def GetAllPredicates(nidm_files_tuple):
pred_set = set()
for file in nidm_files_tuple:
rdf_graph = OpenGraph(file)
predicates = rdf_graph.predicates()
for p in predicates:
pred_set.add(p)
return pred_set
def getSubjectIDfromUUID(nidm_file_tuples, subject_uuid):
for file in nidm_file_tuples:
rdf_graph = OpenGraph(file)
id_generator = rdf_graph.objects(
subject=subject_uuid, predicate=Constants.NDAR['src_subject_id'])
for id in id_generator:
return id
return None
def getProjects(nidm_file_tuples):
projects = []
for file in nidm_file_tuples:
rdf_graph = OpenGraph(file)
#find all the sessions
for (project, p, o) in rdf_graph.triples(
(None, isa, Constants.NIDM['Project'])):
projects.append(project)
return projects
def getActivityData(nidm_file_tuples, acquisition_id):
acquisition_uri = expandID(acquisition_id, Constants.NIIRI)
result = []
category = None
for file in nidm_file_tuples:
rdf_graph = OpenGraph(file)
# find everything generated by the acquisition
for (data_object, p1, o1) in rdf_graph.triples(
(None, Constants.PROV['wasGeneratedBy'], acquisition_uri)):
# make sure this is an acquisition object
if (data_object, isa,
Constants.NIDM['AcquisitionObject']) in rdf_graph:
category = 'instrument'
# iterate over all the items in the acquisition object
for (s, p, o) in rdf_graph.triples((data_object, None, None)):
dti = getDataTypeInfo(rdf_graph, p)
if (dti):
# there is a DataElement describing this predicate
value_type = makeValueTypeFromDataTypeInfo(
value=trimWellKnownURIPrefix(o),
data_type_info_tuple=dti)
result.append(value_type)
else:
#Don't know exactly what this is so just set a label and be done.
if (data_object, isa,
Constants.ONLI['assessment-instrument']
) in rdf_graph:
result.append(
makeValueType(value=trimWellKnownURIPrefix(o),
label=simplifyURIWithPrefix(
nidm_file_tuples, str(p))))
#result[ simplifyURIWithPrefix(nidm_file_list, str(p)) ] = trimWellKnownURIPrefix(o)
else:
result.append(
makeValueType(value=trimWellKnownURIPrefix(o),
label=URITail(str(p))))
# result[ URITail(str(p))] = trimWellKnownURIPrefix(o)
# or maybe it's a stats collection
elif isAStatCollection(nidm_file_tuples, data_object):
category = 'derivative'
for (s, p, o) in rdf_graph.triples((data_object, None, None)):
cde = getDataTypeInfo(rdf_graph, p)
result.append(
makeValueTypeFromDataTypeInfo(
value=str(o), data_type_info_tuple=cde))
# result[ URITail(str(p)) ] = str(o)
return ActivityData(category=category,
uuid=trimWellKnownURIPrefix(acquisition_uri),
data=result)
def getSubjectUUIDsfromID(nidm_file_tuples, sub_id):
uuids = []
for file in nidm_file_tuples:
rdf_graph = OpenGraph(file)
result = rdf_graph.triples(
(None, Constants.NDAR['src_subject_id'], None))
for (s, p, o) in result:
if str(o) == str(sub_id):
uuids.append(URITail(s))
return uuids
def getAcquisitions(nidm_file_tuples, session_id):
session_uri = expandID(session_id, Constants.NIIRI)
acquisitions = []
for file in nidm_file_tuples:
rdf_graph = OpenGraph(file)
#find all the sessions
for (acq, p, o) in rdf_graph.triples((None, isPartOf, session_uri)):
#check if it is a acquisition
if (acq, isa, Constants.NIDM['Acquisition']) in rdf_graph:
acquisitions.append(acq)
return acquisitions
def getActivities(nidm_file_tuples, subject_id):
activities = set([])
subject_uri = expandID(subject_id, Constants.NIIRI)
for file in nidm_file_tuples:
rdf_graph = OpenGraph(file)
for blank_node in rdf_graph.subjects(predicate=Constants.PROV['agent'],
object=subject_uri):
for activity in rdf_graph.subjects(
predicate=Constants.PROV['qualifiedAssociation'],
object=blank_node):
if (activity, isa, Constants.PROV['Activity']) in rdf_graph:
activities.add(activity)
return activities
def getSessions(nidm_file_tuples, project_id):
project_uri = expandID(project_id, Constants.NIIRI)
sessions = []
for file in nidm_file_tuples:
rdf_graph = OpenGraph(file)
#find all the sessions
for (session, p, o) in rdf_graph.triples(
(None, isa, Constants.NIDM['Session'])):
#check if it is part of our project
if (session, isPartOf, project_uri) in rdf_graph:
sessions.append(session)
return sessions
def getActivityData(nidm_file_list, acquisition_id):
acquisition_uri = expandID(acquisition_id, Constants.NIIRI)
result = []
category = None
for file in nidm_file_list:
rdf_graph = OpenGraph(file)
# find everything generated by the acquisition
for (data_object, p1, o1) in rdf_graph.triples(
(None, Constants.PROV['wasGeneratedBy'], acquisition_uri)):
# make sure this is an acquisition object
if (data_object, isa,
Constants.NIDM['AcquisitionObject']) in rdf_graph:
category = 'instrument'
# iterate over all the items in the acquisition object
for (s, p, o) in rdf_graph.triples((data_object, None, None)):
# if this is a onli:assessment-instrument then use the prefix names
if (data_object, isa,
Constants.ONLI['assessment-instrument']
) in rdf_graph:
result.append(
makeValueType(value=trimWellKnownURIPrefix(o),
label=simplifyURIWithPrefix(
nidm_file_list, str(p))))
#result[ simplifyURIWithPrefix(nidm_file_list, str(p)) ] = trimWellKnownURIPrefix(o)
else:
result.append(
makeValueType(value=trimWellKnownURIPrefix(o),
label=URITail(str(p))))
# result[ URITail(str(p))] = trimWellKnownURIPrefix(o)
# or maybe it's a stats collection
elif isAStatCollection(nidm_file_list, data_object):
category = 'derivative'
for (s, p, o) in rdf_graph.triples((data_object, None, None)):
cde = getDataTypeInfo(rdf_graph, p)
result.append(
makeValueType(value=str(o),
label=str(cde['label']),
hasUnit=str(cde['hasUnit']),
datumType=str(cde['datumType']),
measureOf=str(cde['measureOf']),
isAbout=str(cde['isAbout']),
dataElement=URITail(str(p))))
# result[ URITail(str(p)) ] = str(o)
return ActivityData(category=category,
uuid=trimWellKnownURIPrefix(acquisition_uri),
data=result)
def getSubject(nidm_file_tuples, acquisition_id):
acquisition_uri = expandID(acquisition_id, Constants.NIIRI)
subjects = []
for file in nidm_file_tuples:
rdf_graph = OpenGraph(file)
#find all the sessions
for (acq, p, blank) in rdf_graph.triples(
(acquisition_uri, Constants.PROV['qualifiedAssociation'], None)):
for (s2, p2, sub) in rdf_graph.triples(
(blank, Constants.PROV['agent'], None)):
if (blank, Constants.PROV['hadRole'],
Constants.SIO['Subject']) in rdf_graph:
return sub
return None
def GetDataelements(nidm_files_tuple):
result = {"data_elements": {"uuid": [], "label": [], "data_type_info": []}}
found_uris = set()
for file in nidm_files_tuple:
rdf_graph = OpenGraph(file)
#find all the datatypes
for de_uri in rdf_graph.subjects(predicate=isa,
object=Constants.NIDM['DataElement']):
if de_uri not in found_uris: # don't add duplicates
dti = getDataTypeInfo(rdf_graph, de_uri)
result['data_elements']['uuid'].append(
str(dti['dataElementURI']))
result['data_elements']['label'].append(str(dti['label']))
result['data_elements']['data_type_info'].append(dti)
found_uris.add(de_uri)
# find all the datatypes
for de_uri in rdf_graph.subjects(
predicate=isa, object=Constants.NIDM['PersonalDataElement']):
if de_uri not in found_uris: # don't add duplicates
dti = getDataTypeInfo(rdf_graph, de_uri)
result['data_elements']['uuid'].append(
str(dti['dataElementURI']))
result['data_elements']['label'].append(str(dti['label']))
result['data_elements']['data_type_info'].append(dti)
found_uris.add(de_uri)
# now look for any of the CDEs
all_predicates = GetAllPredicates(nidm_files_tuple)
cde_graph = nidm.experiment.CDE.getCDEs()
cde_types = cde_graph.subjects(predicate=Constants.RDFS['subClassOf'],
object=Constants.NIDM['DataElement'])
cde_type_set = set() # i.e. fs:DataElement
known_cde_types = set() # i.e. fs_003579
for t in cde_types:
cde_type_set.add(t)
for s in cde_graph.subjects(predicate=isa, object=t):
known_cde_types.add(s)
for predicate in all_predicates:
if predicate in known_cde_types:
dti = getDataTypeInfo(cde_graph, predicate)
result['data_elements']['uuid'].append(str(dti['dataElementURI']))
result['data_elements']['label'].append(str(dti['label']))
result['data_elements']['data_type_info'].append(dti)
return result
def getNamespaceLookup(nidm_file_tuples):
names = {}
for f in nidm_file_tuples:
rdf_graph = OpenGraph(f)
for (prefix, uri) in rdf_graph.namespace_manager.namespaces():
if not str(uri) in names:
names[str(uri)] = prefix
return names
def GetProjectAttributes(nidm_files_tuple, project_id):
result = {
ACQUISITION_MODALITY: set([]),
IMAGE_CONTRAST_TYPE: set([]),
IMAGE_USAGE_TYPE: set([]),
TASK: set([])
}
project_uuid = expandUUID(project_id)
for file in nidm_files_tuple:
rdf_graph = OpenGraph(file)
#find all the projects
for (project, pred, o) in rdf_graph.triples(
(None, None, Constants.NIDM['Project'])):
#check if it is our project
if str(project) == str(project_uuid):
# get all the basic data from the project
for (proj, predicate, object) in rdf_graph.triples(
(project, None, None)):
result[matchPrefix(str(predicate))] = str(object)
# now drill into the acquisition objects to get some specific
# elements: AcquisitionModality, ImageContrastType, ImageUsageType, Task
sessions = getSessions(nidm_files_tuple, project_id)
for s in sessions:
acquistions = getAcquisitions(nidm_files_tuple, s)
for a in acquistions:
acq_obj = getActivityData(nidm_files_tuple, a)
for de in acq_obj.data:
if de.label == 'hadAcquisitionModality':
result[ACQUISITION_MODALITY].add(de.value)
if de.label == 'hadImageContrastType':
result[IMAGE_CONTRAST_TYPE].add(de.value)
if de.label == 'hadImageUsageType':
result[IMAGE_USAGE_TYPE].add(de.value)
if de.label == 'Task':
result[TASK].add(de.value)
# de-set-ify items so they will play nice with JSON later
result[ACQUISITION_MODALITY] = list(result[ACQUISITION_MODALITY])
result[IMAGE_CONTRAST_TYPE] = list(result[IMAGE_CONTRAST_TYPE])
result[IMAGE_USAGE_TYPE] = list(result[IMAGE_USAGE_TYPE])
result[TASK] = list(result[TASK])
return result
def isAStatCollection(nidm_file_tuples, uri):
for file in nidm_file_tuples:
rdf_graph = OpenGraph(file)
if ((uri, isa, Constants.NIDM['FSStatsCollection']) in rdf_graph ) or \
((uri, isa, Constants.NIDM['FSLStatsCollection']) in rdf_graph) or \
((uri, isa, Constants.NIDM['ANTSStatsCollection']) in rdf_graph) :
return True
return False
def getActivities(nidm_file_tuples, subject_id):
activities = set([])
# if we were passed in a sub_id rather than a UUID, lookup the associated UUID. (we might get multiple!)
if validate_uuid(URITail(subject_id)):
sub_uris = [subject_id]
else:
sub_uris = getSubjectUUIDsfromID(nidm_file_tuples, subject_id)
for file in nidm_file_tuples:
rdf_graph = OpenGraph(file)
for subject_uri in sub_uris:
subject_uri = expandID(subject_uri, Constants.NIIRI)
for blank_node in rdf_graph.subjects(
predicate=Constants.PROV['agent'], object=subject_uri):
for activity in rdf_graph.subjects(
predicate=Constants.PROV['qualifiedAssociation'],
object=blank_node):
if (activity, isa,
Constants.PROV['Activity']) in rdf_graph:
activities.add(activity)
return activities
def GetDataelementDetails(nidm_files_tuple, dataelement):
result = {}
for file in nidm_files_tuple:
rdf_graph = OpenGraph(file)
for de_uri in rdf_graph.subjects(predicate=isa,
object=Constants.NIDM['DataElement']):
dti = getDataTypeInfo(rdf_graph, de_uri)
# check if this is the correct one
if not (dataelement in [
str(dti['label']),
str(dti['dataElement']),
str(dti['dataElementURI'])
]):
continue
for key in dti.keys():
result[key] = dti[key]
result['inProjects'] = set()
# figure out what project the dataelement was used in
uri = dti["dataElementURI"]
a_list = rdf_graph.subjects(predicate=uri)
for a in a_list: # a is an assessment / AcquisitionObject
b_list = rdf_graph.objects(
subject=a, predicate=Constants.PROV['wasGeneratedBy'])
for b in b_list: # b is an Acquisition / Activity
c_list = rdf_graph.objects(
subject=b, predicate=Constants.DCT['isPartOf'])
for c in c_list: # c is a session
d_list = rdf_graph.objects(
subject=c, predicate=Constants.DCT['isPartOf'])
for d in d_list: # d is most likely a project
if d in rdf_graph.subjects(
predicate=isa,
object=Constants.NIDM['Project']):
result['inProjects'].add("{} ({})".format(
str(d), file))
return result # found it, we are done
if result == {}: # didn't find it yet, check the CDEs
cde_graph = nidm.experiment.CDE.getCDEs()
for de_uri in cde_graph.subjects(predicate=isa):
dti = getDataTypeInfo(cde_graph, de_uri)
# check if this is the correct one
if not (dataelement in [
str(dti['label']),
str(dti['dataElement']),
str(dti['dataElementURI'])
]):
continue
for key in dti.keys():
result[key] = dti[key]
result['inProjects'] = set()
result['inProjects'].add("Common Data Element")
for file in nidm_files_tuple:
rdf_graph = OpenGraph(file)
if result['dataElementURI'] in rdf_graph.predicates():
result['inProjects'].add(file)
return result # found it, we are done
return result
