub=stub)
lfp_data = all_channels_lfp_data[:, all_shank_channels]
print('writing LFPs...', flush=True)
# lfp_data[:int(len(lfp_data)/4)]
lfp_ts = ns.write_lfp(nwbfile,
lfp_data,
lfp_fs,
name='lfp',
description='lfp signal for all shank electrodes')
# Read and add special environmental electrodes
for name, channel in special_electrode_dict.items():
ts = TimeSeries(
name=name,
description=
'environmental electrode recorded inline with neural data',
data=all_channels_lfp_data[:, channel],
rate=lfp_fs,
unit='V',
#conversion=np.nan,
resolution=np.nan)
nwbfile.add_acquisition(ts)
# compute filtered LFP
print('filtering LFP...', end='', flush=True)
all_lfp_phases = []
for passband in ('theta', 'gamma'):
lfp_fft = filter_lfp(
lfp_data[:, all_shank_channels == lfp_channel].ravel(),
lfp_fs,
passband=passband)
lfp_phase, _ = hilbert_lfp(lfp_fft)
all_lfp_phases.append(lfp_phase[:, np.newaxis])
data = np.dstack(all_lfp_phases)
print('done.', flush=True)
if include_spike_waveforms:
print('writing waveforms...', end='', flush=True)
nshanks = min((max_shanks, len(ns.get_shank_channels(session_path))))
for shankn in np.arange(nshanks, dtype=int) + 1:
# Get spike activty from .spk file on a per-shank and per-sample basis
ns.write_spike_waveforms(nwbfile, session_path, shankn, stub=stub)
print('done.', flush=True)
# Get the LFP Decomposition Series
decomp_series = DecompositionSeries(
name='LFPDecompositionSeries',
description='Theta and Gamma phase for reference LFP',
data=data,
rate=lfp_fs,
source_timeseries=lfp_ts,
metric='phase',
unit='radians')
decomp_series.add_band(band_name='theta', band_limits=(4, 10))
decomp_series.add_band(band_name='gamma', band_limits=(30, 80))
check_module(nwbfile, 'ecephys',
'contains processed extracellular electrophysiology data'
).add_data_interface(decomp_series)
[nwbfile.add_stimulus(x) for x in ns.get_events(session_path)]
# create epochs corresponding to experiments/environments for the mouse
sleep_state_fpath = os.path.join(session_path,
'{}--StatePeriod.mat'.format(session_id))
exist_pos_data = any(
os.path.isfile(
os.path.join(session_path, '{}__{}.mat'.format(
session_id, task_type['name']))) for task_type in task_types)
if exist_pos_data:
nwbfile.add_epoch_column('label', 'name of epoch')
for task_type in task_types:
label = task_type['name']
file = os.path.join(session_path, session_id + '__' + label + '.mat')
if os.path.isfile(file):
print('loading position for ' + label + '...', end='', flush=True)
pos_obj = Position(name=label + '_position')
matin = loadmat(file)
tt = matin['twhl_norm'][:, 0]
exp_times = find_discontinuities(tt)
if 'conversion' in task_type:
conversion = task_type['conversion']
else:
conversion = np.nan
for pos_type in ('twhl_norm', 'twhl_linearized'):
if pos_type in matin:
pos_data_norm = matin[pos_type][:, 1:]
spatial_series_object = SpatialSeries(
name=label + '_{}_spatial_series'.format(pos_type),
data=H5DataIO(pos_data_norm, compression='gzip'),
reference_frame='unknown',
conversion=conversion,
resolution=np.nan,
timestamps=H5DataIO(tt, compression='gzip'))
pos_obj.add_spatial_series(spatial_series_object)
check_module(
nwbfile, 'behavior',
'contains processed behavioral data').add_data_interface(
pos_obj)
for i, window in enumerate(exp_times):
nwbfile.add_epoch(start_time=window[0],
stop_time=window[1],
label=label + '_' + str(i))
print('done.')
# there are occasional mismatches between the matlab struct and the neuroscope files
# regions: 3: 'CA3', 4: 'DG'
trialdata_path = os.path.join(session_path,
session_id + '__EightMazeRun.mat')
if os.path.isfile(trialdata_path):
trials_data = loadmat(trialdata_path)['EightMazeRun']
trialdatainfo_path = os.path.join(fpath_base, 'EightMazeRunInfo.mat')
trialdatainfo = [
x[0] for x in loadmat(trialdatainfo_path)['EightMazeRunInfo'][0]
]
features = trialdatainfo[:7]
features[:2] = 'start_time', 'stop_time',
[
nwbfile.add_trial_column(x, 'description')
for x in features[4:] + ['condition']
]
for trial_data in trials_data:
if trial_data[3]:
cond = 'run_left'
else:
cond = 'run_right'
nwbfile.add_trial(start_time=trial_data[0],
stop_time=trial_data[1],
condition=cond,
error_run=trial_data[4],
stim_run=trial_data[5],
both_visit=trial_data[6])
"""
mono_syn_fpath = os.path.join(session_path, session_id+'-MonoSynConvClick.mat')
matin = loadmat(mono_syn_fpath)
exc = matin['FinalExcMonoSynID']
inh = matin['FinalInhMonoSynID']
#exc_obj = CatCellInfo(name='excitatory_connections',
# indices_values=[], cell_index=exc[:, 0] - 1, indices=exc[:, 1] - 1)
#module_cellular.add_container(exc_obj)
#inh_obj = CatCellInfo(name='inhibitory_connections',
# indices_values=[], cell_index=inh[:, 0] - 1, indices=inh[:, 1] - 1)
#module_cellular.add_container(inh_obj)
"""
if os.path.isfile(sleep_state_fpath):
matin = loadmat(sleep_state_fpath)['StatePeriod']
table = TimeIntervals(name='states',
description='sleep states of animal')
table.add_column(name='label', description='sleep state')
data = []
for name in matin.dtype.names:
for row in matin[name][0][0]:
data.append({
'start_time': row[0],
'stop_time': row[1],
'label': name
})
[
table.add_row(**row)
for row in sorted(data, key=lambda x: x['start_time'])
]
check_module(nwbfile, 'behavior',
'contains behavioral data').add_data_interface(table)
print('writing NWB file...', end='', flush=True)
if stub:
out_fname = session_path + '_stub.nwb'
else:
out_fname = session_path + '.nwb'
with NWBHDF5IO(out_fname, mode='w') as io:
io.write(nwbfile, cache_spec=cache_spec)
print('done.')
print('testing read...', end='', flush=True)
# test read
with NWBHDF5IO(out_fname, mode='r') as io:
io.read()
print('done.')