def get_files():
with open('dset_config.json', 'r') as fo:
CONFIG = json.load(fo)
DATA_DIR = op.abspath('/home/data/nbc/external-datasets/ds001491/')
all_info = {}
for dset_name in list(CONFIG.keys())[:3]:
layout = BIDSLayout(op.join(DATA_DIR, dset_name))
cfg = CONFIG[dset_name]
task = cfg['task']
dset_info = {}
for sub in layout.get_subjects():
runs = layout.get_runs(subject=sub, task=task)
sub_info = {}
for run in runs:
run_info = {}
run_info['files'] = []
run_info['echo_times'] = []
for echo in sorted(
layout.get_echoes(subject=sub, task=task, run=run)):
raw_files = layout.get(subject=sub,
task=task,
run=run,
echo=echo,
extensions='.nii.gz')
preproc_files = layout.get(subject=sub,
task=task,
run=run,
root='afni-step1',
echo=echo,
extensions='.nii.gz',
desc='realign')
preproc_files = raw_files[:]
if len(preproc_files) != 1:
print(preproc_files)
raise Exception('BAD')
# Replace filename with path when using new version of bids
run_info['files'].append(preproc_files[0].filename)
metadata = layout.get_metadata(preproc_files[0].filename)
run_info['echo_times'].append(metadata['EchoTime'])
sub_info[run] = run_info
dset_info[sub] = sub_info
all_info[dset_name] = dset_info
with open('all_files.json', 'w') as fo:
json.dump(all_info, fo, indent=4, sort_keys=True)
dataDir = '/tmp/Data/ds114'
# Creating the layout object for this BIDS data set
layout = BIDSLayout(dataDir)
# subjects
subjList = layout.get_subjects()
# sessions
sesList = layout.get_sessions()
# tasks
taskList = layout.get_tasks()
# runs
runList = layout.get_runs()
# List of all fMRI data for subject 01
fMRI_sub01 = layout.get(subject='01',
suffix='bold',
extension=['nii', 'nii.gz'],
return_type='file')
# Lets focus on test session
fMRI_sub01_test = layout.get(subject='01',
session='test',
suffix='bold',
extension=['nii', 'nii.gz'],
return_type='file')
# A list of files associated with the covert verb generation
class events_class(object):
def __init__(self, base_dir, task_id, sub_ids=None, verbose=True):
assert type(base_dir) == str, "base_dir should be type(str)"
assert type(task_id) == str, "task_id should be type(str)"
self.base_dir = base_dir
self.layout = BIDSLayout(base_dir)
self.task_id = task_id
if sub_ids == None:
self.sub_ids = self.layout.get_subjects(task=task_id)
else:
assert type(sub_ids) == list, "sub_ids should be type(list)"
for i in sub_ids:
assert type(
i) == str, "elements in sub_ids should be type(str)"
assert any(i == j
for j in self.layout.get_subjects(task=task_id)
), f"cannot find sub_ids with {task} task data"
if verbose:
print(f'{len(self.sub_ids)} subjects in {self.task_id} task')
def get_timing(self,
trial_type_cols=[],
trimTRby=10,
software='AFNI',
write=True):
'''
Throws timing information for each subject into dictionary
'''
assert (software == 'AFNI'
or software == 'SPM'), "software should be AFNI or SPM"
self.events_dict = {}
self.events_dict[self.task_id] = {}
for sub_identifier in self.sub_ids:
self.events_dict[self.task_id][sub_identifier] = {}
run_ids = self.layout.get_runs(subject=sub_identifier,
task=self.task_id)
for run_identifier in run_ids:
__eventFile__ = self.layout.get(suffix='events',
task=self.task_id,
run=run_identifier,
extension='tsv',
return_type='file')
# get df of events information
__trialInfo__ = pd.read_table(__eventFile__[0])
if run_identifier == 1:
onsets = {}
durations = {}
if trial_type_cols == []: #if no columns for trial_types specified
for column in __trialInfo__.columns[2:]:
__conditions__ = sorted(
list(set(__trialInfo__[column])))
# extracting onset and duration information for task, subject, run, trial_type
if run_identifier == 1:
onsets[column] = {}
durations[column] = {}
for itrial in __conditions__:
if run_identifier == 1:
onsets[column][itrial] = {}
durations[column][itrial] = {}
try:
onsets[column][itrial][run_identifier] = (
list(__trialInfo__[__trialInfo__[column] ==
itrial].onset -
trimTRby)
) # subtracting trimTRby due to remove dummy scans
durations[column][itrial][run_identifier] = (
list(__trialInfo__[__trialInfo__[column] ==
itrial].duration))
except KeyError:
onsets[column][itrial] = {}
durations[column][itrial] = {}
onsets[column][itrial][run_identifier] = (
list(__trialInfo__[__trialInfo__[column] ==
itrial].onset -
trimTRby)
) # subtracting trimTRby due to remove dummy scans
durations[column][itrial][run_identifier] = (
list(__trialInfo__[__trialInfo__[column] ==
itrial].duration))
else:
for t1 in trial_type_cols:
assert any(t1 == t2 for t2 in __trialInfo__.columns
), f"{t1} info is not in *events.tsv"
for column in trial_type_cols:
__conditions__ = sorted(
list(set(__trialInfo__[column])))
# extracting onset and duration information for task, subject, run, trial_type
if run_identifier == 1:
onsets[column] = {}
durations[column] = {}
for itrial in __conditions__:
if run_identifier == 1:
onsets[column][itrial] = {}
durations[column][itrial] = {}
try:
onsets[column][itrial][run_identifier] = (
list(__trialInfo__[__trialInfo__[column] ==
itrial].onset -
trimTRby)
) # subtracting trimTRby due to remove dummy scans
durations[column][itrial][run_identifier] = (
list(__trialInfo__[__trialInfo__[column] ==
itrial].duration))
except KeyError:
onsets[column][itrial] = {}
durations[column][itrial] = {}
onsets[column][itrial][run_identifier] = (
list(__trialInfo__[__trialInfo__[column] ==
itrial].onset -
trimTRby)
) # subtracting trimTRby due to remove dummy scans
durations[column][itrial][run_identifier] = (
list(__trialInfo__[__trialInfo__[column] ==
itrial].duration))
self.events_dict[self.task_id][sub_identifier]['onsets'] = onsets
self.events_dict[
self.task_id][sub_identifier]['durations'] = durations
if write == True:
if trial_type_cols == []: #if no columns for trial_types specified
for column in __trialInfo__.columns[2:]:
if column == 'response_time': #cannot handle response_times yet
continue
writeToPath = os.path.join(self.base_dir,
'derivatives', 'timing',
f'sub-{sub_identifier}',
column)
if not os.path.exists(writeToPath):
os.makedirs(writeToPath)
for (onset_key, onset_val), (dur_key, dur_val) in zip(
onsets[column].items(),
durations[column].items()):
assert (
onset_key == dur_key
), "onset conditions and duration conditions do not match!"
with open(
f'{writeToPath}/sub-{sub_identifier}_task-{self.task_id}_timing-{onset_key}.txt',
'w') as filehandle:
for (run_onsets,
run_durs) in zip(onset_val.values(),
dur_val.values()):
filehandle.writelines(
"%f:%f\t" % (onset_time, dur_time)
for (onset_time, dur_time
) in zip(run_onsets, run_durs))
filehandle.write("\n")
else:
for column in trial_type_cols:
if column == 'response_time': #cannot handle response_times yet
continue
writeToPath = os.path.join(self.base_dir,
'derivatives', 'timing',
f'sub-{sub_identifier}',
column)
if not os.path.exists(writeToPath):
os.makedirs(writeToPath)
for (onset_key, onset_val), (dur_key, dur_val) in zip(
onsets[column].items(),
durations[column].items()):
assert (
onset_key == dur_key
), "onset conditions and duration conditions do not match!"
with open(
f'{writeToPath}/sub-{sub_identifier}_task-{self.task_id}_timing-{onset_key}.txt',
'w') as filehandle:
for (run_onsets,
run_durs) in zip(onset_val.values(),
dur_val.values()):
filehandle.writelines(
"%f:%f\t" % (onset_time, dur_time)
for (onset_time, dur_time
) in zip(run_onsets, run_durs))
filehandle.write("\n")
"""Simple workflow component that queries a BIDS dataset and handles files individually"""
from bids.layout import BIDSLayout
from nipype.interfaces.io import BIDSDataGrabber
from nipype import Workflow, Node, Function, MapNode
from niworkflows.interfaces.bids import DerivativesDataSink
import os
# Set project variables and set up workflow
project_dir = os.path.abspath('../..')
data_dir = 'data/ds000171/'
wf = Workflow(name='wf', base_dir = os.getcwd())
# Leverage BIDS to get subject list
layout = BIDSLayout(os.path.join(project_dir, data_dir))
subject_list = layout.get_subjects()[:2]
runs = layout.get_runs()[:2]
"""
==============================
Anatomical Image Input
==============================
"""
grab_anat = Node(BIDSDataGrabber(), name = 'grab_anat')
grab_anat.inputs.base_dir = os.path.join(project_dir, data_dir)
# Iterate through subjects; [:2] for only 2 subjects to keep memory usage low
grab_anat.iterables = ('subject', subject_list)
# Define filetypes to grab, and how ouput will be accesses by other nodes
grab_anat.inputs.output_query = {'T1w': dict(extension=['nii.gz'], suffix='T1w')}
